fix: auto-detect image link column when header varies

remote_sync/skills/dingtalk-excel-image-return/scripts/convert_excel_image_links.py

@@ -27,6 +27,15 @@ CONTENT_TYPE_TO_EXT = {
 }
 
 SUPPORTED_EXTENSIONS = {".png", ".jpg", ".jpeg", ".webp", ".bmp", ".gif"}
+IMAGE_HINT_HEADERS = (
+    "图片链接",
+    "商品图片",
+    "主图链接",
+    "图片url",
+    "image_url",
+    "image url",
+    "图片",
+)
 
 
 @dataclass
@@ -109,11 +118,76 @@ def resolve_link_column(sheet, header_row: int, link_column: str) -> int:
         if str(value).strip() == column_spec:
             return col_idx
 
+    guessed = guess_link_column(sheet, header_row)
+    if guessed is not None:
+        return guessed
+
     raise ValueError(
         f"Sheet '{sheet.title}' cannot find link column '{link_column}' in header row {header_row}."
     )
 
 
+def looks_like_http_url(value: str) -> bool:
+    return str(value or "").lower().startswith(("http://", "https://"))
+
+
+def looks_like_image_url(url: str) -> bool:
+    raw = str(url or "").strip().lower()
+    if not looks_like_http_url(raw):
+        return False
+    parsed = urllib.parse.urlparse(raw)
+    path = parsed.path or ""
+    host = parsed.hostname or ""
+    if any(path.endswith(ext) for ext in SUPPORTED_EXTENSIONS):
+        return True
+    if host.startswith("img.") or "image" in host:
+        return True
+    if any(token in path for token in ("/img", "/image", "/upload")):
+        return True
+    return False
+
+
+def guess_link_column(sheet, header_row: int) -> int | None:
+    max_scan_rows = min(sheet.max_row, header_row + 30)
+    best_col = None
+    best_score = -1
+
+    for col_idx in range(1, sheet.max_column + 1):
+        header_val = str(sheet.cell(row=header_row, column=col_idx).value or "").strip().lower()
+        score = 0
+        if header_val:
+            if header_val in IMAGE_HINT_HEADERS:
+                score += 40
+            if "图片" in header_val:
+                score += 20
+            if "链接" in header_val or "url" in header_val:
+                score += 15
+
+        http_count = 0
+        image_like_count = 0
+        for row in range(header_row + 1, max_scan_rows + 1):
+            cell_value = sheet.cell(row=row, column=col_idx).value
+            if cell_value is None:
+                continue
+            text = str(cell_value).strip()
+            if not text:
+                continue
+            if looks_like_http_url(text):
+                http_count += 1
+            if looks_like_image_url(text):
+                image_like_count += 1
+
+        score += http_count * 3 + image_like_count * 6
+        if image_like_count == 0 and http_count == 0 and score < 20:
+            continue
+
+        if score > best_score:
+            best_score = score
+            best_col = col_idx
+
+    return best_col
+
+
 def choose_sheets(workbook, names: Iterable[str]):
     if not names:
         return list(workbook.worksheets)