webscout 8.2.4__py3-none-any.whl → 8.2.6__py3-none-any.whl

webscout/Provider/typefully.py

@@ -1,10 +1,10 @@
-from typing import Union, Any, Dict
+from typing import Optional, Union, Any, Dict
 import re
 from uuid import uuid4
 
 from webscout.AIutel import Optimizers
 from webscout.AIutel import Conversation
-from webscout.AIutel import AwesomePrompts
+from webscout.AIutel import AwesomePrompts, sanitize_stream # Import sanitize_stream
 from webscout.AIbase import Provider
 from webscout import exceptions
 from webscout.litagent import LitAgent
@@ -114,6 +114,17 @@ class TypefullyAI(Provider):
         )
         self.conversation.history_offset = history_offset
 
+    @staticmethod
+    def _typefully_extractor(chunk: Union[str, Dict[str, Any]]) -> Optional[str]:
+        """Extracts content from the Typefully stream format '0:"..."'."""
+        if isinstance(chunk, str):
+            match = re.search(r'0:"(.*?)"(?=,|$)', chunk) # Look for 0:"...", possibly followed by comma or end of string
+            if match:
+                # Decode potential unicode escapes like \u00e9 and handle escaped quotes/backslashes
+                content = match.group(1).encode().decode('unicode_escape')
+                return content.replace('\\\\', '\\').replace('\\"', '"')
+        return None
+
     def ask(
         self,
         prompt: str,
@@ -174,23 +185,22 @@ class TypefullyAI(Provider):
                     raise exceptions.FailedToGenerateResponseError(
                         f"Failed to generate response - ({response.status_code}, {response.reason}) - {response.text}"
                     )
-                streaming_response = ""
-                # Iterate over bytes and decode manually
-                for line_bytes in response.iter_lines():
-                    if line_bytes:
-                        line = line_bytes.decode('utf-8') # Decode bytes
-                        match = re.search(r'0:"(.*?)"', line)
-                        if match:
-                            # Decode potential unicode escapes
-                            content = match.group(1).encode().decode('unicode_escape') 
-                            streaming_response += content
-                            # Yield dict or raw string
-                            yield content if raw else dict(text=content)
-                        elif line.startswith('e:') or line.startswith('d:'):
-                            # End of response
-                            break
+                streaming_text = ""
+                # Use sanitize_stream with the custom extractor
+                processed_stream = sanitize_stream(
+                    data=response.iter_content(chunk_size=None), # Pass byte iterator
+                    intro_value=None, # No simple prefix
+                    to_json=False,    # Content is not JSON
+                    content_extractor=self._typefully_extractor, # Use the specific extractor
+                    end_marker="e:", # Stop processing if "e:" line is encountered (adjust if needed)
+                )
+
+                for content_chunk in processed_stream:
+                    if content_chunk and isinstance(content_chunk, str):
+                        streaming_text += content_chunk
+                        yield content_chunk if raw else dict(text=content_chunk)
                 # Update history and last response after stream finishes
-                self.last_response.update(dict(text=streaming_response))
+                self.last_response.update(dict(text=streaming_text))
                 self.conversation.update_chat_history(
                     prompt, self.get_message(self.last_response)
                 )
@@ -271,13 +281,12 @@ class TypefullyAI(Provider):
         """
         assert isinstance(response, dict), "Response should be of dict data-type only"
         # Handle potential unicode escapes in the final text
+        # Formatting is now handled by the extractor
         text = response.get("text", "")
         try:
-            # Attempt to decode escapes, return original if fails
-            # Already decoded in ask method, just handle formatting
             formatted_text = text.replace('\\n', '\n').replace('\\n\\n', '\n\n')
             return formatted_text
-        except Exception: # Catch potential errors during formatting
+        except Exception: # Catch potential errors during newline replacement
              return text # Return original text if formatting fails
 
 
@@ -319,4 +328,3 @@ if __name__ == "__main__":
 
         except Exception as e:
             print(f"\r{model:<50} {'✗':<10} {str(e)}")
-

webscout/Provider/typegpt.py

@@ -5,7 +5,7 @@ from typing import Union, Any, Dict, Generator
 
 from webscout.AIutel import Optimizers
 from webscout.AIutel import Conversation
-from webscout.AIutel import AwesomePrompts
+from webscout.AIutel import AwesomePrompts, sanitize_stream # Import sanitize_stream
 from webscout.AIbase import Provider
 from webscout import exceptions
 from webscout.litagent import LitAgent
@@ -137,37 +137,30 @@ class TypeGPT(Provider):
                     f"Network connection failed (CurlError). Check your firewall or antivirus settings. Original error: {ce}"
                 ) from ce
 
-            if not response.ok:
-                raise exceptions.FailedToGenerateResponseError(
-                    f"Failed to generate response - ({response.status_code}, {response.reason}) - {response.text}"
-                )
-            message_load = ""
-            # Iterate over bytes and decode manually
-            for line_bytes in response.iter_lines():
-                if line_bytes:
-                    line = line_bytes.decode("utf-8")
-                    if line.startswith("data: "):
-                        line = line[6:]  # Remove "data: " prefix
-                        # Skip [DONE] message
-                        if line.strip() == "[DONE]":
-                            break
-                        try:
-                            data = json.loads(line)
-                            # Extract and yield only new content
-                            if 'choices' in data and len(data['choices']) > 0:
-                                delta = data['choices'][0].get('delta', {})
-                                if 'content' in delta:
-                                    new_content = delta['content']
-                                    message_load += new_content
-                                    # Yield only the new content
-                                    yield dict(text=new_content) if not raw else new_content
-                                    # Update last_response incrementally for potential non-stream use later
-                                    self.last_response = dict(text=message_load) 
-                        except json.JSONDecodeError:
-                            continue
+            response.raise_for_status() # Check for HTTP errors first
+
+            streaming_text = ""
+            # Use sanitize_stream
+            processed_stream = sanitize_stream(
+                data=response.iter_content(chunk_size=None), # Pass byte iterator
+                intro_value="data:",
+                to_json=True,     # Stream sends JSON
+                skip_markers=["[DONE]"],
+                content_extractor=lambda chunk: chunk.get('choices', [{}])[0].get('delta', {}).get('content') if isinstance(chunk, dict) else None,
+                yield_raw_on_error=False # Skip non-JSON or lines where extractor fails
+            )
+
+            for content_chunk in processed_stream:
+                # content_chunk is the string extracted by the content_extractor
+                if content_chunk and isinstance(content_chunk, str):
+                    streaming_text += content_chunk
+                    yield dict(text=content_chunk) if not raw else content_chunk
+                    # Update last_response incrementally
+                    self.last_response = dict(text=streaming_text)
+
             # Update conversation history after stream finishes
-            if message_load: # Only update if something was received
-                 self.conversation.update_chat_history(prompt, message_load)
+            if streaming_text: # Only update if something was received
+                 self.conversation.update_chat_history(prompt, streaming_text)
 
 
         def for_non_stream():
@@ -185,26 +178,30 @@ class TypeGPT(Provider):
                     f"Network connection failed (CurlError). Check your firewall or antivirus settings. Original error: {ce}"
                 ) from ce
 
-            if not response.ok:
-                raise exceptions.FailedToGenerateResponseError(
-                    f"Request failed - {response.status_code}: {response.text}"
-                )
-            
+            response.raise_for_status() # Check for HTTP errors
+
             try:
-                # curl_cffi response.json() handles decoding
-                response_data = response.json() 
-                # Extract the message content for history and return value
-                if 'choices' in response_data and len(response_data['choices']) > 0:
-                    message = response_data['choices'][0].get('message', {})
-                    content = message.get('content', '')
-                    self.last_response = {"text": content} # Store in expected format
-                    self.conversation.update_chat_history(prompt, content)
-                    return self.last_response
-                else:
-                     # Handle cases where response structure is unexpected
-                     self.last_response = {"text": ""} 
-                     return self.last_response
-            except json.JSONDecodeError as je:
+                response_text = response.text # Get raw text
+
+                # Use sanitize_stream for non-streaming JSON response
+                processed_stream = sanitize_stream(
+                    data=response_text,
+                    to_json=True, # Parse the whole text as JSON
+                    intro_value=None,
+                    # Extractor for non-stream structure
+                    content_extractor=lambda chunk: chunk.get('choices', [{}])[0].get('message', {}).get('content') if isinstance(chunk, dict) else None,
+                    yield_raw_on_error=False
+                )
+
+                # Extract the single result
+                content = ""
+                for extracted_content in processed_stream:
+                    content = extracted_content if isinstance(extracted_content, str) else ""
+
+                self.last_response = {"text": content} # Store in expected format
+                self.conversation.update_chat_history(prompt, content)
+                return self.last_response
+            except (json.JSONDecodeError, Exception) as je: # Catch potential JSON errors or others
                  raise exceptions.FailedToGenerateResponseError(f"Failed to decode JSON response: {je} - Response text: {response.text}")
 
 
@@ -290,4 +287,3 @@ if __name__ == "__main__":
 
         except Exception as e:
             print(f"\r{model:<50} {'✗':<10} {str(e)}")
-

webscout/Provider/uncovr.py

@@ -5,7 +5,7 @@ import uuid
 import re
 from typing import Any, Dict, Optional, Generator, Union
 from webscout.AIutel import Optimizers
-from webscout.AIutel import Conversation
+from webscout.AIutel import Conversation, sanitize_stream # Import sanitize_stream
 from webscout.AIutel import AwesomePrompts
 from webscout.AIbase import Provider
 from webscout import exceptions
@@ -109,6 +109,17 @@ class UncovrAI(Provider):
         )
         self.conversation.history_offset = history_offset
 
+    @staticmethod
+    def _uncovr_extractor(chunk: Union[str, Dict[str, Any]]) -> Optional[str]:
+        """Extracts content from the UncovrAI stream format '0:"..."'."""
+        if isinstance(chunk, str):
+            match = re.match(r'^0:\s*"?(.*?)"?$', chunk) # Match 0: maybe optional quotes
+            if match:
+                # Decode potential unicode escapes like \u00e9 and handle escaped quotes/backslashes
+                content = match.group(1).encode().decode('unicode_escape')
+                return content.replace('\\\\', '\\').replace('\\"', '"')
+        return None
+
     def refresh_identity(self, browser: str = None):
         """
         Refreshes the browser identity fingerprint.
@@ -202,27 +213,21 @@ class UncovrAI(Provider):
                         raise exceptions.FailedToGenerateResponseError(
                             f"Request failed with status code {response.status_code} - {response.text}"
                         )
-                
+
                 streaming_text = ""
-                # Iterate over bytes and decode manually
-                for line_bytes in response.iter_lines():
-                    if line_bytes:
-                        try:
-                            line = line_bytes.decode('utf-8')
-                            # Use regex to match content messages
-                            content_match = re.match(r'^0:\s*"?(.*?)"?$', line)
-                            if content_match:  # Content message
-                                content = content_match.group(1).encode().decode('unicode_escape') # Decode escapes
-                                streaming_text += content
-                                resp = dict(text=content)
-                                yield resp if raw else resp
-                            # Check for error messages
-                            error_match = re.match(r'^2:\[{"type":"error","error":"(.*?)"}]$', line)
-                            if error_match:
-                                error_msg = error_match.group(1)
-                                raise exceptions.FailedToGenerateResponseError(f"API Error: {error_msg}")
-                        except (json.JSONDecodeError, UnicodeDecodeError):
-                            continue
+                # Use sanitize_stream with the custom extractor
+                processed_stream = sanitize_stream(
+                    data=response.iter_content(chunk_size=None), # Pass byte iterator
+                    intro_value=None, # No simple prefix
+                    to_json=False,    # Content is not JSON
+                    content_extractor=self._uncovr_extractor, # Use the specific extractor
+                    yield_raw_on_error=True # Keep yielding even if extractor fails, for potential error messages? (Adjust if needed)
+                )
+
+                for content_chunk in processed_stream:
+                    if content_chunk and isinstance(content_chunk, str):
+                        streaming_text += content_chunk
+                        yield dict(text=content_chunk) if not raw else content_chunk
                 
                 self.last_response = {"text": streaming_text}
                 self.conversation.update_chat_history(prompt, streaming_text)
@@ -262,25 +267,25 @@ class UncovrAI(Provider):
                             f"Request failed with status code {response.status_code} - {response.text}"
                         )
 
-                # Process the non-streamed response content (assuming it's similar line format)
+                response_text = response.text # Get the full response text
+                
+                # Use sanitize_stream to process the non-streaming text
+                # It won't parse as JSON, but will apply the extractor line by line
+                processed_stream = sanitize_stream(
+                    data=response_text.splitlines(), # Split into lines first
+                    intro_value=None,
+                    to_json=False,
+                    content_extractor=self._uncovr_extractor,
+                    yield_raw_on_error=True 
+                )
+
+                # Aggregate the results from the generator
                 full_response = ""
-                # Use response.text which should contain the full body for non-streamed curl_cffi requests
-                for line in response.text.splitlines(): 
-                    if line:
-                        try:
-                            # line is already decoded string
-                            content_match = re.match(r'^0:\s*"?(.*?)"?$', line)
-                            if content_match:
-                                content = content_match.group(1).encode().decode('unicode_escape') # Decode escapes
-                                full_response += content
-                            # Check for error messages
-                            error_match = re.match(r'^2:\[{"type":"error","error":"(.*?)"}]$', line)
-                            if error_match:
-                                error_msg = error_match.group(1)
-                                raise exceptions.FailedToGenerateResponseError(f"API Error: {error_msg}")
-                        except (json.JSONDecodeError): # UnicodeDecodeError less likely here
-                            continue
+                for content in processed_stream:
+                    if content and isinstance(content, str):
+                        full_response += content
 
+                # Check if aggregation resulted in empty response (might indicate error not caught by extractor)
                 self.last_response = {"text": full_response}
                 self.conversation.update_chat_history(prompt, full_response)
                 return {"text": full_response}
@@ -323,7 +328,9 @@ class UncovrAI(Provider):
 
     def get_message(self, response: dict) -> str:
         assert isinstance(response, dict), "Response should be of dict data-type only"
-        return response["text"].replace('\\n', '\n').replace('\\n\\n', '\n\n')
+        # Formatting handled by extractor
+        text = response.get("text", "")
+        return text.replace('\\n', '\n').replace('\\n\\n', '\n\n') # Keep newline replacement
 
 if __name__ == "__main__":
     # Ensure curl_cffi is installed
@@ -359,4 +366,3 @@ if __name__ == "__main__":
             print(f"\r{model:<50} {status:<10} {display_text}")
         except Exception as e:
             print(f"\r{model:<50} {'✗':<10} {str(e)}")
-

webscout/__init__.py

@@ -5,7 +5,6 @@ from .webscout_search_async import AsyncWEBS
 from .version import __version__
 from .DWEBS import *
 from .tempid import *
-from .LLM import VLM, LLM
 from .Provider import *
 from .Provider.TTI import *
 from .Provider.TTS import *

webscout/cli.py

@@ -1,6 +1,8 @@
 import sys
 from .swiftcli import CLI, option
 from .webscout_search import WEBS
+from .DWEBS import GoogleSearch  # Import GoogleSearch from DWEBS
+from .yep_search import YepSearch  # Import YepSearch from yep_search
 from .version import __version__
 
 
@@ -282,6 +284,260 @@ def weather(location: str, language: str, proxy: str = None, timeout: int = 10):
     except Exception as e:
         raise e
 
+@app.command()
+@option("--keywords", "-k", help="Search keywords", required=True)
+@option("--region", "-r", help="Region for search results (ISO country code)", default="all")
+@option("--safesearch", "-s", help="SafeSearch setting (on, moderate, off)", default="moderate")
+@option("--max-results", "-m", help="Maximum number of results", type=int, default=10)
+@option("--start-num", "-start", help="Starting position for pagination", type=int, default=0)
+@option("--unique", "-u", help="Filter duplicate results", type=bool, default=True)
+@option("--timeout", "-timeout", help="Timeout value for requests", type=int, default=10)
+@option("--proxy", "-p", help="Proxy URL to use for requests")
+@option("--impersonate", "-i", help="Browser to impersonate", default="chrome110")
+def google_text(
+    keywords: str,
+    region: str,
+    safesearch: str,
+    max_results: int,
+    start_num: int,
+    unique: bool,
+    timeout: int = 10,
+    proxy: str = None,
+    impersonate: str = "chrome110"
+):
+    """Perform a text search using Google Search."""
+    google = GoogleSearch(
+        timeout=timeout,
+        proxies={"https": proxy, "http": proxy} if proxy else None,
+        verify=True,
+        lang="en",
+        sleep_interval=0.0,
+        impersonate=impersonate
+    )
+    
+    try:
+        results = google.text(
+            keywords=keywords,
+            region=region,
+            safesearch=safesearch,
+            max_results=max_results,
+            start_num=start_num,
+            unique=unique
+        )
+        
+        # Convert SearchResult objects to dictionaries for printing
+        formatted_results = []
+        for result in results:
+            result_dict = {
+                "title": result.title,
+                "url": result.url,
+                "description": result.description,
+            }
+            # Add any metadata to the result dictionary
+            for k, v in result.metadata.items():
+                result_dict[k] = v
+                
+            formatted_results.append(result_dict)
+            
+        _print_data(formatted_results)
+    except Exception as e:
+        raise e
+
+@app.command()
+@option("--keywords", "-k", help="Search keywords", required=True)
+@option("--region", "-r", help="Region for search results (ISO country code)", default="all")
+@option("--safesearch", "-s", help="SafeSearch setting (on, moderate, off)", default="moderate")
+@option("--max-results", "-m", help="Maximum number of results", type=int, default=10)
+@option("--timeout", "-timeout", help="Timeout value for requests", type=int, default=10)
+@option("--proxy", "-p", help="Proxy URL to use for requests")
+@option("--impersonate", "-i", help="Browser to impersonate", default="chrome110")
+def google_news(
+    keywords: str,
+    region: str,
+    safesearch: str,
+    max_results: int,
+    timeout: int = 10,
+    proxy: str = None,
+    impersonate: str = "chrome110"
+):
+    """Perform a news search using Google Search."""
+    google = GoogleSearch(
+        timeout=timeout,
+        proxies={"https": proxy, "http": proxy} if proxy else None,
+        verify=True,
+        lang="en",
+        sleep_interval=0.0,
+        impersonate=impersonate
+    )
+    
+    try:
+        results = google.news(
+            keywords=keywords,
+            region=region,
+            safesearch=safesearch,
+            max_results=max_results
+        )
+        
+        # Convert SearchResult objects to dictionaries for printing
+        formatted_results = []
+        for result in results:
+            result_dict = {
+                "title": result.title,
+                "url": result.url,
+                "description": result.description,
+            }
+            # Add any metadata to the result dictionary
+            for k, v in result.metadata.items():
+                result_dict[k] = v
+                
+            formatted_results.append(result_dict)
+            
+        _print_data(formatted_results)
+    except Exception as e:
+        raise e
+
+@app.command()
+@option("--query", "-q", help="Search query", required=True)
+@option("--region", "-r", help="Region for suggestions (ISO country code)", default="all")
+@option("--timeout", "-timeout", help="Timeout value for requests", type=int, default=10)
+@option("--proxy", "-p", help="Proxy URL to use for requests")
+@option("--impersonate", "-i", help="Browser to impersonate", default="chrome110")
+def google_suggestions(
+    query: str,
+    region: str,
+    timeout: int = 10,
+    proxy: str = None,
+    impersonate: str = "chrome110"
+):
+    """Get search suggestions from Google Search."""
+    google = GoogleSearch(
+        timeout=timeout,
+        proxies={"https": proxy, "http": proxy} if proxy else None,
+        verify=True,
+        lang="en",
+        sleep_interval=0.0,
+        impersonate=impersonate
+    )
+    
+    try:
+        results = google.suggestions(query=query, region=region)
+        
+        # Format suggestions for printing
+        formatted_results = []
+        for i, suggestion in enumerate(results, 1):
+            formatted_results.append({"position": i, "suggestion": suggestion})
+            
+        _print_data(formatted_results)
+    except Exception as e:
+        raise e
+
+@app.command()
+@option("--keywords", "-k", help="Search keywords", required=True)
+@option("--region", "-r", help="Region for search results", default="all")
+@option("--safesearch", "-s", help="SafeSearch setting (on, moderate, off)", default="moderate")
+@option("--max-results", "-m", help="Maximum number of results", type=int, default=10)
+@option("--timeout", "-timeout", help="Timeout value for requests", type=int, default=20)
+@option("--proxy", "-p", help="Proxy URL to use for requests")
+@option("--impersonate", "-i", help="Browser to impersonate", default="chrome110")
+def yep_text(
+    keywords: str,
+    region: str,
+    safesearch: str,
+    max_results: int,
+    timeout: int = 20,
+    proxy: str = None,
+    impersonate: str = "chrome110"
+):
+    """Perform a text search using Yep Search."""
+    yep = YepSearch(
+        timeout=timeout,
+        proxies={"https": proxy, "http": proxy} if proxy else None,
+        verify=True,
+        impersonate=impersonate
+    )
+    
+    try:
+        results = yep.text(
+            keywords=keywords,
+            region=region,
+            safesearch=safesearch,
+            max_results=max_results
+        )
+        
+        _print_data(results)
+    except Exception as e:
+        raise e
+
+@app.command()
+@option("--keywords", "-k", help="Search keywords", required=True)
+@option("--region", "-r", help="Region for search results", default="all")
+@option("--safesearch", "-s", help="SafeSearch setting (on, moderate, off)", default="moderate")
+@option("--max-results", "-m", help="Maximum number of results", type=int, default=10)
+@option("--timeout", "-timeout", help="Timeout value for requests", type=int, default=20)
+@option("--proxy", "-p", help="Proxy URL to use for requests")
+@option("--impersonate", "-i", help="Browser to impersonate", default="chrome110")
+def yep_images(
+    keywords: str,
+    region: str,
+    safesearch: str,
+    max_results: int,
+    timeout: int = 20,
+    proxy: str = None,
+    impersonate: str = "chrome110"
+):
+    """Perform an image search using Yep Search."""
+    yep = YepSearch(
+        timeout=timeout,
+        proxies={"https": proxy, "http": proxy} if proxy else None,
+        verify=True,
+        impersonate=impersonate
+    )
+    
+    try:
+        results = yep.images(
+            keywords=keywords,
+            region=region,
+            safesearch=safesearch,
+            max_results=max_results
+        )
+        
+        _print_data(results)
+    except Exception as e:
+        raise e
+
+@app.command()
+@option("--query", "-q", help="Search query", required=True)
+@option("--region", "-r", help="Region for suggestions", default="all")
+@option("--timeout", "-timeout", help="Timeout value for requests", type=int, default=20)
+@option("--proxy", "-p", help="Proxy URL to use for requests")
+@option("--impersonate", "-i", help="Browser to impersonate", default="chrome110")
+def yep_suggestions(
+    query: str,
+    region: str,
+    timeout: int = 20,
+    proxy: str = None,
+    impersonate: str = "chrome110"
+):
+    """Get search suggestions from Yep Search."""
+    yep = YepSearch(
+        timeout=timeout,
+        proxies={"https": proxy, "http": proxy} if proxy else None,
+        verify=True,
+        impersonate=impersonate
+    )
+    
+    try:
+        results = yep.suggestions(query=query, region=region)
+        
+        # Format suggestions for printing
+        formatted_results = []
+        for i, suggestion in enumerate(results, 1):
+            formatted_results.append({"position": i, "suggestion": suggestion})
+            
+        _print_data(formatted_results)
+    except Exception as e:
+        raise e
+
 def main():
     """Main entry point for the CLI."""
     try: