# -*- coding: utf-8 -*- """ 全网舆情爬虫 — 多格式导出器 v4.3 ================================ 把 SentimentSnapshot 导出为: 1. 对话提示(dialog) — 直接贴入对话窗口 2. Word (.docx) — 含封面 / 目录 / 分类 / 摘要,正面 vs 舆情分章 3. Excel (.xlsx) — 分sheet 正面 / 舆情 / 中性 / 全部 4. JSON / CSV — 机器可读备份 输出位置:默认 data/sentiment_exports/ """ from __future__ import annotations import csv import json import logging from pathlib import Path from typing import Dict, Any, List, Optional from datetime import datetime # 兼容可选依赖 try: from openpyxl import Workbook # type: ignore from openpyxl.styles import Font, PatternFill, Alignment, Border, Side # type: ignore HAS_OPENPYXL = True except ImportError: HAS_OPENPYXL = False try: from docx import Document # type: ignore from docx.shared import Pt, Cm, RGBColor # type: ignore from docx.enum.text import WD_ALIGN_PARAGRAPH # type: ignore HAS_DOCX = True except ImportError: HAS_DOCX = False try: from sentiment_crawler import SentimentSnapshot, SentimentArticle # type: ignore except ImportError: SentimentSnapshot = SentimentArticle = None # type: ignore # 路径 SKILL_DATA_DIR = Path(__file__).resolve().parent.parent / "data" EXPORT_DIR = SKILL_DATA_DIR / "sentiment_exports" EXPORT_DIR.mkdir(parents=True, exist_ok=True) logger = logging.getLogger("sentiment_exporter") logger.setLevel(logging.INFO) # ================================================================ # 1. 对话提示 — 给大模型/用户直接看 # ================================================================ def to_dialog(snapshot: SentimentSnapshot, top_k: int = 30, with_links: bool = True) -> str: """生成可直接贴到对话的精简提示""" if snapshot is None or not snapshot.articles: return "📭 本次全网舆情爬取未获取到任何结果,请尝试放宽条件。" lines: List[str] = [] lines.append(f"📰 **全网舆情快照** `{snapshot.snapshot_id}`") lines.append(f"📅 抓取时间:{snapshot.created_at}") lines.append( f"📊 共 {len(snapshot.articles)} 条 | 正面 {snapshot.positive_count()} | " f"舆情 {snapshot.negative_count()} | 中性 {snapshot.neutral_count()}" ) sev = snapshot.stats.get("by_severity", {}) if sev: sev_str = " | ".join(f"{k}: {v}" for k, v in sev.items() if k != "中性") if sev_str: lines.append(f"⚠️ 风险分布:{sev_str}") # 媒体来源 Top 5 src_dist = snapshot.stats.get("by_source_type", {}) if src_dist: src_top = sorted(src_dist.items(), key=lambda x: -x[1])[:5] src_str = " | ".join(f"{k}:{v}" for k, v in src_top) lines.append(f"🗞️ 媒体类别 Top: {src_str}") # 耗时 elapsed = snapshot.stats.get("elapsed_seconds") if elapsed is not None: lines.append(f"⏱ 耗时 {elapsed:.1f}s" + (" (超时返回)" if snapshot.stats.get("timed_out") else "")) lines.append("") # 正面 pos = [a for a in snapshot.articles if a.sentiment == "positive"][:top_k] if pos: lines.append(f"### ✅ 正面新闻 ({len(pos)})") for i, a in enumerate(pos, 1): lines.append(_dialog_one_line(i, a, with_links)) lines.append("") # 舆情(按严重度) negs = [a for a in snapshot.articles if a.sentiment == "negative"] negs.sort(key=lambda a: (-a.sentiment_score, a.publish_time or "")) if negs: lines.append(f"### ⚠️ 舆情(负面)({len(negs)})") for i, a in enumerate(negs[:top_k], 1): lines.append(_dialog_one_line(i, a, with_links)) lines.append("") # 中性 neu = [a for a in snapshot.articles if a.sentiment == "neutral"][:max(5, top_k // 3)] if neu: lines.append(f"### ℹ️ 中性公告 ({len(neu)})") for i, a in enumerate(neu, 1): lines.append(_dialog_one_line(i, a, with_links)) lines.append("") lines.append("---") lines.append(f"📁 完整快照保存于:{snapshot.extra_path if hasattr(snapshot, 'extra_path') else 'data/sentiment_snapshots/'}") # type: ignore[attr-defined] lines.append(f"💡 提示:可对我说「导出Word」「导出Excel」「按严重度筛选」获得更多处理。") return "\n".join(lines) def _dialog_one_line(idx: int, a: SentimentArticle, with_links: bool) -> str: risk_icon = { "低度关注": "🟡", "中度舆情": "🟠", "高危舆情": "🔴", "低度利好": "🟢", "中度利好": "🟢", "重大利好": "✅", }.get(a.severity, "⚪") pos_icon = {"positive": "🟢", "negative": "🔴", "neutral": "⚪"}.get(a.sentiment, "⚪") fields = [ f"{idx}. {pos_icon}{risk_icon} **[{a.target_name}]**", f"《{a.title}》", f" · 平台:{a.source or '未知'} ({a.source_type or '媒体'})", f" · 时间:{a.publish_time or '未知'}", f" · 摘要:{a.summary or '(无摘要)'}", ] if with_links and a.url: fields.append(f" · 链接:{a.url}") if a.keywords_matched: kw_text = "、".join(a.keywords_matched[:6]) fields.append(f" · 命中关键词:{kw_text}") return "\n".join(fields) # ================================================================ # 2. Excel (.xlsx) — 包含总览 + 分类 # ================================================================ def to_excel(snapshot: SentimentSnapshot, output_path: Optional[Path] = None) -> str: """导出为 xlsx 多 sheet。 Sheet: 概览 / 正面新闻 / 舆情(负面) / 中性 / 全部 """ if not HAS_OPENPYXL: return _placeholder("Excel 导出需要 openpyxl: pip install openpyxl") if snapshot is None: return _placeholder("快照为空") output_path = Path(output_path or EXPORT_DIR / f"{snapshot.snapshot_id}.xlsx") output_path.parent.mkdir(parents=True, exist_ok=True) wb = Workbook() # ---- Sheet 1 概览 ---- ws = wb.active ws.title = "概览" ws["A1"] = "全网舆情快照" ws["A1"].font = Font(size=18, bold=True) ws.merge_cells("A1:E1") rows = [ ("快照ID", snapshot.snapshot_id), ("抓取时间", snapshot.created_at), ("目标过滤", json.dumps(snapshot.target_filter, ensure_ascii=False)), ("媒体过滤", " / ".join(snapshot.source_filter)), ("", ""), ("总条数", snapshot.stats.get("total", 0)), ("正面", snapshot.positive_count()), ("舆情", snapshot.negative_count()), ("中性", snapshot.neutral_count()), ("", ""), ("风险分布", json.dumps(snapshot.stats.get("by_severity", {}), ensure_ascii=False)), ("媒体类别分布", json.dumps(snapshot.stats.get("by_source_type", {}), ensure_ascii=False)), ] for i, (k, v) in enumerate(rows, 3): ws.cell(row=i, column=1, value=k).font = Font(bold=True) ws.cell(row=i, column=2, value=v) ws.column_dimensions["A"].width = 18 ws.column_dimensions["B"].width = 80 # ---- Sheet 2-5 分类明细 ---- for sheet_name, sentiment in [ ("正面新闻", "positive"), ("舆情负面", "negative"), ("中性公告", "neutral"), ("全部文章", "all"), ]: ws = wb.create_sheet(sheet_name) headers = ["序号", "标题", "内容简介", "发布平台", "媒体类型", "发布时间", "页面连接", "目标", "严重等级", "情感", "命中关键词"] for col, h in enumerate(headers, 1): c = ws.cell(row=1, column=col, value=h) c.font = Font(bold=True, color="FFFFFF", name="宋体") # v4.3.2: 添加中文字体 c.fill = PatternFill("solid", fgColor="2E5984") c.alignment = Alignment(horizontal="center") items = snapshot.articles if sentiment == "all" else [a for a in snapshot.articles if a.sentiment == sentiment] items.sort(key=lambda a: (-(1 if a.severity in ("高危舆情", "中度舆情") else 0), a.publish_time or "")) for r, a in enumerate(items, 2): ws.cell(row=r, column=1, value=r - 1) ws.cell(row=r, column=2, value=a.title) ws.cell(row=r, column=3, value=a.summary) ws.cell(row=r, column=4, value=a.source) ws.cell(row=r, column=5, value=a.source_type) ws.cell(row=r, column=6, value=a.publish_time) cell_url = ws.cell(row=r, column=7, value=a.url) cell_url.hyperlink = a.url or None cell_url.font = Font(color="0563C1", underline="single") ws.cell(row=r, column=8, value=f"{a.target_name} ({a.target_type})") sev_cell = ws.cell(row=r, column=9, value=a.severity) if a.severity == "高危舆情": sev_cell.fill = PatternFill("solid", fgColor="FFC7CE") sev_cell.font = Font(bold=True, color="9C0006", name="宋体") elif a.severity == "中度舆情": sev_cell.fill = PatternFill("solid", fgColor="FFE699") elif a.severity == "重大利好": sev_cell.fill = PatternFill("solid", fgColor="C6EFCE") sev_cell.font = Font(bold=True, color="006100", name="宋体") elif a.severity == "中度利好": sev_cell.fill = PatternFill("solid", fgColor="E2EFDA") ws.cell(row=r, column=10, value=a.sentiment) ws.cell(row=r, column=11, value="、".join(a.keywords_matched[:6])) # v4.3.2: 批量设置数据区字体 for row in ws.iter_rows(min_row=2, max_row=ws.max_row, min_col=1, max_col=11): for cell in row: if cell.font and cell.font.name and cell.font.name != "宋体": continue # 保留已自定义的字体(超鏈接、严重色等) if cell.value is not None and (cell.font is None or not cell.font.name): cell.font = Font(name="宋体", size=11) # 设置列宽 widths = [6, 50, 60, 16, 14, 18, 40, 24, 14, 10, 30] for i, w in enumerate(widths, 1): col_letter = chr(64 + i) ws.column_dimensions[col_letter].width = w wb.save(output_path) return str(output_path) # ================================================================ # 3. Word (.docx) — 含封面 / 目录 / 章节 # ================================================================ def to_word(snapshot: SentimentSnapshot, output_path: Optional[Path] = None) -> str: """导出为 .docx,含封面 / 概览 / 分类正文。""" if not HAS_DOCX: return _placeholder("Word 导出需要 python-docx: pip install python-docx") if snapshot is None: return _placeholder("快照为空") output_path = Path(output_path or EXPORT_DIR / f"{snapshot.snapshot_id}.docx") output_path.parent.mkdir(parents=True, exist_ok=True) doc = Document() # v4.3.2 修复:设置默认字体为 宋体(中文)+ Times New Roman(拉丁), # 确保跨平台中文不乱码。 from docx.shared import Pt style = doc.styles["Normal"] font = style.font font.name = "Times New Roman" font.size = Pt(11) rPr = style.element.get_or_add_rPr() NS = "http://schemas.openxmlformats.org/wordprocessingml/2006/main" rFonts = rPr.makeelement(f"{{{NS}}}rFonts", {}) rFonts.set(f"{{{NS}}}ascii", "Times New Roman") rFonts.set(f"{{{NS}}}hAnsi", "Times New Roman") rFonts.set(f"{{{NS}}}eastAsia", "宋体") rFonts.set(f"{{{NS}}}cs", "Times New Roman") rPr.insert(0, rFonts) # ---- 标题 ---- title = doc.add_heading("全网舆情快照", level=0) title.alignment = WD_ALIGN_PARAGRAPH.CENTER # 元信息 p = doc.add_paragraph() p.alignment = WD_ALIGN_PARAGRAPH.CENTER p.add_run(f"快照ID:{snapshot.snapshot_id} 抓取时间:{snapshot.created_at}").italic = True # ---- 概览 ---- doc.add_heading("1. 抓取概览", level=1) table = doc.add_table(rows=0, cols=2) table.style = "Light List" overview_rows = [ ("目标过滤", json.dumps(snapshot.target_filter, ensure_ascii=False)), ("媒体过滤", " / ".join(snapshot.source_filter)), ("总条数", str(snapshot.stats.get("total", 0))), ("正面", str(snapshot.positive_count())), ("舆情(负面)", str(snapshot.negative_count())), ("中性", str(snapshot.neutral_count())), ("风险分布", json.dumps(snapshot.stats.get("by_severity", {}), ensure_ascii=False)), ] for k, v in overview_rows: row = table.add_row().cells row[0].text = k row[1].text = v # ---- 章节 ---- sections = [ ("2. ✅ 正面新闻", "positive", "🟢"), ("3. ⚠️ 舆情(按严重度倒序)", "negative", "🔴"), ("4. ℹ️ 中性公告", "neutral", "⚪"), ] for sec_title, sent_type, icon in sections: doc.add_heading(sec_title, level=1) items = [a for a in snapshot.articles if a.sentiment == sent_type] if sent_type == "negative": items.sort(key=lambda a: (-a.sentiment_score, a.publish_time or "")) else: items.sort(key=lambda a: (a.publish_time or ""), reverse=True) if not items: doc.add_paragraph("(本分类暂无内容)").italic = True continue for idx, a in enumerate(items, 1): heading_text = f"{icon} {idx}. [{a.target_name}] 《{a.title}》" doc.add_heading(heading_text, level=2) t = doc.add_table(rows=0, cols=2) t.style = "Light Grid" for k, v in [ ("标题", a.title), ("内容简介", a.summary or "(无摘要)"), ("发布平台", f"{a.source} ({a.source_type})"), ("发布时间", a.publish_time), ("页面连接", a.url), ("严重等级", a.severity), ("命中关键词", "、".join(a.keywords_matched[:8])), ("情感分数", f"{a.sentiment_score:.1f}"), ]: row = t.add_row().cells row[0].text = k row[1].text = v doc.save(output_path) return str(output_path) # ================================================================ # 4. JSON / CSV — 备份 # ================================================================ def to_json(snapshot: SentimentSnapshot, output_path: Optional[Path] = None) -> str: p = Path(output_path or EXPORT_DIR / f"{snapshot.snapshot_id}.json") p.write_text(json.dumps(snapshot.to_dict(), ensure_ascii=False, indent=2), encoding="utf-8") return str(p) def to_csv(snapshot: SentimentSnapshot, output_path: Optional[Path] = None) -> str: p = Path(output_path or EXPORT_DIR / f"{snapshot.snapshot_id}.csv") with open(p, "w", encoding="utf-8-sig", newline="") as f: writer = csv.writer(f) writer.writerow(["标题", "内容简介", "发布平台", "媒体类型", "发布时间", "页面连接", "目标", "情感", "严重等级", "命中关键词"]) for a in snapshot.articles: writer.writerow([ a.title, a.summary, a.source, a.source_type, a.publish_time, a.url, f"{a.target_name} ({a.target_type})", a.sentiment, a.severity, "、".join(a.keywords_matched[:6]) ]) return str(p) # ================================================================ # 5. 总入口 — auto 模式 # ================================================================ def export(snapshot: SentimentSnapshot, fmt: str = "auto", output_path: Optional[Path] = None) -> Dict[str, str]: """统一导出。 fmt: word / excel / csv / json / dialog / all / auto """ out: Dict[str, str] = {} if snapshot is None: return out if fmt == "dialog": out["dialog"] = to_dialog(snapshot) return out if fmt == "word": out["word"] = to_word(snapshot, output_path) elif fmt == "excel": out["excel"] = to_excel(snapshot, output_path) elif fmt == "csv": out["csv"] = to_csv(snapshot, output_path) elif fmt == "json": out["json"] = to_json(snapshot, output_path) elif fmt in ("all", "auto"): out["word"] = to_word(snapshot) out["excel"] = to_excel(snapshot) out["json"] = to_json(snapshot) out["csv"] = to_csv(snapshot) out["dialog"] = to_dialog(snapshot) return out def _placeholder(msg: str) -> str: return f"⚠️ {msg}" # CLI def _cli(): import argparse parser = argparse.ArgumentParser(description="cn-financial-scraper 舆情快照导出 v4.3") parser.add_argument("snapshot_id", type=str, help="快照ID(或传 json 文件路径)") parser.add_argument("--fmt", type=str, default="auto", help="输出格式 word/excel/csv/json/dialog/all/auto") args = parser.parse_args() snapshot_file = SNAPSHOT_DIR if not Path(args.snapshot_id).exists() else Path(args.snapshot_id) # 简单解析快照 if Path(args.snapshot_id).exists(): raw = json.loads(Path(args.snapshot_id).read_text(encoding="utf-8")) from sentiment_crawler import SentimentSnapshot, SentimentArticle # type: ignore arts = [SentimentArticle(**a) for a in raw.get("articles", [])] snap = SentimentSnapshot( snapshot_id=raw["snapshot_id"], created_at=raw["created_at"], target_filter=raw.get("target_filter", {}), source_filter=raw.get("source_filter", []), articles=arts, stats=raw.get("stats", {}), ) snap.extra_path = args.snapshot_id # type: ignore[attr-defined] else: print(f"⚠️ 快照文件不存在: {args.snapshot_id}") return results = export(snap, fmt=args.fmt) for k, v in results.items(): if k == "dialog": print(v) else: print(f"✅ [{k.upper()}] {v}") if __name__ == "__main__": _cli()