refactor(stage-1): eliminate dead parse_text, simplify task progress SQL, remove prompt hacks

app/reader.py

@@ -286,44 +286,6 @@ def split_long_text(text: str, target_words: int = 3000) -> List[str]:
 
     return chunks
 
-
-def parse_text(content: str, title: str = "Untitled", author: str = "") -> Tuple[str, str, List[Dict[str, str]]]:
-    """
-    Parses plain text or Markdown into readable chapters.
-    Splits by explicit chapter markers (e.g. 'Chapter X', '# Chapter') or chunks of ~3000 words.
-    """
-    content = content.replace("\r\n", "\n").replace("\r", "\n").strip()
-    if not content:
-        return title, author, [{"title": title if title != "Untitled" else "Chapter 1", "content": ""}]
-
-    lines = content.split("\n")
-    first_line = lines[0].strip()
-
-    # Extract title from first line or # Heading / <h1...> if default or generic README
-    if title == "Untitled" or title.lower().startswith("readme"):
-        if first_line.startswith("# "):
-            title = first_line.lstrip("# ").strip()
-            if len(lines) > 1:
-                content = "\n".join(lines[1:]).strip()
-        elif first_line.lower().startswith("title:"):
-            title = first_line.split(":", 1)[1].strip()
-            if len(lines) > 1:
-                content = "\n".join(lines[1:]).strip()
-        else:
-            h1_match = re.search(r"<h1[^>]*>(.*?)</h1>", content[:1500], re.IGNORECASE)
-            if h1_match:
-                detected = re.sub(r"<[^>]+>", "", h1_match.group(1)).strip()
-                if detected:
-                    title = detected
-
-    # Extract author from next line if missing
-    if not author:
-        lines = content.split("\n")
-        if lines and lines[0].lower().startswith(("author:", "by:")):
-            author = lines[0].split(":", 1)[1].strip()
-            if len(lines) > 1:
-                content = "\n".join(lines[1:]).strip()
-
 def detect_text_headings(content: str) -> List[Dict]:
     """
     Detects section and chapter headings with high accuracy across Markdown,