Bump version to 0.8.2; enhance image parameter extraction

kapitanbooru_uploader/ImageBrowser.py

@@ -32,7 +32,7 @@ class ImageBrowser(tk.Tk):
         super().__init__()
         self.title("Kapitanbooru Uploader")
         self.geometry("900x600")
-        self.version = "0.8.1"
+        self.version = "0.8.2"
         self.acknowledged_version = parse_version(self.version)
 
         self.settings = Settings()

kapitanbooru_uploader/locales/en/LC_MESSAGES/messages.po

@@ -1,6 +1,6 @@
 msgid ""
 msgstr ""
-"Project-Id-Version: Kapitanbooru Uploader 0.8.1\n"
+"Project-Id-Version: Kapitanbooru Uploader 0.8.2\n"
 "Report-Msgid-Bugs-To: kapitan@mlesniak.pl\n"
 "POT-Creation-Date: 2025-03-27 20:47+0100\n"
 "Content-Type: text/plain; charset=UTF-8\n"

kapitanbooru_uploader/locales/pl/LC_MESSAGES/messages.po

@@ -1,6 +1,6 @@
 msgid ""
 msgstr ""
-"Project-Id-Version: Kapitanbooru Uploader 0.8.1\n"
+"Project-Id-Version: Kapitanbooru Uploader 0.8.2\n"
 "Report-Msgid-Bugs-To: kapitan@mlesniak.pl\n"
 "POT-Creation-Date: 2025-03-27 20:47+0100\n"
 "Language: pl\n"

kapitanbooru_uploader/tag_processing.py

@@ -52,7 +52,7 @@ def get_artist_tags(tags_repo: TagsRepo):
 
 # Wzorce i ustawienia związane z tagami
 PARAMETERS_PATTERN = re.compile(
-    r"^(?P<prompt>.*?)\s+(?:Negative prompt:\s+(?P<negative_prompt>.*?)\s+)?Steps:\s+(?P<steps>\d+),\s+Sampler:\s+(?P<sampler>.*?),\s+Schedule type:\s+(?P<schedule_type>.*?),\s+CFG scale:\s+(?P<cfg_scale>.*?),\s+Seed:\s+(?P<seed>.*?),\s+Size:\s+(?P<size>.*?),\s+Model hash:\s+(?P<model_hash>.*?),\s+Model:\s+(?P<model>.*?),\s+(?:Denoising strength:\s+(?P<denoising_strength>.*?),\s+)?(?:Clip skip: (?P<clip_skip>.*?),\s+)?(?:Hires CFG Scale:\s+(?P<hires_cfg_scale>.*?),\s+Hires upscale:\s+(?P<hires_upscale>.*?),\s+Hires steps:\s+(?P<hires_steps>.*?),\s+Hires upscaler:\s+(?P<hires_upscaler>.*?),\s+)?Version:\s+(?P<version>.*?)\s*$",
+    r"^(?P<prompt>.*?),?\s+(?:Negative prompt:\s+(?P<negative_prompt>.*?)\s+)?Steps:\s+(?P<steps>\d+),\s+Sampler:\s+(?P<sampler>.*?),\s+Schedule type:\s+(?P<schedule_type>.*?),\s+CFG scale:\s+(?P<cfg_scale>.*?),\s+Seed:\s+(?P<seed>.*?),\s+Size:\s+(?P<size>.*?),\s+Model hash:\s+(?P<model_hash>.*?),\s+Model:\s+(?P<model>.*?),\s+(?:Denoising strength:\s+(?P<denoising_strength>.*?),\s+)?(?:Clip skip: (?P<clip_skip>.*?),\s+)?(?:Hires CFG Scale:\s+(?P<hires_cfg_scale>.*?),\s+Hires upscale:\s+(?P<hires_upscale>.*?),\s+Hires steps:\s+(?P<hires_steps>.*?),\s+Hires upscaler:\s+(?P<hires_upscaler>.*?),\s+)?(?:Lora hashes:\s+\"(?P<lora_hashes>.*?)\",\s+)?Version:\s+(?P<version>.*?)\s*$",
     re.DOTALL,
 )
 COEFFICIENT_PATTERN = re.compile(r"^.*?(:\d+|\d+\.\d+)$")
@@ -122,25 +122,53 @@ def extract_parameters(img: Image, file_path: str) -> str:
     parameters = ""
     lower_path = file_path.lower()
 
-    # For PNG: use the custom "parameters" field.
+    # Dla PNG: korzystamy z niestandardowego pola "parameters".
     if isinstance(img, PngImagePlugin.PngImageFile):
         parameters = img.info.get("parameters", "")
 
-    # For JPEG, WebP, and AVIF: extract EXIF UserComment (EXIF tag 37510).
+    # Dla JPEG, WebP i AVIF: wyciągamy EXIF UserComment (tag 37510) ze zagnieżdżonego słownika (tag 34665).
     elif lower_path.endswith((".jpg", ".jpeg", ".webp", ".avif")):
         exif_data = img.getexif()
-        user_comment = exif_data.get(37510)
+        exif_ifd = exif_data.get_ifd(34665)
+        user_comment = None
+        if exif_ifd:
+            user_comment = exif_ifd.get(37510)
         if user_comment:
-            # UserComment might be stored as bytes; decode to string.
-            if isinstance(user_comment, bytes):
-                try:
-                    parameters = user_comment.decode("utf-8", errors="replace")
-                except Exception:
-                    parameters = str(user_comment)
+            if isinstance(user_comment, bytes) and len(user_comment) >= 8:
+                header = user_comment[:8]
+                comment_bytes = user_comment[8:]
+                if header.startswith(b'ASCII'):
+                    try:
+                        parameters = comment_bytes.decode("ascii", errors="ignore")
+                    except Exception:
+                        parameters = str(comment_bytes)
+                elif header.startswith(b'UNICODE'):
+                    # Sprawdzenie obecności BOM w danych
+                    if comment_bytes[:2] in (b'\xff\xfe', b'\xfe\xff'):
+                        try:
+                            parameters = comment_bytes.decode("utf-16", errors="ignore")
+                        except UnicodeDecodeError:
+                            parameters = comment_bytes.decode("utf-16-be", errors="ignore")
+                    else:
+                        try:
+                            parameters = comment_bytes.decode("utf-16-be", errors="ignore")
+                        except UnicodeDecodeError:
+                            parameters = comment_bytes.decode("utf-16", errors="ignore")
+                elif header.startswith(b'JIS'):
+                    try:
+                        parameters = comment_bytes.decode("shift_jis", errors="ignore")
+                    except Exception:
+                        parameters = str(comment_bytes)
+                else:
+                    # Fallback: próba dekodowania jako ASCII
+                    try:
+                        parameters = comment_bytes.decode("ascii", errors="ignore")
+                    except Exception:
+                        parameters = str(comment_bytes)
             else:
                 parameters = str(user_comment)
 
-    # For GIF: extract the GIF comment.
+    # Dla GIF: wyciągamy komentarz GIF.
     elif lower_path.endswith(".gif"):
         comment = img.info.get("comment")
         if comment: