kmisc 2.1.122__py3-none-any.whl → 2.1.124__py3-none-any.whl

kmisc/__init__.py

@@ -3453,7 +3453,7 @@ def Upper(src,default='org'):
     if default in ['org',{'org'}]: return src
     return default
 
-def web_capture(url,output_file,image_size='1920,1080',wait_time=3,ignore_certificate_error=False,username=None,password=None,auth_fields={'auth':{'type':'name','name':('username','password')},'submit':{'type':'submit','name':None}},next_do={},gpu=False,live_capture=0,capture_method='file',find_string=None,found_space='\n',log=None,ocr_enhance=False,daemon=False,backup=False):
+def web_capture(url,output_file,image_size='full',wait_time=3,ignore_certificate_error=False,username=None,password=None,auth_fields={'auth':{'type':'name','name':('username','password')},'submit':{'type':'submit','name':None}},next_do={},gpu=False,live_capture=0,capture_method='file',capture_type='png',video_file=None,ocr_module='easyocr',find_string=None,found_space='\n',log=None,ocr_enhance=False,daemon=False,backup=False):
     #auth_fields.submit.type : name    : login button with name 
     #                        : id      : login button with id 
     #                        : submit  : submit button without name or id
@@ -3463,6 +3463,8 @@ def web_capture(url,output_file,image_size='1920,1080',wait_time=3,ignore_certif
     #auth_fields.auth.name   : (usernane,password)    : username/password field string for name or id
     #next_do                 : put data and click submit
 
+    _url=url.split('/')
+    url=WEB().url_join(*_url[1:],method=_url[0])
     if isinstance(image_size,str):
         if 'x' in image_size:
             image_size=image_size.split('x')
@@ -3475,8 +3477,9 @@ def web_capture(url,output_file,image_size='1920,1080',wait_time=3,ignore_certif
     if isinstance(image_size,(list,tuple)) and len(image_size) == 2:
         image_size=','.join([str(i) for i in image_size])
     else:
-        #Set it to default image size
-        image_size='1920,1080'
+        if not IsIn(image_size,[None,'full','fullscreen','full_screen','auto']):
+            #Set it to default image size
+            image_size='1920,1080'
 
     if Import('import selenium'):
         return False,'Can not install selenium package'
@@ -3484,10 +3487,13 @@ def web_capture(url,output_file,image_size='1920,1080',wait_time=3,ignore_certif
         if backup:
             Import('filecmp')
             Import('shutil')
+        if capture_type in ['mov','mp4']:
+            print('????load cv2')
+            Import('cv2',install_name='opencv-python')
 
         ocr=None
         if capture_method != 'file':
-            ocr=OCR(enhance=ocr_enhance)
+            ocr=OCR(enhance=ocr_enhance,module=ocr_module)
         # Configure Chrome options for headless mode
         from selenium.webdriver.chrome.options import Options
         from selenium.webdriver.common.by import By
@@ -3497,7 +3503,9 @@ def web_capture(url,output_file,image_size='1920,1080',wait_time=3,ignore_certif
         chrome_options.add_argument('--headless')  # Run in headless mode
         chrome_options.add_argument('--no-sandbox')
         chrome_options.add_argument('--disable-dev-shm-usage')
-        chrome_options.add_argument(f"--window-size={image_size}")  # Set window size
+
+        if image_size.lower() not in ['full','fullscreen','full_screen','auto']:
+            chrome_options.add_argument(f"--window-size={image_size}")  # Set window size
         if not gpu:
             chrome_options.add_argument("--disable-gpu")
         if ignore_certificate_error:
@@ -3505,6 +3513,14 @@ def web_capture(url,output_file,image_size='1920,1080',wait_time=3,ignore_certif
             chrome_options.add_argument('--allow-insecure-localhost')  # gnore-certificate-errors
         # Initialize the Chrome driver
         driver = selenium.webdriver.Chrome(options=chrome_options)
+        if image_size.lower() in ['full','fullscreen','full_screen','auto']:
+            #original_size = driver.get_window_size()
+            #full_width = driver.execute_script("return Math.max(document.body.scrollWidth, document.body.offsetWidth, document.documentElement.clientWidth, document.documentElement.scrollWidth, document.documentElement.offsetWidth);")
+            #full_height = driver.execute_script("return Math.max(document.body.scrollHeight, document.body.offsetHeight, document.documentElement.clientHeight, document.documentElement.scrollHeight, document.documentElement.offsetHeight);")
+            #driver.set_window_size(full_width, full_height)
+            # code save screenshot
+            #driver.set_window_size(original_size['width'], original_size['height']) #restore size
+            driver.maximize_window()
         rc=False,output_file
         try:
             # Navigate to the URL
@@ -3559,7 +3575,7 @@ def web_capture(url,output_file,image_size='1920,1080',wait_time=3,ignore_certif
                     next_do_button = driver.find_element(By.XPATH, "//button[@type='submit']")
                 next_do_button.click()
 
-            def _capture_(live_capture,driver,output_file,wait_time,capture_method,backup,ocr,log,find_string,daemon):
+            def _capture_(live_capture,driver,output_file,wait_time,capture_method,backup,ocr,log,find_string,daemon,video_file):
                 def wait_body(driver,timeout=10):
                     #wait until get screen data
                     try:
@@ -3578,23 +3594,37 @@ def web_capture(url,output_file,image_size='1920,1080',wait_time=3,ignore_certif
                     time.sleep(timeout)
 
                 live_capture=Int(live_capture)
-                wait_time=Int(wait_time,10)
+                if not isinstance(wait_time,(int,float)):
+                    wait_time=Int(wait_time,10)
                 backup_idx=0
                 if backup:
                     backup=Int(backup,2)
                 if isinstance(live_capture,int) and live_capture > wait_time*2:
+                    #Keep capture
                     Time=TIME()
+                    if capture_type in ['mov','mp4']:
+                        frame_rate=1/wait_time
+                        window_size = driver.get_window_size()
+                        width = window_size['width']
+                        height = window_size['height']
+                        fourcc = cv2.VideoWriter_fourcc(*'mp4v') # Codec for MP4
+                        video_writer = cv2.VideoWriter(video_file, fourcc, frame_rate, (width, height))
                     while True:
                         if Time.Out(live_capture):
                             driver.quit()
-                            return False
+                            if capture_type in ['mov','mp4']:
+                                video_writer.release()
+                                return True
+                            else:
+                                #Do something, but not return until timeout. So return False
+                                return False
                         # Capture screenshot
                         if log:
-                            if log in ['screen','log','print',print]:
+                            if IsIn(log,['screen','log','print',print]):
                                 printf(Dot(),direct=True)
                             else:
                                 printf(Dot(),log=log,direct=True)
-                        if backup:
+                        if backup and capture_type not in ['mov','mp4']:
                             save_file='{}.{}'.format(output_file,backup_idx%backup)
                         else:
                             save_file=output_file
@@ -3602,43 +3632,85 @@ def web_capture(url,output_file,image_size='1920,1080',wait_time=3,ignore_certif
                         wait_body(driver,timeout=wait_time)
                         #capture
                         driver.save_screenshot(save_file)
-                        if backup:
-                            if backup == 2:
-                                comp_a=f'{output_file}.0'
-                                comp_b=f'{output_file}.1'
-                                if os.path.isfile(comp_a) and os.path.isfile(comp_b):
-                                    if filecmp.cmp(comp_a,comp_b):
-                                        if log:
-                                            if log in ['screen','log','print',print]:
-                                                printf(Dot(),direct=True)
-                                            else:
-                                                printf(Dot(),log=log,direct=True)
-                                        time.sleep(wait_time)
-                                        backup_idx+=1
-                                        continue
-                            shutil.copy2(save_file,output_file)
-                        if IsIn(capture_method,['log','screen','text']):
-                            found_words=ocr.Text(image_file=save_file)
-                            found_strings=found_space.join(found_words)
-                            printf(found_strings,log=log,mode='d' if log else 's')
-                            if find_string:
-                                if find_string in found_strings:
-                                    driver.quit()
-                                    return True
+                        if capture_type in ['mov','mp4']:
+                            #Make a video file
+                            frame = cv2.imread(save_file)
+                            frame = cv2.resize(frame, (width, height))
+                            video_writer.write(frame)
+                        else:
+                            #do something with picture file
+                            if backup:
+                                if backup == 2:
+                                    comp_a=f'{output_file}.0'
+                                    comp_b=f'{output_file}.1'
+                                    if os.path.isfile(comp_a) and os.path.isfile(comp_b):
+                                        if filecmp.cmp(comp_a,comp_b):
+                                            if log:
+                                                if IsIn(log,['screen','log','print',print]):
+                                                    printf(Dot(),direct=True)
+                                                else:
+                                                    printf(Dot(),log=log,direct=True)
+                                            time.sleep(wait_time)
+                                            backup_idx+=1
+                                            continue
+                                shutil.copy2(save_file,output_file)
+                            if IsIn(capture_method,['log','screen','text']):
+                                found_words=ocr.Text(image_file=save_file)
+                                found_strings=found_space.join(found_words)
+                                if IsIn(log,['screen','log','print',print,None]):
+                                    printf(found_strings,mode='s')
+                                else:
+                                    printf(found_strings,log=log,mode='d')
+                                if find_string:
+                                    if not isinstance(find_string,list):
+                                        find_string=[find_string]
+                                    for ff in find_string:
+                                        if ff in found_strings:
+                                            #Find exit string, So True, So True, So True, So True
+                                            driver.quit()
+                                            return True
+                            backup_idx+=1
+                        #capture interval
                         time.sleep(wait_time)
-                        backup_idx+=1
                 else:
+                    #Single capture
                     #wait
-                    wait_body(driver,timeout=wait_time)
+                    #wait_body(driver,timeout=wait_time)
+                    time.sleep(wait_time)
                     #capture
                     driver.save_screenshot(output_file)
-                driver.quit()
+                    if IsIn(capture_method,['log','screen','text']):
+                        found_words=ocr.Text(image_file=output_file)
+                        found_strings=found_space.join(found_words)
+                        if IsIn(log,['screen','log','print',print,None]):
+                            printf(found_strings,mode='s')
+                        else:
+                            printf(found_strings,log=log,mode='d')
+                        if find_string:
+                            if find_string in found_strings:
+                                driver.quit()
+                                return True
+                    driver.quit()
+                    if IsIn(capture_method,['text']):
+                        return found_strings
+                    return True
+
+            ##
+            if IsIn(capture_type,['mov','mp4']):
+                if not video_file:
+                    video_file='{}.mp4'.format('.'.join(output_file.split('.')[:-1]))
+
+            #Background running
             if daemon:
-                t=kThread(target=_capture_, args=(live_capture,driver,output_file,wait_time,capture_method,backup,ocr,log,find_string,daemon))
+                t=kThread(target=_capture_, args=(live_capture,driver,output_file,wait_time,capture_method,backup,ocr,log,find_string,daemon,video_file))
                 return t
             else:
-                _capture_(live_capture,driver,output_file,wait_time,capture_method,backup,ocr,log,find_string,daemon)
-                return True,output_file
+                #Single process running
+                rc=_capture_(live_capture,driver,output_file,wait_time,capture_method,backup,ocr,log,find_string,daemon,video_file)
+                if IsIn(capture_type,['mov','mp4']):
+                    return rc,video_file
+                else:
+                    return rc,output_file
 
         except Exception as e:
             #print(f"Error capturing screenshot: {str(e)}")
@@ -3650,42 +3722,92 @@ def web_capture(url,output_file,image_size='1920,1080',wait_time=3,ignore_certif
 #        return rc
 
 class OCR:
-    def __init__(self,image_file=None,enhance=False,language=['en'],gpu=False,model_storage_directory=None,**opts):
+    def __init__(self,image_file=None,enhance=False,language=['en'],gpu=False,model_storage_directory=None,ocr_module='easyocr',**opts):
         self.enhance=enhance
         self.image_file=image_file
-        Import('easyocr')
-        if self.enhance:
-            Import('PIL',install_name='Pillow')
-            Import('numpy')
-        self.reader = easyocr.Reader(language,gpu=gpu,model_storage_directory=model_storage_directory)
-        # Suppress Torch pin_memory warning
-        warnings.filterwarnings("ignore", category=UserWarning, module="torch.utils.data.dataloader")
+        self.ocr_module=ocr_module
+        self.language=language
+        self.gpu=gpu
+        self.model_storage_directory=model_storage_directory
+        if self.ocr_module == 'pytesseract':
+            Import('pytesseract')
+            Import('import numpy as np')
+            Import('cv2',install_name='opencv-python')
+        else:
+            Import('easyocr')
+            if self.enhance:
+                Import('PIL',install_name='Pillow')
+                Import('numpy')
+            self.reader = easyocr.Reader(self.language,gpu=gpu,model_storage_directory=model_storage_directory)
+            # Suppress Torch pin_memory warning
+            warnings.filterwarnings("ignore", category=UserWarning, module="torch.utils.data.dataloader")
 
-        # Suppress EasyOCR CPU warning
-        #warnings.filterwarnings("ignore", message="WARNING:easyocr.easyocr:Using CPU. Note: This module is much faster with a GPU.")
+            # Suppress EasyOCR CPU warning
+            #warnings.filterwarnings("ignore", message="WARNING:easyocr.easyocr:Using CPU. Note: This module is much faster with a GPU.")
 
-        # Suppress NetworkX backend warning
-        warnings.filterwarnings("ignore", category=RuntimeWarning, module="networkx.utils.backends")
+            # Suppress NetworkX backend warning
+            warnings.filterwarnings("ignore", category=RuntimeWarning, module="networkx.utils.backends")
 
-    def Text(self,detail=0,low_text=None,contrast_ths=None,image_file=None):
+    def Text(self,detail=0,low_text=None,contrast_ths=None,image_file=None,output=str):
         if not image_file: image_file=self.image_file
         if not image_file: return False
-        opts={}
-        opts['detail']=detail
-        if isinstance(low_text,float): opts['low_text']=low_test
-        if isinstance(contrast_ths,float): opts['contrast_ths']=contrast_ths
-        if self.enhance:
-            image = PIL.Image.open(image_file)
-            image = image.convert('L') #Grayscale
-            image = PIL.ImageEnhance.Contrast(image).enhance(3.0) #high contrast
-            image = PIL.ImageEnhance.Sharpness(image).enhance(2.0)#Sharpen
-            image = image.convert('RGB').point(lambda p: 255 if p > 140 else 0)  # Adjust threshold if needed
-#            image = image.resize((800, int(800 * image.height / image.width)), PIL.Image.Resampling.LANCZOS)
-            image.save(image_file)
-#            image_np = numpy.array(image)
-#            return self.reader.readtext(image_np,**opts)
-#        else:
-        return self.reader.readtext(image_file,**opts)
+        if not os.path.isfile(image_file): return False
+        if self.ocr_module == 'pytesseract':
+            image = cv2.imread(image_file)
+            # Convert to grayscale
+            gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
+
+            # Invert the image to make text black on white (Tesseract prefers this)
+            inverted = cv2.bitwise_not(gray)
+
+            # Light noise reduction with small Gaussian blur (fast)
+            blurred = cv2.GaussianBlur(inverted, (3, 3), 0)
+
+            # Optional CLAHE for contrast (comment out if too slow; it's generally fast)
+            clahe = cv2.createCLAHE(clipLimit=2.0, tileGridSize=(8, 8))
+            enhanced = clahe.apply(blurred)
+
+            # Small upscale (1.5x) for better DPI without heavy computation
+            scale_factor = 1.5
+            resized = cv2.resize(enhanced, None, fx=scale_factor, fy=scale_factor, interpolation=cv2.INTER_LINEAR)  # LINEAR is faster than CUBIC
+
+            # Adaptive thresholding for varying console text quality
+            thresh = cv2.adaptiveThreshold(resized, 255, cv2.ADAPTIVE_THRESH_GAUSSIAN_C, cv2.THRESH_BINARY, 11, 2)
+
+            # Minimal morphological cleanup (small kernel for speed)
+            kernel = np.ones((2, 2), np.uint8)
+            cleaned = cv2.morphologyEx(thresh, cv2.MORPH_CLOSE, kernel, iterations=1)
+            lang=self.language[0] if isinstance(self.language,(list,tuple)) else self.language
+            if not lang or lang == 'en': lang='eng'
+            try:
+                text=pytesseract.image_to_string(cleaned, config=r'--oem 3 --psm 6', lang=lang).strip()
+            except:
+                text=pytesseract.image_to_string(cleaned, config=r'--oem 3 --psm 6', lang='eng').strip()
+            if output is str:
+                return text
+            else:
+                return text.split()
+        else:
+            opts={}
+            opts['detail']=detail
+            if isinstance(low_text,float): opts['low_text']=low_test
+            if isinstance(contrast_ths,float): opts['contrast_ths']=contrast_ths
+            if self.enhance:
+                image = PIL.Image.open(image_file)
+                image = image.convert('L') #Grayscale
+                image = PIL.ImageEnhance.Contrast(image).enhance(3.0) #high contrast
+                image = PIL.ImageEnhance.Sharpness(image).enhance(2.0)#Sharpen
+                image = image.convert('RGB').point(lambda p: 255 if p > 140 else 0)  # Adjust threshold if needed
+#                image = image.resize((800, int(800 * image.height / image.width)), PIL.Image.Resampling.LANCZOS)
+                image.save(image_file)
+#                image_np = numpy.array(image)
+#                return self.reader.readtext(image_np,**opts)
+#            else:
+            text=self.reader.readtext(image_file,**opts)
+            if output is str:
+                return ' '.join(text)
+            else:
+                return text
 
 ############################################
 #Temporary function map for replacement

{kmisc-2.1.122.dist-info → kmisc-2.1.124.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: kmisc
-Version: 2.1.122
+Version: 2.1.124
 Summary: Enginering useful library
 Home-page: https://github.com/kagepark/kmisc
 Author: Kage Park

kmisc-2.1.124.dist-info/RECORD

@@ -0,0 +1,6 @@
+kmisc/__init__.py,sha256=9KNDkVxjqk5_f1h8TRJMwItMXSZFyXBAzwK4MKR-6aM,163758
+kmisc-2.1.124.dist-info/LICENSE,sha256=mn9ekhb34HJxsrVhcxrLXJUzy55T62zg-Gh9Ro0mVJI,1066
+kmisc-2.1.124.dist-info/METADATA,sha256=0Jeai7LsZye2IPSLxdVv8EfoU37IpRSJYJb5D4J1io8,5523
+kmisc-2.1.124.dist-info/WHEEL,sha256=eOLhNAGa2EW3wWl_TU484h7q1UNgy0JXjjoqKoxAAQc,92
+kmisc-2.1.124.dist-info/top_level.txt,sha256=wvdHf5aQTqcGYvxk-F9E_BMWLMhlwC8INBmwO-V6_X4,6
+kmisc-2.1.124.dist-info/RECORD,,

kmisc-2.1.122.dist-info/RECORD

@@ -1,6 +0,0 @@
-kmisc/__init__.py,sha256=nX41gPkw15kBnuav_Xmh11ZBdVZKRJq7OT13qs2Vjn0,156878
-kmisc-2.1.122.dist-info/LICENSE,sha256=mn9ekhb34HJxsrVhcxrLXJUzy55T62zg-Gh9Ro0mVJI,1066
-kmisc-2.1.122.dist-info/METADATA,sha256=-e9Y6YbWJmJ8DT1bcCvIHlwClRtEgMUVHboS5EjxYXw,5523
-kmisc-2.1.122.dist-info/WHEEL,sha256=eOLhNAGa2EW3wWl_TU484h7q1UNgy0JXjjoqKoxAAQc,92
-kmisc-2.1.122.dist-info/top_level.txt,sha256=wvdHf5aQTqcGYvxk-F9E_BMWLMhlwC8INBmwO-V6_X4,6
-kmisc-2.1.122.dist-info/RECORD,,