feat(compose): 标书docx合成支持章节markdown图片行——下载invoke_model产物真图add_picture嵌入(居中+图注),下载失败保留文字占位诚实降级(2026-09-07用户需求:章节配图用文生图真图非字符画)

pipeline_bidding/bid_compose_capability.py

@@ -35,11 +35,57 @@ async def _workspace_dir(sor, project_id):
     return ''
 
 
+def _fetch_image_bytes(url, timeout=30):
+    """下载配图（章节正文里的 markdown 图片 URL → 本地字节流）。
+
+    URL 是 invoke_model 产物经 downloadfile2url 落地的本地持久地址
+    （/idfile 静态路径，上游 24h 时效 URL 已在推理层落地，这里拿到的是
+    平台自身地址）。http(s) 下载；本地绝对路径直读。失败返回 None（诚实
+    降级：docx 里保留文字说明，不假装成功）。
+    """
+    try:
+        u = (url or '').strip()
+        if not u:
+            return None
+        if u.startswith('http://') or u.startswith('https://'):
+            import urllib.request
+            req = urllib.request.Request(u, headers={'User-Agent': 'pipeline-bidding'})
+            with urllib.request.urlopen(req, timeout=timeout) as resp:
+                data = resp.read()
+            return data if data else None
+        if u.startswith('/') and os.path.isfile(u):
+            with open(u, 'rb') as f:
+                return f.read()
+    except Exception as e:
+        logger.warning("compose: 配图下载失败 url=%s: %s", (url or '')[:120], e)
+    return None
+
+
+# markdown 图片行：![alt](url)（章节写作者按角色 prompt 规范嵌入 invoke_model 产物）
+_MD_IMG_RE = None
+
+
+def _parse_md_image(line):
+    """解析一行 markdown 图片语法，返回 (alt, url) 或 None。"""
+    global _MD_IMG_RE
+    if _MD_IMG_RE is None:
+        import re
+        _MD_IMG_RE = re.compile(r'^\s*!\[([^\]]*)\]\(([^)\s]+)\)\s*$')
+    m = _MD_IMG_RE.match(line or '')
+    if not m:
+        return None
+    return m.group(1), m.group(2)
+
+
 def _write_docx(path, title, subtitle, chapters):
-    """用 python-docx 写标书。返回 (ok, page_hint_or_err)。"""
+    """用 python-docx 写标书。返回 (ok, page_hint_or_err)。
+
+    章节正文支持 markdown 图片行 ![alt](url)（写作者经 invoke_model 生成的
+    真实配图）：下载后 add_picture 嵌入；下载失败保留文字占位（诚实降级）。
+    """
     try:
         from docx import Document
-        from docx.shared import Pt
+        from docx.shared import Pt, Inches
         from docx.enum.text import WD_ALIGN_PARAGRAPH
     except ImportError:
         return False, "python-docx 未安装"
@@ -72,12 +118,35 @@ def _write_docx(path, title, subtitle, chapters):
             doc.add_paragraph("%s　%s" % (c.get("chapter_no", ""), c.get("title", "")))
         doc.add_page_break()
 
+        img_ok = img_fail = 0
         for c in chapters:
             doc.add_heading("%s　%s" % (c.get("chapter_no", ""), c.get("title", "")), level=1)
             for line in (c.get("content") or "").split("\n"):
                 t = line.rstrip()
                 if not t:
                     continue
+                img = _parse_md_image(t)
+                if img:
+                    alt, url = img
+                    data = _fetch_image_bytes(url)
+                    if data:
+                        import io
+                        pic_p = doc.add_paragraph()
+                        pic_p.alignment = WD_ALIGN_PARAGRAPH.CENTER
+                        try:
+                            pic_p.add_run().add_picture(io.BytesIO(data),
+                                                        width=Inches(5.5))
+                            if alt:
+                                cap = doc.add_paragraph(alt)
+                                cap.alignment = WD_ALIGN_PARAGRAPH.CENTER
+                            img_ok += 1
+                        except Exception as e:
+                            doc.add_paragraph("［配图 %s 嵌入失败：%s］" % (alt or url[:60], str(e)[:80]))
+                            img_fail += 1
+                    else:
+                        doc.add_paragraph("［配图缺失（下载失败）：%s %s］" % (alt or "", url[:80]))
+                        img_fail += 1
+                    continue
                 if t.startswith("### "):
                     doc.add_heading(t[4:], level=3)
                 elif t.startswith("## "):
@@ -90,6 +159,8 @@ def _write_docx(path, title, subtitle, chapters):
                     doc.add_paragraph(t)
             doc.add_page_break()
         doc.save(path)
+        if img_ok or img_fail:
+            logger.info("compose: docx 配图嵌入 成功%d 失败%d", img_ok, img_fail)
         return True, ""
     except Exception as e:
         return False, "%s: %s" % (type(e).__name__, str(e)[:200])