From 07e2f733b97dbd4d213365e94e83e68c23479d95 Mon Sep 17 00:00:00 2001
From: Ethanfel <ethan.fel@ts-pc.fr>
Date: Sat, 9 May 2026 13:47:48 +0200
Subject: [PATCH] feat: bulk update source paths in train dialog

Add ProcessedDB.update_source_paths() to re-resolve missing or stale
source_path entries by matching filenames against a directory listing
and the current playlist. Exposed as "Update paths" button in the
train dialog next to the video dir field.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
---
 core/db.py | 43 +++++++++++++++++++++++++++++++++++++++++++
 main.py    | 17 +++++++++++++++++
 2 files changed, 60 insertions(+)

diff --git a/core/db.py b/core/db.py
index 74b6477..628f48f 100644
--- a/core/db.py
+++ b/core/db.py
@@ -263,6 +263,49 @@ class ProcessedDB:
             )
             self._con.commit()
 
+    def update_source_paths(self, new_dir: str,
+                            playlist_paths: list[str] | None = None,
+                            profile: str = "") -> int:
+        """Re-resolve source_path for all rows whose current path is missing.
+
+        Checks *new_dir* and *playlist_paths* by filename match.
+        Returns the number of rows updated.
+        """
+        if not self._enabled:
+            return 0
+        lookup: dict[str, str] = {}
+        if playlist_paths:
+            for p in playlist_paths:
+                lookup[os.path.basename(p)] = p
+        if new_dir and os.path.isdir(new_dir):
+            for f in os.listdir(new_dir):
+                fp = os.path.join(new_dir, f)
+                if os.path.isfile(fp):
+                    lookup[f] = fp
+        if not lookup:
+            return 0
+        query = "SELECT DISTINCT filename, source_path FROM processed"
+        params: tuple = ()
+        if profile:
+            query += " WHERE profile = ?"
+            params = (profile,)
+        rows = self._con.execute(query, params).fetchall()
+        updated = 0
+        with self._lock:
+            for fn, sp in rows:
+                if sp and os.path.exists(sp):
+                    continue
+                new_path = lookup.get(fn)
+                if new_path and os.path.isfile(new_path):
+                    self._con.execute(
+                        "UPDATE processed SET source_path = ? WHERE filename = ?",
+                        (new_path, fn),
+                    )
+                    updated += 1
+            if updated:
+                self._con.commit()
+        return updated
+
     def get_labels(self) -> list[str]:
         """Return distinct non-empty labels ordered by most recently used."""
         if not self._enabled:
diff --git a/main.py b/main.py
index be9f9b5..d4036bb 100755
--- a/main.py
+++ b/main.py
@@ -517,6 +517,12 @@ class TrainDialog(QDialog):
         btn_browse.setFixedWidth(30)
         btn_browse.clicked.connect(self._browse_video_dir)
         vid_row.addWidget(btn_browse)
+        self._btn_update_paths = QPushButton("Update paths")
+        self._btn_update_paths.setToolTip(
+            "Re-resolve missing source_path entries using the video dir and playlist")
+        self._btn_update_paths.setFixedWidth(90)
+        self._btn_update_paths.clicked.connect(self._update_source_paths)
+        vid_row.addWidget(self._btn_update_paths)
         self._lbl_video_dir = QLabel("Video dir:")
         self._video_dir_widget = QWidget()
         self._video_dir_widget.setLayout(vid_row)
@@ -558,6 +564,17 @@ class TrainDialog(QDialog):
         if d:
             self._txt_video_dir.setText(d)
 
+    def _update_source_paths(self):
+        video_dir = self._txt_video_dir.text()
+        n = self._db.update_source_paths(
+            video_dir, playlist_paths=self._playlist_paths,
+            profile=self._profile)
+        if n:
+            self._lbl_stats.setText(f"Updated {n} source path(s)")
+            self._debounce.start()
+        else:
+            self._lbl_stats.setText("No paths to update (all resolved or no matches)")
+
     def _manage_negatives(self):
         dlg = HardNegativesDialog(self._db, self._profile, parent=self)
         dlg.exec()