GameSentenceMiner 2.10.10__py3-none-any.whl → 2.10.12__py3-none-any.whl

GameSentenceMiner/config_gui.py

@@ -80,6 +80,7 @@ class ConfigApp:
     def __init__(self, root):
         self.window = root
         self.on_exit = None
+        self.window.tk.call('tk', 'scaling', 1.5)  # Set DPI scaling factor
         # self.window = ttk.Window(themename='darkly')
         self.window.title('GameSentenceMiner Configuration')
         self.window.protocol("WM_DELETE_WINDOW", self.hide)

GameSentenceMiner/gsm.py

@@ -452,6 +452,13 @@ def cleanup():
     if icon:
         icon.stop()
 
+    for video in gsm_state.videos_to_remove:
+        try:
+            if os.path.exists(video):
+                os.remove(video)
+        except Exception as e:
+            logger.error(f"Error removing temporary video file {video}: {e}")
+
     settings_window.window.destroy()
     time.sleep(5)
     logger.info("Cleanup complete.")

GameSentenceMiner/ocr/gsm_ocr_config.py

@@ -44,13 +44,13 @@ class WindowGeometry:
 class OCRConfig:
     scene: str
     rectangles: List[Rectangle]
-    pre_scale_rectangles: List[Rectangle] = None
+    pre_scale_rectangles: Optional[List[Rectangle]] = None
     coordinate_system: str = None
     window_geometry: Optional[WindowGeometry] = None
     window: Optional[str] = None
     language: str = "ja"
 
-    def __post_init__(self):
+    def scale_coords(self):
         self.pre_scale_rectangles = deepcopy(self.rectangles)
         if self.coordinate_system and self.coordinate_system == "percentage" and self.window:
             import pygetwindow as gw

GameSentenceMiner/ocr/owocr_area_selector.py

@@ -20,6 +20,7 @@ except ImportError:
 
 try:
     import tkinter as tk
+    from tkinter import font as tkfont  # NEW: Import for better font control
 
     selector_available = True
 except ImportError:
@@ -206,6 +207,66 @@ class ScreenSelector:
         self.drawn_rect_ids.append(new_rect_id)
         print("Redo: Restored rectangle.")
 
+    # --- NEW METHOD TO DISPLAY INSTRUCTIONS ---
+    def _create_instructions_widget(self, canvas):
+        """Creates a text box with usage instructions on the canvas."""
+        instructions_text = (
+            "How to Use:\n"
+            "  • Left Click + Drag: Create a capture area (green).\n"
+            "  • Shift + Left Click + Drag: Create an exclusion area (orange).\n"
+            "  • Right-Click on a box: Delete it.\n\n"
+            "Hotkeys:\n"
+            "  • Ctrl + S: Save and Quit\n"
+            "  • Ctrl + Z / Ctrl + Y: Undo / Redo\n"
+            "  • M: Toggle background visibility\n"
+            "  • I: Toggle these instructions\n"
+            "  • Esc: Quit without saving"
+            "  "
+        )
+
+        # Use a common, readable font
+        instruction_font = tkfont.Font(family="Segoe UI", size=10, weight="normal")
+
+        # Create the text item first to get its size
+        text_id = canvas.create_text(
+            20, 20,  # Position with a small margin
+            text=instructions_text,
+            anchor=tk.NW,
+            fill='white',
+            font=instruction_font,
+            justify=tk.LEFT
+        )
+
+        # Get the bounding box of the text to draw a background
+        text_bbox = canvas.bbox(text_id)
+
+        # Create a background rectangle with padding
+        rect_id = canvas.create_rectangle(
+            text_bbox[0] - 10,  # left
+            text_bbox[1] - 10,  # top
+            text_bbox[2] + 10,  # right
+            text_bbox[3] + 10,  # bottom
+            fill='#2B2B2B',  # Dark, semi-opaque background
+            outline='white',
+            width=1
+        )
+
+        # Lower the rectangle so it's behind the text
+        canvas.tag_lower(rect_id, text_id)
+
+    def toggle_instructions(self, event=None):
+        canvas = event.widget.winfo_toplevel().winfo_children()[0]
+        # Find all text and rectangle items (assuming only one of each for instructions)
+        text_items = [item for item in canvas.find_all() if canvas.type(item) == 'text']
+        rect_items = [item for item in canvas.find_all() if canvas.type(item) == 'rectangle']
+
+        if text_items and rect_items:
+            current_state = canvas.itemcget(text_items[0], 'state')
+            new_state = tk.NORMAL if current_state == tk.HIDDEN else tk.HIDDEN
+            for item in text_items + rect_items:
+                canvas.itemconfigure(item, state=new_state)
+            print("Toggled instructions visibility.")
+
     def start(self):
         self.root = tk.Tk()
         self.root.withdraw()
@@ -230,6 +291,10 @@ class ScreenSelector:
         canvas.pack(fill=tk.BOTH, expand=True)
         canvas.create_image(0, 0, image=self.photo_image, anchor=tk.NW)
 
+        # --- MODIFIED: CALL THE INSTRUCTION WIDGET CREATOR ---
+        self._create_instructions_widget(canvas)
+        # --- END MODIFICATION ---
+
         # Draw existing rectangles (which were converted to absolute pixels on load)
         for _, abs_coords, is_excluded in self.rectangles:
             x_abs, y_abs, w_abs, h_abs = abs_coords
@@ -275,17 +340,37 @@ class ScreenSelector:
             self.current_rect_id = self.start_x = self.start_y = None
 
         def on_right_click(event):
-            items = canvas.find_closest(event.x, event.y)
-            if items and items[0] in self.drawn_rect_ids:
-                item_id = items[0]
-                idx_to_del = self.drawn_rect_ids.index(item_id)
-                del self.drawn_rect_ids[idx_to_del]
-                del self.rectangles[idx_to_del]
-                self.redo_stack.clear()
-                canvas.delete(item_id)
+            # Iterate through our rectangles in reverse to find the topmost one.
+            for i in range(len(self.rectangles) - 1, -1, -1):
+                _monitor, abs_coords, _is_excluded = self.rectangles[i]
+                x_abs, y_abs, w_abs, h_abs = abs_coords
+                canvas_x1 = x_abs - self.bounding_box['left']
+                canvas_y1 = y_abs - self.bounding_box['top']
+                canvas_x2 = canvas_x1 + w_abs
+                canvas_y2 = canvas_y1 + h_abs
+
+                if canvas_x1 <= event.x <= canvas_x2 and canvas_y1 <= event.y <= canvas_y2:
+                    # --- UNDO/REDO CHANGE ---
+                    # We found the rectangle. Prepare the 'remove' action.
+                    # We need to save the data AND its original index to restore it correctly.
+                    rect_tuple_to_del = self.rectangles[i]
+                    item_id_to_del = self.drawn_rect_ids[i]
+
+                    self.redo_stack.append((*rect_tuple_to_del, i))
+
+                    # Now, perform the deletion
+                    del self.rectangles[i]
+                    del self.drawn_rect_ids[i]
+                    canvas.delete(item_id_to_del)
+                    print("Deleted rectangle.")
+
+                    break  # Stop after deleting the topmost one
 
         def toggle_image_mode(e=None):
-            self.image_mode = not self.image_mode; window.attributes("-alpha", 1.0 if self.image_mode else 0.25)
+            self.image_mode = not self.image_mode
+            # Only change alpha of the main window, not the text widget
+            window.attributes("-alpha", 1.0 if self.image_mode else 0.25)
+            print("Toggled background visibility.")
 
         def on_enter(e=None):
             canvas.focus_set()
@@ -296,13 +381,15 @@ class ScreenSelector:
         canvas.bind('<ButtonRelease-1>', on_release)
         canvas.bind('<Button-3>', on_right_click)
         canvas.bind('<Control-s>', self.save_rects)
-        canvas.bind('<Control-z>', self.undo_last_rect)
         canvas.bind('<Control-y>', self.redo_last_rect)
+        canvas.bind('<Control-z>', self.undo_last_rect)
         canvas.bind("<Escape>", self.quit_app)
         canvas.bind("<m>", toggle_image_mode)
+        canvas.bind("<i>", self.toggle_instructions)
 
         canvas.focus_set()
-        print("Starting UI. Press Esc to quit, Ctrl+S to save, M to toggle background.")
+        # The print message is now redundant but kept for console feedback
+        print("Starting UI. See on-screen instructions. Press Esc to quit, Ctrl+S to save.")
         self.root.mainloop()
 
     def quit_app(self, event=None):
@@ -350,10 +437,6 @@ if __name__ == "__main__":
     target_window_title = "Windowed Projector (Preview)"  # Default
     if len(sys.argv) > 1:
         target_window_title = sys.argv[1]
-    # else:
-    #     print("Usage: python your_script_name.py \"Target Window Title\"", file=sys.stderr)
-    #     print("Example: python selector.py \"Windowed Projector (Preview)\"", file=sys.stderr)
-    #     sys.exit(1)
 
     selection_result = get_screen_selection(target_window_title)
 

GameSentenceMiner/ocr/owocr_helper.py

@@ -178,17 +178,28 @@ class WebsocketServerThread(threading.Thread):
         asyncio.run(main())
 
 
+def compare_ocr_results(prev_text, new_text, threshold=90):
+    if not prev_text or not new_text:
+        return False
+    if isinstance(prev_text, list):
+        prev_text = ''.join([item for item in prev_text if item is not None]) if prev_text else ""
+    if isinstance(new_text, list):
+        new_text = ''.join([item for item in new_text if item is not None]) if new_text else ""
+    similarity = fuzz.ratio(prev_text, new_text)
+    return similarity >= threshold
+
 all_cords = None
 rectangles = None
-last_ocr2_result = ""
+last_ocr2_result = []
 
 def do_second_ocr(ocr1_text, time, img, filtering, ignore_furigana_filter=False):
     global twopassocr, ocr2, last_ocr2_result
     try:
         orig_text, text = run.process_and_write_results(img, None, last_ocr2_result, filtering, None,
                                                         engine=ocr2, furigana_filter_sensitivity=furigana_filter_sensitivity if not ignore_furigana_filter else 0)
-        if fuzz.ratio(last_ocr2_result, orig_text) >= 90:
-            logger.info("Seems like the same text from previous ocr2 result, not sending")
+
+        if compare_ocr_results(last_ocr2_result, orig_text):
+            logger.info("Detected similar text from previous OCR2 result, not sending")
             return
         save_result_image(img)
         last_ocr2_result = orig_text
@@ -241,12 +252,12 @@ def text_callback(text, orig_text, time, img=None, came_from_ss=False, filtering
 
     line_start_time = time if time else datetime.now()
 
-    if not twopassocr:
-        if previous_text and fuzz.ratio(orig_text_string, previous_orig_text) >= 90:
+    if manual or not twopassocr:
+        if compare_ocr_results(previous_orig_text, orig_text_string):
             logger.info("Seems like Text we already sent, not doing anything.")
             return
         save_result_image(img)
-        asyncio.run(send_result(text, time))
+        asyncio.run(send_result(text, line_start_time))
         previous_orig_text = orig_text_string
         previous_text = None
         previous_img = None
@@ -260,13 +271,13 @@ def text_callback(text, orig_text, time, img=None, came_from_ss=False, filtering
         if previous_text and text_stable_start_time:
             stable_time = text_stable_start_time
             previous_img_local = previous_img
-            if previous_text and fuzz.ratio(orig_text_string, previous_orig_text) >= 90:
+            if compare_ocr_results(previous_orig_text, orig_text_string):
                 logger.info("Seems like Text we already sent, not doing anything.")
                 previous_text = None
                 return
             previous_orig_text = orig_text_string
             previous_ocr1_result = previous_text
-            if crop_coords:
+            if crop_coords and optimize_second_scan:
                 previous_img_local.save(os.path.join(get_temporary_directory(), "pre_oneocrcrop.png"))
                 previous_img_local = previous_img_local.crop(crop_coords)
             second_ocr_queue.put((previous_text, stable_time, previous_img_local, filtering))
@@ -389,70 +400,88 @@ def set_force_stable_hotkey():
     print("Press Ctrl+Shift+F to toggle force stable mode.")
 
 if __name__ == "__main__":
-    global ocr1, ocr2, twopassocr, language, ss_clipboard, ss, ocr_config, furigana_filter_sensitivity, area_select_ocr_hotkey, window
-    import sys
-
-    import argparse
-
-    parser = argparse.ArgumentParser(description="OCR Configuration")
-    parser.add_argument("--language", type=str, default="ja", help="Language for OCR (default: ja)")
-    parser.add_argument("--ocr1", type=str, default="oneocr", help="Primary OCR engine (default: oneocr)")
-    parser.add_argument("--ocr2", type=str, default="glens", help="Secondary OCR engine (default: glens)")
-    parser.add_argument("--twopassocr", type=int, choices=[0, 1], default=1, help="Enable two-pass OCR (default: 1)")
-    parser.add_argument("--manual", action="store_true", help="Use screenshot-only mode")
-    parser.add_argument("--clipboard", action="store_true", help="Use clipboard for input")
-    parser.add_argument("--clipboard-output", action="store_true", default=False, help="Use clipboard for output")
-    parser.add_argument("--window", type=str, help="Specify the window name for OCR")
-    parser.add_argument("--furigana_filter_sensitivity", type=float, default=0, help="Furigana Filter Sensitivity for OCR (default: 0)")
-    parser.add_argument("--manual_ocr_hotkey", type=str, default=None, help="Hotkey for manual OCR (default: None)")
-    parser.add_argument("--area_select_ocr_hotkey", type=str, default="ctrl+shift+o", help="Hotkey for area selection OCR (default: ctrl+shift+o)")
-
-    args = parser.parse_args()
-
-    language = args.language
-    ocr1 = args.ocr1
-    ocr2 = args.ocr2 if args.ocr2 else None
-    twopassocr = bool(args.twopassocr)
-    manual = args.manual
-    ss_clipboard = args.clipboard
-    window_name = args.window
-    furigana_filter_sensitivity = args.furigana_filter_sensitivity
-    ss_hotkey = args.area_select_ocr_hotkey.lower()
-    manual_ocr_hotkey = args.manual_ocr_hotkey.lower().replace("ctrl", "<ctrl>").replace("shift", "<shift>").replace("alt", "<alt>") if args.manual_ocr_hotkey else None
-    clipboard_output = args.clipboard_output
-
-    logger.info(f"Received arguments: {vars(args)}")
-    # set_force_stable_hotkey()
-    ocr_config: OCRConfig = get_ocr_config(window=window_name)
-    if ocr_config:
-        if ocr_config.window:
-            start_time = time.time()
-            while time.time() - start_time < 30:
-                window = get_window(ocr_config.window)
-                if window or manual:
-                    break
-                logger.info(f"Window: {ocr_config.window} Could not be found, retrying in 1 second...")
-                time.sleep(1)
-            else:
-                logger.error(f"Window '{ocr_config.window}' not found within 30 seconds.")
-                sys.exit(1)
-        logger.info(f"Starting OCR with configuration: Window: {ocr_config.window}, Rectangles: {ocr_config.rectangles}, Engine 1: {ocr1}, Engine 2: {ocr2}, Two-pass OCR: {twopassocr}")
-    set_dpi_awareness()
-    if manual or ocr_config:
-        rectangles = ocr_config.rectangles if ocr_config and ocr_config.rectangles else []
-        oneocr_threads = []
-        ocr_thread = threading.Thread(target=run_oneocr, args=(ocr_config,rectangles ), daemon=True)
-        ocr_thread.start()
-        if not manual:
-            worker_thread = threading.Thread(target=process_task_queue, daemon=True)
-            worker_thread.start()
-        websocket_server_thread = WebsocketServerThread(read=True)
-        websocket_server_thread.start()
-        add_ss_hotkey(ss_hotkey)
-        try:
-            while not done:
-                time.sleep(1)
-        except KeyboardInterrupt as e:
-            pass
-    else:
-        print("Failed to load OCR configuration. Please check the logs.")
+    try:
+        global ocr1, ocr2, twopassocr, language, ss_clipboard, ss, ocr_config, furigana_filter_sensitivity, area_select_ocr_hotkey, window, optimize_second_scan
+        import sys
+
+        import argparse
+
+        parser = argparse.ArgumentParser(description="OCR Configuration")
+        parser.add_argument("--language", type=str, default="ja", help="Language for OCR (default: ja)")
+        parser.add_argument("--ocr1", type=str, default="oneocr", help="Primary OCR engine (default: oneocr)")
+        parser.add_argument("--ocr2", type=str, default="glens", help="Secondary OCR engine (default: glens)")
+        parser.add_argument("--twopassocr", type=int, choices=[0, 1], default=1,
+                            help="Enable two-pass OCR (default: 1)")
+        parser.add_argument("--manual", action="store_true", help="Use screenshot-only mode")
+        parser.add_argument("--clipboard", action="store_true", help="Use clipboard for input")
+        parser.add_argument("--clipboard-output", action="store_true", default=False, help="Use clipboard for output")
+        parser.add_argument("--window", type=str, help="Specify the window name for OCR")
+        parser.add_argument("--furigana_filter_sensitivity", type=float, default=0,
+                            help="Furigana Filter Sensitivity for OCR (default: 0)")
+        parser.add_argument("--manual_ocr_hotkey", type=str, default=None, help="Hotkey for manual OCR (default: None)")
+        parser.add_argument("--area_select_ocr_hotkey", type=str, default="ctrl+shift+o",
+                            help="Hotkey for area selection OCR (default: ctrl+shift+o)")
+        parser.add_argument("--optimize_second_scan", action="store_true",
+                            help="Optimize second scan by cropping based on first scan results")
+
+        args = parser.parse_args()
+
+        language = args.language
+        ocr1 = args.ocr1
+        ocr2 = args.ocr2 if args.ocr2 else None
+        twopassocr = bool(args.twopassocr)
+        manual = args.manual
+        ss_clipboard = args.clipboard
+        window_name = args.window
+        furigana_filter_sensitivity = args.furigana_filter_sensitivity
+        ss_hotkey = args.area_select_ocr_hotkey.lower()
+        manual_ocr_hotkey = args.manual_ocr_hotkey.lower().replace("ctrl", "<ctrl>").replace("shift",
+                                                                                             "<shift>").replace(
+            "alt", "<alt>") if args.manual_ocr_hotkey else None
+        clipboard_output = args.clipboard_output
+        optimize_second_scan = args.optimize_second_scan
+
+        window = None
+        logger.info(f"Received arguments: {vars(args)}")
+        # set_force_stable_hotkey()
+        ocr_config: OCRConfig = get_ocr_config(window=window_name)
+        if ocr_config:
+            if ocr_config.window:
+                start_time = time.time()
+                while time.time() - start_time < 30:
+                    window = get_window(ocr_config.window)
+                    if window or manual:
+                        if window:
+                            ocr_config.scale_coords()
+                        break
+                    logger.info(f"Window: {ocr_config.window} Could not be found, retrying in 1 second...")
+                    time.sleep(1)
+                else:
+                    logger.error(f"Window '{ocr_config.window}' not found within 30 seconds.")
+                    sys.exit(1)
+            logger.info(
+                f"Starting OCR with configuration: Window: {ocr_config.window}, Rectangles: {ocr_config.rectangles}, Engine 1: {ocr1}, Engine 2: {ocr2}, Two-pass OCR: {twopassocr}")
+        set_dpi_awareness()
+        if manual or ocr_config:
+            rectangles = ocr_config.rectangles if ocr_config and ocr_config.rectangles else []
+            oneocr_threads = []
+            ocr_thread = threading.Thread(target=run_oneocr, args=(ocr_config, rectangles), daemon=True)
+            ocr_thread.start()
+            if not manual:
+                worker_thread = threading.Thread(target=process_task_queue, daemon=True)
+                worker_thread.start()
+            websocket_server_thread = WebsocketServerThread(read=True)
+            websocket_server_thread.start()
+            add_ss_hotkey(ss_hotkey)
+            try:
+                while not done:
+                    time.sleep(1)
+            except KeyboardInterrupt as e:
+                pass
+        else:
+            print("Failed to load OCR configuration. Please check the logs.")
+    except Exception as e:
+        logger.info(e, exc_info=True)
+        logger.debug(e, exc_info=True)
+        logger.info("Closing in 5 seconds...")
+        time.sleep(5)

GameSentenceMiner/owocr/owocr/ocr.py

@@ -1,6 +1,7 @@
 import re
 import os
 import io
+import time
 from pathlib import Path
 import sys
 import platform
@@ -17,8 +18,6 @@ from google.generativeai import GenerationConfig
 from loguru import logger
 import requests
 
-from GameSentenceMiner.util.configuration import get_app_directory, get_temporary_directory
-
 try:
     from manga_ocr import MangaOcr as MOCR
 except ImportError:
@@ -1247,6 +1246,68 @@ class GroqOCR:
     def _preprocess(self, img):
         return base64.b64encode(pil_image_to_bytes(img, png_compression=1)).decode('utf-8')
 
+# class QWENOCR:
+#     name = 'qwenvl'
+#     readable_name = 'Qwen2-VL'
+#     key = 'q'
+#     available = False
+#
+#     def __init__(self, config={}):
+#         try:
+#             import torch
+#             from transformers import Qwen2VLForConditionalGeneration, AutoProcessor
+#             self.model = Qwen2VLForConditionalGeneration.from_pretrained(
+#                 "Qwen/Qwen2-VL-2B-Instruct", torch_dtype="auto", device_map="auto"
+#             )
+#             self.processor = AutoProcessor.from_pretrained("Qwen/Qwen2-VL-2B-Instruct", use_fast=True)
+#             self.device = "cuda" if torch.cuda.is_available() else "cpu"
+#             print(self.device)
+#             self.available = True
+#             logger.info('Qwen2-VL ready')
+#         except Exception as e:
+#             logger.warning(f'Qwen2-VL not available: {e}')
+#
+#     def __call__(self, img, furigana_filter_sensitivity=0):
+#         if not self.available:
+#             return (False, 'Qwen2-VL is not available.')
+#         try:
+#             img = input_to_pil_image(img)
+#             conversation = [
+#                 {
+#                     "role": "user",
+#                     "content": [
+#                         {"type": "image"},
+#                         {"type": "text", "text": "Analyze the image. Extract text *only* from within dialogue boxes (speech bubbles or panels containing character dialogue). If Text appears to be vertical, read the text from top to bottom, right to left. From the extracted dialogue text, filter out any furigana. Ignore and do not include any text found outside of dialogue boxes, including character names, speaker labels, or sound effects. Return *only* the filtered dialogue text. If no text is found within dialogue boxes after applying filters, return nothing. Do not include any other output, formatting markers, or commentary."},
+#                     ],
+#                 }
+#             ]
+#             text_prompt = self.processor.apply_chat_template(conversation, add_generation_prompt=True)
+#             inputs = self.processor(
+#                 text=[text_prompt], images=[img], padding=True, return_tensors="pt"
+#             )
+#             inputs = inputs.to(self.device)
+#             output_ids = self.model.generate(**inputs, max_new_tokens=128)
+#             generated_ids = [
+#                 output_ids[len(input_ids):]
+#                 for input_ids, output_ids in zip(inputs.input_ids, output_ids)
+#             ]
+#             output_text = self.processor.batch_decode(
+#                 generated_ids, skip_special_tokens=True, clean_up_tokenization_spaces=True
+#             )
+#             return (True, output_text[0] if output_text else "")
+#         except Exception as e:
+#             return (False, f'Qwen2-VL inference failed: {e}')
+
+
+# qwenocr = QWENOCR()
+#
+# for i in range(10):
+#     start_time = time.time()
+#     res, text = qwenocr(Image.open('test_furigana.png'), furigana_filter_sensitivity=0)  # Example usage
+#     end_time = time.time()
+#
+#     print(f"Time taken: {end_time - start_time:.2f} seconds")
+#     print(text)
 # class LocalOCR:
 #     name = 'local_ocr'
 #     readable_name = 'Local OCR'

GameSentenceMiner/util/configuration.py

@@ -172,7 +172,7 @@ class Screenshot:
 class Audio:
     enabled: bool = True
     extension: str = 'opus'
-    beginning_offset: float = 0.0
+    beginning_offset: float = -0.5
     end_offset: float = 0.5
     pre_vad_end_offset: float = 0.0
     ffmpeg_reencode_options: str = '-c:a libopus -f opus -af \"afade=t=in:d=0.10\"' if is_windows() else ''
@@ -683,6 +683,7 @@ class GsmAppState:
         self.last_mined_line = None
         self.keep_running = True
         self.current_game = ''
+        self.videos_to_remove = set()
 
 @dataclass_json
 @dataclass

GameSentenceMiner/web/service.py

@@ -21,7 +21,7 @@ def handle_texthooker_button(video_path='', get_audio_from_video=None):
                 if get_config().advanced.audio_player_path:
                     play_audio_in_external(gsm_state.previous_audio)
                 elif get_config().advanced.video_player_path:
-                    play_video_in_external(line, gsm_state.previous_audio)
+                    play_video_in_external(line, video_path)
                 else:
                     import sounddevice as sd
                     data, samplerate = gsm_state.previous_audio
@@ -35,9 +35,7 @@ def handle_texthooker_button(video_path='', get_audio_from_video=None):
                 play_audio_in_external(audio)
                 gsm_state.previous_audio = audio
             elif get_config().advanced.video_player_path:
-                new_video_path = play_video_in_external(line, video_path)
-                gsm_state.previous_audio = new_video_path
-                gsm_state.previous_replay = new_video_path
+                play_video_in_external(line, video_path)
             else:
                 import sounddevice as sd
                 import soundfile as sf
@@ -75,8 +73,8 @@ def handle_texthooker_button(video_path='', get_audio_from_video=None):
         logger.debug(f"Error Playing Audio/Video: {e}", exc_info=True)
         return
     finally:
-        if video_path and get_config().paths.remove_video and os.path.exists(video_path):
-            os.remove(video_path)
+        gsm_state.previous_replay = video_path
+        gsm_state.videos_to_remove.add(video_path)
 
 
 def play_audio_in_external(filepath):
@@ -94,37 +92,28 @@ def play_audio_in_external(filepath):
 
 
 def play_video_in_external(line, filepath):
-    def move_video_when_closed(p, fp):
-        p.wait()
-        os.remove(fp)
-
-    shutil.move(filepath, get_temporary_directory())
-    new_filepath = os.path.join(get_temporary_directory(), os.path.basename(filepath))
-
     command = [get_config().advanced.video_player_path]
 
-    start, _, _, _ = get_video_timings(new_filepath, line)
+    start, _, _, _ = get_video_timings(filepath, line)
 
     if start:
         if "vlc" in get_config().advanced.video_player_path:
             command.extend(["--start-time", convert_to_vlc_seconds(start), '--one-instance'])
         else:
             command.extend(["--start", convert_to_vlc_seconds(start)])
-    command.append(os.path.normpath(new_filepath))
+    command.append(os.path.normpath(filepath))
 
     logger.info(" ".join(command))
 
 
 
     try:
-        proc = subprocess.Popen(command)
-        print(f"Opened {filepath} in {get_config().advanced.video_player_path}.")
-        threading.Thread(target=move_video_when_closed, args=(proc, filepath)).start()
+        subprocess.Popen(command)
+        logger.info(f"Opened {filepath} in {get_config().advanced.video_player_path}.")
     except FileNotFoundError:
-        print("VLC not found. Make sure it's installed and in your PATH.")
+        logger.error("VLC not found. Make sure it's installed and in your PATH.")
     except Exception as e:
-        print(f"An error occurred: {e}")
-    return new_filepath
+        logger.error(f"An error occurred: {e}")
 
 
 def convert_to_vlc_seconds(time_str):

GameSentenceMiner/web/texthooking_page.py

@@ -287,8 +287,8 @@ def get_screenshot():
     if event_id is None:
         return jsonify({'error': 'Missing id'}), 400
     gsm_state.line_for_screenshot = get_line_by_id(event_id)
-    if gsm_state.previous_line_for_screenshot and gsm_state.line_for_screenshot.id == gsm_state.previous_line_for_screenshot.id:
-        handle_texthooker_button()
+    if gsm_state.previous_line_for_screenshot and gsm_state.line_for_screenshot.id == gsm_state.previous_line_for_screenshot.id or gsm_state.previous_line_for_audio:
+        handle_texthooker_button(gsm_state.previous_replay)
     else:
         obs.save_replay_buffer()
     return jsonify({}), 200
@@ -301,8 +301,8 @@ def play_audio():
     if event_id is None:
         return jsonify({'error': 'Missing id'}), 400
     gsm_state.line_for_audio = get_line_by_id(event_id)
-    if gsm_state.previous_line_for_audio and gsm_state.line_for_audio == gsm_state.previous_line_for_audio:
-        handle_texthooker_button()
+    if gsm_state.previous_line_for_audio and gsm_state.line_for_audio == gsm_state.previous_line_for_audio or gsm_state.previous_line_for_screenshot:
+        handle_texthooker_button(gsm_state.previous_replay)
     else:
         obs.save_replay_buffer()
     return jsonify({}), 200

{gamesentenceminer-2.10.10.dist-info → gamesentenceminer-2.10.12.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: GameSentenceMiner
-Version: 2.10.10
+Version: 2.10.12
 Summary: A tool for mining sentences from games. Update: Full UI Re-design
 Author-email: Beangate <bpwhelan95@gmail.com>
 License: MIT License

{gamesentenceminer-2.10.10.dist-info → gamesentenceminer-2.10.12.dist-info}/RECORD

@@ -1,8 +1,8 @@
 GameSentenceMiner/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 GameSentenceMiner/anki.py,sha256=kWw3PV_Jj5-lHcttCB3lRXejHlaAbiJ2Ag_NAGX-RI8,16632
-GameSentenceMiner/config_gui.py,sha256=h-vDxpFCC347iK_mDJAjwKm7Qubeu-NWaxvd9SvzqzY,90942
+GameSentenceMiner/config_gui.py,sha256=Xa_a-sdQzht3kzR-Z9gkLy4qnaPyP1bdVadYTHp5lUQ,91018
 GameSentenceMiner/gametext.py,sha256=6VkjmBeiuZfPk8T6PHFdIAElBH2Y_oLVYvmcafqN7RM,6747
-GameSentenceMiner/gsm.py,sha256=PSL_J723k23SIfgeNhoXgTqlG-V3MQTFJtLDcrZDFqs,24625
+GameSentenceMiner/gsm.py,sha256=p4DVa_Jx1EOsgUxAAdC7st7VXLKWnP2BLDGT78ToO8w,24864
 GameSentenceMiner/obs.py,sha256=ZV9Vk39hrsJLT-AlIxa3qgncKxXaL3Myl33vVJEDEoA,14670
 GameSentenceMiner/vad.py,sha256=G0NkaWFJaIfKQAV7LOFxyKoih7pPNYHDuy4SzeFVCkI,16389
 GameSentenceMiner/ai/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
@@ -16,21 +16,21 @@ GameSentenceMiner/assets/icon512.png,sha256=HxUj2GHjyQsk8NV433256UxU9phPhtjCY-YB
 GameSentenceMiner/assets/icon64.png,sha256=N8xgdZXvhqVQP9QUK3wX5iqxX9LxHljD7c-Bmgim6tM,9301
 GameSentenceMiner/assets/pickaxe.png,sha256=VfIGyXyIZdzEnVcc4PmG3wszPMO1W4KCT7Q_nFK6eSE,1403829
 GameSentenceMiner/ocr/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-GameSentenceMiner/ocr/gsm_ocr_config.py,sha256=0hZmNIvZmlAEcy_NaTukG_ALUORULUT7sQ8q5VlDJU4,4047
+GameSentenceMiner/ocr/gsm_ocr_config.py,sha256=jtTzAWtMAx8GuA1XIJ_BmyNn3aYaO3u_c5Q7m5D4gS8,4056
 GameSentenceMiner/ocr/ocrconfig.py,sha256=_tY8mjnzHMJrLS8E5pHqYXZjMuLoGKYgJwdhYgN-ny4,6466
-GameSentenceMiner/ocr/owocr_area_selector.py,sha256=GEqIIhRc3WCIAx3HunuYo6ayJsCnZWT-x9fwZMCy2e8,16183
-GameSentenceMiner/ocr/owocr_helper.py,sha256=YHhG3PuJsPWP4352TAu4dtdX7itRiOybngzZVT4B50c,20184
+GameSentenceMiner/ocr/owocr_area_selector.py,sha256=boAqarX17jvFscu-7s6C9rqesjQ54s-kfuW0bjCru-M,19834
+GameSentenceMiner/ocr/owocr_helper.py,sha256=VDcuBfyZ1B7TN6yImJVuNxqWY7pr95R2cRM9jgD5Rk8,21670
 GameSentenceMiner/ocr/ss_picker.py,sha256=0IhxUdaKruFpZyBL-8SpxWg7bPrlGpy3lhTcMMZ5rwo,5224
 GameSentenceMiner/owocr/owocr/__init__.py,sha256=87hfN5u_PbL_onLfMACbc0F5j4KyIK9lKnRCj6oZgR0,49
 GameSentenceMiner/owocr/owocr/__main__.py,sha256=XQaqZY99EKoCpU-gWQjNbTs7Kg17HvBVE7JY8LqIE0o,157
 GameSentenceMiner/owocr/owocr/config.py,sha256=qM7kISHdUhuygGXOxmgU6Ef2nwBShrZtdqu4InDCViE,8103
 GameSentenceMiner/owocr/owocr/lens_betterproto.py,sha256=oNoISsPilVVRBBPVDtb4-roJtAhp8ZAuFTci3TGXtMc,39141
-GameSentenceMiner/owocr/owocr/ocr.py,sha256=fWrbKomSrdkSdlEiMGTKb6-F7wCgfaZZNBUo2gCqmuA,52247
+GameSentenceMiner/owocr/owocr/ocr.py,sha256=Mri_zB_COk7x9GmolyhYCINJ-lQlD45GuJ4m4M0IBFM,55328
 GameSentenceMiner/owocr/owocr/run.py,sha256=mZIGDm3fGYrYbSNuFOk7Sbslfgi36YN0YqfC1xYh_eY,55286
 GameSentenceMiner/owocr/owocr/screen_coordinate_picker.py,sha256=Na6XStbQBtpQUSdbN3QhEswtKuU1JjReFk_K8t5ezQE,3395
 GameSentenceMiner/util/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 GameSentenceMiner/util/audio_offset_selector.py,sha256=8Stk3BP-XVIuzRv9nl9Eqd2D-1yD3JrgU-CamBywJmY,8542
-GameSentenceMiner/util/configuration.py,sha256=iXgfrawPPpAmQdXv2zPR-LuZnXx1ORGAwwP55OmLOs8,28778
+GameSentenceMiner/util/configuration.py,sha256=8PZk4IhtWFimfRy7biREcfG1NGkFNzKzFjlOjxNEFd0,28817
 GameSentenceMiner/util/electron_config.py,sha256=3VmIrcXhC-wIMMc4uqV85NrNenRl4ZUbnQfSjWEwuig,9852
 GameSentenceMiner/util/ffmpeg.py,sha256=t0tflxq170n8PZKkdw8fTZIUQfXD0p_qARa9JTdhBTc,21530
 GameSentenceMiner/util/gsm_utils.py,sha256=_279Fu9CU6FEh4cP6h40TWOt_BrqmPgytfumi8y53Ew,11491
@@ -47,8 +47,8 @@ GameSentenceMiner/util/downloader/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeR
 GameSentenceMiner/util/downloader/download_tools.py,sha256=mvnOjDHFlV1AbjHaNI7mdnC5_CH5k3N4n1ezqzzbzGA,8139
 GameSentenceMiner/util/downloader/oneocr_dl.py,sha256=EJbKISaZ9p2x9P4x0rpMM5nAInTTc9b7arraGBcd-SA,10381
 GameSentenceMiner/web/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-GameSentenceMiner/web/service.py,sha256=2o62dZfDSBxBH5zCjrcHCX5yAc3PmGeP2lr07n8-dgo,5779
-GameSentenceMiner/web/texthooking_page.py,sha256=rX2rBFIlVlKmVXB8dseuyWfMzcDcjNNQosncwUolMu8,16054
+GameSentenceMiner/web/service.py,sha256=S7bYf2kSk08u-8R9Qpv7piM-pxfFjYZUvU825xupmuI,5279
+GameSentenceMiner/web/texthooking_page.py,sha256=EmcIBEPGWNgI2LGL3kKUsm0rs2Vn0CPq9PVKKnuIt2g,16183
 GameSentenceMiner/web/static/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 GameSentenceMiner/web/static/apple-touch-icon.png,sha256=OcMI8af_68DA_tweOsQ5LytTyMwm7-hPW07IfrOVgEs,46132
 GameSentenceMiner/web/static/favicon-96x96.png,sha256=lOePzjiKl1JY2J1kT_PMdyEnrlJmi5GWbmXJunM12B4,16502
@@ -62,9 +62,9 @@ GameSentenceMiner/web/templates/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm
 GameSentenceMiner/web/templates/index.html,sha256=n0J-dV8eksj8JXUuaCTIh0fIxIjfgm2EvxGBdQ6gWoM,214113
 GameSentenceMiner/web/templates/text_replacements.html,sha256=tV5c8mCaWSt_vKuUpbdbLAzXZ3ATZeDvQ9PnnAfqY0M,8598
 GameSentenceMiner/web/templates/utility.html,sha256=3flZinKNqUJ7pvrZk6xu__v67z44rXnaK7UTZ303R-8,16946
-gamesentenceminer-2.10.10.dist-info/licenses/LICENSE,sha256=OXLcl0T2SZ8Pmy2_dmlvKuetivmyPd5m1q-Gyd-zaYY,35149
-gamesentenceminer-2.10.10.dist-info/METADATA,sha256=KJtMtM6AUz0qc8xsuSrNxd_I53gcYtAPNWO9VkDGSsY,7355
-gamesentenceminer-2.10.10.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-gamesentenceminer-2.10.10.dist-info/entry_points.txt,sha256=2APEP25DbfjSxGeHtwBstMH8mulVhLkqF_b9bqzU6vQ,65
-gamesentenceminer-2.10.10.dist-info/top_level.txt,sha256=V1hUY6xVSyUEohb0uDoN4UIE6rUZ_JYx8yMyPGX4PgQ,18
-gamesentenceminer-2.10.10.dist-info/RECORD,,
+gamesentenceminer-2.10.12.dist-info/licenses/LICENSE,sha256=OXLcl0T2SZ8Pmy2_dmlvKuetivmyPd5m1q-Gyd-zaYY,35149
+gamesentenceminer-2.10.12.dist-info/METADATA,sha256=YzGn0pkP-I00xGsRHt-5GK8x9pCKpoKR5lHBYL_z8Ho,7355
+gamesentenceminer-2.10.12.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+gamesentenceminer-2.10.12.dist-info/entry_points.txt,sha256=2APEP25DbfjSxGeHtwBstMH8mulVhLkqF_b9bqzU6vQ,65
+gamesentenceminer-2.10.12.dist-info/top_level.txt,sha256=V1hUY6xVSyUEohb0uDoN4UIE6rUZ_JYx8yMyPGX4PgQ,18
+gamesentenceminer-2.10.12.dist-info/RECORD,,