thread fix

backend/app/services/reply_extractor.py

@@ -4,19 +4,13 @@ from html import unescape
 # Patterns that mark the start of quoted previous-message content.
 TEXT_QUOTE_MARKERS = [
     re.compile(r"\n-{2,}\s*Original Message\s*-{2,}", re.IGNORECASE),
-    re.compile(
-        r"\nOn .+? wrote:\s*\n",
-        re.IGNORECASE,
-    ),
-    re.compile(
-        r"\nOn .+?, .+? at .+? .+? wrote:\s*\n",
-        re.IGNORECASE,
-    ),
+    re.compile(r"(?:^|\n)\s*On .+? wrote:\s*", re.IGNORECASE),
     re.compile(
         r"\nFrom:\s*.+\nSent:\s*.+\nTo:\s*.+\nSubject:\s*.+",
         re.IGNORECASE,
     ),
     re.compile(r"\n_{5,}\n"),
+    re.compile(r"\n\s*> ?"),
 ]
 
 HTML_QUOTE_PATTERNS = [
@@ -110,6 +104,46 @@ def extract_new_reply_content(
     }
 
 
+def reply_fields_for_message(
+    message: dict,
+    *,
+    body_text: str | None,
+    body_html: str | None = None,
+) -> dict[str, str | bool]:
+    if "new_reply_content" in message:
+        return {
+            "new_reply_content": message.get("new_reply_content", ""),
+            "quoted_content": message.get("quoted_content", ""),
+            "has_quoted_text": bool(message.get("has_quoted_text") or message.get("quoted_content")),
+        }
+    built = build_reply_display(body_text=body_text, body_html=body_html)
+    return {
+        "new_reply_content": built["new_reply_content"],
+        "quoted_content": built["quoted_content"],
+        "has_quoted_text": built["has_quoted_text"],
+    }
+
+
+def build_reply_display(
+    *,
+    body_text: str | None,
+    body_html: str | None = None,
+    limit: int = 240,
+) -> dict[str, str | bool]:
+    extracted = extract_new_reply_content(body_text=body_text, body_html=body_html)
+    new_content = extracted["new_reply_content"]
+    quoted = extracted["quoted_content"]
+    fallback = re.sub(r"\s+", " ", (body_text or "")).strip()
+    snippet_source = new_content or fallback
+    return {
+        "new_reply_content": new_content,
+        "quoted_content": quoted,
+        "has_quoted_text": bool(quoted),
+        "raw_body": extracted["raw_body"],
+        "snippet": re.sub(r"\s+", " ", snippet_source).strip()[:limit],
+    }
+
+
 def validate_external_reply_body(body: str) -> bool:
     normalized = (body or "").strip()
     if not normalized: