1.0.0.7 - Matching improves, added library page. Removed schedule scans support

server/core/file_scanner.py

@@ -24,7 +24,7 @@ logger.addHandler(handler)
 
 import sys
 sys.path.insert(0, str(Path(__file__).parent))
-from subtitle_processor import parse_srt
+from subtitle_processor import parse_srt, SUBLOGUE_SENTINEL, SUBLOGUE_TOKEN_PATTERN
 
 
 class FileScanner:
@@ -114,7 +114,8 @@ class FileScanner:
             # --------------------------------------------
 
             try:
-                has_plot = cls._check_has_plot(file_path)
+                plot_marker_count = cls._count_plot_markers(file_path)
+                has_plot = plot_marker_count > 0
                 logger.debug(
                     "Plot check for %s: %s",
                     file_path.name,
@@ -148,6 +149,8 @@ class FileScanner:
                 "path": str(file_path),
                 "name": file_path.name,
                 "has_plot": has_plot,
+                "plot_marker_count": plot_marker_count,
+                "duplicate_plot": plot_marker_count > 1,
                 "status": "Has Plot" if has_plot else "Not Loaded",
                 "summary": metadata.get("summary", ""),
                 "plot": metadata.get("summary", ""),
@@ -216,30 +219,25 @@ class FileScanner:
                 )
 
     @classmethod
-    def _check_has_plot(cls, file_path: Path) -> bool:
+    def _count_plot_markers(cls, file_path: Path) -> int:
         """
-        Check first N lines for Sublogue signature.
+        Count Sublogue plot markers to detect duplicates.
         """
-        logger.debug("Scanning for plot marker in %s", file_path.name)
+        logger.debug("Scanning for plot markers in %s", file_path.name)
 
         try:
-            with file_path.open("r", encoding="utf-8", errors="ignore") as f:
-                for i, line in enumerate(f):
-                    if i >= cls.PLOT_SCAN_LINES:
-                        break
-                    if "generated by sublogue" in line.lower():
-                        logger.debug(
-                            "Plot marker found in %s (line %d)",
-                            file_path.name, i + 1
-                        )
-                        return True
+            content = file_path.read_text(encoding="utf-8", errors="ignore")
+            lower_content = content.lower()
+            generated_count = lower_content.count("generated by sublogue")
+            if generated_count > 0:
+                return generated_count
+            return content.count(SUBLOGUE_SENTINEL)
         except Exception as e:
             logger.error(
                 "Error reading file during plot scan: %s (%s)",
                 file_path, e
             )
-
-        return False
+            return 0
 
     @classmethod
     def _extract_metadata(cls, file_path: Path) -> Dict:
@@ -270,6 +268,7 @@ class FileScanner:
 
         plot_text = blocks[1].text
         plot_text = plot_text.split("Generated by Sublogue")[0].strip()
+        plot_text = SUBLOGUE_TOKEN_PATTERN.sub("", plot_text).strip()
         metadata["summary"] = plot_text
 
         # --------------------------------------------
@@ -279,7 +278,9 @@ class FileScanner:
         header_lines = blocks[0].text.split("\n")
 
         if header_lines:
-            first_line = header_lines[0]
+            first_line = header_lines[0].strip()
+            if first_line == SUBLOGUE_SENTINEL and len(header_lines) > 1:
+                first_line = header_lines[1].strip()
             year_match = re.search(r"\((\d{4})\)", first_line)
             if year_match:
                 metadata["year"] = year_match.group(1)