match series to tvdb episodes py

adapt serverlibrary

media_editing/match_series_to_tvdb.py

@@ -1,25 +1,33 @@
 #!/usr/bin/env python3
 
-import re
-import json
 import argparse
+import json
 import logging
+import re
+import sys
 from pathlib import Path
-from typing import Any
 
 weight_jaccard = 0.5
 weight_fuzzy = 1 - weight_jaccard
 
-from serverlibrary import (words, word_similarity, score_match_jaccard, score_match_fuzzy,
+from serverlibrary import (
-                           combined_score, tmdb_get, tvdb_get)
+    combined_score,
+    score_match_fuzzy,
+    score_match_jaccard,
+    score_match_recall_jaccard,
+    tvdb_get,
+    words,
+)
 
 MATCH_THRESHOLD = 0.85
 
+
 class CustomLogger(logging.Logger):
     def trace(self, msg: str, *args, **kwargs) -> None:
         if self.isEnabledFor(TRACE):
             self._log(TRACE, msg, args, **kwargs)
 
+
 TRACE = 5
 
 logging.setLoggerClass(CustomLogger)
@@ -34,12 +42,14 @@ def normalize(text: str) -> str:
         r"\b(1080p|720p|2160p|x264|x265|h264|h265|bluray|web-dl|webrip|dvdrip)\b",
         "",
         text,
-        flags=re.I,
+        flags=re.IGNORECASE,
     )
 
     text = re.sub(r"[._-]", " ", text)
     text = re.sub(r"ß", "ss", text)
     text = re.sub(r"\s+", " ", text)
+    text = re.sub(r"\((\d+)\s+\d+\)", r"(\1)", text)
+    text = re.sub(r"\((\d+)_+\d+\)", r"(\1)", text)
 
     return text.strip()
 
@@ -70,6 +80,7 @@ def get_all_episodes_by_tvdbid(tvdb_id: int):
 
     return episodes
 
+
 def parse_series_name(series_name: str):
     """
     Universe (1987) [66757]
@@ -97,7 +108,8 @@ def parse_series_name(series_name: str):
 
     return title, year, tmdb_id
 
-def get_tvdbid_by_name(series_name: str) -> tuple[int, Any]:
+
+def get_tvdbid_by_name(series_name: str):
     title, year, tmdb_id = parse_series_name(series_name)
 
     params = {
@@ -133,20 +145,19 @@ def get_tvdbid_by_name(series_name: str) -> tuple[int, Any]:
         if filtered:
             results = filtered
 
-    return int(results[0]["tvdb_id"]), results[0]
+    return results[0]
+
 
 def best_episode_match(filename, episodes):
-    search_words = words(filename)
+    search_words = words(normalize(filename))
-
+    logger.debug(f"Search Words: {search_words}")
     best_score = 0.0
     best_episode = None
 
     for episode in episodes:
-        logger.trace(f" Title: {episode["title"]}")
         if not episode["title"] or episode["title"] == "Unkown":
             continue
-        candidate_words = words(episode["title"])
+        candidate_words = words(normalize(episode["title"]))
-
         fuzzy = score_match_fuzzy(
             search_words,
             candidate_words,
@@ -155,6 +166,14 @@ def best_episode_match(filename, episodes):
             search_words,
             candidate_words,
         )
+        recall_jaccard1 = score_match_recall_jaccard(
+            search_words,
+            candidate_words,
+        )
+        recall_jaccard2 = score_match_recall_jaccard(
+            candidate_words,
+            search_words,
+        )
 
         score = combined_score(
             fuzzy,
@@ -162,8 +181,15 @@ def best_episode_match(filename, episodes):
             weight_fuzzy=weight_fuzzy,
             weight_jaccard=weight_jaccard,
         )
+        score = max(score, recall_jaccard1, recall_jaccard2)
+
+        if "hyänen" in search_words and score > 0.6:
+            logger.warning(
+                f"SEARCH WORDS: {search_words}, CANDIDATE WORDS: {candidate_words}"
+            )
 
         if score > best_score:
+            logger.trace(f"{search_words}:\t\t{candidate_words},\t\t{score}")
             best_score = score
             best_episode = episode
 
@@ -173,45 +199,55 @@ def best_episode_match(filename, episodes):
 def process_series(series_path: Path, dry_run: bool = False):
     series_name = series_path.name
     logger.debug(f"Pricess series with name : {series_name}")
-    #tvdb_id, media_data = get_tvdbid_by_name(series_name)
+    if Path(f"{series_name}_episodes.json").is_file():
-    # episodes = get_all_episodes_by_tvdbid(tvdb_id)
+        logger.info("load json files")
-    #with open("media_data.json", "w", encoding="utf-8") as f:
+        with open(f"{series_name}_media_data.json", encoding="utf-8") as f:
-    #    json.dump(media_data, f, ensure_ascii=False, indent=2)
-
-    with open("media_data.json", encoding="utf-8") as f:
             media_data = json.load(f)
-    with open("episodes.json", encoding="utf-8") as f:
+        with open(f"{series_name}_episodes.json", encoding="utf-8") as f:
             episodes = json.load(f)
-
+        tvdb_id = int(media_data["tvdb_id"])
-
+        for ep in episodes:
-    logger.info(f"Found series name {media_data["name"]}")
+            logger.debug(f"{ep.get('title')}")
+    else:
+        logger.info("QUery from TVDB")
+        media_data = get_tvdbid_by_name(series_name)
+        tvdb_id = int(media_data["tvdb_id"])
+        episodes = get_all_episodes_by_tvdbid(tvdb_id)
+        logger.info("Dump to json")
+        with open(f"{series_name}_media_data.json", "w", encoding="utf-8") as f:
+            json.dump(media_data, f, ensure_ascii=False, indent=2)
+        with open(f"{series_name}_episodes.json", "w", encoding="utf-8") as f:
+            json.dump(episodes, f, ensure_ascii=False, indent=2)
+    logger.info(f"Found series name {media_data['name']}")
     extra_episodes_path = series_path / "Extra Episodes"
 
     for file in extra_episodes_path.iterdir():
         if not file.is_file():
             continue
 
-        filename = normalize(re.sub( media_data["name"],"", file.stem))
+        filename = re.sub(media_data["name"], "", file.stem)
+
         episode, score = best_episode_match(
             filename,
             episodes,
         )
-
         if episode is None:
             continue
 
         if score < MATCH_THRESHOLD:
-            logger.warning(f"LOW CONFIDENCE: {filename} {score:.2f}: {episode.get("title")}")
+            logger.warning(
+                f"LOW CONFIDENCE: {filename} {score:.2f}: {normalize(episode.get('title'))}"
+            )
             continue
 
         safe_title = re.sub(
             r'[<>:"/\\|?*]',
-            '',
+            "",
             episode["title"],
         )
 
         new_name = (
-            f"{media_data.get("name")}"
+            f"{media_data.get('name')}"
             f" - S{episode['season']:02d}"
             f"E{episode['episode']:02d}"
             f" - {safe_title}"
@@ -219,12 +255,10 @@ def process_series(series_path: Path, dry_run: bool = False):
         )
 
         target = file.with_name(new_name)
-
+        if score < 1:
-        logger.info(
+            logger.debug(f"{file.name} -> {new_name},  score={score:.3f}")
-            f"{file.name}"
+        elif score == 1:
-            f" -> {new_name}, "
+            logger.trace(f"{file.name} -> {new_name},  score={score:.3f}")
-            f" score={score:.3f}"
-        )
 
         if not dry_run:
             file.rename(target)
@@ -234,7 +268,7 @@ def main():
     parser = argparse.ArgumentParser()
     parser.add_argument(
         "path",
-        help='Either a series folder "Maya the Bee (1975) [10620]" or a root folder containing such folders'
+        help='Either a series folder "Maya the Bee (1975) [10620]" or a root folder containing such folders',
     )
 
     parser.add_argument(
@@ -268,7 +302,7 @@ def main():
         format="%(asctime)s %(levelname)s %(message)s",
     )
 
-    input_path = Path(args.path)
+    input_path = Path(args.path).expanduser().resolve()
 
     if not input_path.is_dir():
         parser.error(f"Directory not found: {input_path}")

serverlibrary/__init__.py

@@ -1,11 +1,14 @@
-from .core import safe_rename
+from .core import (
-from .core import safe_delete
+    combined_score,
-from .core import get_audio_tracks
+    get_audio_tracks,
-from .core import get_subtitle_tracks
+    get_subtitle_tracks,
-from .core import words
+    safe_delete,
-from .core import word_similarity
+    safe_rename,
-from .core import score_match_jaccard
+    score_match_fuzzy,
-from .core import score_match_fuzzy
+    score_match_jaccard,
-from .core import combined_score
+    score_match_recall_jaccard,
-from .core import tmdb_get
+    tmdb_get,
-from .core import tvdb_get
+    tvdb_get,
+    word_similarity,
+    words,
+)

serverlibrary/core.py

@@ -1,11 +1,11 @@
-
-from pathlib import Path
-import shutil
-import subprocess
 import json
 import re
-import requests
+import shutil
+import subprocess
 from difflib import SequenceMatcher
+from pathlib import Path
+
+import requests
 
 
 # -----------------------------
@@ -17,8 +17,15 @@ def safe_delete(path: Path):
     else:
         path.unlink()
 
-def safe_rename(src: Path, dst: Path, dry: bool=False, confirm: bool=False,
+
-                silent=False,overwrite=False) -> (str, bool):
+def safe_rename(
+    src: Path,
+    dst: Path,
+    dry: bool = False,
+    confirm: bool = False,
+    silent=False,
+    overwrite=False,
+) -> tuple[str, bool]:
     """
     Safely rename a file or directory without overwriting existing files.
     Automatically appends _1, _2, ... if needed.
@@ -29,7 +36,7 @@ def safe_rename(src: Path, dst: Path, dry: bool=False, confirm: bool=False,
 
     # nothing to do
     if src.resolve() == dst.resolve():
-        return
+        return "", False
 
     directory = dst.parent
     base = dst.stem
@@ -60,7 +67,10 @@ def safe_rename(src: Path, dst: Path, dry: bool=False, confirm: bool=False,
     src.rename(dst)
     return ret, True
 
-def get_audio_tracks( source: str, title: int | None = None, handbrake: bool = False) -> list[dict]:
+
+def get_audio_tracks(
+    source: str, title: int | None = None, handbrake: bool = False
+) -> list[dict]:
     """
     Return audio track information from a media file or DVD title.
 
@@ -103,8 +113,10 @@ def get_audio_tracks( source: str, title: int | None = None, handbrake: bool = F
         result = subprocess.run(
             [
                 "HandBrakeCLI",
-                "-i", source,
+                "-i",
-                "-t", str(title),
+                source,
+                "-t",
+                str(title),
                 "--scan",
             ],
             stdout=subprocess.PIPE,
@@ -113,22 +125,24 @@ def get_audio_tracks( source: str, title: int | None = None, handbrake: bool = F
             check=True,
         )
 
-        pattern = re.compile(
+        pattern = re.compile(r"\+\s+(\d+),\s+(.+?)\s+\((?:AC3|AAC|MP3|E-AC-3|DTS)")
-            r"\+\s+(\d+),\s+(.+?)\s+\((?:AC3|AAC|MP3|E-AC-3|DTS)"
-        )
 
         for match in pattern.finditer(result.stdout):
-            tracks.append({
+            tracks.append(
+                {
                     "number": int(match.group(1)),
                     "description": match.group(2).strip(),
-            })
+                }
+            )
 
     else:
         result = subprocess.run(
             [
                 "ffprobe",
-                "-v", "quiet",
+                "-v",
-                "-print_format", "json",
+                "quiet",
+                "-print_format",
+                "json",
                 "-show_streams",
                 source,
             ],
@@ -143,31 +157,33 @@ def get_audio_tracks( source: str, title: int | None = None, handbrake: bool = F
             if stream.get("codec_type") != "audio":
                 continue
 
-            tracks.append({
+            tracks.append(
+                {
                     "number": stream["index"],
                     "description": (
                         stream.get("tags", {}).get("title")
                         or stream.get("tags", {}).get("language")
                         or f"Track {stream['index']}"
                     ),
-                "primary": stream.get("codec_type") == "audio" and stream.get("disposition", {}).get("default") == 1
+                    "primary": stream.get("codec_type") == "audio"
-            })
+                    and stream.get("disposition", {}).get("default") == 1,
+                }
+            )
 
     if not tracks:
-        tracks.append({
+        tracks.append({"number": 0, "description": "", "primary": False})
-            "number": 0,
-            "description": "",
-            "primary": False
-        })
 
     return tracks
 
+
 def get_subtitle_tracks(source):
     result = subprocess.run(
         [
             "ffprobe",
-            "-v", "quiet",
+            "-v",
-            "-print_format", "json",
+            "quiet",
+            "-print_format",
+            "json",
             "-show_streams",
             source,
         ],
@@ -184,36 +200,48 @@ def get_subtitle_tracks(source):
         if stream.get("codec_type") != "subtitle":
             continue
 
-        tracks.append({
+        tracks.append(
+            {
                 "number": stream["index"],
                 "description": (
                     stream.get("tags", {}).get("title")
                     or stream.get("tags", {}).get("language")
                     or f"Subtitle {stream['index']}"
                 ),
-            "primary": (
+                "primary": (stream.get("disposition", {}).get("default") == 1),
-                stream.get("disposition", {}).get("default") == 1
+            }
-            ),
+        )
-        })
 
     if not tracks:
-        tracks.append({
+        tracks.append(
+            {
                 "number": 0,
                 "description": "",
                 "primary": False,
-        })
+            }
+        )
 
     return tracks
 
 
 def words(text):
-    return set(filter(None, re.split(r"[^a-z0-9]+", text.lower())))
+    return set(re.findall(r"[^\W_]+", text.lower()))
 
 
 def word_similarity(a: str, b: str) -> float:
     return SequenceMatcher(None, a, b).ratio()
 
 
+def score_match_recall_jaccard(
+    search_words: set[str], candidate_words: set[str]
+) -> float:
+    if not search_words:
+        return 0.0
+
+    intersection = search_words & candidate_words
+    return len(intersection) / len(search_words)
+
+
 def score_match_jaccard(search_words, candidate_words):
     intersection = search_words & candidate_words
     union = search_words | candidate_words
@@ -239,8 +267,9 @@ def score_match_fuzzy(search_words, candidate_words):
     return total / len(search_words)
 
 
-def combined_score(score_fuzzy: float, score_jaccard: float,
+def combined_score(
-                   weight_fuzzy: float, weight_jaccard: float) -> float:
+    score_fuzzy: float, score_jaccard: float, weight_fuzzy: float, weight_jaccard: float
+) -> float:
     return score_fuzzy * weight_fuzzy + score_jaccard * weight_jaccard
 
 
@@ -278,6 +307,7 @@ def tmdb_get(endpoint, params=None):
 
     return response.json()
 
+
 def tvdb_get(endpoint, params=None):
     """Perform a GET request against TVDB."""
     BASE_URL = "https://api4.thetvdb.com/v4"