fix: patch incorrect description type & invalid disambiguation_id refs (#782)

* fix: update DESCRIPTION to the TEXT_BOX type

* fix: remove `disambiguation_id` refs with a tag is deleted

Author: CyanVoxel

tagstudio/src/core/enums.py

@@ -71,4 +71,4 @@ class LibraryPrefs(DefaultEnum):
     IS_EXCLUDE_LIST = True
     EXTENSION_LIST: list[str] = [".json", ".xmp", ".aae"]
     PAGE_SIZE: int = 500
-    DB_VERSION: int = 6
+    DB_VERSION: int = 7

tagstudio/src/core/library/alchemy/fields.py

@@ -113,7 +113,7 @@ class _FieldID(Enum):
     AUTHOR = DefaultField(id=1, name="Author", type=FieldTypeEnum.TEXT_LINE)
     ARTIST = DefaultField(id=2, name="Artist", type=FieldTypeEnum.TEXT_LINE)
     URL = DefaultField(id=3, name="URL", type=FieldTypeEnum.TEXT_LINE)
-    DESCRIPTION = DefaultField(id=4, name="Description", type=FieldTypeEnum.TEXT_LINE)
+    DESCRIPTION = DefaultField(id=4, name="Description", type=FieldTypeEnum.TEXT_BOX)
     NOTES = DefaultField(id=5, name="Notes", type=FieldTypeEnum.TEXT_BOX)
     COLLATION = DefaultField(id=9, name="Collation", type=FieldTypeEnum.TEXT_LINE)
     DATE = DefaultField(id=10, name="Date", type=FieldTypeEnum.DATETIME)

tagstudio/src/core/library/alchemy/library.py

@@ -318,6 +318,7 @@ def open_sqlite_library(self, library_dir: Path, is_new: bool) -> LibraryStatus:
         # https://docs.sqlalchemy.org/en/20/changelog/migration_07.html
         # Under -> sqlite-the-sqlite-dialect-now-uses-nullpool-for-file-based-databases
         poolclass = None if self.storage_path == ":memory:" else NullPool
+        db_version: int = 0
 
         logger.info(
             "[Library] Opening SQLite Library",
@@ -328,11 +329,13 @@ def open_sqlite_library(self, library_dir: Path, is_new: bool) -> LibraryStatus:
         with Session(self.engine) as session:
             # dont check db version when creating new library
             if not is_new:
-                db_version = session.scalar(
+                db_result = session.scalar(
                     select(Preferences).where(Preferences.key == LibraryPrefs.DB_VERSION.name)
                 )
+                if db_result:
+                    db_version = db_result.value  # type: ignore
 
-                if not db_version or db_version.value != LibraryPrefs.DB_VERSION.default:
+                if db_version < 6:  # NOTE: DB_VERSION 6 is the first supported SQL DB version.
                     mismatch_text = Translations.translate_formatted(
                         "status.library_version_mismatch"
                     )
@@ -344,15 +347,14 @@ def open_sqlite_library(self, library_dir: Path, is_new: bool) -> LibraryStatus:
                         success=False,
                         message=(
                             f"{mismatch_text}\n"
-                            f"{found_text} v{0 if not db_version else db_version.value}, "
+                            f"{found_text} v{db_version}, "
                             f"{expected_text} v{LibraryPrefs.DB_VERSION.default}"
                         ),
                     )
 
+            logger.info(f"[Library] DB_VERSION: {db_version}")
             make_tables(self.engine)
 
-            # TODO: Determine a good way of updating built-in data after updates.
-
             # Add default tag color namespaces.
             if is_new:
                 namespaces = default_color_groups.namespaces()
@@ -421,14 +423,52 @@ def open_sqlite_library(self, library_dir: Path, is_new: bool) -> LibraryStatus:
                 )
                 session.add(folder)
                 session.expunge(folder)
-
                 session.commit()
                 self.folder = folder
 
+            # Apply any post-SQL migration patches.
+            if not is_new:
+                # NOTE: DB_VERSION 6 was first used in v9.5.0-pr1
+                if db_version == 6:
+                    self.apply_db6_patches(session)
+                else:
+                    pass
+
+            # Update DB_VERSION
+            self.set_prefs(LibraryPrefs.DB_VERSION, LibraryPrefs.DB_VERSION.default)
+
         # everything is fine, set the library path
         self.library_dir = library_dir
         return LibraryStatus(success=True, library_path=library_dir)
 
+    def apply_db6_patches(self, session: Session):
+        """Apply migration patches to a library with DB_VERSION 6.
+
+        DB_VERSION 6 was first used in v9.5.0-pr1.
+        """
+        logger.info("[Library] Applying patches to DB_VERSION: 6 library...")
+        with session:
+            # Repair "Description" fields with a TEXT_LINE key instead of a TEXT_BOX key.
+            desc_stmd = (
+                update(ValueType)
+                .where(ValueType.key == _FieldID.DESCRIPTION.name)
+                .values(type=FieldTypeEnum.TEXT_BOX.name)
+            )
+            session.execute(desc_stmd)
+            session.flush()
+
+            # Repair tags that may have a disambiguation_id pointing towards a deleted tag.
+            all_tag_ids: set[int] = {tag.id for tag in self.tags}
+            disam_stmt = (
+                update(Tag)
+                .where(Tag.disambiguation_id.not_in(all_tag_ids))
+                .values(disambiguation_id=None)
+            )
+            session.execute(disam_stmt)
+            session.flush()
+
+            session.commit()
+
     @property
     def default_fields(self) -> list[BaseField]:
         with Session(self.engine) as session:
@@ -793,6 +833,14 @@ def remove_tag(self, tag: Tag):
                     session.delete(child_tag)
                     session.expunge(child_tag)
 
+                disam_stmt = (
+                    update(Tag)
+                    .where(Tag.disambiguation_id == tag.id)
+                    .values(disambiguation_id=None)
+                )
+                session.execute(disam_stmt)
+                session.flush()
+
                 session.delete(tag)
                 session.commit()
                 session.expunge(tag)