Fix text similarity searches

Twixes · Nov 13, 2024 · f78b0cd · f78b0cd
1 parent 2bffa9f
commit f78b0cd
Showing 1 changed file with 4 additions and 4 deletions.
diff --git a/plugins/imaging.py b/plugins/imaging.py
@@ -254,10 +254,10 @@ async def robot9000(self, ctx: commands.Context, *, text_query: Optional[str] =
                 with data.session() as session:
                     for image9000, textual_similarity in (
                         session.query(Image9000)
-                        .add_column(Image9000.text.op('<%')(text_query).label("textual_similarity"))
+                        .add_column(func.word_similarity(text_query, Image9000.text).label("textual_similarity"))
                         .filter(
                             Image9000.server_id == ctx.guild.id,
-                            Image9000.text.op('<%')(text_query)
+                            Image9000.text.op('%>')(text_query)
                         )
                         .order_by(desc("textual_similarity"))
                         .limit(20)
@@ -309,7 +309,7 @@ async def robot9000(self, ctx: commands.Context, *, text_query: Optional[str] =
 
                         if base_image9000.text and len(base_image9000.text) >= self.IMAGE9000_TEXTUAL_SIMILARITY_MIN_CHARS:
                             textual_similarity_column = func.word_similarity(
-                                Image9000.text, base_image9000.text
+                             base_image9000.text, Image9000.text
                             ).label("textual_similarity")
                             perceptual_matches = (
                                 server_images
@@ -323,7 +323,7 @@ async def robot9000(self, ctx: commands.Context, *, text_query: Optional[str] =
                                 server_images
                                 .add_column(perceptual_distance_column)
                                 .add_column(textual_similarity_column)
-                                .filter(Image9000.text.op('<%')(base_image9000.text))
+                                .filter(Image9000.text.op('%>')(base_image9000.text))
                             )
                             for other_image9000, perceptual_distance, textual_similarity in (
                                 perceptual_matches.union(textual_matches).distinct()