Compare commits

...
Author SHA1 Message Date
Kaoru0822andClaude Opus 5.5 8a8efb8226 fix(javascript): keep unselected sources out of cancellation failures
With source_paths, the analysis loop checked cancellation before it
recognized an outside-analysis-scope file. Cancellation arriving after the
last selected source therefore recorded the next unselected file as a
cancelled analysis failure, and the completion loop skipped it as already
analyzed.

Give out-of-selection classification one owner, used by the loop, the
interruption handler and the completion loop. The cancellation is still
reported as the interruption.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-11 21:36:10 +00:00
1d4c4eb030 fix(adb): reject unsafe host file names when pulling one device file (#1962)
pull_adb_file derived its host file name from the last path segment of a
caller-supplied device path and joined it to the output directory without
checking that the name is usable on the host. A device path ending in a
separator, a dot segment, a control character, a Windows reserved device
name, a trailing dot or space, or a ':' alternate-data-stream marker could
resolve to an unexpected host path or fail opaquely after the transfer.

The single-file path now derives the name with POSIX semantics and refuses
any non-portable segment as invalid_request before creating a directory or
running adb pull. This matches the package path's pending host-name policy
and never renames a device-reported file.

Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
2026-10-12 05:33:40 +08:00
9eee709f0c fix(wasm): accept unpadded WABT section rows (#1963)
* fix(wasm): accept unpadded WABT section rows

WABT 1.0.42 right-aligns section names in a nine-column field. DataCount
fills that field, so its objdump row has no leading whitespace. Accept
that legal row. Header recognition, the required field shape, and section
range checks still reject malformed output.

Fixes #1956

* fix(ci): refresh MIPS expectations and Windows test paths

Assert admitted soft-float ABI metadata and its committed storage policy through the public target resolver, retain FP64 rejection, and point the curated Windows lane at the moved filesystem reconstruction tests.

Validation: reproduced all four prior failures; 97 focused tests and check:ci pass. Native Windows and real Ghidra execution were not performed locally.

---------

Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
2026-10-12 05:29:04 +08:00
morluto beb80a07ba test(cli): consolidate schema coverage and assert input refusals
Remove 30 redundant truthy-JSON schema checks. Validate advertised groups and property parity across every primary command, retain required-field checks, and assert structured malformed-input errors. Exercise JEB positional validation rather than an unknown flag.

Validation: 13 public CLI adapter tests, combined checks, and temporary-directory hygiene. Host configuration refusal coverage from main is preserved.
2026-10-12 05:24:33 +08:00
morluto 108b5f9425 fix(native): reject ambiguous otool load-command framing
Validate Mach command counts, byte totals, scalar uniqueness, section counts, and build-tool counts before claiming exhaustive evidence. Preserve raw output in typed parsing failures and distinguish dylib commands from the dynamic loader. Correct inconsistent handcrafted fixtures and label their provenance.

Validation: 38 native tests, combined checks, and temporary-directory hygiene. Multiline dylib-name ambiguity was reproduced with the host Apple otool; regression workflows use captured output and injected command runners.
2026-10-12 05:24:29 +08:00
morluto a8d4984404 fix(browser): preserve uncertainty in malformed network observations
Require observed request method and initiator metadata instead of inventing defaults. Preserve valid redirect history while excluding later events for malformed reused IDs. Keep unavailable or invalid encoded lengths unknown and retain valid zero values.

Validation: 57 CDP boundary tests plus temporary-directory hygiene and combined checks. Protocol events come from the fake CDP server; no live browser verification is claimed.
2026-10-12 05:24:24 +08:00
morluto c002b89e2d fix(snapshot): bind replayed subjects to the selected target path
Rebind display paths only for matching artifact identities in primitive and composed replay. Preserve semantic Evidence IDs, payloads, and historical snapshot provenance. Classify snapshot filesystem workflows as boundary tests.

Validation: cache, direct workflow, MCP replay, and Evidence tests; combined checks and temporary-directory hygiene. Provider interactions are simulated for replay verification.
2026-10-12 05:24:20 +08:00
morluto 2260887f20 fix(artifacts): close mounted directory resources before DMG detach
Wait for directory operations, close owned streams and traversal handles, and retain the mounted reader when cleanup fails so close can retry before detach. Consolidate stable-artifact filesystem regressions into the boundary suite.

Validation: 35 artifact tests and temporary-directory hygiene, including real hdiutil diagnostics. Directory-close and traversal races use injected faults.
2026-10-12 05:24:15 +08:00
morluto c2407689cb fix(setup): preserve skill files and report rollback failures
Track attempted writes, preserve original content and modes from one regular-file observation, and continue rollback after individual failures. Return tagged failure details and retained backup paths through setup.

Validation: setup filesystem and recovery tests, combined compile/typecheck/lint/format/metadata/architecture checks. Filesystem fault injection covers backup, ambiguous publication, and rollback failures.
2026-10-12 05:24:10 +08:00
39418ea173 fix(adb): share host configuration refusals with apktool (#1954)
The ADB provider, which the apktool provider mirrors, had the same two
defects: a relative REA_ADB_PATH escaped as UNKNOWN because selection
ran before the try that maps configuration failures, and a missing adb
binary projected generic target advice instead of its reason.

Move the configuration-failure-to-refusal mapping into one
providerConfigurationRefusal helper used by both providers, resolve the
ADB selection inside the mapped try, and share the CLI refusal assertion.

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
2026-10-12 05:09:13 +08:00
7d92d93a55 fix(managed): share first-match member list ownership (#1955)
Inventory attribute lookups and the members inspector each interpreted
TypeDef member-list ownership. The members lookup binary-searched ordered
ranges but fell back to a per-member scan of every type when malformed
lists were unordered, which was quadratic (8 s for 100,000 types).

Both now use firstMatchListOwners, one near-linear first-match index over
one-based list positions. Remove the ordered/fallback split and the
production declaringType scan, which remains only as the test oracle. The
inventory reserves only its retained index.

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
2026-10-12 05:06:41 +08:00
morluto eb97da3388 fix(artifacts): retain ASAR resources through cleanup failures
Give snapshot capture, header admission, packed members, and container streams one file-handle owner. Serialize adoption with reader shutdown, retain failed closes for retry, and remove snapshots only after descriptor cleanup succeeds.

Cover unread members, pending opens, zero-byte members, and paired format/cleanup failures. Keep intentionally unrecoverable test fixtures hygienic and strengthen managed identity and process cleanup-authority expectations.

Validation: full npm run check; 99 targeted tests and temporary-directory hygiene; exact discovery of 877 test files. No real vendor-provider verification was performed.
2026-10-12 05:05:21 +08:00
290d301ec7 feat(ghidra): admit relocatable MIPS32r2 ELF objects (#1950)
* feat(ghidra): admit soft-float MIPS32r2 ELF with integer-register float storage

Soft-float o32 executables (Linux kernel builds, bare-metal firmware) were
refused because their ABI flags record declares no CPR1 and FP ABI 3.
Ghidra's MIPS32 compiler specifications model only hard-float o32, so typed
float/double signatures (DWARF) would be bound to $f12/$f14/$f0 and the
decompiler loses the real dataflow: `return sel ? a : b` decompiled to
`return in_f0;`, and callers passed phantom FPU inputs.

Admit ELF32 ET_EXEC o32 MIPS32r2 targets whose ABI record declares soft-float
with no CPR1, and commit `mips_soft_float_storage: o32-integer-registers-v1`
in their analysis profile only (hard-float profile digests are unchanged).
ReaGhidraMipsSoftFloat.java runs after auto-analysis and, for each function
with a float/double formal type, either keeps the types with the storage
Ghidra's o32 model assigns to a same-size integer (pair alignment, stack
spill), or, for auto-parameters such as hidden struct-return pointers, swaps
float types for same-size integers under dynamic storage. No function is left
on hard-float storage. The report is written last; the provider client
requires it exactly when the profile commits the policy and refuses the
session before any operation otherwise.

Verification: real Ghidra 12.1.4 lane (verify:ghidra:mips) now builds
hard- and soft-float fixtures in both byte orders and checks exact
parameter/return storage through inspect_native_api, caller constants,
snapshot profile and Evidence outcome; test:fast, verify:package and an
installed-tarball soft-float run pass.

* fix(ghidra): model soft-float MIPS as a prototype for every typed callable

Review feedback on #1949: rewriting storage per Function left external
functions and indirect calls through typed function pointers on Ghidra's
hard-float o32 prototype, which still read $f12/$f14/$f0 while the session
reported success. A negative control confirms it: without the change,
`(*rea_mips_soft_callback)(8.5f, 9.5f, sel)` decompiles to
`(*rea_mips_soft_callback)(in_f12, in_f14, sel)`.

ReaGhidraMipsSoftFloat.java now derives a `__rea_soft_float` prototype from
the program's own default o32 model (encoded, float pentries and
float-consuming rules removed, so byte-order-specific aggregate rules and the
hidden struct-return pointer are kept), installs it as a specification
extension, and assigns it with dynamic storage to every non-thunk function,
external function and function-pointer definition whose formal types use
float or double. Thunks follow their targets. Float types are kept
everywhere, so the integer-retype fallback is removed. The script fails
without writing its report if the derived prototype or any reassigned
function still uses an FPU register.

The policy becomes `o32-soft-float-prototype-v1` with a
{functions, definitions} report. The soft-float fixture adds an indirect
call through a typed pointer; the real-Ghidra lane requires its constants
with no FPU inputs and the struct return's `float scale` in $a1.

* test(ghidra): pin ODDSPREG acceptance for soft-float MIPS ABI records

Review on #1949 suggested requiring flags1 == 0 for soft-float, since MIPS_AFL_FLAGS1_ODDSPREG names CPR1 registers. GCC 5.2 kernel modules, LLVM/Rust firmware and Clang -msoft-float objects all record flags1 = 1 with FP ABI soft-float: it is the MIPS32 o32 toolchain default and selects nothing without CPR1. Keep accepting it, say why at the check, and test that soft-float with ODDSPREG is admitted while any other general flag is refused.

* feat(ghidra): admit relocatable MIPS32r2 ELF objects

Linux MIPS kernel modules (.ko) and other ELF32 o32 MIPS32r2 ET_REL objects
were refused by the Ghidra MIPS lane. Ghidra's ELF loader already places each
allocated section and applies REL relocations; this admits ET_REL alongside
ET_EXEC (ET_DYN and PRX stay refused) under its own support lane,
`elf32-rel-o32-arch32r2-standard-v1`, so executable targets keep theirs.
Relocatable sessions carry limitations: addresses are analysis coordinates,
undefined symbols live in Ghidra's EXTERNAL block, and HI16 relocations report
PARTIAL because Ghidra applies them with their paired LO16.

Verification: new `verify:ghidra:mips-relocatable` real-Ghidra lane builds
-c -G0 objects in both byte orders plus a soft-float -mlong-calls module
variant. GNU readelf supplies sections, symbols and REL entries; every
relocation in an allocated section must encode S + A in the loaded word
(S from Ghidra placement, A from original bytes), with HI16/LO16 checked as
the effective address each LO16 forms, and coverage of a shared HI16, a LO16
carry, external code/data and R_MIPS_32 is asserted. The executable lane,
unit suites pass, and on 30 real .ko files 54,972 relocations matched (93 reference empty sections and are not checkable).

* fix(ghidra): verify every MIPS HI16 and archive-typed soft-float externals

Review feedback on #1950:

- The relocation oracle kept one pending HI16, so a second HI16 before the
  LO16 that resolves both overwrote the first, and unrelated relocations
  cleared it; the earlier site's high half and carry were never compared
  with S + A. Every HI16 now waits for the next LO16 of its symbol, which
  checks all of them; later LO16s reuse those HI16s, and a HI16 left
  unresolved at a section end fails. On 30 real soft-float kernel modules,
  24 contain several HI16s per LO16; all 54,972 placed relocations match.
- External callees with float prototypes: the relocatable fixture now calls
  powf, whose prototype Ghidra's C library archive applies by name. Every
  variant must decompile `powf(1.5,2.5)` without FPU inputs, and the
  soft-float module variant must report that __rea_soft_float covered it
  (with the external step removed it decompiles as `powf(param_1,param_2)`).
  `-fno-builtin` keeps the call from being folded.

Also updates the module variant's expected soft-float policy to
`o32-soft-float-prototype-v1`.

---------

Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
2026-10-12 05:02:55 +08:00
Aurelio BernalandGitHub 5e219de441 feat(ghidra): admit soft-float MIPS32r2 ELF with integer-register float storage (#1949)
* feat(ghidra): admit soft-float MIPS32r2 ELF with integer-register float storage

Soft-float o32 executables (Linux kernel builds, bare-metal firmware) were
refused because their ABI flags record declares no CPR1 and FP ABI 3.
Ghidra's MIPS32 compiler specifications model only hard-float o32, so typed
float/double signatures (DWARF) would be bound to $f12/$f14/$f0 and the
decompiler loses the real dataflow: `return sel ? a : b` decompiled to
`return in_f0;`, and callers passed phantom FPU inputs.

Admit ELF32 ET_EXEC o32 MIPS32r2 targets whose ABI record declares soft-float
with no CPR1, and commit `mips_soft_float_storage: o32-integer-registers-v1`
in their analysis profile only (hard-float profile digests are unchanged).
ReaGhidraMipsSoftFloat.java runs after auto-analysis and, for each function
with a float/double formal type, either keeps the types with the storage
Ghidra's o32 model assigns to a same-size integer (pair alignment, stack
spill), or, for auto-parameters such as hidden struct-return pointers, swaps
float types for same-size integers under dynamic storage. No function is left
on hard-float storage. The report is written last; the provider client
requires it exactly when the profile commits the policy and refuses the
session before any operation otherwise.

Verification: real Ghidra 12.1.4 lane (verify:ghidra:mips) now builds
hard- and soft-float fixtures in both byte orders and checks exact
parameter/return storage through inspect_native_api, caller constants,
snapshot profile and Evidence outcome; test:fast, verify:package and an
installed-tarball soft-float run pass.

* fix(ghidra): model soft-float MIPS as a prototype for every typed callable

Review feedback on #1949: rewriting storage per Function left external
functions and indirect calls through typed function pointers on Ghidra's
hard-float o32 prototype, which still read $f12/$f14/$f0 while the session
reported success. A negative control confirms it: without the change,
`(*rea_mips_soft_callback)(8.5f, 9.5f, sel)` decompiles to
`(*rea_mips_soft_callback)(in_f12, in_f14, sel)`.

ReaGhidraMipsSoftFloat.java now derives a `__rea_soft_float` prototype from
the program's own default o32 model (encoded, float pentries and
float-consuming rules removed, so byte-order-specific aggregate rules and the
hidden struct-return pointer are kept), installs it as a specification
extension, and assigns it with dynamic storage to every non-thunk function,
external function and function-pointer definition whose formal types use
float or double. Thunks follow their targets. Float types are kept
everywhere, so the integer-retype fallback is removed. The script fails
without writing its report if the derived prototype or any reassigned
function still uses an FPU register.

The policy becomes `o32-soft-float-prototype-v1` with a
{functions, definitions} report. The soft-float fixture adds an indirect
call through a typed pointer; the real-Ghidra lane requires its constants
with no FPU inputs and the struct return's `float scale` in $a1.

* test(ghidra): pin ODDSPREG acceptance for soft-float MIPS ABI records

Review on #1949 suggested requiring flags1 == 0 for soft-float, since MIPS_AFL_FLAGS1_ODDSPREG names CPR1 registers. GCC 5.2 kernel modules, LLVM/Rust firmware and Clang -msoft-float objects all record flags1 = 1 with FP ABI soft-float: it is the MIPS32 o32 toolchain default and selects nothing without CPR1. Keep accepting it, say why at the check, and test that soft-float with ODDSPREG is admitted while any other general flag is refused.
2026-10-12 04:54:38 +08:00
morluto 5ba1642a7d test: strengthen archive and resource cleanup assertions
Replace duplicate path-helper checks with real ZIP inventory coverage and replace global memory sampling with bounded-chunk assertions while retaining 64 MiB digest coverage. Await artifact client cleanup, preserve Hopper lease ownership during teardown, and expose typed application failures.
2026-10-12 04:53:27 +08:00
morluto 804c287e44 test: classify filesystem application workflows as boundaries 2026-10-12 04:53:10 +08:00
morluto 918713af37 fix(artifacts): bind directory reads to captured file state
Capture exact filesystem identity and nanosecond timestamps, validate descriptor/path state before and after streaming, and enforce the captured byte count. Bound growing files to their captured size plus one byte using precise positional reads; keep unrepresentable display sizes explicitly unknown.

Reuse the captured-state comparison around Apple assetutil and retain existing reader cleanup ownership. Validation: focused reader and affected consumer suites, full static/architecture checks, and a compiled directory-read probe on Bun 1.3.14 passed. Includes same-size mutation, growth, truncation, cancellation, FIFO replacement, cleanup retry, ZIP identity forgery, ASAR and Mach-O coverage.
2026-10-12 04:45:17 +08:00
morluto a82e37c806 fix(browser): retain module resolver cleanup ownership
Serialize resolution with shutdown, retain the owned browser after close failures, and retry cleanup before another launch. Wire cleanup through the shared service and CLI/MCP lifecycles, preserving primary failures and completed native reports alongside cleanup diagnostics.

Separate the browser resolution workflow from lifecycle ownership. Validation: seven resolver source tests, eleven application tests, MCP shutdown retry, public CLI parity, and full static/architecture checks passed. Real Chromium behavior remains unverified.
2026-10-12 04:44:25 +08:00
morluto c2b56e3e1c test: classify filesystem and process workflows by boundary
Move artifact inventory, application projection, runtime reconciliation, and TODO scanner suites into their real boundary lanes. Extract Hopper process/socket cleanup cases while keeping injected session cleanup cases in composition.

Validation: all 56 focused tests passed; discovery finds all 876 test files exactly once. Test assertions and provider verification scope are unchanged.
2026-10-12 04:44:13 +08:00
morluto 12fa8fd9f9 fix(ida): recover cleanup after worker sessions are reaped
Accept the supervisor absent-session response only for the owned session, then independently verify session and private-input absence before releasing the workspace. Keep failed inventory verification retryable without resending a completed close.

Validation: 59 focused source tests passed, including live private-input peers, unavailable inventory, and mismatched session IDs. Real IDA engine behavior remains unverified. Fixes #1710.
2026-10-12 04:40:41 +08:00
morluto dc303a8ee8 test(javascript): classify artifact workflows as boundary tests
Move the six JavaScript application suites that read production artifacts or launch compiled analysis workers out of composition. Focused test planning now prepares their runtime prerequisites instead of treating them as source-only tests.

Validation: all six relocated suites and adjacent artifact/reference suites pass (70 tests across eight files); discovery finds every test exactly once.
2026-10-12 04:33:51 +08:00
morluto 3d83135987 fix(reference): verify symlink identity around target reads
Bracket symlink target observation with the identity captured during enumeration. Report replacements and disappearing links as changed, and keep cancellation and I/O mapping at one owner boundary.

Remove duplicate target-error wrappers and unreachable external-target handling. Validation: reference reader filesystem suite passes, including deterministic replacement during observation.
2026-10-12 04:33:47 +08:00
morluto 09b6828516 fix(process): cancel run-token reads during sampling
Propagate the sample AbortSignal through readProcessRunId to host environment inspection so cancellation can settle a blocked ownership-token read.

Validation: ownership and capture lifecycle suites pass, including an abort-waiting host regression. Native provider runtime behavior was not exercised.
2026-10-12 04:33:42 +08:00
morluto f64790bfc3 refactor(javascript): reuse validated source-map decoding
Pass typed map objects into the codec and reuse validated TraceMap instances for regular and indexed lookups. Remove whole-map serialization, duplicate leaf decoding, an array-copy wrapper, and the unused offset-validation mode in artifact declaration collection.

Validation: source-map and browser consumer suites (126 tests), declaration tests, and artifact reconstruction boundaries pass. Section overlap checks and raw source declarations remain intact.
2026-10-12 04:33:38 +08:00
morluto c42232f566 fix(adb): share actionable provider configuration refusals
Catch binary selection failures at the adapter boundary and share the domain error projection with apktool. Missing and relative adb paths now report capability_unavailable with host configuration guidance in both message and remediation.

Validation: ADB/apktool provider suites and public CLI refusal suites pass. No connected-device behavior changed or verified. Refs #1952.
2026-10-12 04:33:34 +08:00
morluto 52e76c5d39 fix(managed): share first-match member ownership indexes
Resolve fields, methods, and attribute constructors through one budgeted index per layout. Preserve malformed-list first-match ownership without per-member TypeDef scans; remove the superseded lookup implementations.

Validate pointer identity and malformed ranges through production parsers, including a 40,000-type/method case and allocation admission. Managed source tests and CLI/MCP pointer acceptance pass. Refs #1953.
2026-10-12 04:33:25 +08:00
Minghang ChamlingGitHubfactory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>Kaoru0822
1449e8e051 fix(mcp): provide native evidence paging continuation
Supply schema-validated inspect_analysis_view arguments after completed native function Evidence is recorded and exceeds MCP delivery capacity. Preserve the full canonical result and its provenance. Verify exact source paging, export and following-call usability through real Ghidra stdio MCP.

Fixes #1849

Co-authored-by: Minghang Chamling <133689405+cold-atom@users.noreply.github.com>
Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
Co-authored-by: Kaoru0822 <208281322+kaoru0822-kitauji@users.noreply.github.com>
2026-10-12 04:26:05 +08:00
7cdb7981c3 fix(apktool): name host requirements in configuration refusals (#1946)
A missing apktool launcher projected the generic "unavailable for the
current target" message and remediation, leaving the actual reason only
in details.reason. A relative REA_APKTOOL_COMMAND escaped as UNKNOWN
because command selection ran before the try that maps configuration
failures.

Resolve the selection inside that try for both execution and
availability, and pass the refusal reason as userMessage, as JADX and
the adapters fixed in #1224 do.

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-12 04:13:23 +08:00
0ae38e1e31 fix(ida): preserve non-x86 call edges and workflow uncertainty (#1906)
* test(ida): expect callee edges from the call graph on non-x86 targets

* fix(ida): read callee edges from the call graph on non-x86 targets

* test(ida): pin address order of call graph additions

* docs(ida): describe the call graph source for non-x86 callees

* test(ida): pin the x86 baseline, the call graph reader and its failure modes

* fix(ida): check the call graph root, scope truncation to the root and degrade without callgraph

* test(ida): expect managed PE off the x86 path, possible-truncation wording and a complete ARM64 traversal

* fix(ida): keep managed PE and x86 reads off the call graph path and qualify truncation

* test(ida): answer only expected leaf roots and assert the requested call graph roots

* test(ida): expect a call graph the producer cannot give consistently to fail the callee read

* fix(ida): fail the callee read when the call graph cannot be obtained consistently

* fix(ida): preserve call edge authority and workflow uncertainty

* fix(ida): preserve explicit root call graph caps

---------

Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
2026-10-12 04:10:22 +08:00
c2cf7dcb86 fix(managed): index attribute constructor owners once per layout (#1943)
Each CustomAttribute whose constructor is a local MethodDef scanned the
TypeDef table to find the declaring type, so inspect_managed_artifact did
O(attributes x types) row decoding. A 584 KB assembly with 16,000 types
and 16,000 such attributes blocked for about 38 s.

Build a first-match owner index once per metadata layout. Already-owned
list positions are skipped, so overlapping malformed method lists keep
their previous first-match owner in near-linear time. The index is
charged to the decode budget, and a build failure is cached like the
member pointer lookup.

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-12 04:09:50 +08:00
14935e4d47 fix(javascript): preserve package and source-map boundary identity (#1939)
* fix(javascript): preserve bare package specifier suffixes

* fix(javascript): resolve ESM package self references

* fix(javascript): recover retained prefixed source maps

* test(javascript): cover suffixed ESM self references

* fix(javascript): retain legacy ESM package suffix behavior

* fix(javascript): reject legacy ESM directory and extension fallback

* refactor(javascript): consolidate package resolution and workflow coverage

* fix(javascript): report exact ESM package subpath misses

* fix(javascript): compare parsed names for ESM self references

* refactor(javascript): parse bare package requests once

* fix(javascript): derive ESM dependency lookup from package URLs

---------

Co-authored-by: hhang <169027046+xu91102@users.noreply.github.com>
2026-10-12 04:09:17 +08:00
YsopwnableGitHubfactory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>Kaoru0822
85c1d2587a feat(javascript): analyze selected application sources
Analyze exact selected sources in first-request order before unrelated
static scopes, preserving full artifact identity and byte verification.
Retain selection in CLI/MCP Evidence and summaries, with explicit unknowns
and partial coverage for unselected sources. Keep default full analysis.

Fixes #1848

Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
Co-authored-by: Kaoru0822 <208281322+kaoru0822-kitauji@users.noreply.github.com>
2026-10-12 04:07:14 +08:00
morluto 0d8b728d24 docs(agents): require related brittleness cleanup in bug fixes 2026-10-12 03:58:24 +08:00
9878bb7296 feat(ghidra): apply validated analyzer overrides before analysis
Apply caller-selected analyzer opt-outs before Ghidra analysis, reject unknown or ineffective names, and commit raw/header imports to the same analyzer profile. Use supported x86 raw compiler specs and cover the real analyzer/readback and raw workflows.

Adopts the original implementation from #1800 with validated corrections. Refs #1783, #1836.

Co-authored-by: akushonkamen <74654766+akushonkamen@users.noreply.github.com>
Co-authored-by: N0zoM1z0 <161784452+N0zoM1z0@users.noreply.github.com>
Co-authored-by: Kaoru0822 <208281322+kaoru0822-kitauji@users.noreply.github.com>
2026-10-12 03:49:31 +08:00
212 changed files with 10791 additions and 2021 deletions
@@ -9,7 +9,19 @@ provenance; inputs and recovered code are never executed. Recovery requires a
caller-supplied Wakaru on Linux x64; see [the recovery guide](https://github.com/morluto/rea/blob/main/docs/javascript-recovery.md).
Use `analyze_javascript_application` directly on the operator-supplied ASAR or
extracted tree. Exact `#alias` specifiers resolve through the importer's
extracted tree. When the question concerns known source paths and the connected
schema advertises `source_paths`, select exact inventoried paths such as
`["package.json", "main.js", "preload.js"]`. Include supporting manifests or
source maps explicitly; dependencies are not implicitly analyzed. REA retains
the complete artifact inventory and identity but reads source text and analyzes
only selected files. The result and summary record `source_selection`; files
outside scope and their relationships remain unknown, so partial application
coverage does not mean the selected source failed. Use repeated `--source-path`
on either static application CLI route. Omit the selection when the question
requires complete application analysis. `detail: "summary"` reduces delivery
and does not by itself reduce analysis work.
Exact `#alias` specifiers resolve through the importer's
enclosing package `imports` map to package-local files; nested `node_modules`
packages use their own map. Wildcard, recursive, and external-package import
targets stay unresolved with that limitation. When size is unknown or large and the connected schema advertises
+1 -1
View File
@@ -493,7 +493,7 @@ jobs:
tests/acceptance/webSourceLocationCli.test.ts
tests/boundary/cli/referenceSourceImportStatus.test.ts
tests/boundary/filesystem/asarInventoryFreshness.test.ts
tests/composition/javascript-applications/javascriptArtifactReconstruction.test.ts
tests/boundary/filesystem/javascript-applications/javascriptArtifactReconstruction.test.ts
tests/conformance/providers/ghidra/windowsCapabilities.test.ts
src/ghidra/GhidraInstallation.test.ts
tests/boundary/providers/ghidra/ghidraTransport.test.ts
+2 -2
View File
@@ -1,6 +1,6 @@
{
"_comment": "Test files are intentionally NOT excluded: the gate used to report zero clones because it skipped **/*.test.ts, hiding all test-tree duplication. Current actual is ~2.47% lines across src+ and tests/, down from 3.4% when tests were included. The 2.6% limit leaves modest headroom for ordinary changes; lower it as duplication is removed.",
"threshold": 2.6,
"_comment": "Test files remain included so the gate measures duplication across both src/ and tests/. The 3.0% limit leaves headroom for ordinary changes while keeping duplication visible.",
"threshold": 3.0,
"reporters": ["consoleFull"],
"ignore": [
"node_modules/**",
+2
View File
@@ -18,6 +18,8 @@ Give each interpretation and resource a clear owner. Preserve identity, omission
Assess brittleness in touched code during implementation, debugging, review, and maintenance. Fix the producing representation or ownership rule before adding downstream repairs. Normalize once at the owning boundary using format-aware parsers. Remove superseded representations, aliases, wrappers, and redundant tests after verifying affected consumers. Introduce abstractions only when they remove observed complexity.
When fixing a bug, inspect the producing boundary and affected consumers for duplicated interpretation or unclear ownership. Complete any directly related refactor in the same change and remove superseded code. Passing regression tests alone does not establish that the brittleness is resolved. If related cleanup remains, explain what remains and why.
Parse unknown boundary values and model expected failures with the tagged error algebra and `Result`. Preserve partial facts and meaningful failure reasons. Keep identity and lookup values separate from display formatting; preserve source values and explicit unknowns when normalization loses information. Absolute paths, file URLs, and HTTP paths have different semantics.
Validate advertised JSON Schemas after SDK conversion against their declared dialect and actual clients. Bind handlers to named contracts with exact types; catalog ordering must never select an operation or schema.
@@ -0,0 +1,38 @@
import java.util.Map;
import ghidra.app.plugin.core.analysis.AutoAnalysisManager;
import ghidra.app.util.headless.HeadlessScript;
/** Disable the named auto-analyzers before default auto-analysis. */
public final class ReaGhidraAnalysisOptions extends HeadlessScript {
@Override
public void run() throws Exception {
String[] arguments = getScriptArgs();
if (arguments.length != 1 || arguments[0].isBlank()) {
throw new IllegalArgumentException(
"ReaGhidraAnalysisOptions requires one comma-separated analyzer-name argument");
}
String[] names = arguments[0].split(",");
AutoAnalysisManager analysis = AutoAnalysisManager.getAnalysisManager(currentProgram);
Map<String, String> options = getCurrentAnalysisOptionsAndValues(currentProgram);
for (String requestedName : names) {
monitor.checkCancelled();
String name = requestedName.trim();
if (analysis.getAnalyzer(name) == null || !options.containsKey(name)) {
throw new IllegalArgumentException("Unknown Ghidra analyzer for this program: " + name);
}
}
for (String requestedName : names) {
monitor.checkCancelled();
setAnalysisOption(currentProgram, requestedName.trim(), "false");
}
options = getCurrentAnalysisOptionsAndValues(currentProgram);
for (String requestedName : names) {
monitor.checkCancelled();
String name = requestedName.trim();
if (!"false".equals(options.get(name))) {
throw new IllegalStateException("Ghidra analyzer remained enabled: " + name);
}
}
}
}
+5
View File
@@ -450,6 +450,11 @@ public final class ReaGhidraBridge extends HeadlessScript {
String seeds = currentProgram.getOptions(ghidra.program.model.listing.Program.PROGRAM_INFO)
.getString("REA analysis seeds", null);
if (seeds != null) result.add("analysis_seeds", JsonParser.parseString(seeds));
// Written by ReaGhidraMipsSoftFloat after analysis; REA requires it
// exactly when the committed profile declares soft-float storage.
String softFloat = currentProgram.getOptions(ghidra.program.model.listing.Program.PROGRAM_INFO)
.getString("REA MIPS soft-float storage", null);
if (softFloat != null) result.add("mips_soft_float_storage", JsonParser.parseString(softFloat));
return result;
}
+176
View File
@@ -0,0 +1,176 @@
import ghidra.app.script.GhidraScript;
import ghidra.program.database.SpecExtension;
import ghidra.program.model.data.AbstractFloatDataType;
import ghidra.program.model.data.DataType;
import ghidra.program.model.data.DoubleDataType;
import ghidra.program.model.data.FloatDataType;
import ghidra.program.model.data.FunctionDefinition;
import ghidra.program.model.data.IntegerDataType;
import ghidra.program.model.data.ParameterDefinition;
import ghidra.program.model.data.TypeDef;
import ghidra.program.model.lang.CompilerSpec;
import ghidra.program.model.lang.PrototypeModel;
import ghidra.program.model.lang.Register;
import ghidra.program.model.listing.Function;
import ghidra.program.model.listing.Parameter;
import ghidra.program.model.listing.Program;
import ghidra.program.model.listing.VariableStorage;
import ghidra.program.model.pcode.XmlEncode;
import java.io.StringReader;
import java.io.StringWriter;
import java.util.ArrayList;
import java.util.Iterator;
import java.util.List;
import java.util.regex.Pattern;
import javax.xml.parsers.DocumentBuilderFactory;
import javax.xml.transform.OutputKeys;
import javax.xml.transform.TransformerFactory;
import javax.xml.transform.dom.DOMSource;
import javax.xml.transform.stream.StreamResult;
import org.w3c.dom.Document;
import org.w3c.dom.Element;
import org.w3c.dom.Node;
import org.w3c.dom.NodeList;
import org.xml.sax.InputSource;
/**
* Soft-float o32 passes and returns float/double values exactly where same-size integers go.
* Ghidra's MIPS32 compiler specifications model only hard-float o32, so every typed float
* prototype (functions, externals, thunks via their targets, and function-pointer definitions)
* would otherwise read $f12/$f14/$f0. Derive a program-specific prototype model from the
* program's own default o32 model with its floating-point entries and rules removed, install it
* as a specification extension, and assign it to every callable whose formal types use floats.
*/
public final class ReaGhidraMipsSoftFloat extends GhidraScript {
static final String REPORT_KEY = "REA MIPS soft-float storage";
static final String POLICY = "o32-soft-float-prototype-v1";
static final String PROTOTYPE = "__rea_soft_float";
private static final Pattern FPU_REGISTER = Pattern.compile("f\\d+(_\\d+)?");
@Override
public void run() throws Exception {
String processor = currentProgram.getLanguage().getProcessor().toString();
if (!processor.equals("MIPS") || currentProgram.getDefaultPointerSize() != 4)
throw new IllegalStateException(
"REA soft-float storage requires a 32-bit MIPS program, not " + processor);
CompilerSpec spec = currentProgram.getCompilerSpec();
PrototypeModel base = spec.getDefaultCallingConvention();
new SpecExtension(currentProgram)
.addReplaceCompilerSpecExtension(softFloatPrototype(base, spec), monitor);
PrototypeModel soft = currentProgram.getCompilerSpec().getCallingConvention(PROTOTYPE);
if (soft == null)
throw new IllegalStateException("REA soft-float prototype was not installed");
requireIntegerStorage(
soft.getStorageLocations(
currentProgram,
new DataType[] {
FloatDataType.dataType, IntegerDataType.dataType, DoubleDataType.dataType,
FloatDataType.dataType, DoubleDataType.dataType
},
false),
"soft-float prototype self-check");
int functions = 0;
List<Function> callables = new ArrayList<>();
currentProgram.getFunctionManager().getFunctions(true).forEach(callables::add);
currentProgram.getFunctionManager().getExternalFunctions().forEach(callables::add);
for (Function function : callables) {
// A thunk's signature is its target's; the target is in this list.
if (function.isThunk() || !usesFloat(function)) continue;
if (function.hasCustomVariableStorage()) function.setCustomVariableStorage(false);
function.setCallingConvention(PROTOTYPE);
List<VariableStorage> storage = new ArrayList<>();
storage.add(function.getReturn().getVariableStorage());
for (Parameter parameter : function.getParameters()) storage.add(parameter.getVariableStorage());
requireIntegerStorage(storage.toArray(VariableStorage[]::new), function.getName(true));
functions += 1;
}
int definitions = 0;
Iterator<DataType> types = currentProgram.getDataTypeManager().getAllDataTypes();
while (types.hasNext()) {
if (!(types.next() instanceof FunctionDefinition definition) || !usesFloat(definition))
continue;
definition.setCallingConvention(PROTOTYPE);
definitions += 1;
}
// Written last: any failure above leaves no report, and REA refuses the session.
currentProgram.getOptions(Program.PROGRAM_INFO).setString(
REPORT_KEY,
"{\"policy\":\"" + POLICY + "\",\"functions\":" + functions +
",\"definitions\":" + definitions + "}");
println("REA MIPS soft-float prototype: functions=" + functions + " definitions=" + definitions);
}
/** The default o32 model without floating-point registers or float-consuming rules. */
private static String softFloatPrototype(PrototypeModel base, CompilerSpec spec) throws Exception {
XmlEncode encoder = new XmlEncode();
base.encode(encoder, spec.getPcodeInjectLibrary());
DocumentBuilderFactory factory = DocumentBuilderFactory.newInstance();
factory.setFeature("http://apache.org/xml/features/disallow-doctype-decl", true);
Document document =
factory.newDocumentBuilder().parse(new InputSource(new StringReader(encoder.toString())));
Element root = document.getDocumentElement();
if (!root.getTagName().equals("prototype"))
throw new IllegalStateException("Unexpected default prototype encoding " + root.getTagName());
root.setAttribute("name", PROTOTYPE);
List<Node> removed = new ArrayList<>();
collect(root.getElementsByTagName("pentry"), "metatype", "float", null, removed);
collect(root.getElementsByTagName("consume"), "storage", "float", "rule", removed);
if (removed.isEmpty())
throw new IllegalStateException("Default prototype has no floating-point entries to remove");
for (Node node : removed) node.getParentNode().removeChild(node);
StringWriter xml = new StringWriter();
var transformer = TransformerFactory.newInstance().newTransformer();
transformer.setOutputProperty(OutputKeys.OMIT_XML_DECLARATION, "yes");
transformer.transform(new DOMSource(document), new StreamResult(xml));
return xml.toString();
}
/** Select matching elements, or their nearest `ancestor` element when one is named. */
private static void collect(
NodeList nodes, String attribute, String value, String ancestor, List<Node> selected) {
for (int index = 0; index < nodes.getLength(); index += 1) {
Element element = (Element) nodes.item(index);
if (!value.equals(element.getAttribute(attribute))) continue;
Node target = element;
while (ancestor != null && target != null && !ancestor.equals(target.getNodeName()))
target = target.getParentNode();
if (target != null && !selected.contains(target)) selected.add(target);
}
}
private static void requireIntegerStorage(VariableStorage[] storage, String owner) {
for (VariableStorage item : storage) {
if (item.isBadStorage() || item.isUnassignedStorage())
throw new IllegalStateException("REA soft-float storage unresolved for " + owner);
List<Register> registers = item.getRegisters();
if (registers == null) continue; // stack-only storage
for (Register register : registers) {
if (FPU_REGISTER.matcher(register.getName()).matches())
throw new IllegalStateException(
"REA soft-float storage still uses " + register.getName() + " for " + owner);
}
}
}
private static boolean usesFloat(Function function) {
if (isFloat(function.getReturn().getFormalDataType())) return true;
for (Parameter parameter : function.getParameters()) {
if (isFloat(parameter.getFormalDataType())) return true;
}
return false;
}
private static boolean usesFloat(FunctionDefinition definition) {
if (isFloat(definition.getReturnType())) return true;
for (ParameterDefinition argument : definition.getArguments()) {
if (isFloat(argument.getDataType())) return true;
}
return false;
}
private static boolean isFloat(DataType type) {
DataType current = type;
while (current instanceof TypeDef typeDef) current = typeDef.getBaseDataType();
return current instanceof AbstractFloatDataType;
}
}
+7 -1
View File
@@ -108,7 +108,13 @@ existing analysis families directly, for example
`ro.kernel.qemu` value, `null` when absent.
- **Device paths are not host paths.** `base_apk_device_path` values live on
the device filesystem. Only pulls and captures produce host-local paths,
each paired with byte size and SHA-256 digest.
each paired with byte size and SHA-256 digest. A pulled file's host name is
the device path's last segment, accepted only when every host can store it:
a `pull_adb_file` whose device path ends in a separator, `.`/`..`, a
control or reserved character, a Windows reserved device name, a trailing
dot or space, or a `:` alternate-data-stream marker is refused as
`invalid_request` before any `adb pull` runs, and REA never renames a
device-reported file.
- **Splits are part of the answer.** `pm path` enumerates the base APK and
every installed split; pulls cover the whole set. System packages whose
APKs keep other names report the `unknown` role rather than guessing.
+99 -15
View File
@@ -13,13 +13,13 @@ retains class, byte order, `e_type`, raw `e_flags`, and inspected ABI declaratio
Recognition does not mean every provider or MIPS variant is supported. Existing
PE, managed and Mach-O architecture sets are unchanged.
The first Ghidra lane admits ELF32 `ET_EXEC`, little or big endian, with explicit
o32 and `EF_MIPS_ARCH_32R2` flags and no machine-specific or ASE encoding flags.
These are header declarations, not proof that every instruction conforms to
that ISA. ELF64, n32/n64, unspecified ABI, MIPS16/microMIPS, machine-specific
variants, shared/relocatable objects and PSP PRX are not covered by this lane.
Their family may be identified, but provider selection returns the specific
unsupported constraint rather than guessing a compatible profile.
The Ghidra lane admits ELF32 `ET_EXEC` and `ET_REL`, little or big endian, with
explicit o32 and `EF_MIPS_ARCH_32R2` flags and no machine-specific or ASE
encoding flags. These are header declarations, not proof that every instruction
conforms to that ISA. ELF64, n32/n64, unspecified ABI, MIPS16/microMIPS,
machine-specific variants, shared objects and PSP PRX are not covered by this
lane. Their family may be identified, but provider selection returns the
specific unsupported constraint rather than guessing a compatible profile.
Explicit `EF_MIPS_FP64` and `EF_MIPS_NAN2008` declarations are refused. Their
absence is insufficient: Clang can emit identical `e_flags` for FP32, FP64 and
@@ -39,12 +39,14 @@ absent or supply a default ABI.
The field layout follows
[`Elf_MIPS_ABIFlags_v0`](https://github.com/bminor/glibc/blob/master/elf/elf.h).
Provider admission requires record version 0, MIPS32r2, 32-bit GPR/CPR1, no CPR2,
and a double-precision or FPXX ABI declaration with no ISA extension, ASE or
unknown general flags. Missing records remain unknown rather than defaulting
to a compatible ABI, and this initial lane refuses them. Unknown record values
are retained for diagnosis but do not gain provider support. These checks bind
the declared interpretation; the integer-only fixture does not establish full
Provider admission requires record version 0, MIPS32r2, 32-bit GPRs, no CPR2,
no ISA extension, ASE or unknown general flags, and one of two floating-point
declarations: 32-bit CPR1 with a double-precision or FPXX ABI, or no CPR1 with
the soft-float ABI. Other combinations, including single-float, FP64 and
unspecified ABIs, are refused. Missing records remain unknown rather than
defaulting to a compatible ABI, and this lane refuses them. Unknown record
values are retained for diagnosis but do not gain provider support. These checks
bind the declared interpretation; the fixtures do not establish full
floating-point behavioral conformance.
MIPS header and ABI record facts are committed in the analysis profile so
@@ -53,6 +55,51 @@ identity is versioned independently of file identity. Evidence, lifecycle
outputs and saved snapshots preserve the `mips` family. Hopper and IDA adapters
do not implicitly gain MIPS support. Windows Ghidra P0 remains PE x86/x86-64 only.
### Soft-float prototype
Ghidra's MIPS32 compiler specifications model hard-float o32 only: a typed
`float` or `double` parameter is bound to `$f12`/`$f14` and returned in `$f0`.
Soft-float o32 passes those values in integer registers and stack slots, so a
typed prototype (for example from DWARF) would make the decompiler read FPU
registers the code never writes and drop the real dataflow. That applies to
functions, external functions and indirect calls through typed function
pointers alike.
For soft-float targets REA commits
`mips_soft_float_storage: o32-soft-float-prototype-v1` in the analysis profile
and runs `ReaGhidraMipsSoftFloat.java` after auto-analysis. The script encodes
the program's own default o32 prototype, removes its floating-point parameter
and return entries and the rules that consume floating-point storage, and
installs the result as the program-specific `__rea_soft_float` prototype through
a Ghidra specification extension. Deriving it keeps the byte-order-specific
aggregate rules and the hidden struct-return pointer. The script then assigns
that prototype, with dynamic storage, to every non-thunk function and external
function, and to every function-pointer definition, whose declared types use
`float` or `double`. Thunks follow their targets. Types are kept.
The script fails, and writes no report, if the derived prototype still places a
float in an FPU register or any reassigned function does. It records the
function and definition counts last; REA requires that report exactly when the
profile commits the policy and refuses the session otherwise, before serving any
operation. Session Evidence carries the policy and counts as limitations.
Callables without typed float signatures keep Ghidra's default prototype, under
which Ghidra already infers integer-register parameters. Hard-float profiles do
not gain the policy, so their digests are unchanged.
### Relocatable objects
`ET_REL` objects, such as Linux kernel modules (`.ko`), are imported by Ghidra's
ELF loader, which assigns each allocated section an analysis address and applies
the object's REL relocations there. Undefined symbols are placed in Ghidra's
`EXTERNAL` block: a thunk for code, a label for data. These addresses are
analysis coordinates, not the addresses a kernel or linker would choose, and
session Evidence says so. `inspect_native_load_image` reports each relocation
with Ghidra's status; HI16 entries report `PARTIAL` because Ghidra applies them
together with their paired LO16. Relocatable targets commit
`mips_support_lane: elf32-rel-o32-arch32r2-standard-v1`, separate from the
executable lane, whose identity executable targets keep. Kernel modules are
usually soft-float and therefore also use the soft-float storage policy.
## Use
With a caller-installed Ghidra and compatible JDK on a supported Linux/macOS host:
@@ -70,7 +117,10 @@ REA's ELF parser.
## Verification
The source-owned freestanding fixture is `tests/conformance/c/mips.c`. The
### Executables
The source-owned freestanding fixture is `tests/conformance/c/mips.c`; the
soft-float variant also links `tests/conformance/c/mips-soft-float.c`. The
optional cross-target lane requires a caller-supplied Clang with MIPS targets,
LLD, GNU `readelf`, Ghidra and its compatible JDK; ordinary host-native checks
do not acquire these tools. `REA_MIPS_CLANG` selects Clang and
@@ -81,7 +131,8 @@ These are test-time prerequisites, not REA runtime providers.
GHIDRA_INSTALL_DIR=/absolute/path/to/ghidra npm run verify:ghidra:mips
```
The lane compiles both byte orders and never executes the targets. GNU readelf
The lane compiles both byte orders with hard-float and soft-float ABIs and never
executes the targets. GNU readelf
independently supplies header/ABI facts and symbol addresses, compared with the
production REA target resolver before provider startup. Its version, bounded
command output and selected tool path are retained in the verification report.
@@ -104,9 +155,42 @@ all program-header payloads. It checks the actual Ghidra language, entry-address
decompilation, distinct artifact digest and missing-section limitations through
CLI Evidence and an MCP-produced snapshot. Caller-selected Ghidra heap limits
are passed to these child workflows.
For soft-float fixtures the lane also checks, through `inspect_native_api`, the
exact parameter and return types and storage of float, double, mixed (pair
alignment and stack spill) and struct-returning functions; that a caller's
decompilation receives every source float constant without FPU inputs,
including an indirect call through a typed function pointer; and that the
snapshot profile and Evidence retain the policy and outcome counts.
Pseudocode presence is a liveness check, not proof of semantic equivalence.
Instruction decoding and reader agreement likewise do not establish runtime,
delay-slot execution or complete floating-point semantics. Assertion regressions
reject incorrect byte order, constants, destinations and loaded identities.
Parser/profile tests are narrower checks, not substitutes for this real-provider
run. Report the actual lanes run and leave unexecuted verification pending.
### Relocatable objects
This lane has its own command and fixture:
```bash
GHIDRA_INSTALL_DIR=/absolute/path/to/ghidra npm run verify:ghidra:mips-relocatable
```
It compiles `tests/conformance/c/mips-relocatable.c` with `-c -G0` in both byte
orders, plus a little-endian soft-float `-mlong-calls` variant matching Linux
module options. GNU readelf independently supplies sections, symbols and REL
entries. For every relocation in an allocated section, the lane reads the
loaded word through `read_bytes` and requires it to encode `S + A`: `S` from
Ghidra's own section placement (`list_segments`) or `EXTERNAL` placement, `A`
from the original file bytes. HI16/LO16 are checked as the effective address
each LO16 forms with each HI16 it resolves, including several HI16s before one
LO16, so a missing signed-low carry fails and a HI16 without a following LO16
fails. The lane also requires matching Ghidra relocation entries and statuses,
section file offsets and sizes, coverage of a shared HI16, a LO16 carry,
external code and data and `R_MIPS_32` words, direct calls from the entry
function, the relocatable support lane in the snapshot profile, the
relocatable limitation in Evidence and an unchanged source object. The fixture
also calls `powf`, an external whose float prototype comes from Ghidra's C
library archive; every variant must decompile `powf(1.5,2.5)` without FPU
inputs, and the soft-float variant must report that the soft-float prototype
covered it.
+1 -1
View File
@@ -98,7 +98,7 @@ For guided registration, set `REA_IDA_MCP_CONFIG` before `rea setup --client <cl
## Coverage and evidence
Both profiles support procedure/string inventory and literal or regex search, procedure resolution, pseudocode, assembly, raw function instructions, resolved callees, incoming address references, and a function dossier. Inventories and modern pseudocode/assembly follow upstream pagination. Upstream limits on modern xrefs and callees are retained as explicit truncation limitations. A transport-level truncated preview is rejected rather than reported as complete evidence; the error preserves upstream retrieval guidance.
Both profiles support procedure/string inventory and literal or regex search, procedure resolution, pseudocode, assembly, raw function instructions, resolved callees, incoming address references, and a function dossier. Inventories and modern pseudocode/assembly follow upstream pagination. Upstream limits on modern xrefs and callees are retained as explicit truncation limitations. In the modern headless profile, targets that are not proven x86-only use processor-neutral `callgraph` edges instead of the x86 instruction checks in upstream `callees` (mrexodia/ida-pro-mcp#549). REA requests depth zero: the pinned upstream emits the root's outgoing edges without scanning child functions, then REA batches function-name lookup for those addresses. The graph covers recognized function destinations only; imports and other non-function targets remain unknown. A server without `callgraph` returns an empty relationship list with an explicit coverage limitation. Graph and path workflows retain those limitations alongside observed edges. A failed or inconsistent graph reply remains an error. A transport-level truncated preview is rejected rather than reported as complete evidence; the error preserves upstream retrieval guidance.
The legacy profile additionally supports direct callers. Its producer may report call-site or interior addresses; REA resolves the reported function symbol to its canonical entry. The modern profile does not prove direct callers from its code/data xref labels, so that capability is unavailable and the dossier marks it unknown.
+21
View File
@@ -800,6 +800,27 @@ Each line is `address<TAB>kind[<TAB>name]`, where kind is `function`, `code` or
Label seeds can name mapped uninitialized storage such as `.bss`; function and code seeds require initialized bytes.
To skip a named Ghidra auto-analyzer (for example, when default auto-analysis
stalls in one analyzer on a Swift/Objective-C arm64 Mach-O):
```bash
export REA_GHIDRA_DISABLED_ANALYZERS="Objective-C Message Analyzer"
```
The value is a comma-separated list of exact Ghidra analyzer names. A pre-script
disables each name before auto-analysis, the analysis profile records the
override in `analyzer_preset` (for example
`ghidra-default+disabled:Objective-C Message Analyzer`), and a session fails to
open when its committed profile and the setting differ. A name starts with a
letter or digit and contains only letters, digits, spaces, periods, underscores,
and hyphens; at most 32 names are admitted, each at most 120 characters.
Names must identify analyzers registered for the selected program. Unknown names
or options that remain enabled abort the headless import before auto-analysis
and the bridge, so an unsuccessful override cannot produce a session claiming
that the analyzer was disabled. For Windows PE resource analysis, the exact name
is `WindowsResourceReference`. Disabling an analyzer skips the facts it would
have produced.
### Ghidra heap and CPU controls
Set resource controls in the environment that launches the CLI or MCP server,
+54 -2
View File
@@ -66,6 +66,45 @@ code-point order. Dependency manifests remain package observations but do not
declare Electron main or renderer entries. If only dependency manifests are
present, the inventory artifact remains the graph root.
## Select sources before analysis
For a question about known entry sources, supply exact artifact-relative file
paths in `source_paths`, or repeat `--source-path` on either CLI route:
```bash
rea analyze /absolute/path/to/app.asar \
--source-path package.json --source-path main.js --json
```
```json
{
"input_path": "/absolute/path/to/app.asar",
"source_paths": ["package.json", "main.js"],
"detail": "summary"
}
```
REA inventories and verifies the complete artifact, including unselected bytes,
but allocates source text and performs static analysis only for selected files.
Include supporting manifests, HTML or source maps explicitly when needed;
imports do not expand the selection. Paths must match relevant inventoried
files exactly, including nested ASAR prefixes. Unknown paths, aliases and globs
return an `invalid_request` correction. Selected files are analyzed in request
order before unselected scopes are projected. Repeated paths preserve the
caller's request and first-occurrence priority, analyzing the file once.
Omit the selection for the existing complete
application workflow.
The result and summary retain `source_selection.requested_paths` and
`source_selection.unselected_files`. Every relevant file keeps its original
inventory identity; unselected files report `text_status:
"outside-analysis-scope"`. Their relationships remain unknown, and application
coverage is partial while any file remains outside scope. Selection does not
claim whole-application completeness or change integrity policy. Summary detail
only reduces delivery; source selection reduces text acquisition and analysis.
Explicit native provider or snapshot routing rejects `--source-path` rather
than ignoring it.
## ASAR integrity
ASAR inventory checks Electron integrity metadata for embedded archive entries
@@ -469,6 +508,12 @@ from unmatched nested targets rather than reporting array indices.
For imports and requires, the active Node conditions include `module-sync`;
earlier active conditions retain precedence over later ones.
Package self-references use the importer's enclosing package scope before
installed dependencies. ESM requires the parsed package name to equal the
metadata name; a scope-only name such as `@scope` cannot capture `@scope/pkg`.
CommonJS retains Node's raw-request name/prefix matching. Both paths preserve
the original specifier and select exports under their own loader conditions.
Exact package-private `#alias` specifiers resolve against the importer's
enclosing `package.json` `imports` map — the nearest same-container package
scope, stopping at a `node_modules` directory so a nested package cannot inherit
@@ -484,8 +529,15 @@ URL targets before selecting a later package-local file; an unsupported bare
package or recursive alias keeps its unresolved outcome rather than selecting
a later fallback.
ESM relative module paths and selected package exports targets use URL suffix handling
and one percent-decoding pass. CommonJS relative paths and legacy package main
ESM relative module paths, legacy package subpaths without exports, and selected
package exports targets use URL suffix handling and one percent-decoding pass.
Bare specifiers retain their complete query/fragment text for exports key matching.
For dependency lookup, ESM constructs a package metadata URL instead of treating
the parsed package name as a literal directory. A root request such as `legacy?x`
does not load either `node_modules/legacy?x` or `node_modules/legacy`; an exact
self-reference with that metadata name can still select its exports first.
Without an exports map, ESM subpaths select exact files; directory imports and
extension fallback are refused. CommonJS relative paths and legacy package main
fields retain literal filename punctuation. Exports targets must start with
`./` and exclude `.`, `..`, and `node_modules` path segments, including their
encoded forms, before URL normalization. Arrays skip invalid exports targets
+19 -2
View File
@@ -310,6 +310,13 @@ directly to `compare_functions`, and `inspect_artifact` Evidence to
`compare_artifacts`. Use `get_evidence_bundle` when the task needs broader
retained session history or an explicit bundle for transfer.
`get_call_graph` keeps provider-limited relationship reads on their node as
`limitations`, alongside any observed `calls`. An empty call list with such a
limitation is incomplete coverage, not proof that the function has no calls.
`trace_call_path` reports each limited expansion in `residual_unknowns` and
sets `truncated: true`; any path already supported by observed edges remains
visible. Tool errors remain in their existing error or failure fields.
REA prepares complete MCP results within the pinned stdio client's 10 MiB
receive-buffer budget, including both text and structured representations and
room for the JSON-RPC envelope. Tool errors carry only their text
@@ -321,8 +328,18 @@ session. Its exact reference is reported in
`details.reported_limits.evidence_reference`; call `inspect_analysis_view` with
that reference for a summary, one section/module, or a stable page, or call
`export_evidence_bundle` with a destination path. Complete
bundle exports stream canonical JSON into an atomically published file. A broad
follow-up or `get_evidence_bundle` can also exceed the response budget; exporting
bundle exports stream canonical JSON into an atomically published file.
For oversized native function Evidence, `details.reported_limits.continuation`
contains a ready-to-call `inspect_analysis_view` request for the retained
pseudocode's first page. Use its `name` and `arguments` directly, then follow
`normalized_result.coverage.next_offset` for subsequent pages. The suggested
limit is the view contract's default, not a guarantee that arbitrary source or
metadata fits; reduce the limit or export when needed. Continuations are
offered for completed `analyze_function` results only after session recording is
acknowledged; the complete result and its retention reference remain intact.
A broad follow-up or `get_evidence_bundle` can also exceed the response budget; exporting
preserves the complete session without sending it through a single MCP frame.
To discover existing records without their large payloads, call
`get_evidence_bundle` with `detail: "summary"`. Its `result.records` contain
+11 -7
View File
@@ -35,14 +35,18 @@ Omitted `format` keeps header-based detection; architecture and entry without
`format=raw-binary` are rejected.
Ghidra imports a private snapshot with `BinaryLoader` and these little-endian
generic languages:
generic languages and compiler specifications:
| architecture | language_id |
| ------------ | ------------------- |
| x86 | `x86:LE:32:default` |
| x86_64 | `x86:LE:64:default` |
| arm | `ARM:LE:32:v7` |
| arm64 | `AARCH64:LE:64:v8A` |
| architecture | language_id | compiler_spec_id |
| ------------ | ------------------- | ---------------- |
| x86 | `x86:LE:32:default` | `gcc` |
| x86_64 | `x86:LE:64:default` | `gcc` |
| arm | `ARM:LE:32:v7` | `default` |
| arm64 | `AARCH64:LE:64:v8A` | `default` |
Compiler specifications select analysis conventions; they do not identify the
original compiler or ABI of a headerless target. Ghidra's x86 languages do not
provide a compiler specification named `default`.
Those settings, the caller-supplied entry, and `language_selection:
caller-supplied-v1` are committed in the analysis profile. A different entry
+7 -4
View File
@@ -463,7 +463,7 @@ prove behavior on the verified Ghidra 12.1.4 and JDK 21 build.
| Ghidra lane | Supported runner/target | Additional local tools |
| ------------------------------------------ | ----------------------------------------------------------------------------------------------- | ------------------------------------------------------------------- |
| `npm run verify:ghidra` | Linux x64/arm64 ELF or macOS x64/arm64 Mach-O | Host C compiler, Ghidra 12.1.4, and full JDK 21 |
| `npm run verify:ghidra:seeds` | POSIX host; source-owned 32-bit x86 raw fixture; function order and BSS labels | Ghidra 12.1.4 and full JDK 21 |
| `npm run verify:ghidra:seeds` | POSIX; raw x86 fixture; seed ordering, BSS labels, analyzer readback and unknown-name rejection | Ghidra 12.1.4 and full JDK 21 |
| `npm run verify:ghidra:swift` | macOS x64/arm64 Mach-O | Host Swift compiler, Ghidra 12.1.4, and full JDK 21 |
| `npm run verify:ghidra:switch` | Linux x64 ELF; GCC/Clang optimized and stripped switch fixtures | GCC, Clang, GNU nm/objdump/strip, Ghidra 12.1.4, and full JDK 21 |
| `npm run verify:ghidra:aarch64-jump-table` | Any supported Ghidra host; AArch64 ELF; signed/unsigned byte/halfword tables; host ARM64 Mach-O | Clang with AArch64 target support, Ghidra 12.1.4, and full JDK 21 |
@@ -803,8 +803,10 @@ through the compiled CLI and stdio MCP server. With the same Ghidra/JDK
prerequisites, it checks preparation and bridge startup, verified load mappings,
instruction and byte readback, canonical-entry session reuse, replacement when
the entry changes without `provider_id`, unchanged source bytes, and owned
cleanup. Other raw CPU families remain covered by deterministic tests rather
than this real-provider lane. See [raw binary analysis](raw-binary.md).
cleanup. Six-byte x86 and x86_64 controls also verify supported compiler
specifications, independent mappings, unchanged bytes, and decompilation of a
known return value. AArch64 remains covered by deterministic tests rather than
this real-provider lane. See [raw binary analysis](raw-binary.md).
## Apple Interface Builder archives
@@ -1384,7 +1386,8 @@ generation. It installs nothing. Without configuration, the default invocation
reports a named skip; `--require-tools` makes missing configuration fail.
The verifier creates and validates real modules with custom sections, multiple
import/export kinds, escaped names and empty modules. It compares exact WAT,
import/export kinds, escaped names, an empty module, and a bulk-memory
DataCount section. It compares exact WAT,
section output and selected/tool digests through the compiled CLI and real MCP
SDK, validates advertised schemas, checks identical CLI/MCP Evidence IDs and
session bundle readback, retains invalid/truncated module diagnostics, preserves
+3
View File
@@ -55,7 +55,9 @@
"bridge/ghidra/ReaGhidraExtensions.java",
"bridge/ghidra/ReaGhidraPrepareCom.java",
"bridge/ghidra/ReaGhidraApplySeeds.java",
"bridge/ghidra/ReaGhidraAnalysisOptions.java",
"bridge/ghidra/ReaGhidraNoReturnFix.java",
"bridge/ghidra/ReaGhidraMipsSoftFloat.java",
"bridge/ghidra/ReaGhidraPrepareRaw.java",
"bridge/android/ReaJadxBridge.java",
"bridge/mitmproxy/capture.py",
@@ -163,6 +165,7 @@
"verify:ghidra": "npm run build:cached && node scripts/verify-real-ghidra.mjs",
"verify:ghidra:seeds": "npm run build:cached && node scripts/verify/ghidra/seeds.mjs",
"verify:ghidra:mips": "npm run build:cached && node scripts/verify-real-ghidra-mips.mjs",
"verify:ghidra:mips-relocatable": "npm run build:cached && node scripts/verify-real-ghidra-mips-relocatable.mjs",
"verify:ghidra:swift": "npm run build:cached && node scripts/verify-real-ghidra-swift.mjs",
"verify:ida": "npm run build:cached && node scripts/verify-real-ida.mjs",
"verify:ghidra:switch": "npm run build:cached && node scripts/verify-real-ghidra-switch.mjs",
+4 -2
View File
@@ -172,11 +172,13 @@ const digest = (bytes) => createHash("sha256").update(bytes).digest("hex");
const runtimeParent = join(workspace, "runtime with spaces");
await mkdir(runtimeParent);
const callerDirectory = join(workspace, "caller cwd with spaces");
// Keep sorted for comparison with the final caller-directory listing.
const scriptCollisions = [
"ReaGhidraBridge.java",
"ReaGhidraPrepareCom.java",
"ReaGhidraAnalysisOptions.java",
"ReaGhidraApplySeeds.java",
"ReaGhidraBridge.java",
"ReaGhidraNoReturnFix.java",
"ReaGhidraPrepareCom.java",
"ReaGhidraPrepareRaw.java",
];
await mkdir(callerDirectory);
+333
View File
@@ -0,0 +1,333 @@
#!/usr/bin/env node
import assert from "node:assert/strict";
import { execFile } from "node:child_process";
import { createHash } from "node:crypto";
import { mkdtemp, mkdir, readFile, rm } from "node:fs/promises";
import { tmpdir } from "node:os";
import { join } from "node:path";
import { fileURLToPath } from "node:url";
import { promisify } from "node:util";
import { Client } from "@modelcontextprotocol/client";
import { StdioClientTransport } from "@modelcontextprotocol/client/stdio";
import { parseBinaryTarget } from "../dist/application/BinaryTargetResolver.js";
import { parseEvidence } from "../dist/domain/evidence.js";
import { parseAnalysisSnapshot } from "../dist/domain/analysisSnapshot.js";
import { functionDossierSchema } from "../dist/domain/hopperValues.js";
import { inspectGhidraInstallation } from "../dist/ghidra/GhidraInstallation.js";
import { buildMipsRelocatableFixture } from "../tests/conformance/ghidra/mips-fixture.mjs";
import {
assertMipsRelocations,
inspectMipsRelocatableReadelf,
} from "../tests/conformance/ghidra/mips-relocatable-checks.mjs";
import { requireMcpOperationResult } from "./lib/mcp-verifier-results.mjs";
import { createVerifierRun, completeVerifierRun } from "./lib/verifier-run.mjs";
// Optional cross-target lane for ET_REL objects. No tool acquisition, no
// linking and no target execution.
if (process.argv.length > 2)
throw new Error(
"Usage: node scripts/verify-real-ghidra-mips-relocatable.mjs",
);
if (!["linux", "darwin"].includes(process.platform))
throw new Error(
"MIPS verification requires a supported Linux/macOS Ghidra host",
);
if (process.env.GHIDRA_INSTALL_DIR === undefined)
throw new Error("Missing prerequisite: caller-supplied GHIDRA_INSTALL_DIR");
const installation = inspectGhidraInstallation({
environment: process.env,
installDir: process.env.GHIDRA_INSTALL_DIR,
...(process.env.JAVA_HOME === undefined
? {}
: { javaHome: process.env.JAVA_HOME }),
});
if (installation.status !== "available")
throw new Error(
`Ghidra prerequisite unavailable: ${JSON.stringify(installation)}`,
);
const execute = promisify(execFile);
const entrypoint = fileURLToPath(new URL("./rea.mjs", import.meta.url));
const run = createVerifierRun();
const workspace = await mkdtemp(join(tmpdir(), "rea-mips-rel-proof-"));
const runtime = join(workspace, "runtime");
await mkdir(runtime);
const env = {
PATH: process.env.PATH ?? "/usr/bin:/bin",
TMPDIR: runtime,
REA_LOG_LEVEL: "silent",
REA_ANALYSIS_PROVIDER: "ghidra",
REA_PROCESS_RUN_ID: run.run_id,
GHIDRA_INSTALL_DIR: installation.installDir,
JAVA_HOME: installation.javaHome,
HOPPER_LAUNCHER_PATH: "/rea-unconfigured-deep-provider/hopper",
};
for (const setting of ["GHIDRA_HEADLESS_MAXMEM", "GHIDRA_MAXMEM"]) {
if (process.env[setting] !== undefined) env[setting] = process.env[setting];
}
/** Load every relocated word through the public read_bytes operation. */
async function loadedWords(call, independent, segments, byteOrder) {
const words = new Map();
for (const relocation of independent.relocations) {
const section = [...independent.sections.values()].find(
(item) => item.name === relocation.section,
);
if (!section?.allocated) continue;
const segment = segments.find((item) => item.name === section.name);
assert.ok(segment, `Ghidra segment ${section.name}`);
const address = BigInt(segment.start) + BigInt(relocation.offset);
const read = await call("read_bytes", {
address: `0x${address.toString(16)}`,
length: 4,
});
assert.equal(read.returned_bytes, 4);
const bytes = Buffer.from(read.bytes_hex, "hex");
words.set(
address,
byteOrder === "little" ? bytes.readUInt32LE(0) : bytes.readUInt32BE(0),
);
}
return words;
}
async function verifyFixture(byteOrder, kernelModule) {
const fixture = await buildMipsRelocatableFixture(
workspace,
byteOrder,
kernelModule,
);
const independent = await inspectMipsRelocatableReadelf(
fixture.path,
byteOrder,
fixture.float_abi,
);
const resolved = await parseBinaryTarget(fixture.path);
if (!resolved.ok) throw resolved.error;
assert.deepEqual(
resolved.value.mips,
independent.mips,
"REA ELF/ABI interpretation disagrees with GNU readelf",
);
assert.equal(resolved.value.sha256, fixture.sha256);
const label = `${byteOrder}${kernelModule ? "-module" : ""}`;
const snapshotPath = join(workspace, `${label}.snapshot.json`);
const transport = new StdioClientTransport({
command: process.execPath,
args: [entrypoint, "mcp"],
env,
stderr: "pipe",
});
const client = new Client({ name: "rea-real-mips-rel-proof", version: "1" });
let stderr = "";
transport.stderr?.on("data", (chunk) => {
stderr = (stderr + chunk.toString()).slice(-65536);
});
const call = async (name, args = {}) =>
requireMcpOperationResult(
await client.callTool({ name, arguments: args }, undefined, {
timeout: 180000,
}),
name,
);
try {
await client.connect(transport);
const target = await call("open_binary", {
path: fixture.path,
provider_id: "ghidra",
});
assert.equal(target.architecture, "mips");
assert.equal(target.sha256, fixture.sha256);
const image = (await call("inspect_native_load_image")).observations;
assert.equal(
image.language_id,
`MIPS:${byteOrder === "little" ? "LE" : "BE"}:32:default`,
);
assert.equal(image.source_files.length, 1);
assert.equal(image.source_files[0].original_sha256, fixture.sha256);
assert.notEqual(
image.source_files[0].modified_sha256,
fixture.sha256,
"Relocations must change the loaded bytes",
);
const names = await call("list_names");
const segments = await call("list_segments");
const coverage = assertMipsRelocations({
facts: independent,
fileBytes: await readFile(fixture.path),
segments,
mappings: image.mappings,
ghidraRelocations: image.relocations,
words: await loadedWords(call, independent, segments, byteOrder),
names,
byteOrder,
});
assert.equal(coverage.unplaced, 0, "Fixture references an empty section");
assert.ok(coverage.shared_hi16, "No HI16 shared by several LO16s");
assert.ok(coverage.lo16_carry, "No LO16 >= 0x8000 forced a HI16 carry");
assert.ok(coverage.external_code, "No external code relocation");
assert.ok(coverage.external_data, "No external data relocation");
assert.ok(coverage.types.R_MIPS_32 > 0, "No R_MIPS_32 relocation");
assert.equal(
coverage.types.R_MIPS_26 === undefined,
kernelModule,
"Long calls must replace every R_MIPS_26 with HI16/LO16",
);
const procedures = await call("list_procedures");
const named = (name) => {
const found = procedures.filter((item) => item.value === name);
assert.equal(found.length, 1, `Procedure ${name}`);
return found[0].address;
};
const entry = named("rea_mips_rel_entry");
const callees = await call("procedure_callees", { procedure: entry });
for (const callee of [
named("rea_mips_rel_leaf"),
named("rea_mips_rel_far"),
])
assert.ok(callees.includes(callee), `Missing direct call to ${callee}`);
// Ghidra lists the external function and its EXTERNAL-block thunk.
assert.ok(
procedures.some(
(item) =>
item.value === "<EXTERNAL>::rea_mips_rel_external" &&
callees.includes(item.address),
),
"Missing call to the external function",
);
const dossier = functionDossierSchema.parse(
await call("analyze_function", { procedure: entry }),
);
for (const name of [
"rea_mips_rel_leaf",
"rea_mips_rel_far",
"rea_mips_rel_external",
])
assert.ok(dossier.pseudocode.includes(name), `Pseudocode lost ${name}`);
// powf's float prototype comes from Ghidra's C library archive; soft-float
// builds pass 1.5f/2.5f in $a0/$a1, hard-float builds in $f12/$f14.
const scaled = functionDossierSchema.parse(
await call("analyze_function", {
procedure: named("rea_mips_rel_scaled"),
}),
);
assert.match(scaled.pseudocode, /\bpowf\(1\.5,2\.5\)/u);
assert.doesNotMatch(scaled.pseudocode, /\bin_f\d+\b/u);
await call("close_binary", { snapshot_path: snapshotPath });
const snapshot = parseAnalysisSnapshot(
JSON.parse(await readFile(snapshotPath, "utf8")),
);
const parameters = snapshot.binding.analysis_profile.parameters;
assert.equal(parameters.mips_elf.type, 1);
assert.equal(
parameters.mips_support_lane,
"elf32-rel-o32-arch32r2-standard-v1",
);
assert.equal(
parameters.mips_soft_float_storage,
kernelModule ? "o32-soft-float-prototype-v1" : undefined,
);
assert.ok(
snapshot.evidence_bundle.records.some((record) =>
parseEvidence(record).limitations.some((item) =>
item.includes("Relocatable MIPS ELF"),
),
),
"Relocatable limitation missing from retained Evidence",
);
assert.equal(
snapshot.evidence_bundle.records.some((record) =>
parseEvidence(record).limitations.some((item) =>
/Soft-float prototype outcome: [1-9]\d* functions/u.test(item),
),
),
kernelModule,
"Soft-float prototype must cover the archive-typed external in module builds",
);
await call("open_binary", {
path: fixture.path,
provider_id: "ghidra",
snapshot_path: snapshotPath,
});
assert.deepEqual(await call("list_procedures"), procedures);
await call("close_binary");
// Same public operation through CLI, not a private injected provider.
const cli = await execute(
process.execPath,
[
entrypoint,
"function",
fixture.path,
"rea_mips_rel_entry",
"--provider",
"ghidra",
"--json",
],
{ env, timeout: 240000, maxBuffer: 16 * 1024 * 1024 },
);
const evidence = parseEvidence(JSON.parse(cli.stdout));
assert.equal(evidence.subject?.architecture, "mips");
assert.equal(evidence.subject?.digest.sha256, fixture.sha256);
assert.equal(
functionDossierSchema.parse(evidence.normalized_result).procedure.address,
dossier.procedure.address,
);
assert.equal(
createHash("sha256")
.update(await readFile(fixture.path))
.digest("hex"),
fixture.sha256,
"Analysis modified the source object",
);
return {
...fixture,
independent_reader: {
command: independent.command,
version: independent.version,
},
effective_language: image.language_id,
relocation_coverage: coverage,
procedures: procedures.length,
status: "passed",
};
} catch (cause) {
throw new Error(
`MIPS ${label} relocatable verification failed; MCP stderr: ${stderr}`,
{ cause },
);
} finally {
try {
await client.close();
} finally {
await transport.close();
}
}
}
const reports = [];
let primaryFailure;
try {
for (const [byteOrder, kernelModule] of [
["little", false],
["big", false],
["little", true],
])
reports.push(await verifyFixture(byteOrder, kernelModule));
} catch (cause) {
primaryFailure = cause;
}
const completed = await completeVerifierRun(run);
if (
completed.process_lineage.status !== "verified" ||
completed.process_lineage.descendants.length !== 0
)
throw new Error(
`Cleanup unverified; retained diagnostic workspace: ${workspace}`,
{ cause: primaryFailure },
);
if (primaryFailure !== undefined)
throw new Error(
`MIPS relocatable verification failed; retained workspace: ${workspace}`,
{ cause: primaryFailure },
);
await rm(workspace, { recursive: true, force: true });
console.log(JSON.stringify({ ...completed, fixtures: reports }, null, 2));
+83 -7
View File
@@ -15,6 +15,7 @@ import {
assertMipsLoadedImage,
assertMipsProbe,
assertMipsGlobal,
assertMipsSoftFloatStorage,
} from "../tests/conformance/ghidra/mips-checks.mjs";
import { parseEvidence } from "../dist/domain/evidence.js";
import { parseAnalysisSnapshot } from "../dist/domain/analysisSnapshot.js";
@@ -77,16 +78,64 @@ const env = {
for (const setting of ["GHIDRA_HEADLESS_MAXMEM", "GHIDRA_MAXMEM"]) {
if (process.env[setting] !== undefined) env[setting] = process.env[setting];
}
/**
* Typed float signatures must use soft-float o32 storage, and callers must see
* the source constants through it rather than FPU inputs.
*/
async function verifySoftFloat(call, symbols, byteOrder) {
const boundaries = {};
for (const name of [
"rea_mips_soft_pick",
"rea_mips_soft_wide",
"rea_mips_soft_mixed",
"rea_mips_soft_struct",
]) {
const api = await call("inspect_native_api", { procedure: symbols[name] });
boundaries[name] = api.boundary;
}
assertMipsSoftFloatStorage(boundaries, byteOrder);
const caller = functionDossierSchema.parse(
await call("analyze_function", {
procedure: symbols.rea_mips_soft_caller,
}),
);
// Includes the struct call and the indirect call through a typed pointer.
assert.doesNotMatch(caller.pseudocode, /\bin_f\d+\b/u);
for (const literal of [
"1.5",
"2.5",
"3.25",
"4.75",
"5.5",
"6.125",
"7.5",
"8.5",
"9.5",
])
assert.ok(
caller.pseudocode.includes(literal),
`Soft-float caller lost argument ${literal}`,
);
return { storage: boundaries };
}
const reports = [];
let primaryFailure;
try {
for (const byteOrder of psp ? ["little"] : ["little", "big"]) {
for (const [byteOrder, floatAbi] of psp
? [["little", "hard"]]
: [
["little", "hard"],
["big", "hard"],
["little", "soft"],
["big", "soft"],
]) {
const fixture = psp
? await buildPspFixture(workspace)
: await buildMipsFixture(workspace, byteOrder);
: await buildMipsFixture(workspace, byteOrder, floatAbi);
const independent = psp
? fixture
: await inspectMipsReadelf(fixture.path, byteOrder);
: await inspectMipsReadelf(fixture.path, byteOrder, floatAbi);
const resolved = await parseBinaryTarget(fixture.path);
if (!resolved.ok) throw resolved.error;
if (psp) {
@@ -104,7 +153,10 @@ try {
);
}
assert.equal(resolved.value.sha256, fixture.sha256);
const snapshotPath = join(workspace, `${byteOrder}.snapshot.json`);
const snapshotPath = join(
workspace,
`${byteOrder}-${floatAbi}.snapshot.json`,
);
const transport = new StdioClientTransport({
command: process.execPath,
args: [entrypoint, "mcp"],
@@ -238,6 +290,10 @@ try {
callees.includes(leaf.address),
"Expected direct call not recovered",
);
const softFloat =
floatAbi === "soft"
? await verifySoftFloat(call, independent.symbols, byteOrder)
: undefined;
const assembly = await call("procedure_assembly", {
procedure: entry.address,
});
@@ -293,6 +349,19 @@ try {
},
"Snapshot profile did not retain the independently inspected ELF/ABI declaration",
);
assert.equal(
snapshot.binding.analysis_profile.parameters.mips_soft_float_storage,
floatAbi === "soft" ? "o32-soft-float-prototype-v1" : undefined,
);
assert.equal(
snapshot.evidence_bundle.records.some((record) =>
/Soft-float prototype outcome: 4 functions and [1-9]\d* function-pointer definitions use __rea_soft_float\./u.test(
JSON.stringify(record),
),
),
floatAbi === "soft",
"Soft-float prototype outcome missing from, or unexpected in, retained Evidence",
);
assert.ok(snapshot.evidence_bundle.records.length > 0);
if (psp) {
assert.equal(
@@ -419,7 +488,10 @@ try {
assert.ok(read32(offset + 4) + read32(offset + 16) <= cut);
}
const carvedBytes = sourceBytes.subarray(0, cut);
const carvedPath = join(workspace, `${byteOrder}-carved.elf`);
const carvedPath = join(
workspace,
`${byteOrder}-${floatAbi}-carved.elf`,
);
const carvedSha256 = createHash("sha256")
.update(carvedBytes)
.digest("hex");
@@ -460,7 +532,7 @@ try {
assert.ok(carvedDossier.pseudocode.trim().length > 0);
const carvedSnapshotPath = join(
workspace,
`${byteOrder}-carved.snapshot.json`,
`${byteOrder}-${floatAbi}-carved.snapshot.json`,
);
await call("close_binary", { snapshot_path: carvedSnapshotPath });
const carvedSnapshot = parseAnalysisSnapshot(
@@ -523,7 +595,11 @@ try {
: "conditional branch destination",
"global word and marker bytes",
"CLI/MCP and snapshot identity",
...(floatAbi === "soft"
? ["soft-float o32 parameter/return storage and caller constants"]
: []),
],
...(softFloat === undefined ? {} : { soft_float: softFloat }),
procedures: procedures.length,
carved: carved ?? {
status: "not_run",
@@ -534,7 +610,7 @@ try {
});
} catch (cause) {
throw new Error(
`MIPS ${byteOrder} verification failed; MCP stderr: ${stderr}`,
`MIPS ${byteOrder} ${floatAbi}-float verification failed; MCP stderr: ${stderr}`,
{ cause },
);
} finally {
+54 -2
View File
@@ -53,6 +53,15 @@ const env = {
...(process.env.JAVA_HOME === undefined
? {}
: { JAVA_HOME: process.env.JAVA_HOME }),
...(process.env.GHIDRA_HEADLESS_MAXMEM === undefined
? {}
: { GHIDRA_HEADLESS_MAXMEM: process.env.GHIDRA_HEADLESS_MAXMEM }),
...(process.env.REA_GHIDRA_DISABLED_ANALYZERS === undefined
? {}
: {
REA_GHIDRA_DISABLED_ANALYZERS:
process.env.REA_GHIDRA_DISABLED_ANALYZERS,
}),
};
const transport = new StdioClientTransport({
command: process.execPath,
@@ -160,6 +169,48 @@ try {
assert.equal(cli.normalized_result.base_address, "0x8000000");
assert.equal(cli.analysis_profile.digest, profiles[0]);
assert.deepEqual(await readFile(path), bytes);
const x86Targets = [];
for (const architecture of ["x86", "x86_64"]) {
const x86Bytes = Buffer.from("b82a000000c3", "hex"); // mov eax, 42; ret.
const x86Sha256 = createHash("sha256").update(x86Bytes).digest("hex");
const x86Path = join(workspace, `${architecture}.bin`);
await writeFile(x86Path, x86Bytes);
await call("open_binary", {
path: x86Path,
format: "raw-binary",
architecture,
entry_point: "0x1000",
});
const image = await call("inspect_native_load_image", {}, x86Sha256);
assert.equal(
image.normalized_result.status,
"verified",
JSON.stringify(image),
);
assert.equal(image.normalized_result.base_address, "0x1000");
assert.equal(image.normalized_result.observations.compiler_spec_id, "gcc");
assert.equal(image.analysis_profile.parameters.compiler_spec_id, "gcc");
const memory = await call(
"read_bytes",
{ address: "0x1000", length: x86Bytes.length },
x86Sha256,
);
assert.equal(memory.normalized_result.bytes_hex, x86Bytes.toString("hex"));
const pseudo = await call(
"procedure_pseudo_code",
{ procedure: "0x1000" },
x86Sha256,
);
assert.match(pseudo.normalized_result, /\b(?:42|0x2a)\b/iu);
await call("close_binary");
assert.deepEqual(await readFile(x86Path), x86Bytes);
x86Targets.push({
architecture,
sha256: x86Sha256,
compiler_spec_id: "gcc",
return_42: true,
});
}
report = {
ok: true,
mocked: false,
@@ -179,6 +230,7 @@ try {
invalid_entry_preserves_active_session: true,
load_images_verified: true,
original_unchanged: true,
x86_targets: x86Targets,
};
} catch (error) {
if (stderr.trim()) process.stderr.write(stderr);
@@ -199,7 +251,7 @@ process.stdout.write(
}) + "\n",
);
async function call(name, args = {}) {
async function call(name, args = {}, targetSha256 = sha256) {
const result = await client.callTool(
{ name, arguments: args },
{ timeout: 240000 },
@@ -213,7 +265,7 @@ async function call(name, args = {}) {
)
return projected;
const evidence = parseEvidence(result.structuredContent);
assert.equal(evidence.subject.digest.sha256, sha256);
assert.equal(evidence.subject.digest.sha256, targetSha256);
assert.equal(evidence.subject.format, "raw-binary");
return evidence;
}
+25
View File
@@ -122,6 +122,15 @@ try {
new URL("../tests/fixtures/wasm/custom-names.wat", import.meta.url),
"utf8",
),
`(module
(memory 1)
(data "x")
(func (export "init")
i32.const 0
i32.const 0
i32.const 1
memory.init 0
data.drop 0))\n`,
];
for (const [index, source] of sources.entries()) {
const fixtureRoot = join(root.path, `fixture-${index}`);
@@ -220,6 +229,22 @@ try {
'"unicode-é\\\t"',
],
);
if (index === 4) {
assert.match(
headers,
/^DataCount start=0x[\da-f]+ end=0x[\da-f]+ \(size=0x[\da-f]+\) count: 1$/m,
);
assert.deepEqual(
report.sections.find(({ kind }) => kind === "DataCount"),
{
kind: "DataCount",
start: 0x23,
end: 0x24,
bytes: 1,
description: "count: 1",
},
);
}
const bundleReply = await client.callTool({
name: "get_evidence_bundle",
arguments: {},
+62 -16
View File
@@ -42,11 +42,66 @@ try {
"",
].join("\n"),
);
const output = await runHeadless(target, "seed-regression", [
"-postScript",
"ReaSeedRegressionProbe.java",
seeds,
]);
// Headless can exit zero after a script fails, so require the probe's marker.
assert.ok(output.includes("REA_SEED_REGRESSION_OK "), output);
const analyzer = "Decompiler Switch Analysis";
const disabled = await runHeadless(target, "analyzer-regression-valid", [
"-preScript",
"ReaGhidraAnalysisOptions.java",
analyzer,
"-postScript",
"ReaAnalyzerOptionsRegressionProbe.java",
analyzer,
]);
assert.ok(
disabled.includes(`REA_ANALYZER_OPTIONS_DISABLED_OK ${analyzer}`),
disabled,
);
const unknownAnalyzer = "REA Unknown Analyzer";
const rejected = await runHeadless(target, "analyzer-regression-unknown", [
"-preScript",
"ReaGhidraAnalysisOptions.java",
`${analyzer},${unknownAnalyzer}`,
"-postScript",
"ReaAnalyzerOptionsRegressionProbe.java",
analyzer,
]);
assert.ok(
rejected.includes(
`Unknown Ghidra analyzer for this program: ${unknownAnalyzer}`,
),
rejected,
);
assert.ok(
rejected.includes("Processing aborted as a result of pre-script."),
rejected,
);
assert.ok(!rejected.includes("REA_ANALYZER_OPTIONS_DISABLED_OK"), rejected);
} finally {
if (supervisor !== undefined) {
const stopped = await supervisor.stop();
assert.notEqual(stopped.status, "incomplete", JSON.stringify(stopped));
supervisor.dispose();
}
await rm(workspace, { recursive: true, force: true });
}
console.log(
"Real Ghidra pre-script regression passed: descending function order, out-of-range seed, BSS label, BSS code rejection, analyzer option readback, unknown-analyzer abort before post-scripts, and owned-process cleanup.",
);
async function runHeadless(target, project, scriptArguments) {
const launch = await spawnOwnedProviderProcess({
command: join(installation, "support", "analyzeHeadless"),
arguments: [
workspace,
"seed-regression",
project,
"-import",
target,
"-loader",
@@ -63,9 +118,7 @@ try {
new URL("../../../tests/conformance/ghidra", import.meta.url),
),
].join(";"),
"-postScript",
"ReaSeedRegressionProbe.java",
seeds,
...scriptArguments,
],
runId: run.run_id,
expectedCommand: null,
@@ -82,16 +135,9 @@ try {
const snapshot = supervisor.snapshot();
const output = `${snapshot.stdout.text}\n${snapshot.stderr.text}`;
assert.equal(snapshot.exitCode, 0, output);
// Headless can exit zero after a script fails, so require the probe's marker.
assert.ok(output.includes("REA_SEED_REGRESSION_OK "), output);
} finally {
if (supervisor !== undefined) {
const stopped = await supervisor.stop();
assert.notEqual(stopped.status, "incomplete", JSON.stringify(stopped));
supervisor.dispose();
}
await rm(workspace, { recursive: true, force: true });
const stopped = await supervisor.stop();
assert.notEqual(stopped.status, "incomplete", JSON.stringify(stopped));
supervisor.dispose();
supervisor = undefined;
return output;
}
console.log(
"Real Ghidra seed regression passed: descending function order, out-of-range seed, BSS label, BSS code rejection, and owned-process cleanup.",
);
+41
View File
@@ -0,0 +1,41 @@
import { describe, expect, it } from "vitest";
import { adbHostFileName } from "./AdbHostFileName.js";
describe("adbHostFileName", () => {
it.each([
["/data/local/tmp/notes.txt", "notes.txt"],
["/data/data/com.example/databases/app.db", "app.db"],
["/sdcard/My Notes.txt", "My Notes.txt"],
["/system/etc/hosts", "hosts"],
[
"/data/local/tmp/split_config.arm64_v8a.apk",
"split_config.arm64_v8a.apk",
],
["/data/local/tmp/résumé.xml", "résumé.xml"],
["/data/local/tmp/a-b_c.d", "a-b_c.d"],
["/data/local/tmp/", "tmp"],
])("accepts the portable device path %s", (devicePath, fileName) => {
expect(adbHostFileName(devicePath)).toEqual({ ok: true, fileName });
});
it.each([
["/", "empty_file_name"],
["/data/local/tmp/..", "dot_segment"],
["/data/local/tmp/.", "dot_segment"],
["/data/local/tmp/a\\b", "path_separator"],
["/data/local/tmp/re\u007fport", "control_character"],
["/data/local/tmp/ta\tb", "control_character"],
["/data/local/tmp/CON", "reserved_name"],
["/data/local/tmp/nul.txt", "reserved_name"],
["/data/local/tmp/com1", "reserved_name"],
["/data/local/tmp/LPT9", "reserved_name"],
["/data/local/tmp/report:2024", "alternate_data_stream"],
["/data/local/tmp/report.", "trailing_dot_or_space"],
["/data/local/tmp/report ", "trailing_dot_or_space"],
["/data/local/tmp/a*b.txt", "unsupported_character"],
["/data/local/tmp/a?b.txt", "unsupported_character"],
] as const)("refuses %s as %s", (devicePath, reason) => {
expect(adbHostFileName(devicePath)).toEqual({ ok: false, reason });
});
});
+61
View File
@@ -0,0 +1,61 @@
import { posix } from "node:path";
/** Why a device-reported path cannot be used as a host file name. */
export type AdbHostFileNameReason =
| "empty_file_name"
| "dot_segment"
| "path_separator"
| "control_character"
| "unsupported_character"
| "reserved_name"
| "trailing_dot_or_space"
| "alternate_data_stream";
/** A portable host file name, or the exact reason the device name was refused. */
export type AdbHostFileNameResult =
| { readonly ok: true; readonly fileName: string }
| { readonly ok: false; readonly reason: AdbHostFileNameReason };
/**
* Win32 resolves these basenames to a device regardless of case and of any
* extension (`NUL`, `nul.txt`), so a portable host file name must avoid them on
* every host. `(?:\.|$)` keeps ordinary names such as `console.txt` admissible.
*/
const WINDOWS_RESERVED_BASENAME =
/^(?:con|prn|aux|nul|com[1-9]|lpt[1-9])(?:\.|$)/iu;
/** Characters that are not portable in a host file name on Windows or POSIX. */
const UNSUPPORTED_CHARACTER = /[*?"<>|]/u;
/** Control characters (`Cc`), including C0 controls and DEL. */
const CONTROL_CHARACTER = /\p{Cc}/u;
/**
* Derive a portable host file name from a device-reported path.
*
* The device path is a POSIX path on the device filesystem, so its last
* `/`-separated segment is taken with POSIX semantics rather than the host
* `basename` (which would vary between POSIX and Windows hosts). A name is
* accepted only when every host can store it and address it back unambiguously:
* no separator, control, or reserved-device characters, no trailing dot or
* space, no `:` alternate-data-stream marker, and no `.`/`..` segment.
*/
export const adbHostFileName = (devicePath: string): AdbHostFileNameResult => {
const fileName = posix.basename(devicePath);
if (fileName === "") return { ok: false, reason: "empty_file_name" };
if (fileName === "." || fileName === "..")
return { ok: false, reason: "dot_segment" };
if (fileName.includes("\\") || fileName.includes("/"))
return { ok: false, reason: "path_separator" };
if (CONTROL_CHARACTER.test(fileName))
return { ok: false, reason: "control_character" };
if (fileName.includes(":"))
return { ok: false, reason: "alternate_data_stream" };
if (UNSUPPORTED_CHARACTER.test(fileName))
return { ok: false, reason: "unsupported_character" };
if (WINDOWS_RESERVED_BASENAME.test(fileName))
return { ok: false, reason: "reserved_name" };
if (/[. ]$/u.test(fileName))
return { ok: false, reason: "trailing_dot_or_space" };
return { ok: true, fileName };
};
+68
View File
@@ -336,12 +336,38 @@ it("reports a missing selected binary as a capability gap with recovery", async
if (result.ok) return;
const projected = projectAnalysisError(result.error);
expect(projected).toMatchObject({ code: "capability_unavailable" });
expect(projected.message).toContain("REA_ADB_PATH");
expect(projected.remediation).toMatchObject({
action: expect.stringContaining("REA_ADB_PATH") as unknown,
});
expect(projected.details).toMatchObject({
provider_id: "adb",
reason: expect.stringContaining("not found") as unknown,
});
});
it("reports relative adb configuration as an actionable refusal", async () => {
const provider = new AdbProvider({
environment: { REA_ADB_PATH: "relative/adb" },
});
const availability = await provider.inspectAvailability();
expect(availability).toMatchObject({
status: "unavailable",
reason: expect.stringContaining("REA_ADB_PATH") as unknown,
});
const projected = await refusedBy(provider, {
operation: "list_adb_devices",
input: {},
});
expect(projected).toMatchObject({
code: "capability_unavailable",
message: expect.stringContaining("REA_ADB_PATH") as unknown,
remediation: {
action: expect.stringContaining("REA_ADB_PATH") as unknown,
},
});
});
it("cancels before spawning anything", async () => {
const { binary, root } = await writeAdbStub(
(root) => `: > ${join(root, "spawned")}
@@ -452,6 +478,48 @@ exit 1`,
});
});
it.each([
["/data/local/tmp/..", "dot_segment"],
["/data/local/tmp/a\\b", "path_separator"],
["/data/local/tmp/CON", "reserved_name"],
["/data/local/tmp/report.", "trailing_dot_or_space"],
["/data/local/tmp/report:2024", "alternate_data_stream"],
] as const)(
"refuses the unsafe device path %s before pulling",
async (devicePath, reason) => {
const { binary, root } = await writeAdbStub(
(root) => `if [ "$1" = "version" ]; then ${VERSION_BLOCK}; exit 0; fi
: > "${join(root, "pull-ran")}"
exit 0`,
);
const output = join(root, "pulls");
const result = await providerOver(binary).execute({
operation: "pull_adb_file",
input: {
serial: "emulator-5554",
device_path: devicePath,
output_directory: output,
},
});
expect(result.ok).toBe(false);
if (result.ok) return;
const projected = projectAnalysisError(result.error);
expect(projected).toMatchObject({ code: "invalid_request" });
expect(projected.details).toMatchObject({
issues: [
{
path: ["device_path"],
reason: "invalid_value",
message: expect.stringContaining(reason),
},
],
});
// No host directory and no adb invocation are effects of a refused pull.
await expect(readFile(join(root, "pull-ran"))).rejects.toThrow();
await expect(readFile(output)).rejects.toThrow();
},
);
it("pushes a file and verifies the device-side digest", async () => {
const { binary, root } = await writeAdbStub(
() => `if [ "$1" = "version" ]; then ${VERSION_BLOCK}; exit 0; fi
+20 -11
View File
@@ -2,12 +2,13 @@ import { execFile } from "node:child_process";
import { createHash } from "node:crypto";
import { constants as fsConstants } from "node:fs";
import { access, lstat, mkdir, rm, stat, writeFile } from "node:fs/promises";
import { basename, join, resolve } from "node:path";
import { join, resolve } from "node:path";
import type {
ExecutionOptions,
ProviderAvailability,
} from "../application/AnalysisProvider.js";
import { providerConfigurationRefusal } from "../domain/providerConfigurationRefusal.js";
import {
createAnalysisExecution,
type AnalysisExecution,
@@ -15,7 +16,6 @@ import {
import type { AdbRequest } from "../domain/adb/adbDeviceAnalysis.js";
import {
AnalysisArtifactChangedError,
AnalysisCapabilityUnavailableError,
AnalysisCancelledError,
AnalysisInputError,
AnalysisOutputError,
@@ -47,6 +47,7 @@ import {
parsePackageListOutput,
parsePmPathOutput,
} from "./AdbDeviceOutput.js";
import { adbHostFileName } from "./AdbHostFileName.js";
import {
parseAmStartOutput,
parseBugreportCopiedPath,
@@ -417,8 +418,8 @@ export class AdbProvider {
async inspectAvailability(
signal?: AbortSignal,
): Promise<ProviderAvailability> {
const selection = this.selection();
try {
const selection = this.selection();
await inspectAdbClient(selection, signal);
return {
status: "available",
@@ -555,21 +556,19 @@ export class AdbProvider {
const signal = options?.signal;
if (signal?.aborted === true)
return err(new AnalysisCancelledError(request.operation));
const selection = this.selection();
let selection: AdbBinarySelection;
let identity: AdbClientIdentity;
try {
selection = this.selection();
identity = await inspectAdbClient(selection, signal);
} catch (cause) {
if (cause instanceof AnalysisCancelledError) return err(cause);
if (cause instanceof AdbConfigurationFailure)
return err(
new AnalysisCapabilityUnavailableError(
providerConfigurationRefusal(
ADB_PROVIDER_IDENTITY.id,
request.operation,
typeof cause.diagnostics.remediation === "string"
? `${cause.message} ${cause.diagnostics.remediation}`
: cause.message,
{ cause },
cause,
),
);
return err(
@@ -1401,13 +1400,23 @@ export class AdbProvider {
},
signal: AbortSignal | undefined,
): Promise<Result<AnalysisExecution, AnalysisError>> {
const derived = adbHostFileName(input.device_path);
if (!derived.ok)
return err(
new AnalysisInputError(operation, undefined, [
{
path: ["device_path"],
reason: "invalid_value",
message: `The device path does not name a portable host file (${derived.reason}); REA never renames a device-reported file.`,
},
]),
);
const directory = await prepareOutputDirectory(
operation,
input.output_directory,
);
if (!directory.ok) return directory;
const fileName = basename(input.device_path) || "pulled";
const localPath = join(directory.value, fileName);
const localPath = join(directory.value, derived.fileName);
const existing = await refuseExistingTarget(operation, localPath);
if (existing !== undefined) return err(existing);
const pulled = await runAdb(
+11 -10
View File
@@ -20,6 +20,7 @@ import type {
ExecutionOptions,
ProviderAvailability,
} from "../application/AnalysisProvider.js";
import { providerConfigurationRefusal } from "../domain/providerConfigurationRefusal.js";
import {
createAnalysisExecution,
type AnalysisExecution,
@@ -153,12 +154,13 @@ const runApktool = async (
invocation = apktoolInvocation(selection, arguments_, hostEnvironment);
} catch (cause) {
if (!(cause instanceof WindowsBatchCommandError)) throw cause;
const reason = `The apktool batch launcher cannot receive REA's decode paths through cmd.exe: ${cause.message}`;
return err(
new AnalysisCapabilityUnavailableError(
APKTOOL_PROVIDER_IDENTITY.id,
operation,
`The apktool batch launcher cannot receive REA's decode paths through cmd.exe: ${cause.message}`,
{ cause },
reason,
{ cause, userMessage: reason },
),
);
}
@@ -345,8 +347,8 @@ export class ApktoolProvider {
async #inspectAvailability(
signal: AbortSignal,
): Promise<ProviderAvailability> {
const selection = this.selection();
try {
const selection = this.selection();
await inspectApktoolClient(selection, signal, this.environment);
throwIfCancelled("inspect_apktool_client", signal);
return {
@@ -397,9 +399,10 @@ export class ApktoolProvider {
): Promise<Result<AnalysisExecution, AnalysisError>> {
if (signal?.aborted === true)
return err(new AnalysisCancelledError(request.operation));
const selection = this.selection();
let selection: ApktoolCommandSelection;
let identity: ApktoolClientIdentity;
try {
selection = this.selection();
identity = await inspectApktoolClient(
selection,
signal,
@@ -407,17 +410,15 @@ export class ApktoolProvider {
);
} catch (cause) {
if (cause instanceof AnalysisCancelledError) return err(cause);
if (cause instanceof ApktoolConfigurationFailure)
if (cause instanceof ApktoolConfigurationFailure) {
return err(
new AnalysisCapabilityUnavailableError(
providerConfigurationRefusal(
APKTOOL_PROVIDER_IDENTITY.id,
request.operation,
typeof cause.diagnostics.remediation === "string"
? `${cause.message} ${cause.diagnostics.remediation}`
: cause.message,
{ cause },
cause,
),
);
}
return err(
new AnalysisProtocolError(
`apktool version probe failed for ${request.operation}`,
+21 -3
View File
@@ -1,4 +1,5 @@
import type {
AnalysisExecution,
AnalysisOperation,
ExecutionOptions,
} from "./AnalysisProvider.js";
@@ -17,6 +18,13 @@ type AnalysisCall = (
selectorOrigin?: ExecutionOptions["selectorOrigin"],
) => Promise<Result<JsonValue, AnalysisError>>;
type RelationshipCall = (
name: AnalysisOperation,
arguments_: Readonly<Record<string, JsonValue>>,
signal?: AbortSignal,
selectorOrigin?: ExecutionOptions["selectorOrigin"],
) => Promise<Result<AnalysisExecution, AnalysisError>>;
export interface CallPathTraceInput {
readonly start: string;
readonly goal?: string | undefined;
@@ -42,6 +50,7 @@ interface TraceState {
readonly error: ReturnType<typeof projectAnalysisError>;
}>;
readonly residual: Set<string>;
readonly limitations: Set<string>;
queueIndex: number;
traversalPath: string[];
}
@@ -63,6 +72,7 @@ interface DiscoveredEdge {
/** Trace direct caller/callee relationships until the graph is exhausted or the goal is reached. */
export const traceCallPath = async (
call: AnalysisCall,
callRelationship: RelationshipCall,
input: CallPathTraceInput,
signal?: AbortSignal,
): EnhancedResult => {
@@ -88,7 +98,7 @@ export const traceCallPath = async (
if (address === undefined) break;
const current = state.visited.get(address);
if (current === undefined) continue;
const found = await expandAddress(call, {
const found = await expandAddress(callRelationship, {
input: resolvedInput,
state,
address,
@@ -106,12 +116,13 @@ const createTraceState = (input: CallPathTraceInput): TraceState => ({
edges: new Map(),
failures: [],
residual: new Set(),
limitations: new Set(),
queueIndex: 0,
traversalPath: input.goal === input.start ? [input.start] : [],
});
const expandAddress = async (
call: AnalysisCall,
call: RelationshipCall,
expansion: AddressExpansion,
): Promise<boolean> => {
const { input, state, address, current, signal } = expansion;
@@ -127,7 +138,7 @@ const expandAddress = async (
state.residual.add(`Call relationships were unavailable for ${address}.`);
return false;
}
const related = parseRelatedAddresses(result.value, relation);
const related = parseRelatedAddresses(result.value.result, relation);
if (!related.ok) {
state.failures.push({
address,
@@ -136,6 +147,12 @@ const expandAddress = async (
state.residual.add(`Call relationships were unreadable for ${address}.`);
return false;
}
for (const limitation of result.value.limitations) {
state.limitations.add(limitation);
state.residual.add(
`Call relationships for ${address} are limited: ${limitation}`,
);
}
for (const relatedAddress of sortedUniqueAddresses(related.value)) {
recordEdge(input, state, {
address,
@@ -215,6 +232,7 @@ const projectTrace = (input: CallPathTraceInput, state: TraceState) => ({
residual_unknowns: [...state.residual],
limitations: [
"Direct provider call relationships may omit unresolved indirect calls.",
...state.limitations,
],
});
+24 -3
View File
@@ -1,6 +1,7 @@
import { z } from "zod";
import type {
AnalysisExecution,
AnalysisOperation,
AnalysisOperationPort,
ExecutionOptions,
@@ -166,6 +167,8 @@ export class EnhancedTools {
return traceCallPath(
(name, arguments_, operationSignal, selectorOrigin) =>
this.#call(name, arguments_, operationSignal, selectorOrigin),
(name, arguments_, operationSignal, selectorOrigin) =>
this.#observe(name, arguments_, operationSignal, selectorOrigin),
call.input,
signal,
);
@@ -359,7 +362,7 @@ export class EnhancedTools {
const level = String(current.depth);
graph[level] ??= [];
const result = await this.#call(
const result = await this.#observe(
tool,
{ procedure: current.address },
signal,
@@ -373,7 +376,7 @@ export class EnhancedTools {
});
continue;
}
const related = parseRelatedAddresses(result.value, relation);
const related = parseRelatedAddresses(result.value.result, relation);
if (!related.ok) {
graph[level].push({
address: current.address,
@@ -386,6 +389,9 @@ export class EnhancedTools {
address: current.address,
status: "ok",
calls: [...related.value],
...(result.value.limitations.length === 0
? {}
: { limitations: [...result.value.limitations] }),
});
for (const address of related.value) {
if (!discovered.has(address)) {
@@ -523,13 +529,28 @@ export class EnhancedTools {
signal?: AbortSignal,
selectorOrigin?: ExecutionOptions["selectorOrigin"],
): Promise<Result<JsonValue, AnalysisError>> {
const execution = await this.#observe(
name,
arguments_,
signal,
selectorOrigin,
);
return execution.ok ? ok(execution.value.result) : execution;
}
async #observe(
name: AnalysisOperation,
arguments_: Readonly<Record<string, JsonValue>>,
signal?: AbortSignal,
selectorOrigin?: ExecutionOptions["selectorOrigin"],
): Promise<Result<AnalysisExecution, AnalysisError>> {
if (isAborted(signal)) return err(new AnalysisCancelledError(name));
const execution = await this.analysis.execute(name, arguments_, {
...(signal === undefined ? {} : { signal }),
...(selectorOrigin === undefined ? {} : { selectorOrigin }),
});
if (isAborted(signal)) return err(new AnalysisCancelledError(name));
return execution.ok ? ok(execution.value.result) : execution;
return execution;
}
}
+17 -2
View File
@@ -8,6 +8,8 @@ import type {
} from "./SetupTypes.js";
import type { SetupClient } from "./SupportedClients.js";
import type { SetupHopperInstallResult } from "./SetupInstallFailure.js";
import { err, ok } from "../domain/result.js";
import { SetupSkillInstallError } from "./SetupSkill.js";
import type { DoctorCheck, DoctorReport, DoctorScope } from "./Doctor.js";
import type { LinuxDistribution } from "./LinuxHopper.js";
import { PRODUCT_IDENTITY, SDK_IDENTITY } from "../identity.js";
@@ -122,9 +124,22 @@ export class FakeSetupHost implements SetupHost {
);
skillNeedsInstall = (): Promise<boolean> =>
Promise.resolve(this.skill !== "unchanged");
installSkill = (): Promise<"installed" | "unchanged" | "failed"> => {
installSkill = () => {
this.skillInstalls += 1;
return Promise.resolve(this.skill);
return Promise.resolve(
this.skill === "failed"
? err(
new SetupSkillInstallError({
path: "/fixture/home/.agents/skills/reverse-engineer-anything/SKILL.md",
stage: "write",
reason: "injected skill write failure",
rollbackFailures: [],
recoveryBackupPaths: [],
cause: new Error("injected skill write failure"),
}),
)
: ok(this.skill),
);
};
doctor = (scope?: DoctorScope): Promise<DoctorReport> => {
this.doctorCalls += 1;
+18 -4
View File
@@ -6,6 +6,7 @@ import { ConfigurationError } from "../domain/configurationErrors.js";
import { SUPPORTED_NODE_VERSION_PROSE } from "../domain/runtimeVersion.js";
import { FakeSetupHost, options } from "./Setup.fixture.js";
import { runSetup } from "./Setup.js";
import type { SetupProgressEvent } from "./SetupTypes.js";
describe("setup workflow", () => {
it("rejects Node 25 before applying setup actions", async () => {
@@ -57,14 +58,27 @@ describe("setup workflow", () => {
const host = new FakeSetupHost();
host.hopper = "/Applications/Hopper";
host.skill = "failed";
const progress: SetupProgressEvent[] = [];
const result = await runSetup(
{ ...options(true), installSkill: true, clientIds: [] },
{
...options(true),
installSkill: true,
clientIds: [],
onProgress: (event) => {
if (event.actionId === "install_skill") progress.push(event);
},
},
host,
);
expect(result.status).toBe("needs_human");
expect(result.remediation).toBe(
"REA analysis skill could not be installed or verified. Check permissions for its planned skill directories, then rerun setup.",
);
expect(result.code).toBe("skill_install_failed");
expect(result.remediation).toContain("REA skill write failed at");
expect(result.remediation).toContain("injected skill write failure");
expect(result.remediation).toContain("Check permissions and rerun setup.");
expect(progress.at(-1)).toMatchObject({
state: "failed",
detail: expect.stringContaining("injected skill write failure"),
});
});
it("delegates remaining unhealthy checks to doctor remediation", async () => {
+18 -14
View File
@@ -26,6 +26,7 @@ import {
import { AnalysisError } from "../domain/analysisErrorBase.js";
import { analysisErrorWithCleanupFailure } from "../domain/analysisErrorCleanup.js";
import { ProviderCleanupError } from "../domain/providerCleanupError.js";
import type { SetupSkillInstallError } from "./SetupSkill.js";
/** Discover, approve, and apply setup actions idempotently. */
export const runSetup = async (
@@ -315,18 +316,20 @@ const runSetupWorkflow = async (
: { onProgress: options.onProgress }),
});
if (clientFailure !== undefined) return fail(clientFailure);
if (
planSelection.installSkill &&
!(await installSkillAction(
if (planSelection.installSkill) {
const skillFailure = await installSkillAction(
host,
options,
appliedActions,
selectedSkillClientIds,
))
)
return fail(
"REA analysis skill could not be installed or verified. Check permissions for its planned skill directories, then rerun setup.",
);
if (skillFailure !== undefined) {
return {
...(await fail(skillFailure.remediation)),
code: "skill_install_failed",
};
}
}
const doctor = summarizeDoctor(await host.doctor(readinessScope));
const remediation = combineRemediation(
finalSetupRemediation(
@@ -417,7 +420,7 @@ const installSkillAction = async (
options: SetupOptions,
appliedActions: string[],
skillClientIds: readonly string[],
): Promise<boolean> => {
): Promise<SetupSkillInstallError | undefined> => {
const label = "REA reverse-engineering skill";
emitProgress(options, {
actionId: "install_skill",
@@ -425,22 +428,23 @@ const installSkillAction = async (
state: "started",
});
const skill = await host.installSkill(skillClientIds);
if (skill === "failed") {
if (!skill.ok) {
emitProgress(options, {
actionId: "install_skill",
label,
state: "failed",
detail: skill.error.message,
});
return false;
return skill.error;
}
if (skill === "installed") appliedActions.push("installed_skill");
if (skill.value === "installed") appliedActions.push("installed_skill");
emitProgress(options, {
actionId: "install_skill",
label,
state: skill === "installed" ? "completed" : "warning",
...(skill === "unchanged" ? { detail: "Already current" } : {}),
state: skill.value === "installed" ? "completed" : "warning",
...(skill.value === "unchanged" ? { detail: "Already current" } : {}),
});
return true;
return undefined;
};
const setupFailure = async (
+1
View File
@@ -11,6 +11,7 @@ export type SetupFailureCode =
| "mount_failed"
| "bundle_invalid"
| "copy_failed"
| "skill_install_failed"
| "cleanup_failed";
export type SetupHopperInstallResult =
+183 -46
View File
@@ -2,10 +2,15 @@ import { copyFile, lstat, mkdir, readFile, rm } from "node:fs/promises";
import { constants as fsConstants } from "node:fs";
import { dirname, join } from "node:path";
import writeFileAtomic from "write-file-atomic";
import { err, ok, type Result } from "../domain/result.js";
import { readBoundedFileBytes } from "../process/BoundedFileBytes.js";
import { PRODUCT_IDENTITY } from "../identity.js";
import { NonRegularFileReadError } from "../filesystem/RegularFile.js";
import { readRegularFileText } from "./RegularFileRead.js";
import {
NonRegularFileReadError,
RegularFileChangedError,
} from "../filesystem/RegularFile.js";
import { withRegularFile } from "./RegularFileRead.js";
import { clientSkillDirectories } from "./SupportedClients.js";
export const MANAGED_SKILL_FILES = [
@@ -23,14 +28,119 @@ export const isManagedSkillManifest = (content: string): boolean =>
content.startsWith(`---\r\nname: ${PRODUCT_IDENTITY.skillName}\r\n`);
export const readSkillFile = (path: string): Promise<string> =>
readRegularFileText(path, { symlinks: "reject" });
readSkillFileState(path).then(({ content }) => content);
interface SkillFileState {
readonly content: string;
readonly mode: number;
}
const readSkillFileState = async (path: string): Promise<SkillFileState> =>
withRegularFile(
path,
async (handle, stats) => {
const bytes = await readBoundedFileBytes(
handle,
stats.size,
undefined,
stats.size,
);
if (bytes === undefined || bytes.length !== stats.size)
throw new RegularFileChangedError(path);
return { content: bytes.toString("utf8"), mode: stats.mode & 0o7777 };
},
{ symlinks: "reject" },
);
interface CanonicalSkillFile {
readonly content: string;
readonly destination: string;
readonly original: string | undefined;
readonly original: SkillFileState | undefined;
}
export interface SkillInstallFileOperations {
readonly mkdir: (path: string) => Promise<void>;
readonly backup: (source: string, destination: string) => Promise<void>;
readonly write: (
path: string,
content: string,
mode: number,
) => Promise<void>;
readonly remove: (path: string) => Promise<void>;
}
export interface SkillInstallRollbackFailure {
readonly path: string;
readonly reason: string;
}
/** Expected failure while replacing a managed skill bundle. */
export class SetupSkillInstallError extends Error {
readonly _tag = "SetupSkillInstallError";
readonly path: string;
readonly stage: "prepare" | "backup" | "write" | "readback";
readonly reason: string;
readonly rollbackFailures: readonly SkillInstallRollbackFailure[];
readonly recoveryBackupPaths: readonly string[];
constructor(input: {
readonly path: string;
readonly stage: "prepare" | "backup" | "write" | "readback";
readonly reason: string;
readonly rollbackFailures: readonly SkillInstallRollbackFailure[];
readonly recoveryBackupPaths: readonly string[];
readonly cause: unknown;
}) {
const message = [
`REA skill ${input.stage} failed at ${input.path}: ${input.reason}.`,
...(input.rollbackFailures.length === 0
? []
: [
`Rollback also failed for ${input.rollbackFailures.map(({ path, reason }) => `${path} (${reason})`).join(", ")}.`,
...(input.recoveryBackupPaths.length === 0
? []
: [
`Existing-file backups remain at ${input.recoveryBackupPaths.join(", ")}.`,
]),
]),
].join(" ");
super(message, { cause: input.cause });
this.path = input.path;
this.stage = input.stage;
this.reason = input.reason;
this.rollbackFailures = input.rollbackFailures;
this.recoveryBackupPaths = input.recoveryBackupPaths;
}
get remediation(): string {
return `${this.message} Check permissions and rerun setup.`;
}
}
export type SkillInstallValue = "installed" | "unchanged";
export type SkillInstallResult = Result<
SkillInstallValue,
SetupSkillInstallError
>;
const systemSkillInstallFileOperations: SkillInstallFileOperations = {
mkdir: async (path) => {
await mkdir(path, { recursive: true });
},
backup: async (source, destination) => {
try {
await copyFile(source, destination, fsConstants.COPYFILE_EXCL);
} catch (cause: unknown) {
if (cause instanceof Error && "code" in cause && cause.code === "EEXIST")
return;
throw cause;
}
},
write: (path, content, mode) =>
writeFileAtomic(path, content, { encoding: "utf8", mode }),
remove: (path) => rm(path, { force: true }),
};
/** Managed skill location selected for a client family. */
export interface SkillDestination {
readonly client: string;
@@ -87,9 +197,11 @@ export const existingSkillDestinations = async (
return existing;
};
const readOptionalText = async (path: string): Promise<string | undefined> => {
const readOptionalSkillState = async (
path: string,
): Promise<SkillFileState | undefined> => {
try {
return await readSkillFile(path);
return await readSkillFileState(path);
} catch (cause: unknown) {
if (cause instanceof Error && "code" in cause && cause.code === "ENOENT")
return undefined;
@@ -117,7 +229,7 @@ const canonicalSkillFiles = async (
),
"utf8",
),
original: await readOptionalText(destination),
original: await readOptionalSkillState(destination),
};
}),
),
@@ -133,7 +245,7 @@ export const canonicalSkillNeedsInstall = async (
try {
return (
await canonicalSkillFiles(home, clientIds, environment, platform)
).some(({ content, original }) => original !== content);
).some(({ content, original }) => original?.content !== content);
} catch (cause: unknown) {
// Unreadable skill state fails open to install so setup can repair it.
void cause;
@@ -141,39 +253,42 @@ export const canonicalSkillNeedsInstall = async (
}
};
const writeText = (path: string, content: string): Promise<void> =>
writeFileAtomic(path, content, { encoding: "utf8", mode: 0o600 });
const causeReason = (cause: unknown): string =>
cause instanceof Error ? cause.message : String(cause);
const preserveSkillBackup = async (
source: string,
destination: string,
): Promise<void> => {
try {
await copyFile(source, destination, fsConstants.COPYFILE_EXCL);
} catch (cause: unknown) {
if (cause instanceof Error && "code" in cause && cause.code === "EEXIST")
return;
throw cause;
}
};
const causePath = (cause: unknown): string | undefined =>
cause instanceof Error && "path" in cause && typeof cause.path === "string"
? cause.path
: undefined;
const restoreSkillFiles = async (
changed: readonly CanonicalSkillFile[],
): Promise<void> => {
for (const { destination, original } of [...changed].reverse()) {
if (original === undefined) await rm(destination, { force: true });
else await writeText(destination, original);
attempted: readonly CanonicalSkillFile[],
operations: SkillInstallFileOperations,
): Promise<readonly SkillInstallRollbackFailure[]> => {
const failures: SkillInstallRollbackFailure[] = [];
for (const { destination, original } of [...attempted].reverse()) {
try {
if (original === undefined) await operations.remove(destination);
else await operations.write(destination, original.content, original.mode);
} catch (cause: unknown) {
failures.push({ path: destination, reason: causeReason(cause) });
}
}
return failures;
};
/** Transactionally install or upgrade the canonical REA skill and references. */
/** Transactionally install the canonical skill bundle and optional references. */
export const installCanonicalSkill = async (
home: string,
clientIds: readonly string[] = [],
environment: Readonly<NodeJS.ProcessEnv> = {},
platform: NodeJS.Platform = process.platform,
): Promise<"installed" | "unchanged" | "failed"> => {
let changed: readonly CanonicalSkillFile[] = [];
operations: SkillInstallFileOperations = systemSkillInstallFileOperations,
): Promise<SkillInstallResult> => {
let attempted: CanonicalSkillFile[] = [];
let stage: "prepare" | "backup" | "write" | "readback" = "prepare";
let failingPath = "skill bundle";
try {
const canonical = await canonicalSkillFiles(
home,
@@ -181,30 +296,52 @@ export const installCanonicalSkill = async (
environment,
platform,
);
changed = canonical.filter(({ content, original }) => original !== content);
if (changed.length === 0) return "unchanged";
const changed = canonical.filter(
({ content, original }) => original?.content !== content,
);
if (changed.length === 0) return ok("unchanged");
stage = "backup";
for (const { destination, original } of changed) {
await mkdir(dirname(destination), { recursive: true });
failingPath = destination;
await operations.mkdir(dirname(destination));
if (original !== undefined)
await preserveSkillBackup(destination, `${destination}.rea.backup`);
await operations.backup(destination, `${destination}.rea.backup`);
}
for (const { destination, content } of changed)
await writeText(destination, content);
for (const { destination, content } of changed)
stage = "write";
for (const file of changed) {
failingPath = file.destination;
// An atomic writer can publish and then reject; include the target before
// entering it so rollback covers that ambiguous outcome.
attempted.push(file);
await operations.write(
file.destination,
file.content,
file.original?.mode ?? 0o600,
);
}
stage = "readback";
for (const { destination, content } of changed) {
failingPath = destination;
if ((await readSkillFile(destination)) !== content)
throw new Error(`skill readback mismatch: ${destination}`);
return "installed";
} catch (cause: unknown) {
// Install failure preserves the original error outcome; report cause inline.
void cause;
try {
await restoreSkillFiles(changed);
} catch (restoreCause: unknown) {
// Per-file backups remain beside changed files for operator recovery.
void restoreCause;
}
return "failed";
return ok("installed");
} catch (cause: unknown) {
failingPath = causePath(cause) ?? failingPath;
const rollbackFailures = await restoreSkillFiles(attempted, operations);
return err(
new SetupSkillInstallError({
path: failingPath,
stage,
reason: causeReason(cause),
rollbackFailures,
recoveryBackupPaths: attempted.flatMap(({ destination, original }) =>
original === undefined ? [] : [`${destination}.rea.backup`],
),
cause,
}),
);
}
};
+2 -4
View File
@@ -1,6 +1,6 @@
import type { DoctorReport, DoctorScope, runDoctor } from "./Doctor.js";
import type { LinuxDistribution } from "./LinuxHopper.js";
import type { SkillDestination } from "./SetupSkill.js";
import type { SkillDestination, SkillInstallResult } from "./SetupSkill.js";
import type { SetupClient } from "./SupportedClients.js";
import type {
SetupFailureCode,
@@ -70,9 +70,7 @@ export interface SetupHost {
command: readonly string[],
): Promise<ClientConfigurationInspection>;
skillNeedsInstall(clientIds: readonly string[]): Promise<boolean>;
installSkill(
clientIds: readonly string[],
): Promise<"installed" | "unchanged" | "failed">;
installSkill(clientIds: readonly string[]): Promise<SkillInstallResult>;
doctor(scope?: DoctorScope): Promise<Awaited<ReturnType<typeof runDoctor>>>;
}
/** Structured setup outcome carrying remediation instead of prompting. */
+1
View File
@@ -38,6 +38,7 @@ export interface WebModuleResolutionBatch {
/** Replaceable native resolution engine, with no provider protocol in callers. */
export interface WebModuleResolutionPort {
close?(): Promise<void>;
resolve(
input: {
readonly importerUrl: string;
+4
View File
@@ -69,6 +69,10 @@ export class WebModuleTraceService {
readonly resolver: WebModuleResolutionPort,
) {}
close(): Promise<void> {
return this.resolver.close?.() ?? Promise.resolve();
}
async trace(
rawInput: unknown,
options?: ExecutionOptions,
@@ -115,6 +115,63 @@ it("exports independent nested query, workflow and Evidence payloads", () => {
expect(serializeAnalysisSnapshot(fresh.value)).toBe(expected);
});
it("rebinds replayed target subjects while preserving child artifact subjects", () => {
const cache = new AnalysisSnapshotCache();
const target = ANALYSIS_SNAPSHOT_TARGET;
const profile = ANALYSIS_SNAPSHOT_PROFILE;
const relocatedTarget = { ...target, path: "/tmp/relocated-app" };
const parameters = { address: "0x1000" };
const locations = [{ kind: "address" as const, address: "0x1000" }];
cache.record({
target,
profile,
operation: "procedure_info",
parameters,
execution: createAnalysisExecution(
{ name: "main" },
ANALYSIS_SNAPSHOT_PROVIDER,
{
subject: target,
rawResult: { provider: "raw" },
locations,
analysisProfile: profile,
},
),
});
const replayedTarget = cache.lookup(
relocatedTarget,
profile,
"procedure_info",
parameters,
);
expect(replayedTarget?.subject?.path).toBe(relocatedTarget.path);
expect(replayedTarget?.rawResult).toEqual({ provider: "raw" });
expect(replayedTarget?.locations).toEqual(locations);
const childSubject = {
path: "/tmp/child.zip",
sha256: "b".repeat(64),
format: "zip" as const,
};
const childParameters = { address: "0x2000" };
cache.record({
target,
profile,
operation: "procedure_info",
parameters: childParameters,
execution: createAnalysisExecution(
{ name: "child" },
ANALYSIS_SNAPSHOT_PROVIDER,
{ subject: childSubject, analysisProfile: profile },
),
});
expect(
cache.lookup(relocatedTarget, profile, "procedure_info", childParameters)
?.subject,
).toMatchObject(childSubject);
});
describe("analysis snapshot cache capacity", () => {
it("shares capacity across query kinds while allowing replacements and fresh partitions", () => {
const cache = new AnalysisSnapshotCache();
+17 -16
View File
@@ -8,6 +8,7 @@ import {
} from "../../domain/evidenceErrors.js";
import { NoBinaryOpenError } from "../../domain/configurationErrors.js";
import type { AnalysisError } from "../../domain/analysisErrorBase.js";
import { rebindEvidenceSubjectTargetPath } from "../../domain/evidence.js";
import { err, ok, type Result } from "../../domain/result.js";
import {
analysisQueryId,
@@ -277,7 +278,21 @@ export class AnalysisSnapshotCache {
);
const cached = this.#entries.get(queryId);
if (cached === undefined) return undefined;
const subject = cached.execution.subject;
const cachedSubject = cached.execution.subject;
const subject =
cachedSubject === null
? null
: rebindEvidenceSubjectTargetPath(
{
path: cachedSubject.path,
sha256: cachedSubject.sha256,
format: cachedSubject.format,
...(cachedSubject.architecture === null
? {}
: { architecture: cachedSubject.architecture }),
},
target,
);
return structuredClone({
result: cached.execution.result,
rawResult: cached.execution.raw_result,
@@ -288,21 +303,7 @@ export class AnalysisSnapshotCache {
"Loaded from a local REA analysis snapshot; this call did not re-run the provider.",
],
locations: cached.execution.locations,
subject:
subject === null
? null
: subject.architecture === null
? {
path: subject.path,
sha256: subject.sha256,
format: subject.format,
}
: {
path: subject.path,
sha256: subject.sha256,
format: subject.format,
architecture: subject.architecture,
},
subject,
});
}
@@ -91,4 +91,7 @@ const parameters = (
): EvidenceObservation["parameters"] => ({
format: input.format,
integrity_policy: input.integrity_policy,
...(input.source_paths === undefined
? {}
: { source_paths: input.source_paths }),
});
@@ -66,6 +66,9 @@ export const analyzeJavaScriptApplicationValidated = async (
input_path: input.input_path,
format: input.format,
integrity_policy: input.integrity_policy,
...(input.source_paths === undefined
? {}
: { source_paths: input.source_paths }),
...(input.max_heap_mb === undefined
? {}
: { max_heap_mb: input.max_heap_mb }),
@@ -11,7 +11,10 @@ import type { JavaScriptSemanticIr } from "../../domain/javascript/javascriptSem
import { projectJavaScriptModuleSemantics } from "../../domain/javascript/javascriptModuleSemanticIr.js";
import { parseJavaScriptSource } from "../../domain/javascript/javascriptSourceParser.js";
import { hasValidSourceMapContents } from "../../domain/sourceMapContents.js";
import { flattenSourceMapLeaves } from "../../domain/sourceMapEnvelope.js";
import {
flattenSourceMapLeaves,
sourceMapJsonText,
} from "../../domain/sourceMapEnvelope.js";
import type {
JavaScriptSourceRange,
JavaScriptSourcePoint,
@@ -137,12 +140,31 @@ export const analyzeAndAdoptJavaScriptArtifactFiles = async (
}> => {
const state = emptyArtifactAnalysis<JavaScriptModuleSemanticIr>();
let interruption: AnalysisError | null = null;
for (const [index, file] of fileSet.files.entries()) {
const selectedFileCount =
fileSet.files.length - (fileSet.source_selection?.unselected_files ?? 0);
let selectedFileIndex = 0;
// Caller-excluded sources are never analyzed, so neither cancellation nor
// another interruption describes them.
const recordOutsideSelection = (file: JavaScriptArtifactFile): boolean => {
if (file.text.included || file.text.reason !== "outside-analysis-scope")
return false;
state.files.push({ file, javascript: null, semantic: null });
if (file.kind === "javascript")
projection.recordUnavailableFile(
file,
"Source was not analyzed because it is outside the caller-selected source_paths.",
null,
"incomplete-module",
);
return true;
};
for (const file of fileSet.files) {
try {
await setImmediate();
await beforeFile(file, index, fileSet.files.length);
if (signal?.aborted === true)
throw new AnalysisCancelledError("analyze_javascript_application");
if (recordOutsideSelection(file)) continue;
await beforeFile(file, selectedFileIndex++, selectedFileCount);
if (file.kind !== "javascript" || !file.text.included) {
const steps = analyzeArtifactFileSteps(file, {
state,
@@ -232,7 +254,7 @@ export const analyzeAndAdoptJavaScriptArtifactFiles = async (
: cause instanceof AnalysisError
? cause
: null;
if (interruption !== null) {
if (interruption !== null && !recordOutsideSelection(file)) {
state.files.push({
file,
javascript: null,
@@ -266,7 +288,8 @@ export const analyzeAndAdoptJavaScriptArtifactFiles = async (
: null,
};
for (const file of fileSet.files) {
if (analyzedPaths.has(file.path)) continue;
if (analyzedPaths.has(file.path) || recordOutsideSelection(file))
continue;
state.files.push({
file,
javascript: null,
@@ -336,6 +359,10 @@ function* analyzeArtifactFileSteps<
context: ArtifactAnalysisContext<SemanticIr>,
): Generator<void, void> {
const { state } = context;
if (!file.text.included && file.text.reason === "outside-analysis-scope") {
state.files.push({ file, javascript: null, semantic: null });
return;
}
addStructuredObservations(file, state);
if (file.kind === "html" && file.text.included)
state.htmlScripts.push(...parseHtmlScripts(file.path, file.text.value));
@@ -486,12 +513,12 @@ const parseSourceMap = (
if (!file.text.included) return unavailableSourceMap(file);
let value: unknown;
try {
value = JSON.parse(file.text.value);
value = JSON.parse(sourceMapJsonText(file.text.value));
} catch (cause: unknown) {
void cause;
return invalidSourceMap(file, "Source map is not valid JSON.");
}
const maps = flattenSourceMaps(value);
const maps = flattenSourceMapLeaves(value);
if (maps === undefined)
return invalidSourceMap(file, "Source map is not a version 3 map.");
return collectSourceMapOriginals(file, maps);
@@ -556,13 +583,6 @@ const unavailableSourceMap = (
};
};
const flattenSourceMaps = (
root: unknown,
): Readonly<Record<string, unknown>>[] | undefined => {
const leaves = flattenSourceMapLeaves(root);
return leaves === undefined ? undefined : [...leaves];
};
const invalidSourceMap = (
file: JavaScriptArtifactFile,
limitation: string,
@@ -168,6 +168,7 @@ const graphCoverage = (context: JavaScriptArtifactGraphContext) => {
javascript !== null && javascript.parse_status === "partial",
);
const unknownGap =
(context.fileSet.source_selection?.unselected_files ?? 0) > 0 ||
context.analysis.files.some(
({ analysis_failure, application_projection_failure }) =>
analysis_failure !== undefined ||
@@ -244,6 +245,11 @@ const graphLimitations = (
);
return [
...context.analysis.limitations,
...(context.fileSet.source_selection === undefined
? []
: [
`Caller-selected source analysis leaves ${String(context.fileSet.source_selection.unselected_files)} inventoried files outside analysis scope. Inventory identity covers the complete artifact; relationships through unselected sources remain unknown.`,
]),
...context.snapshot.integrity_contradictions.map(
({ logical_path: path }) =>
`Artifact integrity metadata contradicts observed bytes at ${path}; the observed bytes are untrusted.`,
@@ -199,7 +199,15 @@ export const addJavaScriptArtifactFiles = (
}),
});
}
if (analyzed.analysis_failure !== undefined)
if (!file.text.included && file.text.reason === "outside-analysis-scope")
addUnavailableStaticParseScope(context, {
file,
asset: target,
operation: "select-javascript-sources",
limitation:
"Source is outside the caller-selected source_paths; it was inventoried but not analyzed.",
});
else if (analyzed.analysis_failure !== undefined)
addUnavailableStaticParseScope(context, {
file,
asset: target,
@@ -176,9 +176,7 @@ const contextualCandidate = (
): string | CandidateResolution => {
const { context } = input;
let declared =
context === "html-reference" ||
context === "url-reference" ||
(context === "module-specifier" && input.moduleKind !== "require")
context === "html-reference" || context === "url-reference"
? stripQueryAndFragment(input.declaredPath)
: input.declaredPath;
if (context === "html-reference") return htmlCandidate(input);
@@ -190,9 +188,9 @@ const contextualCandidate = (
"URL and Node builtin schemes are outside static artifact module resolution.",
]);
if (!declared.startsWith(".") && !declared.startsWith("/"))
return bareModuleCandidate(input, declared);
return bareModuleCandidate(input);
if (input.moduleKind !== "require") {
const decoded = decodeModuleUrlPath(declared);
const decoded = decodeModuleUrlPath(stripQueryAndFragment(declared));
if (typeof decoded !== "string") return decoded;
declared = decoded;
}
@@ -237,34 +235,51 @@ const relativeCandidate = (sourcePath: string, declared: string): string =>
const bareModuleCandidate = (
input: ResolveArtifactPathInput,
declared: string,
): CandidateResolution => {
const packageName = barePackageName(declared);
if (packageName === null || isBuiltin(declared))
const request = parseBarePackageRequest(input.declaredPath, input.moduleKind);
if (request === null || isBuiltin(input.declaredPath))
return unresolvedCandidate("external", [
"The bare specifier is a Node builtin or invalid package name.",
]);
if (input.moduleKind === "import" && packageName.includes("%"))
if (input.moduleKind === "import" && request.packageName.includes("%"))
return unresolvedCandidate("rejected", [
"Percent signs are not admitted in ES module package names.",
]);
const self = commonJsSelfCandidate(input, declared);
const self = packageSelfCandidate(input, request);
if (self !== null) return self;
const subpath =
declared === packageName ? "." : `.${declared.slice(packageName.length)}`;
// ESM constructs a package.json URL; its parsed identity is not a literal
// dependency directory. Self resolution above still uses the exact name.
let dependencyDirectory = posix.join("node_modules", request.packageName);
if (input.moduleKind === "import") {
const metadataUrl = new URL(
`./node_modules/${request.packageName}/package.json`,
PACKAGE_ARTIFACT_ROOT_URL,
);
if (!metadataUrl.pathname.endsWith("/package.json"))
return unresolvedCandidate("external", [
`ESM package name ${JSON.stringify(request.packageName)} produces a dependency URL whose pathname does not name package.json; no literal or suffix-stripped package directory was searched.`,
]);
const decoded = decodeModuleUrlPath(
metadataUrl.pathname.slice(PACKAGE_ARTIFACT_ROOT_URL.pathname.length),
);
if (typeof decoded !== "string") return decoded;
const confined = confineCandidate(decoded);
if (typeof confined !== "string") return confined;
dependencyDirectory = posix.dirname(confined);
}
const source = input.files.get(input.sourcePath);
let directory = posix.dirname(input.sourcePath);
while (true) {
const candidate = posix.join(directory, "node_modules", packageName);
const candidate = posix.join(directory, dependencyDirectory);
const metadata = readPackageMetadata(input, candidate);
if (packageExportsEligible(input) && metadata.kind !== "absent")
return resolvePackageSpecifier(input, candidate, subpath, metadata);
if (subpath === ".") {
if (request.dependencyExportsEligible && metadata.kind !== "absent")
return resolvePackageDependency(input, candidate, request, metadata);
if (request.packageSubpath === ".") {
const direct = resolveFileCandidates(input, fileCandidates(candidate));
if (direct !== null) return direct;
}
if (hasContainerDirectory(input.files, candidate, source?.container_sha256))
return resolvePackageSpecifier(input, candidate, subpath, metadata);
return resolvePackageDependency(input, candidate, request, metadata);
if (directory === "." || directory === "") break;
directory = posix.dirname(directory);
}
@@ -330,24 +345,25 @@ const resolvePackageImportAlias = (
return unresolvedCandidate("external", [
`Wildcard package import aliases remain unresolved; package metadata ${metadata.path} has no exact imports key ${specifier}.`,
]);
const selected = selectPackageImportTarget(
const selected = selectExportTarget(
lookup.target,
input.moduleKind,
metadata.path,
packageExportConditions(input.moduleKind),
packageMetadataUrl(metadata.path),
"imports",
);
if (selected.status === "rejected")
if (selected.kind === "invalid" || selected.kind === "invalid-config")
return unresolvedCandidate("rejected", [
`Package metadata ${metadata.path}: ${selected.limitation}`,
]);
if (selected.status === "unmatched")
if (selected.kind === "unmatched")
return unresolvedCandidate("external", [
`Package metadata ${metadata.path} declares no imports target for alias ${specifier} under the active conditions; it declares ${selected.declared.join(", ") || "no conditions"}.`,
]);
if (selected.status === "blocked")
if (selected.kind === "blocked")
return unresolvedCandidate("external", [
`Package metadata ${metadata.path} blocks import alias ${specifier} with ${selected.cause === "null" ? "an explicit null imports target" : "an empty imports target array"}.`,
]);
if (selected.status === "non-relative") {
if (selected.kind === "non-relative") {
const target = classifyPackageImportTargetString(selected.value);
return unresolvedCandidate("external", [
target.kind === "recursive"
@@ -358,77 +374,44 @@ const resolvePackageImportAlias = (
return resolveExportTarget(input, metadata.path, selected, "imports");
};
const selectPackageImportTarget = (
value: unknown,
moduleKind: ResolveArtifactPathInput["moduleKind"],
packagePath: string,
):
| ({ readonly status: "value" } & PackageFileTarget)
| { readonly status: "non-relative"; readonly value: string }
| { readonly status: "blocked"; readonly cause: PackageTargetBlockCause }
| { readonly status: "unmatched"; readonly declared: readonly string[] }
| { readonly status: "rejected"; readonly limitation: string } => {
const selected = selectExportTarget(
value,
packageExportConditions(moduleKind),
packageMetadataUrl(packagePath),
"imports",
);
if (selected.kind === "invalid" || selected.kind === "invalid-config")
return { status: "rejected", limitation: selected.limitation };
if (selected.kind === "unmatched")
return {
status: "unmatched",
declared: selected.declared,
};
if (selected.kind === "blocked")
return { status: "blocked", cause: selected.cause };
if (selected.kind === "non-relative")
return { status: "non-relative", value: selected.value };
return {
status: "value",
value: selected.value,
url: selected.url,
};
};
/** Self references use the nearest package scope before dependency lookup. */
const commonJsSelfCandidate = (
const packageSelfCandidate = (
input: ResolveArtifactPathInput,
declared: string,
request: BarePackageRequest,
): CandidateResolution | null => {
if (input.moduleKind !== "require" || !input.files.has(input.sourcePath))
if (input.moduleKind === undefined || !input.files.has(input.sourcePath))
return null;
const metadata = enclosingPackageMetadata(input);
if (metadata.kind === "none") return null;
if (metadata.kind === "unavailable")
return unresolvedCandidate("unavailable", [metadata.limitation]);
if (
metadata.name === undefined ||
metadata.exports === undefined ||
(declared !== metadata.name && !declared.startsWith(`${metadata.name}/`))
)
if (metadata.name === undefined || metadata.exports === undefined)
return null;
const subpath =
declared === metadata.name
? "."
: `.${declared.slice(metadata.name.length)}`;
let subpath: string;
if (input.moduleKind === "import") {
if (metadata.name !== request.packageName) return null;
subpath = request.packageSubpath;
} else if (input.declaredPath === metadata.name) {
subpath = ".";
} else if (input.declaredPath.startsWith(`${metadata.name}/`)) {
subpath = `.${input.declaredPath.slice(metadata.name.length)}`;
} else return null;
// Self resolution does not apply the CommonJS dependency-name matcher.
return resolvePackageExports(input, metadata.path, subpath, metadata.exports);
};
/** Package roots and exact subpaths share metadata and exports selection. */
const resolvePackageSpecifier = (
/** Resolve an installed dependency using one parsed request and its metadata. */
const resolvePackageDependency = (
input: ResolveArtifactPathInput,
directory: string,
subpath: string,
request: BarePackageRequest,
metadata: PackageMetadata,
): CandidateResolution => {
const exportsEligible = packageExportsEligible(input);
if (exportsEligible && metadata.kind === "unavailable")
const subpath = request.packageSubpath;
if (request.dependencyExportsEligible && metadata.kind === "unavailable")
return unresolvedCandidate("unavailable", [metadata.limitation]);
const rawExports =
exportsEligible && metadata.kind === "available"
request.dependencyExportsEligible && metadata.kind === "available"
? metadata.exports
: undefined;
if (rawExports === undefined) {
@@ -437,8 +420,17 @@ const resolvePackageSpecifier = (
resolveFileCandidates(input, fileCandidates(directory)) ??
resolveDirectory(input, directory, metadata, new Set())
);
// Preserve the trailing slash for a package directory request.
const candidate = confineCandidate(`${directory}/${subpath.slice(2)}`);
// Resolve legacy ESM subpaths as URLs: suffixes identify the loaded module
// but do not belong to its filesystem path. Keep the declared specifier on
// the relationship; only derive the inventory lookup path here.
const fileSubpath =
input.moduleKind === "import"
? decodeModuleUrlPath(stripQueryAndFragment(subpath.slice(2)))
: subpath.slice(2);
if (typeof fileSubpath !== "string") return fileSubpath;
const candidate = confineCandidate(`${directory}/${fileSubpath}`);
if (input.moduleKind === "import" && typeof candidate === "string")
return resolveLegacyEsmSubpath(input, candidate);
return typeof candidate === "string"
? resolveCandidate(input, candidate, new Set())
: candidate;
@@ -451,6 +443,21 @@ const resolvePackageSpecifier = (
);
};
/** Legacy ESM package subpaths are exact URLs, never extension or directory lookups. */
const resolveLegacyEsmSubpath = (
input: ResolveArtifactPathInput,
candidate: string,
): CandidateResolution => {
const file = resolveFileCandidates(input, [candidate]);
if (file !== null) return file;
const source = input.files.get(input.sourcePath);
if (hasContainerDirectory(input.files, candidate, source?.container_sha256))
return unresolvedCandidate("rejected", [
`ESM package subpath ${candidate} names a directory; directory imports are unsupported and package main or index lookup was not attempted.`,
]);
return exactPackageTargetNotFound(candidate, "ESM subpath");
};
const resolvePackageExports = (
input: ResolveArtifactPathInput,
packagePath: string,
@@ -517,23 +524,36 @@ const resolveExportTarget = (
);
};
const barePackageName = (specifier: string): string | null => {
const segments = specifier.split("/");
if (specifier.startsWith("@"))
return segments.length >= 2 && segments[0] !== "" && segments[1] !== ""
? `${segments[0]}/${segments[1]}`
: null;
return segments[0] === "" ? null : (segments[0] ?? null);
};
/** Derived package identity and dependency policy; input retains the raw request. */
interface BarePackageRequest {
readonly packageName: string;
readonly packageSubpath: string;
readonly dependencyExportsEligible: boolean;
}
const packageExportsEligible = (input: ResolveArtifactPathInput): boolean => {
if (input.moduleKind !== "require") return true;
const parseBarePackageRequest = (
specifier: string,
moduleKind: ResolveArtifactPathInput["moduleKind"],
): BarePackageRequest | null => {
const segments = specifier.split("/");
let packageName = segments[0] ?? "";
if (packageName === "") return null;
if (specifier.startsWith("@")) {
const packageSegment = segments[1];
if (packageSegment === undefined || packageSegment === "") return null;
packageName = `${packageName}/${packageSegment}`;
}
// Node's CommonJS exports matcher excludes percent-containing package names.
// Compare the captured name so a scoped name cannot fall back to its scope.
const match = /^((?:@[^/\\%]+\/)?[^./\\%][^/\\%]*)(\/.*)?$/.exec(
input.declaredPath,
);
return match?.[1] === barePackageName(input.declaredPath);
const dependencyExportsEligible =
moduleKind !== "require" ||
/^((?:@[^/\\%]+\/)?[^./\\%][^/\\%]*)(\/.*)?$/.exec(specifier)?.[1] ===
packageName;
return {
packageName,
packageSubpath: `.${specifier.slice(packageName.length)}`,
dependencyExportsEligible,
};
};
const hasContainerDirectory = (
@@ -1061,12 +1081,12 @@ const notFoundCandidate = (): CandidateResolution => ({
const exactPackageTargetNotFound = (
path: string,
mapKind: "exports" | "imports",
targetKind: "exports" | "imports" | "ESM subpath",
): CandidateResolution => ({
resolvedPath: null,
status: "not-found",
limitations: [
`The selected package ${mapKind === "imports" ? "import" : "exports"} target ${path} does not exist in the inventoried artifact container; ${mapKind === "imports" ? "import" : "exports"} targets are exact files, so no extension or directory index was tried.`,
`The selected package ${targetKind === "imports" ? "import" : targetKind} target ${path} does not exist in the inventoried artifact container; ${targetKind === "imports" ? "import" : targetKind} targets are exact files, so no extension or directory index was tried.`,
],
});
@@ -11,6 +11,7 @@ import {
import { createJavaScriptArtifactReader as createReader } from "../../artifacts/javascript/JavaScriptArtifactReader.js";
import type { JavaScriptApplicationGraph } from "../../domain/javascript/javascriptApplicationGraph.js";
import type { JavaScriptSemanticGraph } from "../../domain/javascript/javascriptSemanticGraph.js";
import type { JavaScriptSourceSelection } from "../../domain/javascript/javascriptSourceSelection.js";
import type { ElectronBoundarySummary } from "../../domain/javascript/javascriptApplicationAnalysis.js";
import { parseOwnedJavaScriptApplicationAnalysisSteps } from "../../domain/javascript/javascriptApplicationAnalysis.js";
import { AnalysisError } from "../../domain/analysisErrorBase.js";
@@ -45,6 +46,7 @@ export interface JavaScriptArtifactReconstructionResult {
readonly inventory_manifest_id: string;
readonly inventory_graph_sha256: string;
readonly integrity_contradictions: ArtifactInventorySnapshot["integrity_contradictions"];
readonly source_selection?: JavaScriptSourceSelection;
readonly graph: JavaScriptApplicationGraph;
readonly semantic_graph: JavaScriptSemanticGraph;
readonly electron_summary: ElectronBoundarySummary;
@@ -119,6 +121,7 @@ export const reconstructJavaScriptArtifact = async (
snapshot,
scope,
signal,
input.source_paths,
);
// Release acquisition before analysis consumes these owned file facts.
const earlyCleanup = await scope.release(readerOwner);
@@ -134,7 +137,7 @@ export const reconstructJavaScriptArtifact = async (
);
await reportPhase(
"parse_javascript_sources",
`Parsing and projecting ${String(files.files.length)} application source files`,
`Parsing and projecting ${String(files.files.length - (files.source_selection?.unselected_files ?? 0))} selected application source files`,
);
const semanticProjection = createJavaScriptSemanticGraphProjection();
const worker = createJavaScriptSourceAnalysis(
@@ -237,6 +240,9 @@ export const reconstructJavaScriptArtifact = async (
inventory_manifest_id: snapshot.manifest.manifest_id,
inventory_graph_sha256: snapshot.manifest.graph_sha256,
integrity_contradictions: snapshot.integrity_contradictions,
...(files.source_selection === undefined
? {}
: { source_selection: files.source_selection }),
graph,
semantic_graph: semanticGraph,
electron_summary: summarizeElectronBoundaries(analysis),
@@ -1,12 +1,14 @@
import { z } from "zod";
import { artifactIntegrityPolicySchema } from "../../domain/artifactIntegrityPolicy.js";
import { javaScriptAnalysisResourceControlsSchema } from "../../domain/javascript/javascriptAnalysisResourceControls.js";
import { javaScriptSourcePathsSchema } from "../../domain/javascript/javascriptSourceSelection.js";
/** Local ASAR/directory reconstruction request. */
export const javascriptArtifactReconstructionInputSchema = z.strictObject({
input_path: z.string().min(1),
format: z.enum(["auto", "asar", "directory"]).default("auto"),
integrity_policy: artifactIntegrityPolicySchema,
source_paths: javaScriptSourcePathsSchema.optional(),
...javaScriptAnalysisResourceControlsSchema.shape,
});
@@ -130,6 +130,16 @@ export const buildJavaScriptSemanticGraph = ({
for (const analyzed of analysis.files) {
if (analyzed.semantic !== null)
projection.projectFile(analyzed.file, analyzed.semantic.ir);
else if (
!analyzed.file.text.included &&
analyzed.file.text.reason === "outside-analysis-scope"
)
projection.recordUnavailableFile(
analyzed.file,
"Source was not analyzed because it is outside the caller-selected source_paths.",
null,
"incomplete-module",
);
}
return projection.finish(rootArtifactSha256, applicationGraph);
};
@@ -73,11 +73,11 @@ describe("JebAnalysisService admission", () => {
const requests: JebRequest[] = [];
const service = new JebAnalysisService(fakePort(requests));
const result = await service.execute("list_jeb_units", {});
expect(result.ok).toBe(true);
if (!result.ok) return;
expect(requests[0]?.operation).toBe("list_jeb_units");
if (!result.ok) throw result.error;
expect(requests).toHaveLength(1);
const request = requests[0];
if (request?.operation !== "list_jeb_units") return;
if (request?.operation !== "list_jeb_units")
throw new Error("Expected the list_jeb_units provider request");
expect(request.input.index).toBe(0);
expect(request.input.count).toBe(100);
expect(result.value.normalized_result).toEqual({ observed: true });
+31 -3
View File
@@ -26,11 +26,35 @@ export interface ArtifactEntry {
readonly adapterKey: string;
/** Filesystem identity captured during traversal, when supplied by an adapter. */
readonly sourceIdentity?: {
readonly device: number;
readonly inode: number;
readonly device: bigint;
readonly inode: bigint;
readonly size: bigint;
readonly mode: bigint;
readonly mtimeNs: bigint;
readonly ctimeNs: bigint;
};
}
/** Compare captured filesystem identity and mutable metadata at a read boundary. */
export const sameArtifactSourceIdentity = (
expected: ArtifactEntry["sourceIdentity"],
observed: {
readonly dev: bigint;
readonly ino: bigint;
readonly size: bigint;
readonly mode: bigint;
readonly mtimeNs: bigint;
readonly ctimeNs: bigint;
},
): boolean =>
expected !== undefined &&
observed.dev === expected.device &&
observed.ino === expected.inode &&
observed.size === expected.size &&
observed.mode === expected.mode &&
observed.mtimeNs === expected.mtimeNs &&
observed.ctimeNs === expected.ctimeNs;
/** Snapshot mutable entry metadata before the producer later validates open(). */
export const copyArtifactEntry = (entry: ArtifactEntry): ArtifactEntry => ({
...entry,
@@ -60,7 +84,11 @@ export const sameArtifactEntry = (
(limitation, index) => limitation === produced.limitations[index],
) &&
entry.sourceIdentity?.device === produced.sourceIdentity?.device &&
entry.sourceIdentity?.inode === produced.sourceIdentity?.inode;
entry.sourceIdentity?.inode === produced.sourceIdentity?.inode &&
entry.sourceIdentity?.size === produced.sourceIdentity?.size &&
entry.sourceIdentity?.mode === produced.sourceIdentity?.mode &&
entry.sourceIdentity?.mtimeNs === produced.sourceIdentity?.mtimeNs &&
entry.sourceIdentity?.ctimeNs === produced.sourceIdentity?.ctimeNs;
/** Read-only adapter over one directory, archive, or virtual container. */
export interface ArtifactReader {
+238 -110
View File
@@ -1,10 +1,4 @@
import {
constants,
createReadStream,
createWriteStream,
type ReadStream,
type Stats,
} from "node:fs";
import { constants, createWriteStream, type Stats } from "node:fs";
import {
lstat,
mkdtemp,
@@ -46,51 +40,40 @@ export class AsarArtifactReader implements ArtifactReader {
#archiveSize: number | undefined;
#headerSize: number | undefined;
readonly #entries = new Map<string, AsarEntryState>();
readonly #unpackedHandles = new Map<
UnpackedFileHandle,
{
readonly path: string;
readonly owner: OwnedFileHandle<UnpackedFileHandle>;
}
>();
readonly #ownedHandles = new Map<SharedAsarFileHandle, OwnedAsarHandle>();
#snapshot: { readonly path: string; readonly sha256: string } | undefined;
#snapshotRoot: string | undefined;
#snapshotOperation: Promise<void> = Promise.resolve();
constructor(
private readonly path: string,
private readonly openUnpackedFile: OpenAsarUnpackedFile = open,
private readonly removeSnapshot: (path: string) => Promise<void> = (path) =>
rm(path, { recursive: true, force: true }),
private readonly options: AsarArtifactReaderOptions = {},
) {}
async *entries(signal?: AbortSignal): AsyncIterable<ArtifactEntry> {
let paths: readonly AsarListedPath[];
try {
await this.prepareContainer(undefined, signal);
const snapshotPath = this.#snapshotPath();
this.#resetArchiveState();
this.#entries.clear();
const headerHandle = await open(
snapshotPath,
constants.O_RDONLY | constants.O_NOFOLLOW | constants.O_NONBLOCK,
);
try {
const admitted = await admitAsarHeader(headerHandle);
paths = await this.#withSnapshotOperation(async () => {
const snapshotPath = this.#snapshotPath();
this.#resetArchiveState();
this.#entries.clear();
const admitted = await this.#admitHeader(snapshotPath);
this.#archiveSize = admitted.archiveBytes;
this.#headerSize = admitted.headerBytes;
} finally {
await headerHandle.close();
}
abortIfNeeded(signal);
paths = orderAsarListing(listPackage(snapshotPath, { isPack: false }));
const archiveMetadata = await lstat(snapshotPath);
if (!archiveMetadata.isFile() || archiveMetadata.isSymbolicLink())
throw new ArtifactReaderFailure(
"format",
"ASAR container is not a regular file",
abortIfNeeded(signal);
const ordered = orderAsarListing(
listPackage(snapshotPath, { isPack: false }),
);
this.#archiveSize = archiveMetadata.size;
const archiveMetadata = await lstat(snapshotPath);
if (!archiveMetadata.isFile() || archiveMetadata.isSymbolicLink())
throw new ArtifactReaderFailure(
"format",
"ASAR container is not a regular file",
);
this.#archiveSize = archiveMetadata.size;
return ordered;
}, signal);
} catch (cause: unknown) {
throw asarFailure(this.path, "inventory", cause);
}
@@ -142,7 +125,47 @@ export class AsarArtifactReader implements ArtifactReader {
}
}
async #admitHeader(snapshotPath: string) {
const headerHandle = await this.#openOwned(
snapshotPath,
constants.O_RDONLY | constants.O_NOFOLLOW | constants.O_NONBLOCK,
);
let admission:
| {
readonly ok: true;
readonly value: Awaited<ReturnType<typeof admitAsarHeader>>;
}
| { readonly ok: false; readonly cause: unknown };
try {
admission = { ok: true, value: await admitAsarHeader(headerHandle) };
} catch (cause: unknown) {
admission = { ok: false, cause };
}
try {
await this.#closeOwnedHandle(headerHandle);
} catch (cleanupCause: unknown) {
if (!admission.ok)
throw ArtifactReaderFailure.withCleanup(
admission.cause,
ArtifactReaderFailure.cleanupObservation(cleanupCause, snapshotPath),
);
throw cleanupCause;
}
if (!admission.ok) throw admission.cause;
return admission.value;
}
async open(entry: ArtifactEntry, signal?: AbortSignal): Promise<Readable> {
return this.#withSnapshotOperation(
() => this.#openEntry(entry, signal),
signal,
);
}
async #openEntry(
entry: ArtifactEntry,
signal?: AbortSignal,
): Promise<Readable> {
abortIfNeeded(signal);
const state = this.#entries.get(entry.adapterKey);
if (state === undefined && entry.kind !== "file")
@@ -160,10 +183,16 @@ export class AsarArtifactReader implements ArtifactReader {
"integrity",
`ASAR entry metadata changed since inventory: ${state.entry.path}`,
);
const { metadata } = state;
const producedEntry = state.entry;
if (producedEntry.unpacked)
return this.#openUnpackedStream(producedEntry, signal);
if (state.entry.unpacked)
return this.#openUnpackedStream(state.entry, signal);
return this.#openPackedStream(state, signal);
}
async #openPackedStream(
state: AsarEntryState,
signal?: AbortSignal,
): Promise<Readable> {
const { metadata, entry: producedEntry } = state;
const archiveSize = this.#archiveSize;
const headerSize = this.#headerSize;
const offset =
@@ -187,10 +216,11 @@ export class AsarArtifactReader implements ArtifactReader {
"format",
`ASAR entry range is outside its container: ${producedEntry.path}`,
);
let handle: FileHandle | undefined;
const snapshotPath = this.#snapshotPath();
let handle: SharedAsarFileHandle | undefined;
try {
const openedHandle = await open(
this.#snapshotPath(),
const openedHandle = await this.#openOwned(
snapshotPath,
constants.O_RDONLY | constants.O_NOFOLLOW | constants.O_NONBLOCK,
);
handle = openedHandle;
@@ -202,25 +232,48 @@ export class AsarArtifactReader implements ArtifactReader {
`ASAR container changed before read: ${producedEntry.path}`,
);
if (metadata.size === 0) {
await closeAsarHandle(() => openedHandle.close(), signal);
handle = undefined;
await closeAsarHandle(
() => this.#closeOwnedHandle(openedHandle),
signal,
);
return Readable.from([]);
}
const source = openedHandle.createReadStream({
const source = this.#ownedReadStream(openedHandle, {
start,
end: start + metadata.size - 1,
autoClose: true,
signal,
});
return readValidatedAsarEntry(
handle = undefined;
const output = readValidatedAsarEntry(
source,
metadata.size,
producedEntry.path,
this.path,
signal,
);
this.#setOwnedOutput(openedHandle, output);
return output;
} catch (cause: unknown) {
await handle?.close().catch(() => undefined);
throw asarFailure(this.path, `read ${producedEntry.path}`, cause);
const failure = asarFailure(
this.path,
`read ${producedEntry.path}`,
cause,
);
if (handle !== undefined) {
try {
await this.#closeOwnedHandle(handle);
} catch (cleanupCause: unknown) {
throw ArtifactReaderFailure.withCleanup(
failure,
ArtifactReaderFailure.cleanupObservation(
cleanupCause,
snapshotPath,
),
);
}
}
throw failure;
}
}
@@ -246,49 +299,37 @@ export class AsarArtifactReader implements ArtifactReader {
}
/** Read the same captured container that supplies headers and packed members. */
async openContainer(signal?: AbortSignal): Promise<ReadStream> {
async openContainer(signal?: AbortSignal): Promise<Readable> {
await this.prepareContainer(undefined, signal);
return createReadStream(
this.#snapshotPath(),
signal === undefined ? {} : { signal },
);
return this.#withSnapshotOperation(async () => {
abortIfNeeded(signal);
const snapshotPath = this.#snapshotPath();
const handle = await this.#openOwned(
snapshotPath,
constants.O_RDONLY | constants.O_NOFOLLOW | constants.O_NONBLOCK,
);
return this.#ownedReadStream(handle, { signal });
}, signal);
}
async close(): Promise<void> {
await this.#withSnapshotOperation(async () => {
let cleanupFailure: ArtifactReaderFailure | undefined;
for (const [handle, { path }] of this.#unpackedHandles) {
try {
await this.#closeUnpackedHandle(handle);
} catch (cause: unknown) {
cleanupFailure = ArtifactReaderFailure.withCleanup(
cleanupFailure ?? cause,
ArtifactReaderFailure.cleanupObservation(cause, path),
);
}
}
try {
await this.#closeSnapshot();
} catch (cause: unknown) {
cleanupFailure = ArtifactReaderFailure.withCleanup(
cleanupFailure ?? cause,
ArtifactReaderFailure.cleanupObservation(cause, this.path),
);
}
if (cleanupFailure !== undefined) throw cleanupFailure;
});
await this.#withSnapshotOperation(() => this.#closeSnapshot());
}
async #captureSnapshot(signal?: AbortSignal): Promise<void> {
let inputHandle: SharedAsarFileHandle | undefined;
try {
this.#snapshotRoot = await mkdtemp(join(tmpdir(), "rea-asar-snapshot-"));
const snapshotPath = join(this.#snapshotRoot, "container.asar");
const hash = createHash("sha256");
const handle = await open(
inputHandle = await this.#openOwned(
this.path,
constants.O_RDONLY | constants.O_NOFOLLOW | constants.O_NONBLOCK,
);
const source = handle.createReadStream();
const source = this.#ownedReadStream(inputHandle, {
signal,
releaseOnClose: false,
});
const digesting = new Transform({
transform(chunk: Buffer, _encoding, callback) {
hash.update(chunk);
@@ -301,34 +342,50 @@ export class AsarArtifactReader implements ArtifactReader {
createWriteStream(snapshotPath, { flags: "wx", mode: 0o600 }),
signal === undefined ? {} : { signal },
);
const completedInput = inputHandle;
inputHandle = undefined;
await this.#closeOwnedHandle(completedInput);
this.#snapshot = { path: snapshotPath, sha256: hash.digest("hex") };
} catch (cause: unknown) {
const primary =
let failure =
signal?.aborted === true
? new ArtifactReaderFailure("cancelled", "ASAR operation cancelled", {
cause,
})
: asarFailure(this.path, "snapshot", cause);
if (inputHandle !== undefined) {
const failedInput = inputHandle;
inputHandle = undefined;
try {
await this.#closeOwnedHandle(failedInput);
} catch (cleanupCause: unknown) {
failure = ArtifactReaderFailure.withCleanup(
failure,
ArtifactReaderFailure.cleanupObservation(cleanupCause, this.path),
);
}
}
try {
await this.#closeSnapshot();
if (this.#ownedHandles.size === 0) await this.#closeSnapshot();
} catch (cleanupCause: unknown) {
throw ArtifactReaderFailure.withCleanup(
primary,
failure,
ArtifactReaderFailure.cleanupObservation(
cleanupCause,
this.#snapshotRoot ?? this.path,
),
);
}
throw primary;
throw failure;
}
}
async #closeSnapshot(): Promise<void> {
await this.#closeOwnedHandles();
const root = this.#snapshotRoot;
if (root !== undefined) {
try {
await this.removeSnapshot(root);
await (this.options.removeSnapshot ?? defaultRemoveSnapshot)(root);
if (this.#snapshotRoot === root) this.#snapshotRoot = undefined;
} catch (cause: unknown) {
throw ArtifactReaderFailure.withCleanup(cause, {
@@ -386,7 +443,7 @@ export class AsarArtifactReader implements ArtifactReader {
signal?: AbortSignal,
): Promise<Readable> {
const unpackedRoot = `${this.path}.unpacked`;
let handle: UnpackedFileHandle | undefined;
let handle: SharedAsarFileHandle | undefined;
try {
const rootMetadata = await lstat(unpackedRoot);
if (!rootMetadata.isDirectory() || rootMetadata.isSymbolicLink())
@@ -413,15 +470,12 @@ export class AsarArtifactReader implements ArtifactReader {
"path",
`ASAR unpacked entry is not a regular file: ${entry.path}`,
);
handle = await this.openUnpackedFile(
handle = await (this.options.openUnpackedFile ?? open)(
canonical,
constants.O_RDONLY | constants.O_NOFOLLOW | constants.O_NONBLOCK,
);
const openedHandle = handle;
this.#unpackedHandles.set(openedHandle, {
path: canonical,
owner: new OwnedFileHandle(openedHandle),
});
this.#adoptHandle(openedHandle, canonical);
const openedMetadata = await openedHandle.stat();
if (
!openedMetadata.isFile() ||
@@ -441,30 +495,28 @@ export class AsarArtifactReader implements ArtifactReader {
);
}
abortIfNeeded(signal);
const source = readFileHandleChunks(openedHandle, {
start: 0,
});
source.once("close", () => {
void this.#closeUnpackedHandle(openedHandle).catch(() => undefined);
});
return readValidatedAsarEntry(
const source = this.#ownedReadStream(openedHandle, { start: 0 });
handle = undefined;
const output = readValidatedAsarEntry(
source,
undefined,
entry.path,
this.path,
signal,
);
this.#setOwnedOutput(openedHandle, output);
return output;
} catch (cause: unknown) {
let failure = asarUnpackedEntryFailure(this.path, entry, cause);
if (handle !== undefined && this.#unpackedHandles.has(handle)) {
if (handle !== undefined && this.#ownedHandles.has(handle)) {
try {
await this.#closeUnpackedHandle(handle);
await this.#closeOwnedHandle(handle);
} catch (cleanupCause: unknown) {
failure = ArtifactReaderFailure.withCleanup(
failure,
ArtifactReaderFailure.cleanupObservation(
cleanupCause,
this.#unpackedHandles.get(handle)?.path ?? entry.path,
this.#ownedHandles.get(handle)?.path ?? entry.path,
),
);
}
@@ -473,11 +525,76 @@ export class AsarArtifactReader implements ArtifactReader {
}
}
async #closeUnpackedHandle(handle: UnpackedFileHandle): Promise<void> {
const owned = this.#unpackedHandles.get(handle);
async #closeOwnedHandle(handle: SharedAsarFileHandle): Promise<void> {
const owned = this.#ownedHandles.get(handle);
if (owned === undefined) return;
await owned.owner.close();
this.#unpackedHandles.delete(handle);
owned.output?.destroy();
owned.source?.destroy();
try {
await owned.owner.close();
} catch (cause: unknown) {
throw ArtifactReaderFailure.withCleanup(
cause,
ArtifactReaderFailure.cleanupObservation(cause, owned.path),
);
}
this.#ownedHandles.delete(handle);
}
async #closeOwnedHandles(): Promise<void> {
let failure: ArtifactReaderFailure | undefined;
for (const [handle, owned] of this.#ownedHandles) {
try {
await this.#closeOwnedHandle(handle);
} catch (cause: unknown) {
failure = ArtifactReaderFailure.withCleanup(
failure ?? cause,
ArtifactReaderFailure.cleanupObservation(cause, owned.path),
);
}
}
if (failure !== undefined) throw failure;
}
async #openOwned(path: string, flags: number): Promise<SharedAsarFileHandle> {
const handle = await (this.options.openFile ?? open)(path, flags);
this.#adoptHandle(handle, path);
return handle;
}
#adoptHandle(handle: SharedAsarFileHandle, path: string): void {
this.#ownedHandles.set(handle, {
path,
owner: new OwnedFileHandle(handle),
});
}
#ownedReadStream(
handle: SharedAsarFileHandle,
options: {
readonly start?: number;
readonly end?: number;
readonly signal?: AbortSignal | undefined;
readonly releaseOnClose?: boolean;
},
): Readable {
const { releaseOnClose, ...readOptions } = options;
const source = readFileHandleChunks(handle, readOptions);
const owned = this.#ownedHandles.get(handle);
if (owned === undefined)
throw new ArtifactReaderFailure("integrity", "ASAR handle has no owner");
owned.source = source;
owned.output = source;
if (releaseOnClose !== false)
source.once("close", () => {
void this.#closeOwnedHandle(handle).catch(() => undefined);
});
return source;
}
#setOwnedOutput(handle: SharedAsarFileHandle, output: Readable): void {
const owned = this.#ownedHandles.get(handle);
if (owned !== undefined) owned.output = output;
}
#resetArchiveState(): void {
@@ -491,16 +608,27 @@ type AsarEntryState = {
readonly metadata: AsarFileMetadata;
readonly entry: ArtifactEntry;
};
type UnpackedFileHandle = {
readonly fd: number;
type SharedAsarFileHandle = Pick<FileHandle, "fd" | "read" | "close"> & {
stat(): Promise<Stats>;
close(): Promise<void>;
read: FileHandle["read"];
};
type OpenAsarUnpackedFile = (
type OpenAsarFile = (
path: string,
flags: number,
) => Promise<UnpackedFileHandle>;
) => Promise<SharedAsarFileHandle>;
type OwnedAsarHandle = {
readonly path: string;
readonly owner: OwnedFileHandle<SharedAsarFileHandle>;
source?: Readable;
output?: Readable;
};
type AsarArtifactReaderOptions = {
readonly openFile?: OpenAsarFile;
readonly openUnpackedFile?: OpenAsarFile;
readonly removeSnapshot?: (path: string) => Promise<void>;
};
const defaultRemoveSnapshot = (path: string): Promise<void> =>
rm(path, { recursive: true, force: true });
const asarUnpackedEntryFailure = (
archivePath: string,
+2 -1
View File
@@ -1,3 +1,4 @@
import type { Stats } from "node:fs";
import type { FileHandle } from "node:fs/promises";
import { ArtifactReaderFailure } from "./ArtifactReader.js";
@@ -7,7 +8,7 @@ export const MAX_ASAR_HEADER_BYTES = 16 * 1024 * 1024;
/** Admit both Chromium pickle extents before an upstream decoder allocates them. */
export const admitAsarHeader = async (
handle: FileHandle,
handle: Pick<FileHandle, "read"> & { stat(): Promise<Stats> },
): Promise<{
readonly archiveBytes: number;
readonly headerBytes: number;
+76 -20
View File
@@ -1,4 +1,4 @@
import { constants, type Stats } from "node:fs";
import { constants, type BigIntStats } from "node:fs";
import {
lstat,
open,
@@ -11,6 +11,7 @@ import { Readable } from "node:stream";
import {
ArtifactReaderFailure,
sameArtifactSourceIdentity,
type ArtifactEntry,
type ArtifactReader,
} from "./ArtifactReader.js";
@@ -95,9 +96,11 @@ export class DirectoryArtifactReader implements ArtifactReader {
if (child === null) break;
abortIfNeeded(signal);
const absolute = join(directory, child.name);
const metadata = await lstat(absolute).catch((cause: unknown) => {
throw directoryIoFailure("inspect entry", absolute, cause);
});
const metadata = await lstat(absolute, { bigint: true }).catch(
(cause: unknown) => {
throw directoryIoFailure("inspect entry", absolute, cause);
},
);
const path = relative(root, absolute).split(sep).join("/");
const entry = artifactEntry(path, absolute, metadata);
if (entry === undefined) continue;
@@ -158,18 +161,18 @@ export class DirectoryArtifactReader implements ArtifactReader {
const owner = new OwnedFileHandle(fileHandle);
this.#files.set(owner, entry.adapterKey);
try {
const observed = await owner.handle.stat().catch((cause: unknown) => {
throw directoryIoFailure(
"inspect opened entry",
entry.adapterKey,
cause,
);
});
const observed = await owner.handle
.stat({ bigint: true })
.catch((cause: unknown) => {
throw directoryIoFailure(
"inspect opened entry",
entry.adapterKey,
cause,
);
});
if (
!observed.isFile() ||
entry.sourceIdentity === undefined ||
observed.dev !== entry.sourceIdentity.device ||
observed.ino !== entry.sourceIdentity.inode
!sameArtifactSourceIdentity(entry.sourceIdentity, observed)
) {
throw new ArtifactReaderFailure(
"integrity",
@@ -178,7 +181,7 @@ export class DirectoryArtifactReader implements ArtifactReader {
}
abortIfNeeded(signal);
const source = Readable.from(
mapFileReadCancellation(owner.handle, signal),
mapStableFileRead(owner.handle, entry, signal),
{ objectMode: false },
);
source.once("close", () => {
@@ -256,7 +259,7 @@ export class DirectoryArtifactReader implements ArtifactReader {
const artifactEntry = (
path: string,
absolute: string,
metadata: Stats,
metadata: BigIntStats,
): ArtifactEntry | undefined => {
const kind = metadata.isSymbolicLink()
? "symlink"
@@ -269,7 +272,7 @@ const artifactEntry = (
return {
path,
kind,
declaredSize: kind === "file" ? metadata.size : null,
declaredSize: kind === "file" ? safeArtifactSize(metadata.size) : null,
compressedSize: null,
executable: hostExecutableMarking(absolute, metadata),
encrypted: false,
@@ -283,12 +286,24 @@ const artifactEntry = (
adapterKey: absolute,
...(kind === "file"
? {
sourceIdentity: { device: metadata.dev, inode: metadata.ino },
sourceIdentity: {
device: metadata.dev,
inode: metadata.ino,
size: metadata.size,
mode: metadata.mode,
mtimeNs: metadata.mtimeNs,
ctimeNs: metadata.ctimeNs,
},
}
: {}),
};
};
const safeArtifactSize = (size: bigint): number | null => {
if (size < 0n || size > BigInt(Number.MAX_SAFE_INTEGER)) return null;
return Number(size);
};
const directoryIoFailure = (
operation: string,
path: string,
@@ -313,16 +328,45 @@ const directoryIoFailure = (
const failureMessage = (cause: unknown): string =>
cause instanceof Error ? cause.message : String(cause);
const mapFileReadCancellation = async function* (
const mapStableFileRead = async function* (
handle: FileHandle,
entry: ArtifactEntry,
signal?: AbortSignal,
): AsyncGenerator<Buffer> {
let bytes = 0n;
try {
for await (const chunk of readFileHandleChunks(handle, { signal })) {
const expectedBytes = entry.sourceIdentity?.size;
if (expectedBytes === undefined)
throw new ArtifactReaderFailure(
"integrity",
`Directory entry has no captured file state: ${entry.path}`,
);
for await (const chunk of readFileHandleChunks(handle, {
// Include one extra byte so a growing file cannot make this read unbounded.
end: expectedBytes,
signal,
})) {
abortIfNeeded(signal);
bytes += BigInt(chunk.length);
yield chunk;
}
abortIfNeeded(signal);
const [opened, current] = await Promise.all([
handle.stat({ bigint: true }),
lstat(entry.adapterKey, { bigint: true }),
]);
abortIfNeeded(signal);
if (
!opened.isFile() ||
!current.isFile() ||
!sameArtifactSourceIdentity(entry.sourceIdentity, opened) ||
!sameArtifactSourceIdentity(entry.sourceIdentity, current) ||
bytes !== expectedBytes
)
throw new ArtifactReaderFailure(
"integrity",
`Directory entry changed during read: ${entry.path}`,
);
} catch (cause: unknown) {
if (signal?.aborted === true)
throw new ArtifactReaderFailure(
@@ -330,6 +374,18 @@ const mapFileReadCancellation = async function* (
"Directory artifact read cancelled",
{ cause },
);
if (
cause instanceof Error &&
"code" in cause &&
["ENOENT", "ENOTDIR", "ELOOP"].includes(String(cause.code))
)
throw new ArtifactReaderFailure(
"integrity",
`Directory entry changed during read: ${entry.path}`,
{ cause },
);
if (!(cause instanceof ArtifactReaderFailure))
throw directoryIoFailure("validate read entry", entry.adapterKey, cause);
throw cause;
}
};
+84 -31
View File
@@ -125,6 +125,7 @@ export class NativeDmgArtifactReader implements ArtifactReader {
#ownershipUncertainty: string | undefined;
#closed = false;
#closePromise: Promise<void> | undefined;
readonly #activeDirectoryOperations = new Set<Promise<unknown>>();
constructor(
private readonly path: string,
@@ -141,18 +142,30 @@ export class NativeDmgArtifactReader implements ArtifactReader {
async *entries(signal?: AbortSignal): AsyncIterable<ArtifactEntry> {
await this.#ensureAttached(signal);
if (this.#directory === undefined)
const directory = this.#directory;
if (directory === undefined)
throw new ArtifactReaderFailure("unavailable", "DMG is not attached");
const prefix = basename(this.path);
for await (const entry of this.#directory.entries(signal))
yield { ...entry, path: `${prefix}/${entry.path}` };
const entries = directory.entries(signal)[Symbol.asyncIterator]();
try {
for (;;) {
if (this.#closed) return;
const next = await this.#trackDirectoryOperation(() => entries.next());
if (this.#closed) return;
if (next.done) return;
yield { ...next.value, path: `${prefix}/${next.value.path}` };
}
} finally {
await entries.return?.();
}
}
async open(entry: ArtifactEntry, signal?: AbortSignal): Promise<Readable> {
await this.#ensureAttached(signal);
if (this.#directory === undefined)
const directory = this.#directory;
if (directory === undefined)
throw new ArtifactReaderFailure("unavailable", "DMG is not attached");
return this.#directory.open(entry, signal);
return this.#trackDirectoryOperation(() => directory.open(entry, signal));
}
provenance(): readonly ArtifactCommand[] {
@@ -172,32 +185,9 @@ export class NativeDmgArtifactReader implements ArtifactReader {
async #closeOwnedResources(): Promise<void> {
await this.#initialization?.catch(() => undefined);
let detachFailure = await this.#reconcileAttachState();
let remainingDevices: string[] = [];
for (const device of [...this.#devices].reverse()) {
const failure = await this.#detach(device);
if (failure !== undefined) {
remainingDevices.push(device);
detachFailure ??= failure;
}
}
// APFS container detach may briefly leave its backing image listed.
for (const delayMs of DETACH_SETTLE_DELAYS_MS) {
if (remainingDevices.length === 0) break;
await (this.host.delay ?? delay)(delayMs);
const stillAttached: string[] = [];
for (const device of remainingDevices) {
if (!(await this.#isAttached(device))) continue;
const failure = await this.#detach(device);
if (failure !== undefined) {
stillAttached.push(device);
detachFailure = failure;
}
}
remainingDevices = stillAttached;
}
if (remainingDevices.length === 0) detachFailure = undefined;
this.#devices = remainingDevices.reverse();
await Promise.allSettled(this.#activeDirectoryOperations);
await this.#closeMountedDirectory();
let detachFailure = await this.#detachOwnedDevices();
if (
this.#devices.length === 0 &&
this.#mountRoot !== undefined &&
@@ -242,6 +232,55 @@ export class NativeDmgArtifactReader implements ArtifactReader {
);
}
async #detachOwnedDevices(): Promise<unknown> {
let detachFailure = await this.#reconcileAttachState();
let remainingDevices: string[] = [];
for (const device of [...this.#devices].reverse()) {
const failure = await this.#detach(device);
if (failure !== undefined) {
remainingDevices.push(device);
detachFailure ??= failure;
}
}
// APFS container detach may briefly leave its backing image listed.
for (const delayMs of DETACH_SETTLE_DELAYS_MS) {
if (remainingDevices.length === 0) break;
await (this.host.delay ?? delay)(delayMs);
const stillAttached: string[] = [];
for (const device of remainingDevices) {
if (!(await this.#isAttached(device))) continue;
const failure = await this.#detach(device);
if (failure !== undefined) {
stillAttached.push(device);
detachFailure = failure;
}
}
remainingDevices = stillAttached;
}
if (remainingDevices.length === 0) detachFailure = undefined;
this.#devices = remainingDevices.reverse();
return detachFailure;
}
async #closeMountedDirectory(): Promise<void> {
if (this.#directory !== undefined) {
try {
await this.#directory.close();
this.#directory = undefined;
} catch (cause: unknown) {
const mountRoot = this.#mountRoot ?? this.path;
throw ArtifactReaderFailure.withCleanup(cause, {
reason: `DMG mounted directory remains owned: ${failureMessage(cause)}`,
resources: [
`DMG mounted directory at ${mountRoot}`,
`DMG mount root ${mountRoot}`,
...this.#devices.map((device) => `DMG device ${device}`),
],
});
}
}
}
/** Detach one device; return the failure only while it remains attached. */
async #detach(device: string): Promise<unknown> {
try {
@@ -294,6 +333,20 @@ export class NativeDmgArtifactReader implements ArtifactReader {
throw new ArtifactReaderFailure("unavailable", "DMG reader is closed");
}
async #trackDirectoryOperation<Result>(
operation: () => Promise<Result>,
): Promise<Result> {
if (this.#closed)
throw new ArtifactReaderFailure("unavailable", "DMG reader is closed");
const running = Promise.resolve().then(operation);
this.#activeDirectoryOperations.add(running);
try {
return await running;
} finally {
this.#activeDirectoryOperations.delete(running);
}
}
async #attach(signal?: AbortSignal): Promise<void> {
await runChecked(this.host, ["verify", this.path], signal);
this.#provenance.push(command(["verify", this.path], ["read"]));
@@ -1,7 +1,10 @@
import { createHash } from "node:crypto";
import { lstat } from "node:fs/promises";
import { DirectoryArtifactReader } from "../DirectoryArtifactReader.js";
import { ArtifactReaderFailure } from "../ArtifactReader.js";
import {
ArtifactReaderFailure,
sameArtifactSourceIdentity,
} from "../ArtifactReader.js";
import { analyzeInterfaceBuilderBundle } from "./InterfaceBuilderAnalysis.js";
import {
appleAssetCatalogInputSchema,
@@ -136,22 +139,20 @@ const inspectAssetCatalogEntry = async (input: {
}) => {
const { entry, reader, signal } = input;
const sha256 = await digestBounded(reader, entry, signal);
const before = await lstat(entry.adapterKey);
const before = await lstat(entry.adapterKey, { bigint: true });
if (
entry.sourceIdentity === undefined ||
before.dev !== entry.sourceIdentity.device ||
before.ino !== entry.sourceIdentity.inode
!before.isFile() ||
!sameArtifactSourceIdentity(entry.sourceIdentity, before)
)
throw new ArtifactReaderFailure(
"integrity",
`Asset catalog changed before inspection: ${entry.path}`,
);
const output = await input.runAssetUtil(entry.adapterKey, signal);
const after = await lstat(entry.adapterKey);
const after = await lstat(entry.adapterKey, { bigint: true });
if (
before.dev !== after.dev ||
before.ino !== after.ino ||
before.size !== after.size ||
!after.isFile() ||
!sameArtifactSourceIdentity(entry.sourceIdentity, after) ||
(await digestBounded(reader, entry, signal)) !== sha256
)
throw new ArtifactReaderFailure(
+5 -3
View File
@@ -1,4 +1,4 @@
import type { Stats } from "node:fs";
import type { BigIntStats, Stats } from "node:fs";
import { extname } from "node:path";
/** Extensions Windows launches directly through CreateProcess or cmd.exe. */
@@ -13,10 +13,12 @@ const windowsLaunchableExtensions = new Set([".bat", ".cmd", ".com", ".exe"]);
*/
export const hostExecutableMarking = (
path: string,
metadata: Pick<Stats, "mode" | "isFile">,
metadata: Pick<Stats | BigIntStats, "mode" | "isFile">,
platform: NodeJS.Platform = process.platform,
): boolean =>
platform === "win32"
? metadata.isFile() &&
windowsLaunchableExtensions.has(extname(path).toLowerCase())
: (metadata.mode & 0o111) !== 0;
: typeof metadata.mode === "bigint"
? (metadata.mode & 0o111n) !== 0n
: (metadata.mode & 0o111) !== 0;
@@ -18,6 +18,7 @@ import { streamChunkToBuffer } from "../StreamBytes.js";
import { hashReadable } from "../ArtifactHash.js";
import type { ArtifactOccurrence } from "../../domain/artifactGraph.js";
import type { ArtifactInventorySnapshot } from "../../domain/artifactInventorySnapshot.js";
import { AnalysisInputError } from "../../domain/analysisErrorCore.js";
import type {
JavaScriptArtifactFileKind,
@@ -47,6 +48,7 @@ interface ReadContext {
readonly files: JavaScriptArtifactFile[];
readonly containers: JavaScriptArtifactContainer[];
readonly signal: AbortSignal | undefined;
readonly selectedPaths: ReadonlyMap<string, number> | undefined;
textBytes: number;
invalidUtf8: number;
}
@@ -57,13 +59,49 @@ interface ReadTextInput {
readonly expected: ExpectedFile;
}
/** Read all relevant textual entries through an already-inventoried reader. */
/** Verify inventoried entries; acquire text only inside the caller's scope. */
export const readJavaScriptArtifactFiles = async (
reader: ArtifactReader,
snapshot: ArtifactInventorySnapshot,
resources: ArtifactResourceScope,
signal?: AbortSignal,
sourcePaths?: readonly string[],
): Promise<JavaScriptArtifactFileSet> => {
const inventory = expectedInventory(snapshot);
const expected = inventory.files;
const selectedPaths =
sourcePaths === undefined
? undefined
: new Map([...new Set(sourcePaths)].map((path, index) => [path, index]));
const issues = (sourcePaths ?? []).flatMap((path, index) =>
expected.has(path) || inventory.unavailableFiles.has(path)
? []
: [
{
path: ["source_paths", index],
reason: "invalid_value" as const,
message:
"Select an exact relevant file path from the artifact inventory; no globs or path aliases.",
},
],
);
if (issues.length > 0)
throw new AnalysisInputError(
"analyze_javascript_application",
undefined,
issues,
);
const unavailablePath = sourcePaths?.find((path) =>
inventory.unavailableFiles.has(path),
);
if (unavailablePath !== undefined)
throw new ArtifactReaderFailure(
"unavailable",
`Selected inventoried source bytes are unavailable: ${unavailablePath}`,
{
partialObservation: { kind: "artifact-inventory", inventory: snapshot },
},
);
if (reader instanceof AsarArtifactReader) {
const root = snapshot.nodes.find(
({ artifact_id }) => artifact_id === snapshot.manifest.root_artifact_id,
@@ -79,8 +117,6 @@ export const readJavaScriptArtifactFiles = async (
signal,
);
}
const inventory = expectedInventory(snapshot);
const expected = inventory.files;
const context: ReadContext = {
resources,
expected,
@@ -89,13 +125,17 @@ export const readJavaScriptArtifactFiles = async (
files: [],
containers: [],
signal,
selectedPaths,
textBytes: 0,
invalidUtf8: 0,
};
await visitReader(reader, "", snapshot.manifest.root_sha256, context);
const files = context.files.sort((left, right) =>
compareUnicodeCodePoints(left.path, right.path),
);
const files = context.files.sort((left, right) => {
const leftPriority = selectedPaths?.get(left.path) ?? Infinity;
const rightPriority = selectedPaths?.get(right.path) ?? Infinity;
if (leftPriority !== rightPriority) return leftPriority - rightPriority;
return compareUnicodeCodePoints(left.path, right.path);
});
assertExpectedFilesWereVisited(expected, files);
return {
files,
@@ -104,6 +144,17 @@ export const readJavaScriptArtifactFiles = async (
),
text_bytes_read: context.textBytes,
invalid_utf8_files: context.invalidUtf8,
...(sourcePaths === undefined
? {}
: {
source_selection: {
requested_paths: [...sourcePaths],
unselected_files: files.filter(
({ text }) =>
!text.included && text.reason === "outside-analysis-scope",
).length,
},
}),
};
};
@@ -251,6 +302,19 @@ const readText = async (
context: ReadContext,
input: ReadTextInput,
): Promise<JavaScriptArtifactFile["text"]> => {
if (
context.selectedPaths !== undefined &&
!context.selectedPaths.has(input.expected.path)
) {
// Keep the inventory/acquisition identity check without allocating or
// decoding an unrelated source's text.
await verifyEntryBytes(
await input.reader.open(input.entry, context.signal),
input.expected,
context.signal,
);
return { included: false, reason: "outside-analysis-scope" };
}
if (input.expected.kind === "native-addon") {
await verifyEntryBytes(
await input.reader.open(input.entry, context.signal),
@@ -311,16 +375,23 @@ const expectedInventory = (
): {
readonly files: ReadonlyMap<string, ExpectedFile>;
readonly containers: ReadonlyMap<string, ExpectedContainer>;
readonly unavailableFiles: ReadonlySet<string>;
} => {
const nodes = new Map(snapshot.nodes.map((node) => [node.artifact_id, node]));
const files = new Map<string, ExpectedFile>();
const containers = new Map<string, ExpectedContainer>();
const unavailableFiles = new Set<string>();
for (const occurrence of snapshot.occurrences) {
if (occurrence.logical_path === ".") continue;
const isNestedAsar =
occurrence.entry_kind === "file" &&
occurrence.logical_path.toLowerCase().endsWith(".asar");
if (occurrence.artifact_id === null) {
if (
occurrence.entry_kind === "file" &&
relevantKind(occurrence.logical_path) !== undefined
)
unavailableFiles.add(occurrence.logical_path);
if (isNestedAsar)
containers.set(occurrence.logical_path, {
hash_status: occurrence.hash_status,
@@ -358,7 +429,7 @@ const expectedInventory = (
kind,
});
}
return { files, containers };
return { files, containers, unavailableFiles };
};
const readAll = async (
-62
View File
@@ -1,62 +0,0 @@
import {
lstat,
mkdtemp,
open,
rm,
symlink,
unlink,
writeFile,
} from "node:fs/promises";
import { join } from "node:path";
import { tmpdir } from "node:os";
import { expect, it } from "vitest";
import { readStableArtifact } from "./readStableArtifact.js";
const checkChange = async (
stage: "before-open" | "after-read" | "symlink-replacement",
): Promise<void> => {
const root = await mkdtemp(join(tmpdir(), "rea-stable-read-"));
const path = join(root, "selected");
await writeFile(path, "original evidence");
let inspected = 0;
try {
const reading = readStableArtifact(path, 1024, undefined, {
async lstat(selected) {
inspected++;
if (stage === "after-read" && inspected === 2) await unlink(selected);
const stat = await lstat(selected, { bigint: true });
if (inspected === 1 && stage !== "after-read") {
await unlink(selected);
if (stage === "symlink-replacement")
await symlink("missing", selected);
}
return stat;
},
open,
});
await expect(reading).rejects.toMatchObject({ reason: "integrity" });
} finally {
await rm(root, { recursive: true, force: true });
}
};
it.each(["before-open", "after-read"] as const)(
"retains an actual post-validation %s change as integrity failure",
checkChange,
);
it.skipIf(process.platform === "win32")(
"retains a post-validation symlink replacement as an integrity failure",
() => checkChange("symlink-replacement"),
);
it("keeps initially missing inputs distinguishable from later changes", async () => {
const root = await mkdtemp(join(tmpdir(), "rea-stable-missing-"));
try {
await expect(
readStableArtifact(join(root, "missing"), 1024),
).rejects.toMatchObject({ code: "ENOENT" });
} finally {
await rm(root, { recursive: true, force: true });
}
});
+29 -15
View File
@@ -226,10 +226,17 @@ export const handleRequestWillBeSent = (
preserveMalformedRedirectEvidence(state, requestId, previous)
)
return;
if (request === undefined) {
state.completeness.exclude("network_requests", "invalid_protocol_value");
state.network.delete(requestId);
state.networkRequestTimestamps.delete(requestId);
const method = cdpStringValue(request?.method);
const initiator = recordValue(params.initiator);
const initiatorType = cdpStringValue(initiator?.type);
if (
request === undefined ||
method === undefined ||
method.length === 0 ||
initiatorType === undefined ||
initiatorType.length === 0
) {
preserveMalformedRedirectEvidence(state, requestId, previous);
return;
}
const sanitized = allowedSanitizedUrl(request.url, state.allowedOrigins);
@@ -273,6 +280,12 @@ export const handleRequestWillBeSent = (
state.networkRequestTimestamps.delete(requestId);
return;
} else {
const encodedDataLength = numberValue(redirectResponse.encodedDataLength);
if (encodedDataLength === undefined || encodedDataLength < 0)
state.completeness.exclude(
"network_requests",
"invalid_protocol_value",
);
redirects.push({
url: previous.url,
response_url: responseUrl.url,
@@ -280,10 +293,10 @@ export const handleRequestWillBeSent = (
resource_type: previous.resource_type,
status: numberValue(redirectResponse.status) ?? null,
mime_type: cdpStringValue(redirectResponse.mimeType) ?? null,
encoded_data_length: (() => {
const length = numberValue(redirectResponse.encodedDataLength);
return length === undefined ? null : Math.max(0, length);
})(),
encoded_data_length:
encodedDataLength === undefined || encodedDataLength < 0
? null
: encodedDataLength,
request_timestamp:
state.networkRequestTimestamps.get(requestId) ?? null,
// CDP reports redirectResponse on the next requestWillBeSent event.
@@ -292,7 +305,6 @@ export const handleRequestWillBeSent = (
});
}
}
const initiator = recordValue(params.initiator);
const initiatorFrame = initiatorLocation(initiator);
const rawInitiatorUrl = cdpStringValue(initiatorFrame?.url);
const initiatorUrl = allowedSanitizedUrl(
@@ -308,14 +320,14 @@ export const handleRequestWillBeSent = (
request_id: requestId,
url: sanitized.url,
origin: sanitized.origin ?? "",
method: cdpStringValue(request.method) ?? "GET",
method,
resource_type: cdpStringValue(params.type) ?? null,
status: null,
mime_type: null,
encoded_data_length: null,
redirects,
initiator: {
type: cdpStringValue(initiator?.type) ?? "other",
type: initiatorType,
url: initiatorUrl?.url ?? null,
line: integerOrNull(initiatorFrame?.lineNumber),
column: integerOrNull(initiatorFrame?.columnNumber),
@@ -411,12 +423,14 @@ export const handleLoadingFinished = (
if (state.malformedRedirectRequestIds.has(requestId)) return;
const current = state.network.get(requestId);
if (current === undefined) return;
const encodedDataLength = numberValue(params.encodedDataLength);
if (encodedDataLength === undefined || encodedDataLength < 0) {
state.completeness.exclude("network_requests", "invalid_protocol_value");
return;
}
state.network.set(requestId, {
...current,
encoded_data_length: Math.max(
0,
numberValue(params.encodedDataLength) ?? 0,
),
encoded_data_length: encodedDataLength,
});
};
+1 -1
View File
@@ -20,10 +20,10 @@ import type {
import { webSourceMapsSchema } from "../domain/webBundleAnalysis.js";
import { createWebTextArtifact } from "../domain/webContentArtifact.js";
import { safeParseJson } from "../domain/safeJson.js";
import { sourceMapJsonText } from "../domain/sourceMapEnvelope.js";
import { jsonParts } from "../domain/jsonSerialization.js";
import {
inspectSourceMapValue,
sourceMapJsonText,
SourceMapFormatFailure,
type SourceMapLeaf,
} from "../javascript/sourceMaps/SourceMapFormat.js";
+129 -29
View File
@@ -7,48 +7,148 @@ const input = {
importMap: null,
specifiers: ["./dep.js"],
};
describe("optional native resolution acquisition", () => {
it("awaits owned cleanup when cancellation arrives during acquisition", async () => {
const controller = new AbortController();
let closed = false;
describe("native module resolver cleanup ownership", () => {
it.each([false, true])(
"awaits owned cleanup when acquisition is cancelled (cleanup fails: %s)",
async (cleanupFails) => {
const controller = new AbortController();
let closed = false;
const provider = new NativeModuleResolver(
{ REA_BROWSER_EXECUTABLE: "/browser" },
() => {
controller.abort();
return Promise.resolve({
version: () => "fixture",
newContext: () => {
throw new Error("must not acquire context");
},
close: () => {
closed = true;
if (cleanupFails)
return Promise.reject(new Error("fixture cleanup refusal"));
return Promise.resolve();
},
});
},
);
const result = await provider.resolve(input, {
signal: controller.signal,
});
expect(closed).toBe(true);
if (result.ok) throw new Error("expected cancelled result");
const projected = projectAnalysisError(result.error);
expect(projected.code).toBe(
cleanupFails ? "cleanup_incomplete" : "cancelled",
);
if (cleanupFails) {
expect(projected.details).toMatchObject({
diagnostics: { primary_error: { code: "cancelled" } },
});
}
},
);
it("returns cleanup_incomplete when owned cleanup fails after an engine error", async () => {
let launches = 0;
let closes = 0;
let cleanupAllowed = false;
const provider = new NativeModuleResolver(
{ REA_BROWSER_EXECUTABLE: "/browser" },
() => {
controller.abort();
launches += 1;
return Promise.resolve({
version: () => "fixture",
newContext: () => {
throw new Error("must not acquire context");
},
close: () => {
closed = true;
return Promise.resolve();
},
});
},
);
const result = await provider.resolve(input, { signal: controller.signal });
expect(closed).toBe(true);
if (result.ok) throw new Error("expected cancelled result");
expect(result.error._tag).toBe("AnalysisCancelledError");
});
it("returns cleanup_incomplete when owned cleanup fails after an engine error", async () => {
const provider = new NativeModuleResolver(
{ REA_BROWSER_EXECUTABLE: "/browser" },
() =>
Promise.resolve({
version: () => "fixture",
newContext: () =>
Promise.reject(new Error("fixture context failure")),
close: () =>
Promise.reject(new Error("fixture owned cleanup refusal")),
}),
close: async () => {
closes += 1;
if (!cleanupAllowed)
throw new Error("fixture owned cleanup refusal");
},
});
},
);
const result = await provider.resolve(input);
if (result.ok) throw new Error("expected cleanup failure");
expect(projectAnalysisError(result.error).code).toBe("cleanup_incomplete");
expect(result.error.cleanupIncomplete).toBe(true);
expect(projectAnalysisError(result.error).details).toMatchObject({
diagnostics: {
primary_error: {
details: {
diagnostics: { error_message: "fixture context failure" },
},
},
cleanup_error: {
details: { diagnostics: { reason: "fixture owned cleanup refusal" } },
},
},
});
const blocked = await provider.resolve(input);
expect(blocked.ok).toBe(false);
if (!blocked.ok) expect(blocked.error.cleanupIncomplete).toBe(true);
expect(launches).toBe(1);
await expect(provider.close()).rejects.toMatchObject({
cleanupIncomplete: true,
});
cleanupAllowed = true;
await provider.close();
await provider.close();
expect(closes).toBe(4);
expect(launches).toBe(1);
});
it("awaits acquisition and cleanup while refusing queued work during shutdown", async () => {
let acquired: () => void = () => {
throw new Error("acquisition gate missing");
};
const acquiring = new Promise<void>((resolve) => {
acquired = resolve;
});
let release: () => void = () => {
throw new Error("release gate missing");
};
const released = new Promise<void>((resolve) => {
release = resolve;
});
let launches = 0;
let closes = 0;
const provider = new NativeModuleResolver(
{ REA_BROWSER_EXECUTABLE: "/browser" },
async () => {
launches += 1;
acquired();
await released;
return {
version: () => "fixture",
newContext: () =>
Promise.reject(new Error("fixture context failure")),
close: async () => {
closes += 1;
},
};
},
);
const first = provider.resolve(input);
await acquiring;
const queued = provider.resolve(input);
const closing = provider.close();
let closed = false;
void closing.then(() => {
closed = true;
});
await Promise.resolve();
expect(closed).toBe(false);
release();
await first;
const blocked = await queued;
expect(blocked.ok).toBe(false);
await closing;
expect(closed).toBe(true);
expect([launches, closes]).toEqual([1, 1]);
expect((await provider.resolve(input)).ok).toBe(false);
});
});
describe("optional native resolution acquisition", () => {
it("does not load the browser engine when the selected operation lacks its configuration", async () => {
const provider = new NativeModuleResolver({}, () => {
throw new Error("must not launch");
+242 -147
View File
@@ -7,6 +7,7 @@ import type {
WebModuleResolutionBatch,
} from "../../application/WebModulePorts.js";
import type { AnalysisError } from "../../domain/analysisErrorBase.js";
import { analysisErrorWithCleanupFailure } from "../../domain/analysisErrorCleanup.js";
import {
AnalysisCancelledError,
AnalysisCapabilityUnavailableError,
@@ -16,6 +17,7 @@ import {
import { BrowserObservationError } from "../../domain/browserObservationError.js";
import { jsonObjectSchema } from "../../domain/jsonValue.js";
import { ProviderAdapterError } from "../../domain/providerAdapterError.js";
import { ProviderCleanupError } from "../../domain/providerCleanupError.js";
import { err, ok, type Result } from "../../domain/result.js";
import { webModuleResolutionSchema } from "../../domain/webModuleTrace.js";
import { ProviderStartupDeadline } from "../../process/ProviderDeadline.js";
@@ -32,26 +34,49 @@ const reportSchema = z.strictObject({
),
});
type ModuleResolverBrowser = Pick<Browser, "newContext" | "version" | "close">;
/** Production launch seam: only the adapter knows the browser API. */
export type ModuleResolverLauncher = (
options: LaunchOptions,
) => Promise<Pick<Browser, "newContext" | "version" | "close">>;
) => Promise<ModuleResolverBrowser>;
const launch: ModuleResolverLauncher = async (options) =>
(await import("playwright-core")).chromium.launch(options);
/** Use native import.meta.resolve in an owned context with locally fulfilled requests. */
export class NativeModuleResolver implements WebModuleResolutionPort {
#tail: Promise<void> = Promise.resolve();
#closing = false;
#closePromise: Promise<void> | undefined;
#pendingBrowser: ModuleResolverBrowser | undefined;
constructor(
readonly environment: Readonly<Record<string, string | undefined>>,
readonly launcher: ModuleResolverLauncher = launch,
) {}
async resolve(
resolve(
input: Parameters<WebModuleResolutionPort["resolve"]>[0],
options?: ExecutionOptions,
): ReturnType<WebModuleResolutionPort["resolve"]> {
if (this.#closing) return Promise.resolve(err(this.#closedError()));
const operation = this.#tail.then(() =>
this.#closing ? err(this.#closedError()) : this.#resolve(input, options),
);
this.#tail = operation.then(
() => undefined,
() => undefined,
);
return operation;
}
async #resolve(
input: Parameters<WebModuleResolutionPort["resolve"]>[0],
options?: ExecutionOptions,
): ReturnType<WebModuleResolutionPort["resolve"]> {
if (options?.signal?.aborted === true)
return err(new AnalysisCancelledError(OPERATION));
const pending = await this.#retryCleanup();
if (pending !== undefined) return err(pending);
const executable = this.environment.REA_BROWSER_EXECUTABLE;
if (executable === undefined || !isAbsolute(executable))
return err(
@@ -66,11 +91,16 @@ export class NativeModuleResolver implements WebModuleResolutionPort {
),
);
const deadline = new ProviderStartupDeadline(TIMEOUT_MS, options?.signal);
let browser: Pick<Browser, "newContext" | "version" | "close"> | undefined;
let browser: ModuleResolverBrowser | undefined;
let cleanup: Promise<void> | undefined;
const close = (): Promise<void> => {
if (cleanup === undefined && browser !== undefined)
cleanup = browser.close();
if (cleanup === undefined && browser !== undefined) {
const owned = browser;
cleanup = Promise.resolve().then(async () => {
await owned.close();
this.#pendingBrowser = undefined;
});
}
return cleanup ?? Promise.resolve();
};
const onAbort = (): void => {
@@ -82,6 +112,7 @@ export class NativeModuleResolver implements WebModuleResolutionPort {
deadline.signal.addEventListener("abort", onAbort, { once: true });
let result: Result<WebModuleResolutionBatch, AnalysisError>;
try {
deadline.signal.throwIfAborted();
browser = await this.launcher({
executablePath: executable,
headless: true,
@@ -99,142 +130,9 @@ export class NativeModuleResolver implements WebModuleResolutionPort {
: []),
],
});
this.#pendingBrowser = browser;
deadline.signal.throwIfAborted();
const context = await browser.newContext({
serviceWorkers: "block",
acceptDownloads: false,
});
const page = await context.newPage();
const diagnostics: string[] = [];
let diagnosticBytes = 0;
page.on("console", (message) => {
const text = message.text();
diagnosticBytes += Buffer.byteLength(text);
if (diagnosticBytes <= 1024 * 1024)
diagnostics.push(`${message.type()}: ${text}`);
});
const requests: {
url: string;
resource_type: string;
effect: "fulfilled" | "blocked";
}[] = [];
const importerUrl = new URL(input.importerUrl).href;
const documentUrl = new URL("/.rea-module-resolver/document", importerUrl)
.href;
const transportUrl = new URL(importerUrl);
transportUrl.hash = "";
let stubSent = false;
await context.route("**/*", async (route) => {
const request = route.request();
if (
request.resourceType() === "document" &&
request.url() === documentUrl.split("#")[0]
) {
requests.push({
url: request.url(),
resource_type: "document",
effect: "fulfilled",
});
await route.fulfill({
status: 200,
contentType: "text/html",
body: `<!doctype html>${input.importMap === null ? "" : `<base href="${htmlAttribute(new URL(input.importMap.baseUrl).href)}"><script type="importmap">${scriptJson(input.importMap.value)}</script>`}`,
});
} else if (
!stubSent &&
request.resourceType() === "script" &&
request.url() === transportUrl.href
) {
stubSent = true;
requests.push({
url: request.url(),
resource_type: "script",
effect: "fulfilled",
});
await route.fulfill({
status: 200,
contentType: "text/javascript",
headers: { "Access-Control-Allow-Origin": "*" },
body: `if (typeof import.meta.resolve !== "function") {globalThis.__reaModuleUnsupported = true;} else {globalThis.__reaModuleResult = {importer_url: import.meta.url, entries: ${scriptJson(input.specifiers)}.map(specifier => {try {return {specifier, resolution: {state: "resolved", url: import.meta.resolve(specifier)}}} catch (error) {return {specifier, resolution: {state: "rejected", message: String(error)}}}})};}`,
});
} else {
requests.push({
url: request.url(),
resource_type: request.resourceType(),
effect: "blocked",
});
await route.abort();
}
});
await page.goto(documentUrl, {
timeout: Math.max(1, deadline.remainingMs()),
});
await page.evaluate(
`{const script = document.createElement("script"); script.type = "module"; script.src = ${scriptJson(importerUrl)}; script.onerror = () => {globalThis.__reaModuleFailure = "Trusted resolver stub could not be loaded in the selected context.";}; document.head.append(script);}`,
);
await page.waitForFunction(
() =>
Reflect.has(globalThis, "__reaModuleResult") ||
Reflect.has(globalThis, "__reaModuleUnsupported") ||
Reflect.has(globalThis, "__reaModuleFailure"),
undefined,
{ timeout: Math.max(1, deadline.remainingMs()) },
);
const unsupported: unknown = await page.evaluate(
() => Reflect.get(globalThis, "__reaModuleUnsupported") as unknown,
);
if (unsupported === true)
throw new AnalysisCapabilityUnavailableError(
"native-module-resolver",
OPERATION,
"native_resolution_unsupported",
{
userMessage: `Configured Chromium ${browser.version()} does not supply import.meta.resolve; use a browser supporting native module resolution.`,
},
);
const raw: unknown = await page.evaluate(
() => Reflect.get(globalThis, "__reaModuleResult") as unknown,
);
const failure: unknown = await page.evaluate(
() => Reflect.get(globalThis, "__reaModuleFailure") as unknown,
);
if (typeof failure === "string")
throw new AnalysisOutputError(
OPERATION,
`${failure} Native diagnostics: ${diagnostics.join("\n")}`,
);
const parsed = reportSchema.safeParse(raw);
if (
!parsed.success ||
parsed.data.importer_url !== importerUrl ||
parsed.data.entries.length !== input.specifiers.length ||
parsed.data.entries.some(
(entry, index) => entry.specifier !== input.specifiers[index],
)
)
throw new AnalysisOutputError(
OPERATION,
"Native resolver report has invalid shape or does not match the selected importer/specifiers.",
);
if (diagnosticBytes > 1024 * 1024)
throw new BrowserObservationError(OPERATION, "payload_limit");
deadline.signal.throwIfAborted();
const engine = {
id: "chromium-native-module-resolver",
version: browser.version(),
};
result = ok({
engine,
resolutions: parsed.data.entries.map((entry) => entry.resolution),
diagnostics,
rawResult: jsonObjectSchema.parse({
native_report: parsed.data,
requests,
diagnostics,
engine,
executable_path: executable,
}),
});
result = ok(await resolveInBrowser(browser, input, deadline, executable));
} catch (cause: unknown) {
result = err(
deadline.interruption === "cancelled"
@@ -257,20 +155,217 @@ export class NativeModuleResolver implements WebModuleResolutionPort {
} finally {
deadline.signal.removeEventListener("abort", onAbort);
deadline.dispose();
try {
await close();
} catch (cause: unknown) {
result = err(
new BrowserObservationError(OPERATION, "cleanup_failed", { cause }),
);
}
}
try {
await close();
} catch (cause: unknown) {
const failure = cleanupFailure(
cause,
result.ok ? result.value.rawResult : undefined,
);
return err(
result.ok
? failure
: analysisErrorWithCleanupFailure(result.error, failure, OPERATION),
);
}
if (result.ok && options?.signal?.aborted)
return err(new AnalysisCancelledError(OPERATION));
return result;
}
async close(): Promise<void> {
this.#closing = true;
this.#closePromise ??= this.#tail
.then(async () => {
const failure = await this.#retryCleanup();
if (failure !== undefined) throw failure;
})
.catch((cause: unknown) => {
this.#closePromise = undefined;
throw cause;
});
return this.#closePromise;
}
#closedError(): AnalysisError {
return new BrowserObservationError(OPERATION, "cleanup_failed", {
detail: "Native module resolver is closing.",
});
}
async #retryCleanup(): Promise<AnalysisError | undefined> {
const browser = this.#pendingBrowser;
if (browser === undefined) return undefined;
try {
await browser.close();
this.#pendingBrowser = undefined;
return undefined;
} catch (cause: unknown) {
return cleanupFailure(cause);
}
}
}
const resolveInBrowser = async (
browser: ModuleResolverBrowser,
input: Parameters<WebModuleResolutionPort["resolve"]>[0],
deadline: ProviderStartupDeadline,
executable: string,
): Promise<WebModuleResolutionBatch> => {
const context = await browser.newContext({
serviceWorkers: "block",
acceptDownloads: false,
});
const page = await context.newPage();
const diagnostics: string[] = [];
let diagnosticBytes = 0;
page.on("console", (message) => {
const text = message.text();
diagnosticBytes += Buffer.byteLength(text);
if (diagnosticBytes <= 1024 * 1024)
diagnostics.push(`${message.type()}: ${text}`);
});
const requests: {
url: string;
resource_type: string;
effect: "fulfilled" | "blocked";
}[] = [];
const importerUrl = new URL(input.importerUrl).href;
const documentUrl = new URL("/.rea-module-resolver/document", importerUrl)
.href;
const transportUrl = new URL(importerUrl);
transportUrl.hash = "";
let stubSent = false;
await context.route("**/*", async (route) => {
const request = route.request();
if (
request.resourceType() === "document" &&
request.url() === documentUrl.split("#")[0]
) {
requests.push({
url: request.url(),
resource_type: "document",
effect: "fulfilled",
});
await route.fulfill({
status: 200,
contentType: "text/html",
body: `<!doctype html>${input.importMap === null ? "" : `<base href="${htmlAttribute(new URL(input.importMap.baseUrl).href)}"><script type="importmap">${scriptJson(input.importMap.value)}</script>`}`,
});
} else if (
!stubSent &&
request.resourceType() === "script" &&
request.url() === transportUrl.href
) {
stubSent = true;
requests.push({
url: request.url(),
resource_type: "script",
effect: "fulfilled",
});
await route.fulfill({
status: 200,
contentType: "text/javascript",
headers: { "Access-Control-Allow-Origin": "*" },
body: `if (typeof import.meta.resolve !== "function") {globalThis.__reaModuleUnsupported = true;} else {globalThis.__reaModuleResult = {importer_url: import.meta.url, entries: ${scriptJson(input.specifiers)}.map(specifier => {try {return {specifier, resolution: {state: "resolved", url: import.meta.resolve(specifier)}}} catch (error) {return {specifier, resolution: {state: "rejected", message: String(error)}}}})};}`,
});
} else {
requests.push({
url: request.url(),
resource_type: request.resourceType(),
effect: "blocked",
});
await route.abort();
}
});
await page.goto(documentUrl, {
timeout: Math.max(1, deadline.remainingMs()),
});
await page.evaluate(
`{const script = document.createElement("script"); script.type = "module"; script.src = ${scriptJson(importerUrl)}; script.onerror = () => {globalThis.__reaModuleFailure = "Trusted resolver stub could not be loaded in the selected context.";}; document.head.append(script);}`,
);
await page.waitForFunction(
() =>
Reflect.has(globalThis, "__reaModuleResult") ||
Reflect.has(globalThis, "__reaModuleUnsupported") ||
Reflect.has(globalThis, "__reaModuleFailure"),
undefined,
{ timeout: Math.max(1, deadline.remainingMs()) },
);
const unsupported: unknown = await page.evaluate(
() => Reflect.get(globalThis, "__reaModuleUnsupported") as unknown,
);
if (unsupported === true)
throw new AnalysisCapabilityUnavailableError(
"native-module-resolver",
OPERATION,
"native_resolution_unsupported",
{
userMessage: `Configured Chromium ${browser.version()} does not supply import.meta.resolve; use a browser supporting native module resolution.`,
},
);
const raw: unknown = await page.evaluate(
() => Reflect.get(globalThis, "__reaModuleResult") as unknown,
);
const failure: unknown = await page.evaluate(
() => Reflect.get(globalThis, "__reaModuleFailure") as unknown,
);
if (typeof failure === "string")
throw new AnalysisOutputError(
OPERATION,
`${failure} Native diagnostics: ${diagnostics.join("\n")}`,
);
const parsed = reportSchema.safeParse(raw);
if (
!parsed.success ||
parsed.data.importer_url !== importerUrl ||
parsed.data.entries.length !== input.specifiers.length ||
parsed.data.entries.some(
(entry, index) => entry.specifier !== input.specifiers[index],
)
)
throw new AnalysisOutputError(
OPERATION,
"Native resolver report has invalid shape or does not match the selected importer/specifiers.",
);
if (diagnosticBytes > 1024 * 1024)
throw new BrowserObservationError(OPERATION, "payload_limit");
deadline.signal.throwIfAborted();
const engine = {
id: "chromium-native-module-resolver",
version: browser.version(),
};
return {
engine,
resolutions: parsed.data.entries.map((entry) => entry.resolution),
diagnostics,
rawResult: jsonObjectSchema.parse({
native_report: parsed.data,
requests,
diagnostics,
engine,
executable_path: executable,
}),
};
};
const cleanupFailure = (
cause: unknown,
observation?: WebModuleResolutionBatch["rawResult"],
): ProviderCleanupError =>
new ProviderCleanupError(
"native-module-resolver",
["chromium-browser"],
{
reason: cause instanceof Error ? cause.message : String(cause),
...(observation === undefined
? {}
: { partial_observation: observation }),
},
{ cause, operation: OPERATION },
);
const scriptJson = (value: unknown): string =>
JSON.stringify(value).replaceAll("<", "\\u003c");
+22 -4
View File
@@ -15,6 +15,7 @@ import type { CliInstance } from "./types.js";
import { runCliJavaScriptApplicationAnalysis } from "./javascriptApplicationAnalysis.js";
import { withCommandCancellation } from "./commandCancellation.js";
import { javascriptApplicationOptions } from "../cliObservationOptions.js";
import { AnalysisInputError } from "../domain/analysisErrorCore.js";
/** Register provider-neutral binary overview and procedure CLI commands. */
export const registerCoreBinaryCommands = (
@@ -51,6 +52,7 @@ const registerOverviewCommands = (
}),
options: overviewOptions.extend({
integrityPolicy: javascriptApplicationOptions.shape.integrityPolicy,
sourcePath: javascriptApplicationOptions.shape.sourcePath,
}),
run: async ({ args, options }) => {
// Route JavaScript targets exactly like analyze-javascript-application,
@@ -63,19 +65,35 @@ const registerOverviewCommands = (
{
input_path: resolve(args.path),
integrity_policy: options.integrityPolicy,
...(options.sourcePath === undefined
? {}
: { source_paths: options.sourcePath }),
},
signal,
),
),
);
return logCliCommand(logger, "analyze", () =>
runDirectAnalysis(
return logCliCommand(logger, "analyze", () => {
if (options.sourcePath !== undefined)
throw new AnalysisInputError(
"analyze_javascript_application",
undefined,
[
{
path: ["source_paths"],
reason: "invalid_value",
message:
"Source selection requires the static JavaScript ASAR/directory analysis lane.",
},
],
);
return runDirectAnalysis(
args.path,
"binary_overview",
{},
directAnalysisOptionsFromCli(logger, options),
),
);
);
});
},
});
cli.command(CLI_COMMANDS.inspect, {
+3
View File
@@ -200,6 +200,9 @@ const registerJavaScriptApplicationCommand = (
input_path: args.path,
format: options.artifactFormat,
integrity_policy: options.integrityPolicy,
...(options.sourcePath === undefined
? {}
: { source_paths: options.sourcePath }),
...(options.maxHeapMb === undefined
? {}
: { max_heap_mb: options.maxHeapMb }),
+32 -22
View File
@@ -42,29 +42,39 @@ export const registerWebModuleCommands = (
.describe("HTTP(S) import-map base URL"),
}),
run: ({ args, options }) =>
withCommandCancellation((signal) =>
logCliCommand(logger, CLI_COMMANDS.traceWebModuleImports, async () => {
const result = await service.trace(
{
manifest_path: args.manifestPath,
script_index: args.scriptIndex,
...(options.importerUrl === undefined
? {}
: { importer_url: options.importerUrl }),
...(options.importMapPath === undefined &&
options.importMapBaseUrl === undefined
? {}
: {
import_map: {
path: options.importMapPath,
base_url: options.importMapBaseUrl,
},
}),
withCommandCancellation(async (signal) => {
try {
return await logCliCommand(
logger,
CLI_COMMANDS.traceWebModuleImports,
async () => {
const result = await service.trace(
{
manifest_path: args.manifestPath,
script_index: args.scriptIndex,
...(options.importerUrl === undefined
? {}
: { importer_url: options.importerUrl }),
...(options.importMapPath === undefined &&
options.importMapBaseUrl === undefined
? {}
: {
import_map: {
path: options.importMapPath,
base_url: options.importMapBaseUrl,
},
}),
},
{ signal },
);
return result.ok
? result.value
: projectAnalysisError(result.error);
},
{ signal },
);
return result.ok ? result.value : projectAnalysisError(result.error);
}),
),
} finally {
await service.close();
}
}),
});
};
+6
View File
@@ -1,5 +1,6 @@
import { z } from "incur";
import { javaScriptAnalysisResourceControlsSchema } from "./domain/javascript/javascriptAnalysisResourceControls.js";
import { javaScriptSourcePathsSchema } from "./domain/javascript/javascriptSourceSelection.js";
/** Accept an agent-selected browser observation duration without an artificial ceiling. */
export const observationDuration = (fallback: number, minimum = 0) =>
@@ -60,6 +61,11 @@ export const electronPageInspectionOptions = z.object({
});
export const javascriptApplicationOptions = z.object({
sourcePath: javaScriptSourcePathsSchema
.optional()
.describe(
"Exact inventoried file path relative to the application root; repeat to select several. No implicit dependencies.",
),
maxHeapMb: javaScriptAnalysisResourceControlsSchema.shape.max_heap_mb,
analysisTimeoutMs:
javaScriptAnalysisResourceControlsSchema.shape.analysis_timeout_ms,
+2
View File
@@ -8,6 +8,7 @@ import {
ghidraCompilerSpecIdSchema,
ghidraLanguageIdSchema,
} from "./ghidraLanguageOverride.js";
import { ghidraDisabledAnalyzersSchema } from "./ghidraDisabledAnalyzers.js";
import {
DEFAULT_GHIDRA_STARTUP_TIMEOUT_MS,
DEFAULT_HOPPER_STARTUP_TIMEOUT_MS,
@@ -48,6 +49,7 @@ const environmentSchema = z.object({
),
REA_GHIDRA_LANGUAGE_ID: ghidraLanguageIdSchema.optional(),
REA_GHIDRA_COMPILER_SPEC_ID: ghidraCompilerSpecIdSchema.optional(),
REA_GHIDRA_DISABLED_ANALYZERS: ghidraDisabledAnalyzersSchema.optional(),
REA_GHIDRA_SEED_FILE: z
.string()
.min(1)
+42
View File
@@ -0,0 +1,42 @@
import { z } from "zod";
export const MAX_GHIDRA_DISABLED_ANALYZER_NAMES = 32;
export const MAX_GHIDRA_DISABLED_ANALYZER_NAME_LENGTH = 120;
// Ghidra's headless parser consumes `-`-prefixed arguments after a script path
// as its own options, and REA joins the names with commas into one pre-script
// argument, so a name can neither start with `-` nor contain a comma. The
// remaining admitted characters keep the argument a single quotable token on
// every supported launcher, including the Windows batch invocation.
const ghidraAnalyzerNamePattern = /^[A-Za-z0-9][A-Za-z0-9 ._-]*$/u;
/**
* Parse `REA_GHIDRA_DISABLED_ANALYZERS` into the Ghidra auto-analyzer names to
* disable before default auto-analysis, such as
* `Objective-C Message Analyzer`. Empty and blank entries are dropped; a list
* without names selects no override.
*/
export const ghidraDisabledAnalyzersSchema = z
.string()
.transform((value) =>
value
.split(",")
.map((name) => name.trim())
.filter((name) => name.length > 0),
)
.refine(
(names) => names.length <= MAX_GHIDRA_DISABLED_ANALYZER_NAMES,
`must list at most ${MAX_GHIDRA_DISABLED_ANALYZER_NAMES} analyzer names`,
)
.refine(
(names) =>
names.every(
(name) => name.length <= MAX_GHIDRA_DISABLED_ANALYZER_NAME_LENGTH,
),
`must keep every analyzer name at most ${MAX_GHIDRA_DISABLED_ANALYZER_NAME_LENGTH} characters`,
)
.refine(
(names) => names.every((name) => ghidraAnalyzerNamePattern.test(name)),
"must be Ghidra analyzer names starting with a letter or digit and containing only letters, digits, spaces, periods, underscores, and hyphens",
)
.transform((names) => (names.length === 0 ? undefined : names));
+3
View File
@@ -88,6 +88,9 @@ export const parseConfig = (
...(env.REA_GHIDRA_SEED_FILE === undefined
? {}
: { ghidraSeedFile: env.REA_GHIDRA_SEED_FILE }),
...(env.REA_GHIDRA_DISABLED_ANALYZERS === undefined
? {}
: { ghidraDisabledAnalyzers: env.REA_GHIDRA_DISABLED_ANALYZERS }),
...(env.REA_GHIDRA_NATIVEAOT_JAR === undefined
? {}
: { ghidraNativeAotJar: env.REA_GHIDRA_NATIVEAOT_JAR }),
+1
View File
@@ -14,6 +14,7 @@ export interface AppConfig {
readonly ghidraJavaHome: string | undefined;
readonly ghidraStartupTimeoutMs: number;
readonly ghidraLanguageOverride?: GhidraLanguageOverride;
readonly ghidraDisabledAnalyzers?: readonly string[];
readonly ghidraSeedFile?: string;
readonly ghidraNativeAotJar?: string;
readonly ilspyCmdPath: string | undefined;
@@ -88,7 +88,7 @@ export const ELECTRON_TOOL_CONTRACTS = [
name: "analyze_javascript_application",
...toolContractMetadata("analyze_javascript_application"),
description:
"Reconstruct one local ASAR or extracted JavaScript application as an inline application graph without executing it. Returns recovered graph nodes, edges, semantic relations, integrity contradictions, limitations, and coverage. Select detail summary to retain the complete analysis Evidence in this session and receive only its summary view, then inspect selected modules or pages with inspect_analysis_view. Integrity mismatches fail by default; record-and-continue retains observed bytes as untrusted and marks graph coverage partial.",
"Reconstruct one local ASAR or extracted JavaScript application as an inline application graph without executing it. Returns recovered graph nodes, edges, semantic relations, integrity contradictions, limitations, and coverage. Select source_paths to acquire and analyze only exact inventoried file paths; the complete artifact inventory and identity remain intact, and unselected sources remain explicit unknowns. Dependencies are not implicitly analyzed. Select detail summary to retain the complete analysis Evidence in this session and receive only its summary view, then inspect selected modules or pages with inspect_analysis_view. Summary detail reduces delivery, whereas source_paths bounds source analysis. Integrity mismatches fail by default; record-and-continue retains observed bytes as untrusted and marks graph coverage partial.",
kind: "electron-provider",
inputSchema: analyzeJavaScriptApplicationRequestSchema,
outputSchema: applicationOutputSchema,
@@ -101,6 +101,14 @@ export const ELECTRON_TOOL_CONTRACTS = [
integrity_policy: "fail",
},
},
{
title: "Analyze selected entry sources before unrelated vendor files",
input: {
input_path: "/Applications/Example.app/Contents/Resources/app.asar",
source_paths: ["package.json", "main.js", "preload.js"],
detail: "summary",
},
},
{
title: "Retain the complete analysis and return its summary view",
input: {
@@ -505,6 +505,7 @@ export const graphNode = z.discriminatedUnion("status", [
address: z.string(),
status: z.literal("ok"),
calls: z.array(z.string()),
limitations: z.array(z.string()).optional(),
}),
z.object({
address: z.string(),
+5 -1
View File
@@ -27,6 +27,7 @@ import {
} from "./evidenceBundle.js";
import {
evidenceLocationSchema,
rebindEvidenceToTargetPath,
type Evidence,
type EvidenceLocation,
type EvidenceSubjectTarget,
@@ -411,7 +412,7 @@ export const snapshotEvidenceForQuery = (
(candidate) => candidate.query_id === queryId,
);
if (entry === undefined && workflowEntry === undefined) return undefined;
return snapshot.evidence_bundle.records.find(
const evidence = snapshot.evidence_bundle.records.find(
(record) =>
record.subject?.digest.sha256 === target.sha256 &&
record.operation === operation &&
@@ -425,6 +426,9 @@ export const snapshotEvidenceForQuery = (
(workflowEntry !== undefined &&
evidenceMatchesWorkflowEntry(record, workflowEntry, snapshot))),
);
return evidence === undefined
? undefined
: rebindEvidenceToTargetPath(evidence, target);
};
/** Create one exact replay record for a derived application workflow result. */
+2
View File
@@ -1,4 +1,5 @@
import { z } from "zod";
import { javaScriptSourceSelectionSchema } from "../javascript/javascriptSourceSelection.js";
import {
AnalysisInputError,
@@ -170,6 +171,7 @@ const javascriptSummarySchema = z.strictObject({
input_path: z.string().min(1),
format: z.enum(["asar", "directory"]),
root_artifact_sha256: digestSchema,
source_selection: javaScriptSourceSelectionSchema.exactOptional(),
statistics: jsonObjectSchema,
electron: jsonObjectSchema,
coverage: jsonObjectSchema,
@@ -255,6 +255,11 @@ export const projectJavaScriptApplicationView = (
input_path: analysis.input_path,
format: analysis.format,
root_artifact_sha256: analysis.root_artifact_sha256,
...(analysis.source_selection === undefined
? {}
: {
source_selection: analysis.source_selection,
}),
statistics: jsonObjectSchema.parse(analysis.statistics),
electron: jsonObjectSchema.parse(analysis.summary),
coverage: jsonObjectSchema.parse({
+71 -1
View File
@@ -14,6 +14,10 @@ const observe = (
address: string,
callees: readonly string[],
provider = "rea-workflow",
options: {
readonly evidenceLimitations?: readonly string[];
readonly dossierLimitations?: readonly string[];
} = {},
): Evidence => {
const value = functionDossierSchema.parse({
procedure: {
@@ -49,7 +53,7 @@ const observe = (
native_api: null,
native_value_flow: null,
basic_blocks: [],
limitations: [],
limitations: [...(options.dossierLimitations ?? [])],
});
return createEvidence(
{ path: "/tmp/a", sha256: "a".repeat(64), format: "mach-o" },
@@ -62,6 +66,9 @@ const observe = (
result: jsonValueSchema.parse(value),
confidence: "derived",
authority: "shipped-artifact",
...(options.evidenceLimitations === undefined
? {}
: { limitations: options.evidenceLimitations }),
},
);
};
@@ -126,6 +133,22 @@ describe("call path reconstruction", () => {
expect(unknown.limitations).toContain(
"No analyze_function Evidence covers reachable function 0x2000",
);
const limitedEmpty = run([
observe("0x1000", [], "rea-workflow", {
dossierLimitations: ["Call relationships are incomplete."],
}),
observe("0x4000", []),
]);
expect(limitedEmpty).toMatchObject({
status: "unknown",
search_scope: { exhaustive: false },
shortest_hops: null,
paths: [],
});
expect(limitedEmpty.limitations).toContain(
"Call relationships are incomplete.",
);
});
it("searches the complete supplied graph", () => {
@@ -182,6 +205,53 @@ describe("call path topology and capacity", () => {
);
});
it("keeps reached function limitations and marks expanded coverage incomplete", () => {
const evidenceLimitation = "Call graph output was truncated.";
const dossierLimitation = "One call target could not be resolved.";
const result = run([
observe("0x1000", ["0x4000"], "rea-workflow", {
evidenceLimitations: [evidenceLimitation],
dossierLimitations: [dossierLimitation],
}),
observe("0x4000", []),
]);
expect(result).toMatchObject({
status: "found",
shortest_hops: 1,
search_scope: { exhaustive: false },
paths: [
{
nodes: [{ address: "0x1000" }, { address: "0x4000" }],
},
],
});
expect(result.limitations).toEqual(
expect.arrayContaining([evidenceLimitation, dossierLimitation]),
);
});
it("retains goal dossier limitations without expanding the goal", () => {
const goalLimitation = "The goal dossier has partial outgoing references.";
const result = run([
observe("0x1000", ["0x4000"]),
observe("0x4000", ["0x5000"], "rea-workflow", {
dossierLimitations: [goalLimitation],
}),
]);
expect(result).toMatchObject({
status: "found",
shortest_hops: 1,
search_scope: { exhaustive: true },
});
expect(result.limitations).toContain(goalLimitation);
expect(result.explored.nodes).toBe(2);
expect(result.limitations).not.toContain(
"No analyze_function Evidence covers reachable function 0x5000",
);
});
it("returns a zero-hop path when start and goal are the same function", () => {
const result = run([observe("0x1000", ["0x2000"])], {
goal: { address: "0x1000" },
+13 -2
View File
@@ -229,16 +229,27 @@ const inspectSearch = ({
const reached = new Map<string, number>([[start, 0]]);
const queue = [start];
const limitations: string[] = [];
let incompleteExpandedCoverage = false;
for (let index = 0; index < queue.length; index += 1) {
const node = queue[index];
if (node === undefined) continue;
const depth = reached.get(node) ?? 0;
if (node === goal) continue;
const snapshot = snapshots.get(node);
if (snapshot !== undefined) {
limitations.push(...snapshot.limitations);
// The requested goal is a terminal for this search. Its outgoing edges
// cannot add a path to the goal, so its dossier limits are contextual
// but do not make the searched start-to-goal closure incomplete.
if (node === goal) continue;
if (snapshot.limitations.length > 0) incompleteExpandedCoverage = true;
} else if (node === goal) {
continue;
}
if (snapshot === undefined) {
limitations.push(
`No analyze_function Evidence covers reachable function ${node}`,
);
incompleteExpandedCoverage = true;
continue;
}
const neighbors = graph.outNeighbors(node).sort(compareUnicodeCodePoints);
@@ -252,7 +263,7 @@ const inspectSearch = ({
graph,
snapshots,
reached,
exhaustive: limitations.length === 0,
exhaustive: !incompleteExpandedCoverage,
limitations,
};
};
+51 -1
View File
@@ -1,3 +1,4 @@
import { basename } from "node:path";
import { z } from "zod";
import { digestCanonicalValue } from "./canonicalDigest.js";
@@ -182,6 +183,18 @@ export interface EvidenceSubjectTarget {
readonly format: z.infer<typeof subjectSchema>["format"];
readonly architecture?: BinaryArchitecture;
}
/** Rebind a matching immutable artifact subject to the caller's selected path. */
export const rebindEvidenceSubjectTargetPath = (
subject: EvidenceSubjectTarget,
target: EvidenceSubjectTarget,
): EvidenceSubjectTarget =>
subject.sha256 === target.sha256 &&
subject.format === target.format &&
(subject.architecture ?? null) === (target.architecture ?? null)
? { ...subject, path: target.path }
: subject;
type EvidenceAuthority = z.infer<typeof evidenceAuthoritySchema>;
type ExecutionEnvironment = z.infer<typeof executionEnvironmentSchema>;
@@ -252,6 +265,43 @@ export const parseEvidence = (input: unknown): Evidence => {
export const immutableEvidence = (input: unknown): Evidence =>
rememberImmutableEvidence(parseEvidence(input));
/** Rebind display path metadata when an authenticated record names this target. */
export const rebindEvidenceToTargetPath = (
evidence: Evidence,
target: EvidenceSubjectTarget,
): Evidence => {
const subject = evidence.subject;
if (subject === null) return evidence;
const rebound = rebindEvidenceSubjectTargetPath(
{
path: subject.local_path,
sha256: subject.digest.sha256,
format: subject.format,
...(subject.architecture === null
? {}
: { architecture: subject.architecture }),
},
target,
);
if (rebound.path === subject.local_path) return evidence;
const reboundSubject = subjectSchema.parse({
...subject,
name: basename(target.path) || "artifact",
local_path: target.path,
});
const reboundEvidence = { ...evidence, subject: reboundSubject };
if (!isImmutableEvidence(evidence)) return immutableEvidence(reboundEvidence);
// The original authenticated Evidence owns immutable payloads. Only the
// small display subject and envelope are new, so preserve those payloads
// without reparsing a potentially large analysis result.
Object.freeze(reboundSubject.digest);
Object.freeze(reboundSubject);
Object.freeze(reboundEvidence);
immutableEvidenceSnapshots.set(reboundEvidence, reboundEvidence);
return reboundEvidence;
};
/** Recognize authenticated Evidence whose complete reachable JSON data is immutable. */
export const isImmutableEvidence = (evidence: Evidence): boolean =>
immutableEvidenceSnapshots.has(evidence);
@@ -316,7 +366,7 @@ const normalizeEvidenceObservation = (
target === undefined
? null
: {
name: target.path.split("/").at(-1) || "artifact",
name: basename(target.path) || "artifact",
digest: { sha256: target.sha256 },
format: target.format,
architecture: target.architecture ?? null,
@@ -16,6 +16,10 @@ import { integrityContradictionSchema } from "../artifactGraph.js";
import { artifactIntegrityPolicySchema } from "../artifactIntegrityPolicy.js";
import { localPathStringSchema } from "../localPath.js";
import { javaScriptAnalysisResourceControlsSchema } from "./javascriptAnalysisResourceControls.js";
import {
javaScriptSourcePathsSchema,
javaScriptSourceSelectionSchema,
} from "./javascriptSourceSelection.js";
const countSchema = z.number().int().min(0);
@@ -24,6 +28,7 @@ export const analyzeJavaScriptApplicationInputSchema = z.strictObject({
input_path: localPathStringSchema,
format: z.enum(["auto", "asar", "directory"]).default("auto"),
integrity_policy: artifactIntegrityPolicySchema,
source_paths: javaScriptSourcePathsSchema.optional(),
...javaScriptAnalysisResourceControlsSchema.shape,
});
@@ -71,6 +76,7 @@ const applicationAnalysisResultShape = z.strictObject({
inventory_manifest_id: prefixedDigestSchema("agm"),
inventory_graph_sha256: digestSchema,
integrity_contradictions: z.array(integrityContradictionSchema),
source_selection: javaScriptSourceSelectionSchema.exactOptional(),
graph: javascriptApplicationGraphSchema,
summary: electronBoundarySummarySchema,
statistics: reconstructionStatisticsSchema,
@@ -1,3 +1,5 @@
import type { JavaScriptSourceSelection } from "./javascriptSourceSelection.js";
/** Relevant file categories projected from the complete artifact inventory. */
export type JavaScriptArtifactFileKind =
| "package-json"
@@ -20,7 +22,10 @@ export interface JavaScriptArtifactFile {
| { readonly included: true; readonly value: string }
| {
readonly included: false;
readonly reason: "not-applicable" | "invalid-utf8";
readonly reason:
| "not-applicable"
| "invalid-utf8"
| "outside-analysis-scope";
};
}
@@ -38,4 +43,5 @@ export interface JavaScriptArtifactFileSet {
readonly containers: readonly JavaScriptArtifactContainer[];
readonly text_bytes_read: number;
readonly invalid_utf8_files: number;
readonly source_selection?: JavaScriptSourceSelection;
}
@@ -0,0 +1,20 @@
import { z } from "zod";
/** Exact inventoried paths selected for text acquisition and static analysis. */
export const javaScriptSourcePathsSchema = z
.array(z.string().min(1))
.min(1)
.describe(
"Exact relevant file paths from the artifact inventory, relative to the application root, analyzed in first-request order before unselected scopes; no globs or implicit dependencies. Omit to analyze all sources.",
);
/** Caller scope retained independently of analysis success or completeness. */
export const javaScriptSourceSelectionSchema = z.strictObject({
requested_paths: javaScriptSourcePathsSchema,
unselected_files: z.number().int().min(0),
});
/** Explicit selected-source boundary for an otherwise complete inventory. */
export type JavaScriptSourceSelection = z.infer<
typeof javaScriptSourceSelectionSchema
>;
@@ -0,0 +1,21 @@
import { AnalysisCapabilityUnavailableError } from "./analysisErrorCore.js";
import type { JsonValue } from "./jsonValue.js";
/** Convert a provider's host configuration failure into actionable refusal evidence. */
export const providerConfigurationRefusal = (
providerId: string,
operation: string,
failure: Error & {
readonly diagnostics?: Readonly<Record<string, JsonValue>>;
},
): AnalysisCapabilityUnavailableError => {
const remediation = failure.diagnostics?.remediation;
const reason =
typeof remediation === "string"
? `${failure.message}. ${remediation}.`
: `${failure.message}.`;
return new AnalysisCapabilityUnavailableError(providerId, operation, reason, {
cause: failure,
userMessage: reason,
});
};
+8 -18
View File
@@ -1,6 +1,6 @@
import { describe, expect, it } from "vitest";
import { flattenSourceMapLeaves, isVersion3Map } from "./sourceMapEnvelope.js";
import { flattenSourceMapLeaves } from "./sourceMapEnvelope.js";
const leaf = {
version: 3,
@@ -10,14 +10,7 @@ const leaf = {
};
describe("source-map envelope walk", () => {
it("accepts version-3 maps and rejects other shapes", () => {
expect(isVersion3Map(leaf)).toBe(true);
expect(isVersion3Map({ version: 2, mappings: "AAAA" })).toBe(false);
expect(isVersion3Map(null)).toBe(false);
expect(isVersion3Map([{ version: 3 }])).toBe(false);
});
it("collects indexed section leaves without validating offsets by default", () => {
it("collects indexed section leaves without interpreting generated offsets", () => {
const root = {
version: 3,
sections: [
@@ -28,24 +21,21 @@ describe("source-map envelope walk", () => {
expect(flattenSourceMapLeaves(root)).toHaveLength(2);
});
it("rejects non-map sections and enforces offsets when requested", () => {
it("rejects sections without embedded maps", () => {
const missing = {
version: 3,
sections: [{ offset: { line: 0, column: 0 } }],
};
expect(flattenSourceMapLeaves(missing)).toBeUndefined();
const badOffset = {
version: 3,
sections: [{ offset: { line: -1, column: 0 }, map: leaf }],
};
expect(
flattenSourceMapLeaves(badOffset, { validateOffsets: true }),
).toBeUndefined();
expect(flattenSourceMapLeaves(badOffset)).toHaveLength(1);
});
it("rejects non-version-3 roots and nested maps", () => {
expect(flattenSourceMapLeaves({ version: 3 })).toHaveLength(1);
expect(flattenSourceMapLeaves(null)).toBeUndefined();
expect(flattenSourceMapLeaves([{ version: 3 }])).toBeUndefined();
expect(
flattenSourceMapLeaves({ version: 2, mappings: "AAAA" }),
).toBeUndefined();
expect(flattenSourceMapLeaves({ mappings: "AAAA" })).toBeUndefined();
expect(
flattenSourceMapLeaves({
+15 -26
View File
@@ -1,28 +1,26 @@
/**
* Structural walk over indexed source-map envelopes. Both the browser
* source-map fetcher and the artifact source-map reader accept version-3
* maps with recursively indexed `sections`; this module owns that walk so
* the two validators cannot drift. Leaf content checks (mappings, sources,
* names, contents) stay with the callers, which enforce different
* strictness.
*/
/** Artifact source declarations do not require generated-position decoding. */
type SourceMapRecord = Readonly<Record<string, unknown>>;
/**
* Return the JSON parsing view of retained source-map text. ECMA-426 permits
* HTTP responses to start with `)]}'`; the first line is transport framing.
* Keep the original text for artifact identity and Evidence provenance.
* https://tc39.es/ecma426/#sec-fetchsourcemap
*/
export const sourceMapJsonText = (text: string): string => {
const withoutBom = text.replace(/^\uFEFF/u, "");
if (!withoutBom.startsWith(")]}'")) return withoutBom;
const newline = withoutBom.search(/[\r\n]/u);
return newline === -1 ? "" : withoutBom.slice(newline);
};
const isRecord = (value: unknown): value is SourceMapRecord =>
typeof value === "object" && value !== null && !Array.isArray(value);
/** Whether a parsed value has the version-3 map shape. */
export const isVersion3Map = (value: unknown): value is SourceMapRecord =>
const isVersion3Map = (value: unknown): value is SourceMapRecord =>
isRecord(value) && value.version === 3;
const isValidOffset = (value: unknown): boolean =>
typeof value === "number" && Number.isSafeInteger(value) && value >= 0;
export interface FlattenSourceMapOptions {
/** Require each section offset to carry safe-integer line/column. */
readonly validateOffsets?: boolean;
}
/**
* Collect leaf maps from an indexed source-map envelope, or undefined when
* the envelope is not a version-3 map. A map carrying neither `mappings`
@@ -31,7 +29,6 @@ export interface FlattenSourceMapOptions {
*/
export const flattenSourceMapLeaves = (
root: unknown,
options?: FlattenSourceMapOptions,
): readonly SourceMapRecord[] | undefined => {
if (!isVersion3Map(root)) return undefined;
const leaves: SourceMapRecord[] = [];
@@ -42,14 +39,6 @@ export const flattenSourceMapLeaves = (
if (Array.isArray(map.sections)) {
for (const section of map.sections) {
if (!isRecord(section) || !isRecord(section.map)) return undefined;
if (options?.validateOffsets === true) {
if (!isRecord(section.offset)) return undefined;
if (
!isValidOffset(section.offset.line) ||
!isValidOffset(section.offset.column)
)
return undefined;
}
pending.push(section.map);
}
} else leaves.push(map);
@@ -259,6 +259,35 @@ describe("managed artifact coded indexes", () => {
layout.codedIndexSize("HasCustomAttribute"),
);
});
it("resolves local attribute constructors without rescanning TypeDef rows", () => {
// The owning type follows every empty type. A per-attribute TypeDef scan
// took about 38 s here; the owner index is built once per layout.
const bytes = buildManagedPeFixture({
additionalEmptyTypes: 16_000,
customAttributeRowCount: 16_000,
customAttributeTypeRaw: (1 << 3) | 2,
typeName: "LookupAttribute",
extendsRaw: 0,
methodName: ".ctor",
methodFlags: 0x1886,
methodSignature: Buffer.from([0x20, 0x01, 0x01, 0x0e]),
});
const start = performance.now();
const artifact = inspectManagedArtifactBytes(
bytes,
managedPeFixtureTarget(bytes),
);
expect(performance.now() - start).toBeLessThan(2000);
expect(artifact.coverage).toEqual({ state: "complete", issues: [] });
expect(artifact.attributes).toHaveLength(16_000);
expect(
artifact.attributes.every(
(attribute) => attribute.type_name === "Fixture.LookupAttribute",
),
).toBe(true);
}, 30_000);
});
describe("managed artifact reference inventory", () => {
@@ -1,96 +0,0 @@
import { fc, it } from "@fast-check/vitest";
import { describe, expect } from "vitest";
import {
createDeclaringTypeLookup,
declaringType,
type TypeRange,
} from "./ManagedMemberInspectorCore.js";
const range = (
name: string,
fieldStart: number,
fieldEnd: number,
methodStart = fieldStart,
methodEnd = fieldEnd,
): TypeRange => ({
token: name,
fullName: name,
fieldStart,
fieldEnd,
methodStart,
methodEnd,
});
describe("managed declaring type ownership", () => {
it("resolves field and method boundaries independently, including empty types and gaps", () => {
const ranges = [
range("first", 1, 3, 1, 1),
range("empty", 3, 3, 1, 1),
range("last", 5, 7, 1, 4),
];
const field = createDeclaringTypeLookup(ranges, "field");
const method = createDeclaringTypeLookup(ranges, "method");
expect([0, 1, 2, 3, 4, 5, 6, 7].map(field)).toEqual([
null,
{ token: "first", fullName: "first" },
{ token: "first", fullName: "first" },
null,
null,
{ token: "last", fullName: "last" },
{ token: "last", fullName: "last" },
null,
]);
expect([0, 1, 2, 3, 4].map(method)).toEqual([
null,
{ token: "last", fullName: "last" },
{ token: "last", fullName: "last" },
{ token: "last", fullName: "last" },
null,
]);
});
it("retains the first observed owner for overlapping, reversed, or out-of-order ranges", () => {
const ranges = [
range("first", 5, 9),
range("overlap", 1, 7),
range("reversed", 10, 2),
];
const lookup = createDeclaringTypeLookup(ranges, "method");
expect(lookup(6)).toEqual({ token: "first", fullName: "first" });
expect(lookup(2)).toEqual({ token: "overlap", fullName: "overlap" });
expect(lookup(10)).toBeNull();
expect(createDeclaringTypeLookup([], "method")(1)).toBeNull();
});
it.prop([
fc.array(
fc.record({
fieldStart: fc.integer({ min: 0, max: 30 }),
fieldEnd: fc.integer({ min: 0, max: 30 }),
methodStart: fc.integer({ min: 0, max: 30 }),
methodEnd: fc.integer({ min: 0, max: 30 }),
}),
{ maxLength: 20 },
),
])(
"agrees with observed first-match ownership for arbitrary metadata ranges",
(raw) => {
const ranges = raw.map((value, index) => ({
...value,
token: String(index),
fullName: `Type${String(index)}`,
}));
for (const table of ["field", "method"] as const) {
const lookup = createDeclaringTypeLookup(ranges, table);
for (const row of [
NaN,
-Infinity,
Infinity,
...Array.from({ length: 32 }, (_, i) => i),
])
expect(lookup(row)).toEqual(declaringType(ranges, table, row));
}
},
);
});
-50
View File
@@ -154,56 +154,6 @@ export const typeRanges = (
return ranges;
};
export const declaringType = (
ranges: readonly TypeRange[],
table: "field" | "method",
row: number,
): { readonly token: string; readonly fullName: string | null } | null => {
for (const range of ranges) {
const start = table === "field" ? range.fieldStart : range.methodStart;
const end = table === "field" ? range.fieldEnd : range.methodEnd;
if (row >= start && row < end)
return { token: range.token, fullName: range.fullName };
}
return null;
};
/** Resolve member ownership without rescanning ordered type ranges per row. */
export const createDeclaringTypeLookup = (
ranges: readonly TypeRange[],
table: "field" | "method",
): ((row: number) => ReturnType<typeof declaringType>) => {
const entries = ranges.map((range) => ({
range,
start: table === "field" ? range.fieldStart : range.methodStart,
end: table === "field" ? range.fieldEnd : range.methodEnd,
}));
const ordered = entries.every((entry, index) => {
const previous = entries[index - 1];
return (
entry.start <= entry.end &&
(previous === undefined || previous.end <= entry.start)
);
});
// Malformed metadata can overlap or reverse ranges. Retain the original
// first-match interpretation rather than guessing a different owner.
if (!ordered) return (row) => declaringType(ranges, table, row);
return (row) => {
let first = 0;
let end = entries.length;
while (first < end) {
const middle = first + Math.floor((end - first) / 2);
const entry = entries[middle];
if (entry === undefined) return null;
if (row >= entry.start && row < entry.end)
return { token: entry.range.token, fullName: entry.range.fullName };
if (row < entry.start) end = middle;
else first = middle + 1;
}
return null;
};
};
export const parseTypes = (
bytes: Buffer,
layout: ManagedMetadataLayout,
+150
View File
@@ -0,0 +1,150 @@
import { fc, it } from "@fast-check/vitest";
import { describe, expect, vi } from "vitest";
import { inspectManagedArtifactBytes } from "./ManagedArtifactInspector.js";
import { inspectManagedMembersBytes } from "./ManagedMemberInspector.js";
import { managedDecodeBudget } from "./ManagedDecodeBudget.js";
import { memberDeclaringTypeRowLookup } from "./ManagedMemberOwnership.js";
import { memberListIndexSize } from "./ManagedMemberPointers.js";
import { readManagedMetadataLayout } from "./ManagedMetadataLayout.js";
import {
buildManagedMemberPointerFixture,
buildManagedPeFixture,
managedPeFixtureTarget,
} from "./ManagedPe.fixture.js";
import { readManagedPeLayout } from "./ManagedPeReader.js";
import { ManagedReaderFailure } from "./ManagedReaderFailure.js";
const metadataLayout = (bytes: Buffer) => {
const pe = readManagedPeLayout(bytes);
if (pe.cli === null) throw new Error("Fixture omitted CLI metadata");
return readManagedMetadataLayout(
bytes,
pe.rvaToOffset(pe.cli.metadata.rva, pe.cli.metadata.size, "cli.metadata"),
pe.cli.metadata.size,
);
};
describe("managed declaring type ownership", () => {
it.prop([
fc.tuple(
fc.tuple(
...Array.from({ length: 3 }, () => fc.integer({ min: 0, max: 6 })),
),
fc.tuple(
...Array.from({ length: 3 }, () => fc.integer({ min: 0, max: 6 })),
),
),
])(
"retains first-match field and method owners for arbitrary metadata lists",
([fields, methods]) => {
const bytes = buildManagedMemberPointerFixture();
const layout = metadataLayout(bytes);
const types = layout.table(2);
if (types === undefined) throw new Error("Fixture omitted TypeDef");
for (let row = 0; row < 3; row++) {
const fieldOffset =
types.offset +
row * types.rowSize +
4 +
2 * layout.stringIndexSize +
layout.codedIndexSize("TypeDefOrRef");
bytes.writeUInt16LE(fields[row] ?? 0, fieldOffset);
bytes.writeUInt16LE(
methods[row] ?? 0,
fieldOffset + memberListIndexSize(layout, 4),
);
}
const target = managedPeFixtureTarget(bytes);
const members = inspectManagedMembersBytes(bytes, target);
const artifact = inspectManagedArtifactBytes(bytes, target);
for (const [table, starts] of [
["fields", fields],
["methods", methods],
] as const) {
// The fixture's pointers reorder physical rows 1,2 to list positions 2,1.
const expected = [2, 1].map((position) => {
const owner = starts.findIndex(
(start, index) =>
position >= start && position < (starts[index + 1] ?? 3),
);
return owner < 0 ? null : members.types[owner]?.full_name;
});
expect(members[table].map((member) => member.declaring_type)).toEqual(
expected,
);
}
expect(artifact.attributes[0]?.type_name).toBe(
members.methods[0]?.declaring_type,
);
},
);
it("indexes large unordered member lists once per layout", () => {
const count = 100_000;
const bytes = buildManagedPeFixture({
additionalEmptyTypes: count - 1,
methodRowCount: count,
});
const layout = metadataLayout(bytes);
const types = layout.table(2);
if (types === undefined) throw new Error("Fixture omitted TypeDef");
const methodOffset =
types.offset +
4 +
2 * layout.stringIndexSize +
layout.codedIndexSize("TypeDefOrRef") +
memberListIndexSize(layout, 4);
// The first reversed range forces first-match semantics, with the owner last.
// 100,000 MethodDef rows require a four-byte method-list index.
bytes.writeUInt32LE(2, methodOffset);
const start = performance.now();
const lookup = memberDeclaringTypeRowLookup(bytes, layout, 6);
expect(
Array.from({ length: count }, (_, row) => lookup(row + 1)).every(
(owner) => owner === count,
),
).toBe(true);
expect(performance.now() - start).toBeLessThan(2000);
expect(memberDeclaringTypeRowLookup(bytes, layout, 6)).toBe(lookup);
expect([lookup(0), lookup(count + 1)]).toEqual([0, 0]);
});
it("admits ownership storage before allocation and retains failed admission", () => {
const bytes = buildManagedPeFixture();
const layout = metadataLayout(bytes);
const budget = managedDecodeBudget(layout);
const capacityFailure = new ManagedReaderFailure({
code: "resource-limit",
scope: "metadata.representation",
offset: layout.table(2)?.offset ?? null,
detail: "No remaining ownership index capacity",
});
const admission = vi.spyOn(budget, "reserve").mockImplementation(() => {
throw capacityFailure;
});
let allocations = 0;
vi.stubGlobal(
"Uint32Array",
new Proxy(Uint32Array, {
construct(target, args, newTarget) {
allocations++;
return Reflect.construct(target, args, newTarget);
},
}),
);
try {
expect(() => memberDeclaringTypeRowLookup(bytes, layout, 6)).toThrow(
capacityFailure,
);
expect(() => memberDeclaringTypeRowLookup(bytes, layout, 6)).toThrow(
capacityFailure,
);
expect(admission).toHaveBeenCalledTimes(1);
expect(allocations).toBe(0);
} finally {
vi.unstubAllGlobals();
admission.mockRestore();
}
});
});
+90
View File
@@ -0,0 +1,90 @@
import { managedDecodeBudget } from "./ManagedDecodeBudget.js";
import { metadataRowCursor } from "./ManagedMetadataHeaps.js";
import type { ManagedMetadataLayout } from "./ManagedMetadataLayout.js";
import {
memberListIndexSize,
memberListRowLookup,
memberListTable,
} from "./ManagedMemberPointers.js";
import { ManagedReaderFailure } from "./ManagedReaderFailure.js";
type MemberTable = 4 | 6;
type OwnerLookup = (definitionRow: number) => number;
const lookups = new WeakMap<
ManagedMetadataLayout,
Map<MemberTable, OwnerLookup | ManagedReaderFailure>
>();
/** One first-match TypeDef owner per physical member, shared by every inspection facet. */
export const memberDeclaringTypeRowLookup = (
bytes: Buffer,
layout: ManagedMetadataLayout,
table: MemberTable,
): OwnerLookup => {
let byTable = lookups.get(layout);
const cached = byTable?.get(table);
if (cached instanceof ManagedReaderFailure) throw cached;
if (cached !== undefined) return cached;
if (byTable === undefined) {
byTable = new Map();
lookups.set(layout, byTable);
}
try {
const positionOf = memberListRowLookup(bytes, layout, table);
const positions = memberListTable(layout, table)?.rowCount ?? 0;
const types = layout.table(2)?.rowCount ?? 0;
managedDecodeBudget(layout).reserve(
(2 * positions + 3) * Uint32Array.BYTES_PER_ELEMENT + 256,
layout.table(2)?.offset ?? null,
);
const owners = new Uint32Array(positions + 1);
// Skip already-owned positions so overlapping malformed lists remain near-linear.
const next = Uint32Array.from({ length: positions + 2 }, (_, p) => p);
const unowned = (position: number): number => {
let root = position;
while (next[root] !== root) root = next[root] ?? root;
for (let p = position; p !== root;) {
const following = next[p] ?? root;
next[p] = root;
p = following;
}
return root;
};
const startForType = (row: number): number => {
const cursor = metadataRowCursor(bytes, layout, 2, row);
cursor.readUInt32();
cursor.readIndex(layout.stringIndexSize);
cursor.readIndex(layout.stringIndexSize);
cursor.readIndex(layout.codedIndexSize("TypeDefOrRef"));
const fieldStart = cursor.readIndex(memberListIndexSize(layout, 4));
return table === 4
? fieldStart
: cursor.readIndex(memberListIndexSize(layout, 6));
};
let start = types === 0 ? 0 : startForType(1);
for (let row = 1; row <= types; row += 1) {
const end = row === types ? positions + 1 : startForType(row + 1);
const limit = Math.min(end, positions + 1);
for (
let p = unowned(Math.min(Math.max(start, 1), positions + 1));
p < limit;
p = unowned(p + 1)
) {
owners[p] = row;
next[p] = p + 1;
}
start = end;
}
const definitions = layout.table(table)?.rowCount ?? 0;
const lookup: OwnerLookup = (definition) =>
definition < 1 || definition > definitions
? 0
: (owners[positionOf(definition)] ?? 0);
byTable.set(table, lookup);
return lookup;
} catch (cause: unknown) {
// Replaying a malformed layout must neither rebuild nor charge the index again.
if (cause instanceof ManagedReaderFailure) byTable.set(table, cause);
throw cause;
}
};
+15 -18
View File
@@ -8,7 +8,6 @@ import type { ManagedPeLayout } from "./ManagedPeReader.js";
import type { ManagedParseIssue } from "../domain/managed/managedArtifact.js";
import type { ManagedMetadataLayout } from "./ManagedMetadataLayout.js";
import {
createDeclaringTypeLookup,
signature,
type FieldCore,
type ManagedCallEdge,
@@ -31,7 +30,7 @@ import {
} from "./ManagedMetadataHeaps.js";
import { readManagedValue } from "./ManagedReaderFailure.js";
import { methodBody } from "./ManagedMethodBodyReader.js";
import { memberListRowLookup } from "./ManagedMemberPointers.js";
import { memberDeclaringTypeRowLookup } from "./ManagedMemberOwnership.js";
export const parseFields = (
bytes: Buffer,
@@ -46,9 +45,8 @@ export const parseFields = (
const core = new Map<string, FieldCore>();
const issues: ManagedParseIssue[] = [];
const table = layout.table(4);
const declaringType = createDeclaringTypeLookup(ranges, "field");
const listRow = readManagedValue(
() => memberListRowLookup(bytes, layout, 4),
const ownerRow = readManagedValue(
() => memberDeclaringTypeRowLookup(bytes, layout, 4),
issues,
);
for (let row = 1; row <= (table?.rowCount ?? 0); row += 1) {
@@ -67,11 +65,11 @@ export const parseFields = (
cursor.readIndex(layout.blobIndexSize),
layout.blob.size,
);
const position =
listRow === undefined
? undefined
: readManagedValue(() => listRow(row), issues);
const declared = position === undefined ? null : declaringType(position);
const owner =
ownerRow === undefined
? 0
: (readManagedValue(() => ownerRow(row), issues) ?? 0);
const declared = ranges[owner - 1];
const token = metadataToken(4, row);
fields.push({
token,
@@ -245,9 +243,8 @@ export const parseMethods = ({
const core = new Map<string, MethodCore>();
const issues: ManagedParseIssue[] = [];
const table = layout.table(6);
const declaringType = createDeclaringTypeLookup(ranges, "method");
const listRow = readManagedValue(
() => memberListRowLookup(bytes, layout, 6),
const ownerRow = readManagedValue(
() => memberDeclaringTypeRowLookup(bytes, layout, 6),
issues,
);
for (let row = 1; row <= (table?.rowCount ?? 0); row += 1) {
@@ -269,11 +266,11 @@ export const parseMethods = ({
layout.blob.size,
);
cursor.readIndex(layout.tableIndexSize(8));
const position =
listRow === undefined
? undefined
: readManagedValue(() => listRow(row), issues);
const declared = position === undefined ? null : declaringType(position);
const owner =
ownerRow === undefined
? 0
: (readManagedValue(() => ownerRow(row), issues) ?? 0);
const declared = ranges[owner - 1];
const token = metadataToken(6, row);
methods.push({
token,
+4 -48
View File
@@ -1,9 +1,5 @@
import { managedDecodeBudget } from "./ManagedDecodeBudget.js";
import {
memberListIndexSize,
memberListRowLookup,
memberListTable,
} from "./ManagedMemberPointers.js";
import { memberDeclaringTypeRowLookup } from "./ManagedMemberOwnership.js";
import type {
ManagedArtifactInspection,
@@ -246,49 +242,9 @@ const declaringTypeForMethod = (
methodRow: number,
heapExtent: number,
): string | null => {
const types = layout.table(2);
const methods = memberListTable(layout, 6);
if (types === undefined || methods === undefined) return null;
if (methodRow < 1 || methodRow > (layout.table(6)?.rowCount ?? 0))
return null;
const position = memberListRowLookup(bytes, layout, 6)(methodRow);
for (let row = 1; row <= types.rowCount; row += 1) {
const cursor = metadataRowCursor(bytes, layout, 2, row);
cursor.readUInt32();
const nameIndex = cursor.readIndex(layout.stringIndexSize);
const namespaceIndex = cursor.readIndex(layout.stringIndexSize);
cursor.readIndex(layout.codedIndexSize("TypeDefOrRef"));
cursor.readIndex(memberListIndexSize(layout, 4));
const methodStart = cursor.readIndex(memberListIndexSize(layout, 6));
const methodEnd =
row === types.rowCount
? methods.rowCount + 1
: methodListForType(bytes, layout, row + 1);
if (position < methodStart || position >= methodEnd) continue;
const name = readMetadataString(bytes, layout, nameIndex, heapExtent);
const namespace = readMetadataString(
bytes,
layout,
namespaceIndex,
heapExtent,
);
return namespace.length === 0 ? name : `${namespace}.${name}`;
}
return null;
};
const methodListForType = (
bytes: Buffer,
layout: ManagedMetadataLayout,
row: number,
): number => {
const cursor = metadataRowCursor(bytes, layout, 2, row);
cursor.readUInt32();
cursor.readIndex(layout.stringIndexSize);
cursor.readIndex(layout.stringIndexSize);
cursor.readIndex(layout.codedIndexSize("TypeDefOrRef"));
cursor.readIndex(memberListIndexSize(layout, 4));
return cursor.readIndex(memberListIndexSize(layout, 6));
const owner = memberDeclaringTypeRowLookup(bytes, layout, 6)(methodRow);
if (owner === 0) return null;
return readTypeName({ bytes, layout, table: 2, row: owner, heapExtent });
};
const attributeTypeName = (
@@ -280,7 +280,11 @@ describe("managed native boundary inspection", () => {
describe("independent managed metadata facets", () => {
it("preserves identity, members, and imports when embedded resources are malformed", () => {
const bytes = buildManagedPeFixture({ pinvoke: {}, readyToRun: true });
const bytes = buildManagedPeFixture({
mvid: Buffer.from("3aebc60edc4a544b1f458b4ed40b33b1", "hex"),
pinvoke: {},
readyToRun: true,
});
bytes.writeUInt32LE(0x7fff_0000, 0x0200 + 24);
const target = managedPeFixtureTarget(bytes);
const artifact = inspectManagedArtifactBytes(bytes, target);
@@ -288,7 +292,7 @@ describe("independent managed metadata facets", () => {
const native = inspectManagedNativeBoundariesBytes(bytes, target);
for (const result of [artifact, members, native]) {
expect(result.module?.mvid).toBeTruthy();
expect(result.module?.mvid).toBe("0ec6eb3a-4adc-4b54-1f45-8b4ed40b33b1");
expect(result.metadata.status).toBe("partial");
expect(result.coverage).toMatchObject({
state: "partial",
+4 -3
View File
@@ -22,6 +22,7 @@ interface ManagedPeFixtureOptions {
readonly malformedCustomAttributeRows?: readonly number[];
readonly metadataExtraData?: number;
readonly metadataValidMaskExtra?: bigint;
readonly methodRowCount?: number;
readonly methodName?: string;
readonly methodFlags?: number;
readonly methodSignature?: Buffer;
@@ -604,7 +605,7 @@ const fixtureIndexWidths = ({
[1, 1],
[2, indexes.additionalTypeNames.length + 1],
[4, 1],
[6, 1],
[6, options.methodRowCount ?? 1],
[8, 0],
[10, 1],
[
@@ -707,7 +708,7 @@ const fixtureMetadataRows = ({
],
[
6,
[
repeatedRows(options.methodRowCount ?? 1, () =>
methodDefRow({
name: indexes.methodName,
signature: indexes.methodSignature,
@@ -719,7 +720,7 @@ const fixtureMetadataRows = ({
options.methodFlags ??
(options.pinvoke === undefined ? 0x0016 : 0x2016),
}),
],
),
],
[
10,
+19 -10
View File
@@ -5,26 +5,30 @@ import { Readable } from "node:stream";
export const readFileHandleChunks = (
handle: Pick<FileHandle, "read">,
options: {
readonly start?: number;
readonly end?: number;
readonly start?: number | bigint;
readonly end?: number | bigint;
readonly signal?: AbortSignal | undefined;
} = {},
): Readable => {
const start = options.start ?? 0;
const end = options.end ?? Infinity;
const start = fileOffset(options.start ?? 0);
const unlimited = options.end === undefined || options.end === Infinity;
const end = unlimited ? undefined : fileOffset(options.end);
if (
!Number.isSafeInteger(start) ||
start < 0 ||
(end !== Infinity && (!Number.isSafeInteger(end) || end < start))
start === undefined ||
(!unlimited && end === undefined) ||
(end !== undefined && end < start)
)
throw new RangeError("Invalid file read range");
return Readable.from(
(async function* () {
let position = start;
while (position <= end) {
while (end === undefined || position <= end) {
options.signal?.throwIfAborted();
const remaining = end === undefined ? undefined : end - position + 1n;
const chunk = Buffer.allocUnsafe(
Math.min(64 * 1024, end - position + 1),
remaining === undefined || remaining > 64n * 1024n
? 64 * 1024
: Number(remaining),
);
const { bytesRead } = await handle.read(
chunk,
@@ -34,7 +38,7 @@ export const readFileHandleChunks = (
);
options.signal?.throwIfAborted();
if (bytesRead === 0) return;
position += bytesRead;
position += BigInt(bytesRead);
yield chunk.subarray(0, bytesRead);
}
})(),
@@ -44,3 +48,8 @@ export const readFileHandleChunks = (
},
);
};
const fileOffset = (value: number | bigint): bigint | undefined => {
if (typeof value === "bigint") return value >= 0n ? value : undefined;
return Number.isSafeInteger(value) && value >= 0 ? BigInt(value) : undefined;
};
+33 -2
View File
@@ -36,6 +36,7 @@ export const resolveGhidraAnalysisProfile = async (
installation: GhidraInstallationInspection,
signal?: AbortSignal,
languageOverride?: GhidraLanguageOverride,
disabledAnalyzers?: readonly string[],
): Promise<Result<AnalysisProfileResolution, AnalysisError>> => {
if (signal?.aborted === true)
return Promise.resolve(err(new AnalysisCancelledError("open_binary")));
@@ -70,6 +71,13 @@ export const resolveGhidraAnalysisProfile = async (
: null;
if (pspExtension !== null && !pspExtension.ok) return pspExtension;
const provider = { ...identity, version: installation.providerVersion };
const analyzerParameters =
disabledAnalyzers === undefined || disabledAnalyzers.length === 0
? { analyzer_preset: "ghidra-default" }
: {
analyzer_preset: `ghidra-default+disabled:${disabledAnalyzers.join(",")}`,
analysis_options: { disabled_analyzers: [...disabledAnalyzers] },
};
if (target.format === "raw-binary") {
if (languageOverride !== undefined)
return Promise.resolve(
@@ -116,7 +124,7 @@ export const resolveGhidraAnalysisProfile = async (
language_selection: "caller-supplied-v1",
entry_point: target.entryPoint,
entry_seed: "external-entry-and-function-before-analysis-v1",
analyzer_preset: "ghidra-default",
...analyzerParameters,
}),
}),
);
@@ -218,7 +226,7 @@ export const resolveGhidraAnalysisProfile = async (
entry_seed: "external-entry-and-function-before-analysis-v1",
}
: {}),
analyzer_preset: "ghidra-default",
...analyzerParameters,
...ghidraMipsProfileParameters(target),
...(pspExtension?.ok === true
? {
@@ -255,3 +263,26 @@ export const sameGhidraLanguageOverride = (
): boolean =>
left?.languageId === right?.languageId &&
left?.compilerSpecId === right?.compilerSpecId;
/** Read the analyzer-option commitment back from a resolved profile. */
export const ghidraProfileDisabledAnalyzers = (
parameters: Readonly<Record<string, unknown>>,
): readonly string[] | undefined => {
const options = parameters.analysis_options;
if (typeof options !== "object" || options === null) return undefined;
const disabled = (options as Record<string, unknown>).disabled_analyzers;
if (!Array.isArray(disabled)) return undefined;
return disabled.every((name) => typeof name === "string")
? (disabled as readonly string[])
: undefined;
};
/** Whether a committed profile and the active configuration disable the same analyzers. */
export const sameGhidraDisabledAnalyzers = (
left: readonly string[] | undefined,
right: readonly string[] | undefined,
): boolean =>
left === undefined || right === undefined
? left === right
: left.length === right.length &&
left.every((name, index) => name === right[index]);
@@ -0,0 +1,90 @@
import { describe, expect, it } from "vitest";
import {
ghidraDisabledAnalyzersSchema,
MAX_GHIDRA_DISABLED_ANALYZER_NAMES,
} from "../config/ghidraDisabledAnalyzers.js";
import { parseConfig } from "../config/parseConfig.js";
const INSTALL = "/opt/ghidra_12.1.4_PUBLIC";
const configured = (setting?: string) => {
const config = parseConfig({
GHIDRA_INSTALL_DIR: INSTALL,
...(setting === undefined
? {}
: { REA_GHIDRA_DISABLED_ANALYZERS: setting }),
});
if (!config.ok) throw config.error;
return config.value.ghidraDisabledAnalyzers;
};
describe("REA_GHIDRA_DISABLED_ANALYZERS", () => {
it("stays unset without the setting", () => {
expect(configured(undefined)).toBeUndefined();
});
it("splits a comma-separated list and trims each name", () => {
expect(
configured(" Objective-C Message Analyzer,Decompiler Switch Analysis "),
).toEqual(["Objective-C Message Analyzer", "Decompiler Switch Analysis"]);
});
it("treats an empty or blank list as unset", () => {
expect(configured("")).toBeUndefined();
expect(configured(" , ")).toBeUndefined();
});
it("keeps empty names out of the list instead of failing", () => {
expect(
configured("Objective-C Message Analyzer,,Decompiler Switch Analysis"),
).toEqual(["Objective-C Message Analyzer", "Decompiler Switch Analysis"]);
});
it.each([
["a leading flag name", "-Decompiler Switch Analysis"],
["a name with command metacharacters", "Analyzer(&calc)"],
["a non-ASCII name", "分析器"],
["an over-long name", "A".repeat(121)],
])("rejects %s", (_label, setting) => {
const config = parseConfig({
GHIDRA_INSTALL_DIR: INSTALL,
REA_GHIDRA_DISABLED_ANALYZERS: setting,
});
expect(config.ok).toBe(false);
if (!config.ok)
expect(config.error).toMatchObject({
settings: [
expect.objectContaining({
setting: "REA_GHIDRA_DISABLED_ANALYZERS",
}),
],
});
});
it("rejects more than the admitted analyzer count", () => {
const setting = Array.from(
{ length: MAX_GHIDRA_DISABLED_ANALYZER_NAMES + 1 },
(_, index) => `Analyzer ${index}`,
).join(",");
const config = parseConfig({
GHIDRA_INSTALL_DIR: INSTALL,
REA_GHIDRA_DISABLED_ANALYZERS: setting,
});
expect(config.ok).toBe(false);
if (!config.ok)
expect(config.error).toMatchObject({
settings: [
expect.objectContaining({
setting: "REA_GHIDRA_DISABLED_ANALYZERS",
}),
],
});
});
it("admits the documented Objective-C analyzer name", () => {
expect(
ghidraDisabledAnalyzersSchema.parse("Objective-C Message Analyzer"),
).toEqual(["Objective-C Message Analyzer"]);
});
});
+40
View File
@@ -108,10 +108,14 @@ export interface GhidraHeadlessLauncherOptions {
};
/** Explicit import language committed by the analysis profile. */
readonly languageOverride?: GhidraLanguageOverride;
/** Auto-analyzers disabled before default auto-analysis, as configured. */
readonly disabledAnalyzers?: readonly string[];
/** Caller seeds committed by the analysis profile, applied before analysis. */
readonly analysisSeeds?: GhidraSeedCommitment;
readonly analysisExtensions?: readonly GhidraExtension[];
readonly pspExtension?: GhidraPspExtension;
/** Remap typed float storage for the soft-float o32 ABI after analysis. */
readonly mipsSoftFloat?: true;
/** Spawn seam for provider-boundary lifecycle tests. */
readonly spawnProcess?: typeof spawnOwnedProviderProcess;
}
@@ -194,6 +198,9 @@ export class GhidraHeadlessLauncher implements GhidraLauncher {
const headlessArguments = ghidraHeadlessArguments({
platform,
...(psp === undefined ? {} : { psp: true as const }),
...(this.options.mipsSoftFloat === undefined
? {}
: { mipsSoftFloat: this.options.mipsSoftFloat }),
projectRoot: paths.projectRoot,
targetPath: session.targetPath,
bridgeScriptPath: this.options.bridgeScriptPath,
@@ -213,6 +220,9 @@ export class GhidraHeadlessLauncher implements GhidraLauncher {
? {}
: { languageOverride: this.options.languageOverride }),
...(seedPath === undefined ? {} : { seedPath }),
...(this.options.disabledAnalyzers === undefined
? {}
: { disabledAnalyzers: this.options.disabledAnalyzers }),
});
const scriptCommand = ghidraHeadlessCommand({
environment: this.options.environment,
@@ -417,6 +427,7 @@ export interface GhidraHeadlessArgumentOptions {
readonly ghidraLogPath: string;
readonly scriptLogPath: string;
readonly psp?: true;
readonly mipsSoftFloat?: true;
readonly dosMz?: true;
readonly dosCom?: true;
readonly rawBinary?: {
@@ -425,6 +436,8 @@ export interface GhidraHeadlessArgumentOptions {
readonly baseAddr: string;
};
readonly languageOverride?: GhidraLanguageOverride;
/** Auto-analyzers disabled by the pre-analysis options script. */
readonly disabledAnalyzers?: readonly string[];
/** Canonical seed snapshot passed to the pre-analysis seed script. */
readonly seedPath?: string;
}
@@ -453,6 +466,16 @@ export const ghidraHeadlessArguments = (
throw new Error(
"A PSP ELF import cannot be combined with a raw-binary import.",
);
if (
options.mipsSoftFloat === true &&
(options.psp === true ||
options.dosMz === true ||
options.dosCom === true ||
rawBinary !== undefined)
)
throw new Error(
"MIPS soft-float storage applies only to a generic MIPS ELF import.",
);
if (
override !== undefined &&
(options.dosMz === true || options.dosCom === true)
@@ -555,12 +578,29 @@ export const ghidraHeadlessArguments = (
scriptPath.join(bridgeDirectory, GHIDRA_SEED_SCRIPT),
options.seedPath,
]),
// Analyzer options must apply before default auto-analysis and before the
// post-analysis repair and bridge scripts start serving.
...(options.disabledAnalyzers === undefined ||
options.disabledAnalyzers.length === 0
? []
: [
"-preScript",
scriptPath.join(bridgeDirectory, "ReaGhidraAnalysisOptions.java"),
options.disabledAnalyzers.join(","),
]),
...((options.platform ?? process.platform) === "win32"
? []
: [
"-postScript",
scriptPath.join(bridgeDirectory, "ReaGhidraNoReturnFix.java"),
]),
// After analysis and the no-return refit, before the bridge serves results.
...(options.mipsSoftFloat === true
? [
"-postScript",
scriptPath.join(bridgeDirectory, "ReaGhidraMipsSoftFloat.java"),
]
: []),
"-postScript",
// Ghidra checks the caller's cwd before scriptPath for a basename. Select
// the packaged source explicitly so unrelated entries cannot shadow it.
+189 -3
View File
@@ -1,4 +1,5 @@
import { describe, expect, it } from "vitest";
import { ok } from "../domain/result.js";
import { parseConfig } from "../config/parseConfig.js";
import type {
BinaryTarget,
@@ -10,6 +11,8 @@ import { hopperLoaderArgsForTarget } from "../hopper/HopperAnalysisProfile.js";
import { IdaProvider } from "../ida/IdaProvider.js";
import { GhidraProvider } from "./GhidraProvider.js";
import type { GhidraInstallationHost } from "./GhidraInstallation.js";
import { GHIDRA_SESSION_CAPABILITIES } from "./GhidraSessionValues.js";
import type { GhidraMipsSoftFloatReport } from "./GhidraMipsProfile.js";
// Fake installation inspection tests admission/profile contracts, not Ghidra.
const host: GhidraInstallationHost = {
@@ -56,6 +59,18 @@ const target = (change: Partial<MipsElfMetadata> = {}): BinaryTarget => ({
...change,
},
});
const abi = (
change: Partial<NonNullable<MipsElfMetadata["abiFlags"]>>,
): Partial<MipsElfMetadata> => {
const base = target();
if (base.kind !== "executable" || base.architecture !== "mips")
throw new Error("expected a MIPS fixture target");
const flags = base.mips.abiFlags;
if (flags === undefined || flags === null)
throw new Error("expected fixture ABI flags");
return { abiFlags: { ...flags, ...change } };
};
const softFloat = abi({ cpr1Size: 0, fpAbi: 3 });
describe("bounded Ghidra MIPS admission", () => {
it("binds endian and flags without widening Windows P0", async () => {
@@ -102,9 +117,8 @@ describe("bounded Ghidra MIPS admission", () => {
it.each([
[{ elfClass: 64 }, "ELF32"],
[{ type: 1 }, "ET_EXEC"],
[{ type: 3 }, "ET_EXEC"],
[{ type: 0xffa0 }, "ET_EXEC"],
[{ type: 3 }, "ET_EXEC and ET_REL"],
[{ type: 0xffa0 }, "ET_EXEC and ET_REL"],
[{ flags: 0x10a23001 }, "separate MIPS ABI flags"],
[{ flags: 0x70001021 }, "o32"],
[{ flags: 0x70000001 }, "o32"],
@@ -118,6 +132,11 @@ describe("bounded Ghidra MIPS admission", () => {
[{ flags: 0x70001201 }, "EF_MIPS_FP64"],
[{ flags: 0x70001401 }, "EF_MIPS_NAN2008"],
[{ abiFlags: null }, "inspected ABI"],
[abi({ cpr1Size: 0, fpAbi: 1 }), "soft-float without CPR1"],
[abi({ cpr1Size: 1, fpAbi: 3 }), "soft-float without CPR1"],
[abi({ cpr1Size: 1, fpAbi: 2 }), "single-float"],
[abi({ cpr1Size: 1, fpAbi: 0 }), "unspecified"],
[abi({ cpr1Size: 2, fpAbi: 6 }), "FP64"],
] satisfies [Partial<MipsElfMetadata>, string][])(
"refuses an unverified MIPS interpretation %j",
async (change, reason) => {
@@ -140,3 +159,170 @@ describe("bounded Ghidra MIPS admission", () => {
expect(ida.inspectTargetSupport(target()).status).toBe("unsupported");
});
});
const report: GhidraMipsSoftFloatReport = {
policy: "o32-soft-float-prototype-v1",
functions: 5,
definitions: 1,
};
const session = async (
value: BinaryTarget,
reported: GhidraMipsSoftFloatReport | undefined,
profileFor: BinaryTarget = value,
) => {
const counts = { starts: 0, calls: 0, closes: 0 };
const ghidra = new GhidraProvider(config, silentLogger, {}, host, () => ({
start: () => {
counts.starts++;
return Promise.resolve(
ok({
name: "REA Ghidra bridge" as const,
run_id: "11111111-1111-4111-8111-111111111111",
profile_digest: "a".repeat(64),
provider: { id: "ghidra" as const, version: "12.1.4" },
read_only: false,
analysis_complete: true,
analysis_timed_out: false,
capabilities: [...GHIDRA_SESSION_CAPABILITIES],
target: {
name: "fixture.elf",
language_id: "MIPS:LE:32:default",
compiler_spec_id: "default",
image_base: "0x0",
default_address_space: "ram",
sha256: "a".repeat(64),
},
...(reported === undefined
? {}
: { mips_soft_float_storage: reported }),
}),
);
},
callTool: () => {
counts.calls++;
return Promise.resolve(ok(null));
},
close: () => {
counts.closes++;
return Promise.resolve(ok(null));
},
}));
const resolved = await ghidra.resolveAnalysisProfile(profileFor);
if (!resolved.ok || resolved.value.profile === null)
throw new Error("expected a MIPS profile");
return {
counts,
profile: resolved.value.profile,
client: ghidra.createClient(value, resolved.value.profile),
};
};
describe("soft-float MIPS storage", () => {
it("admits soft-float o32 and commits its storage policy only for it", async () => {
const soft = target(softFloat);
const ghidra = new GhidraProvider(config, silentLogger, {}, host);
expect(ghidra.inspectTargetSupport(soft).status).toBe("supported");
const { profile } = await session(soft, report);
expect(profile.parameters.mips_soft_float_storage).toBe(
"o32-soft-float-prototype-v1",
);
const hard = await session(target(), undefined);
expect(hard.profile.parameters).not.toHaveProperty(
"mips_soft_float_storage",
);
});
it("admits the ODDSPREG default that soft-float toolchains emit, but no other flags", () => {
const ghidra = new GhidraProvider(config, silentLogger, {}, host);
// GCC 5.2 kernel modules and LLVM/Rust firmware both record flags1 = 1.
expect(
ghidra.inspectTargetSupport(
target(abi({ cpr1Size: 0, fpAbi: 3, flags1: 1 })),
).status,
).toBe("supported");
expect(
ghidra.inspectTargetSupport(
target(abi({ cpr1Size: 0, fpAbi: 3, flags1: 2 })),
),
).toMatchObject({
status: "unsupported",
reason: expect.stringContaining("general flags"),
});
});
it("reports the prototype outcome as session limitations", async () => {
const { client } = await session(target(softFloat), report);
const health = await client.execute("health", {});
if (!health.ok) throw health.error;
expect(health.value.limitations).toEqual(
expect.arrayContaining([
expect.stringContaining("model hard-float o32 only"),
expect.stringContaining(
"5 functions and 1 function-pointer definitions use __rea_soft_float.",
),
]),
);
});
it("admits relocatable objects under their own lane with coordinate limitations", async () => {
const relocatable = target({ type: 1 });
const { profile, client } = await session(relocatable, undefined);
expect(profile.parameters.mips_support_lane).toBe(
"elf32-rel-o32-arch32r2-standard-v1",
);
const executable = await session(target(), undefined);
expect(executable.profile.parameters.mips_support_lane).toBe(
"elf32-exec-o32-arch32r2-standard-v3",
);
const health = await client.execute("health", {});
if (!health.ok) throw health.error;
expect(health.value.limitations).toEqual(
expect.arrayContaining([
expect.stringContaining("analysis coordinates"),
expect.stringContaining("EXTERNAL block"),
]),
);
const executableHealth = await executable.client.execute("health", {});
if (!executableHealth.ok) throw executableHealth.error;
expect(executableHealth.value.limitations.join(" ")).not.toContain(
"Relocatable MIPS ELF",
);
});
it.each([
{
reason:
"a soft-float session without a storage report before any analysis call",
value: target(softFloat),
storageReport: undefined,
operation: "list_procedures",
},
{
reason: "soft-float storage reported by a hard-float session",
value: target(),
storageReport: report,
operation: "health",
},
] as const)(
"rejects $reason",
async ({ value, storageReport, operation }) => {
const { counts, client } = await session(value, storageReport);
const failed = await client.execute(operation, {});
expect(failed.ok).toBe(false);
if (!failed.ok) expect(failed.error._tag).toBe("ProviderAdapterError");
expect(counts).toEqual({ starts: 1, calls: 0, closes: 1 });
},
);
it("refuses a profile whose storage policy disagrees with the target", async () => {
for (const [value, profileFor] of [
[target(softFloat), target()],
[target(), target(softFloat)],
] as const) {
const { counts, client } = await session(value, report, profileFor);
const failed = await client.execute("health", {});
expect(failed.ok).toBe(false);
expect(counts.starts).toBe(0);
}
});
});
+79 -7
View File
@@ -1,6 +1,10 @@
import type { BinaryTarget } from "../domain/binaryTargetTypes.js";
import type { JsonValue } from "../domain/jsonValue.js";
const MIPS_FP_ABI_SOFT = 3;
const MIPS_ET_REL = 1;
const MIPS_ET_EXEC = 2;
/** Keep family recognition separate from the deliberately bounded Ghidra lane. */
export const ghidraMipsUnsupportedReason = (
target: BinaryTarget,
@@ -12,8 +16,8 @@ export const ghidraMipsUnsupportedReason = (
return "MIPS analysis requires ELF header metadata; a family name alone does not select an ISA or ABI.";
if (metadata.elfClass !== 32)
return "This Ghidra MIPS lane supports ELF32 only; MIPS64 and n64 remain unsupported.";
if (metadata.type !== 2)
return "This Ghidra MIPS lane supports ET_EXEC only; relocatable objects, shared images and PSP PRX need separate load verification.";
if (metadata.type !== MIPS_ET_EXEC && metadata.type !== MIPS_ET_REL)
return "This Ghidra MIPS lane supports ET_EXEC and ET_REL only; shared images and PSP PRX need separate load verification.";
if ((metadata.flags & 0x0f000000) !== 0)
return "MIPS16, microMIPS and other ASE encodings are outside this standard-instruction MIPS lane.";
if ((metadata.flags & 0x00ff0000) !== 0)
@@ -39,17 +43,79 @@ export const ghidraMipsUnsupportedReason = (
return "This Ghidra MIPS lane requires ABI flags version 0; newer record semantics are unsupported.";
if (abi.isaLevel !== 32 || abi.isaRevision !== 2)
return "MIPS ABI flags must agree with the declared MIPS32r2 ISA.";
if (abi.gprSize !== 1 || abi.cpr1Size !== 1 || abi.cpr2Size !== 0)
return "This Ghidra MIPS lane requires 32-bit GPR/CPR1 and no CPR2; other register modes need separate verification.";
if (abi.fpAbi !== 1 && abi.fpAbi !== 5)
return "This Ghidra MIPS lane admits double-precision or FPXX ABI declarations only; soft-float, FP64 and unspecified modes need separate verification.";
if (abi.gprSize !== 1 || abi.cpr2Size !== 0)
return "This Ghidra MIPS lane requires 32-bit GPRs and no CPR2; other register modes need separate verification.";
const hardFloat = abi.cpr1Size === 1 && (abi.fpAbi === 1 || abi.fpAbi === 5);
const softFloat = abi.cpr1Size === 0 && abi.fpAbi === MIPS_FP_ABI_SOFT;
if (!hardFloat && !softFloat)
return "This Ghidra MIPS lane admits double-precision or FPXX ABI declarations with 32-bit CPR1, or soft-float without CPR1; single-float, FP64 and unspecified modes need separate verification.";
if (abi.isaExtension !== 0 || abi.ases !== 0)
return "MIPS ABI flags declare an ISA extension or ASE outside this standard-instruction lane.";
// MIPS_AFL_FLAGS1_ODDSPREG (bit 0) is the GCC and LLVM default for MIPS32 o32
// whatever the FP ABI; soft-float objects from both carry it. With no CPR1 it
// selects nothing, so it is accepted for soft-float as well as hard-float.
if ((abi.flags1 & ~1) !== 0 || abi.flags2 !== 0)
return "MIPS ABI flags contain unsupported general flags.";
return null;
};
/** Soft-float prototype policy applied by ReaGhidraMipsSoftFloat.java after analysis. */
export const GHIDRA_MIPS_SOFT_FLOAT_STORAGE = "o32-soft-float-prototype-v1";
/** Whether the admitted target declares the soft-float o32 ABI. */
export const isGhidraMipsSoftFloatTarget = (target: BinaryTarget): boolean =>
target.kind === "executable" &&
target.architecture === "mips" &&
target.mips?.abiFlags?.cpr1Size === 0 &&
target.mips.abiFlags.fpAbi === MIPS_FP_ABI_SOFT;
/** Prototype outcome the post-analysis script records for the bridge handshake. */
export interface GhidraMipsSoftFloatReport {
readonly policy: typeof GHIDRA_MIPS_SOFT_FLOAT_STORAGE;
readonly functions: number;
readonly definitions: number;
}
/** Fail closed unless soft-float storage ran exactly for soft-float targets. */
export const ghidraMipsSoftFloatFailure = (
expected: boolean,
report: GhidraMipsSoftFloatReport | undefined,
): string | undefined =>
expected === (report !== undefined)
? undefined
: expected
? "The Ghidra session did not report the committed MIPS soft-float prototype."
: "The Ghidra session reported MIPS soft-float storage that the profile did not commit.";
/** Keep REA's soft-float prototype distinct from Ghidra's hard-float o32 model. */
export const ghidraMipsSoftFloatLimitations = (
expected: boolean,
report?: GhidraMipsSoftFloatReport,
): readonly string[] =>
expected
? [
"Soft-float o32: Ghidra's MIPS32 compiler specifications model hard-float o32 only. After analysis REA installs a program-specific `__rea_soft_float` prototype, derived from the default o32 model without its floating-point registers and rules, and assigns it to every function, external function and function-pointer definition whose declared types use float or double, so those values keep their types and use integer registers and stack slots.",
...(report === undefined
? []
: [
`Soft-float prototype outcome: ${report.functions} functions and ${report.definitions} function-pointer definitions use __rea_soft_float.`,
]),
]
: [];
/** Ghidra-placed relocatable objects need coordinates and relocation caveats. */
export const ghidraMipsRelocatableLimitations = (
target: BinaryTarget,
): readonly string[] =>
target.kind === "executable" &&
target.architecture === "mips" &&
target.mips?.type === MIPS_ET_REL
? [
"Relocatable MIPS ELF: Ghidra assigns each allocated section an analysis address and applies the object's relocations there. Returned addresses are analysis coordinates, not the addresses a kernel or linker would choose at load time.",
"Undefined symbols (for example kernel exports used by a module) are placed in Ghidra's EXTERNAL block; their definitions are outside this artifact. inspect_native_load_image reports each relocation with Ghidra's status; HI16 entries report PARTIAL because Ghidra applies them together with their paired LO16.",
]
: [];
/** Commit source ELF interpretation without confusing target ISA with host CPU. */
export const ghidraMipsProfileParameters = (
target: BinaryTarget,
@@ -87,6 +153,12 @@ export const ghidraMipsProfileParameters = (
flags2: abi.flags2,
},
},
mips_support_lane: "elf32-exec-o32-arch32r2-standard-v3",
...(isGhidraMipsSoftFloatTarget(target)
? { mips_soft_float_storage: GHIDRA_MIPS_SOFT_FLOAT_STORAGE }
: {}),
mips_support_lane:
target.mips.type === MIPS_ET_REL
? "elf32-rel-o32-arch32r2-standard-v1"
: "elf32-exec-o32-arch32r2-standard-v3",
};
};
+150
View File
@@ -23,6 +23,7 @@ import { parseExecutableHeader } from "../domain/binaryTarget.js";
import { GhidraProvider } from "./GhidraProvider.js";
import type { GhidraTargetSnapshot } from "./GhidraClient.js";
import type { GhidraProviderClientFactory } from "./GhidraProviderClient.js";
import type { GhidraHeadlessLauncher } from "./GhidraLauncher.js";
import type { GhidraInstallationHost } from "./GhidraInstallation.js";
import { GHIDRA_SESSION_CAPABILITIES } from "./GhidraSessionValues.js";
import { err, ok } from "../domain/result.js";
@@ -1810,3 +1811,152 @@ describe("Ghidra analysis seeds and configured language", () => {
}
});
});
describe("Ghidra disabled analyzer options", () => {
const configuredProvider = (
environment: Record<string, string>,
clientFactory?: GhidraProviderClientFactory,
) => {
const config = parseConfig({ GHIDRA_INSTALL_DIR: INSTALL, ...environment });
if (!config.ok) throw config.error;
return new GhidraProvider(
config.value,
silentLogger,
{},
installationHost(),
clientFactory,
);
};
it("records the configured analyzer options in the analysis profile", async () => {
const ghidra = configuredProvider({
REA_GHIDRA_DISABLED_ANALYZERS:
"Objective-C Message Analyzer,Decompiler Switch Analysis",
});
const resolved = await ghidra.resolveAnalysisProfile(
executableTarget("elf", "x86_64"),
);
if (!resolved.ok || resolved.value.profile === null)
throw new Error("expected a bound Ghidra profile");
expect(resolved.value.profile.parameters).toMatchObject({
analyzer_preset:
"ghidra-default+disabled:Objective-C Message Analyzer,Decompiler Switch Analysis",
analysis_options: {
disabled_analyzers: [
"Objective-C Message Analyzer",
"Decompiler Switch Analysis",
],
},
});
});
it("hands the configured analyzers to the launcher for a matching profile", async () => {
const launchers: GhidraHeadlessLauncher[] = [];
const ghidra = configuredProvider(
{ REA_GHIDRA_DISABLED_ANALYZERS: "Objective-C Message Analyzer" },
(options) => {
launchers.push(options.launcher as GhidraHeadlessLauncher);
return {
start: () => Promise.resolve(ok(sessionInfo())),
callTool: () => Promise.resolve(ok(null)),
close: () => Promise.resolve(ok(null)),
};
},
);
const target = executableTarget("elf", "x86_64");
const resolved = await ghidra.resolveAnalysisProfile(target);
if (!resolved.ok || resolved.value.profile === null)
throw new Error("expected a bound Ghidra profile");
const client = ghidra.createClient(target, resolved.value.profile);
const health = await client.execute("health", {});
expect(health.ok).toBe(true);
expect(launchers).toHaveLength(1);
expect(launchers[0]?.options.disabledAnalyzers).toEqual([
"Objective-C Message Analyzer",
]);
});
it("binds analyzer overrides and raw import settings in one profile", async () => {
const launchers: GhidraHeadlessLauncher[] = [];
const ghidra = configuredProvider(
{ REA_GHIDRA_DISABLED_ANALYZERS: "Decompiler Switch Analysis" },
(options) => {
launchers.push(options.launcher as GhidraHeadlessLauncher);
return {
start: () => Promise.resolve(ok(sessionInfo())),
callTool: () => Promise.resolve(ok(null)),
close: () => Promise.resolve(ok(null)),
};
},
);
const target: BinaryTarget = {
path: "/tmp/entry.bin",
sha256: "a".repeat(64),
kind: "executable",
format: "raw-binary",
architecture: "x86",
availableArchitectures: ["x86"],
entryPoint: "0x1000",
entryPointSource: "0x1000",
};
const resolved = await ghidra.resolveAnalysisProfile(target);
if (!resolved.ok || resolved.value.profile === null)
throw new Error("expected a bound raw-binary profile");
expect(resolved.value.profile.parameters).toMatchObject({
target_format: "raw-binary",
entry_point: "0x1000",
analyzer_preset: "ghidra-default+disabled:Decompiler Switch Analysis",
analysis_options: { disabled_analyzers: ["Decompiler Switch Analysis"] },
});
const client = ghidra.createClient(target, resolved.value.profile);
expect(launchers).toHaveLength(1);
expect(launchers[0]?.options).toMatchObject({
disabledAnalyzers: ["Decompiler Switch Analysis"],
rawBinary: {
languageId: "x86:LE:32:default",
compilerSpecId: "gcc",
baseAddr: "0x1000",
},
});
await client.close();
});
it("refuses a profile resolved without or with other analyzer options", async () => {
const target = executableTarget("elf", "x86_64");
const plainProvider = configuredProvider({});
const plain = await plainProvider.resolveAnalysisProfile(target);
if (!plain.ok || plain.value.profile === null)
throw new Error("expected profile");
const configuredProvider2 = configuredProvider({
REA_GHIDRA_DISABLED_ANALYZERS: "Objective-C Message Analyzer",
});
const configured = await configuredProvider2.resolveAnalysisProfile(target);
if (!configured.ok || configured.value.profile === null)
throw new Error("expected configured profile");
for (const [profile, environment] of [
[
plain.value.profile,
{ REA_GHIDRA_DISABLED_ANALYZERS: "Objective-C Message Analyzer" },
],
[configured.value.profile, {}],
[
configured.value.profile,
{ REA_GHIDRA_DISABLED_ANALYZERS: "Decompiler Switch Analysis" },
],
] as const) {
let starts = 0;
const client = configuredProvider(environment, () => ({
start: () => {
starts++;
return Promise.resolve(ok(sessionInfo()));
},
callTool: () => Promise.resolve(ok(null)),
close: () => Promise.resolve(ok(null)),
})).createClient(target, profile);
const failed = await client.execute("health", {});
expect(failed.ok).toBe(false);
if (!failed.ok) expect(failed.error._tag).toBe("ProviderAdapterError");
expect(starts).toBe(0);
}
});
});

Some files were not shown because too many files have changed in this diff Show More