Merge pull request #450 from g0ldyy/feat/rtn-filter-logs

feat: add `RTN_FILTER_DEBUG` setting to enable verbose logging for torrent filtering rejections

.env-sample

@@ -244,6 +244,7 @@ TORRENT_DISABLED_STREAM_URL=https://comet.looks.legal # Optional URL included in
 # Content Filtering              #
 # ============================== #
 REMOVE_ADULT_CONTENT=False
+RTN_FILTER_DEBUG=False # Set to True to log why torrents are excluded by RTN (debug only, verbose!)
 DIGITAL_RELEASE_FILTER=False # Filter unreleased content
 TMDB_READ_ACCESS_TOKEN= # Optional: Provide your own TMDB Read Access Token to avoid using the shared default key
 

comet/core/logger.py

@@ -403,6 +403,7 @@ def log_startup_info(settings):
     )
 
     logger.log("COMET", f"Remove Adult Content: {bool(settings.REMOVE_ADULT_CONTENT)}")
+    logger.log("COMET", f"RTN Filter Debug: {bool(settings.RTN_FILTER_DEBUG)}")
     logger.log(
         "COMET", f"Digital Release Filter: {bool(settings.DIGITAL_RELEASE_FILTER)}"
     )

comet/core/models.py

@@ -133,6 +133,7 @@ class AppSettings(BaseSettings):
     PROXY_ETHOS: Optional[str] = "always"
     RATELIMIT_MAX_RETRIES: Optional[int] = 3
     RATELIMIT_RETRY_BASE_DELAY: Optional[float] = 1.0
+    RTN_FILTER_DEBUG: Optional[bool] = False
 
     @field_validator("INDEXER_MANAGER_TYPE")
     def set_indexer_manager_type(cls, v, values):

comet/scrapers/comet.py

@@ -22,7 +22,7 @@ class CometScraper(BaseScraper):
 
                 try:
                     title = title_full.split("\n")[0].split("📄 ")[1]
-                except:
+                except Exception:
                     continue
                 seeders = (
                     int(title_full.split("👤 ")[1].split(" ")[0])

comet/services/filtering.py

@@ -1,30 +1,53 @@
 from RTN import parse, title_match
 
+from comet.core.logger import logger
+from comet.core.models import settings
+
+if settings.RTN_FILTER_DEBUG:
+
+    def _log_exclusion(msg):
+        logger.log("FILTER", msg)
+else:
+
+    def _log_exclusion(msg):
+        pass
+
 
 def filter_worker(torrents, title, year, year_end, aliases, remove_adult_content):
-    # todo: log reason for each filtered torrent
     results = []
+
     for torrent in torrents:
         torrent_title = torrent["title"]
         if "sample" in torrent_title.lower() or torrent_title == "":
+            _log_exclusion(f"🚫 Rejected (Sample/Empty) | {torrent_title}")
             continue
 
         parsed = parse(torrent_title)
 
         if remove_adult_content and parsed.adult:
+            _log_exclusion(f"🔞 Rejected (Adult) | {torrent_title}")
             continue
 
         if not parsed.parsed_title or not title_match(
             title, parsed.parsed_title, aliases=aliases
         ):
+            _log_exclusion(
+                f"❌ Rejected (Title Mismatch) | {torrent_title} | Parsed: {parsed.parsed_title} | Expected: {title}"
+            )
             continue
 
         if year and parsed.year:
             if year_end is not None:
                 if not (year <= parsed.year <= year_end):
+                    _log_exclusion(
+                        f"📅 Rejected (Year Mismatch) | {torrent_title} | Year: {parsed.year} | Expected: {year}-{year_end}"
+                    )
                     continue
             else:
                 if year < (parsed.year - 1) or year > (parsed.year + 1):
+                    _log_exclusion(
+                        f"📅 Rejected (Year Mismatch) | {torrent_title} | Year: {parsed.year} | Expected: ~{year}"
+                    )
                     continue
 
         torrent["parsed"] = parsed