webscout 8.3.1__py3-none-any.whl → 8.3.3__py3-none-any.whl

webscout/Provider/AISEARCH/__init__.py

@@ -1,3 +1,4 @@
+from .stellar_search import *
 from .felo_search import *
 from .DeepFind import *
 from .genspark_search import *

webscout/Provider/AISEARCH/genspark_search.py

@@ -201,12 +201,12 @@ class Genspark(AISearch):
                     json={},
                     stream=True,
                     timeout=self.timeout,
-                ) as SearchResponse:
-                    if not SearchResponse.ok:
+                ) as resp:
+                    if not resp.ok:
                         raise exceptions.APIConnectionError(
-                            f"Failed to generate SearchResponse - ({SearchResponse.status_code}, {SearchResponse.reason}) - {SearchResponse.text}"
+                            f"Failed to generate SearchResponse - ({resp.status_code}, {resp.reason}) - {resp.text}"
                         )
-                    for line in SearchResponse.iter_lines(decode_unicode=True):
+                    for line in resp.iter_lines(decode_unicode=True):
                         if not line or not line.startswith("data: "):
                             continue
                         try:
@@ -287,7 +287,7 @@ class Genspark(AISearch):
                                     yield processed_event_payload
                         except json.JSONDecodeError:
                             continue
-            except cloudscraper.exceptions as e:
+            except cloudscraper.exceptions.CloudflareException as e:
                 raise exceptions.APIConnectionError(f"Request failed due to Cloudscraper issue: {e}")
             except requests.exceptions.RequestException as e:
                 raise exceptions.APIConnectionError(f"Request failed: {e}")
@@ -315,8 +315,8 @@ if __name__ == "__main__":
     from rich import print
     ai = Genspark()
     try:
-        SearchResponse = ai.search(input(">>> "), stream=True, raw=False)
-        for chunk in SearchResponse:
+        search_result_stream = ai.search(input(">>> "), stream=True, raw=False)
+        for chunk in search_result_stream:
             print(chunk, end="", flush=True)
     except KeyboardInterrupt:
         print("\nSearch interrupted by user.")

webscout/Provider/AISEARCH/stellar_search.py

@@ -0,0 +1,132 @@
+import requests
+import re
+from typing import Dict, Optional, Generator, Union, Any
+from webscout.AIbase import AISearch, SearchResponse
+from webscout import exceptions
+from webscout.litagent import LitAgent
+from webscout.AIutel import sanitize_stream
+
+class Stellar(AISearch):
+    """AI Search provider for stellar.chatastra.ai"""
+    def __init__(self, timeout: int = 30, proxies: Optional[dict] = None):
+        self.api_endpoint = "https://stellar.chatastra.ai/search/x1GUVzl"
+        self.timeout = timeout
+        self.proxies = proxies
+        self.session = requests.Session()
+        self.headers = {
+            "accept": "text/x-component",
+            "accept-encoding": "gzip, deflate, br, zstd",
+            "accept-language": "en-US,en;q=0.9,en-IN;q=0.8",
+            "content-type": "multipart/form-data; boundary=----WebKitFormBoundaryQsWD5Qs3QqDkNBPH",
+            "dnt": "1",
+            "next-action": "efc2643ed9bafe182a010b58ebea17f068ad3985",
+            "next-router-state-tree": "%5B%22%22%2C%7B%22children%22%3A%5B%22__PAGE__%22%2C%7B%7D%2C%22%2F%22%2C%22refresh%22%5D%7D%2Cnull%2Cnull%2Ctrue%5D",
+            "origin": "https://stellar.chatastra.ai",
+            "priority": "u=1, i",
+            "referer": "https://stellar.chatastra.ai/search/x1GUVzl",
+            "sec-ch-ua": '"Microsoft Edge";v="137", "Chromium";v="137", "Not/A)Brand";v="24"',
+            "sec-ch-ua-mobile": "?0",
+            "sec-ch-ua-platform": '"Windows"',
+            "sec-fetch-dest": "empty",
+            "sec-fetch-mode": "cors",
+            "sec-fetch-site": "same-origin",
+            "sec-gpc": "1",
+            "user-agent": LitAgent().random(),
+            "cookie": "__client_uat=0; __client_uat_K90aduOv=0",
+        }
+        self.session.headers.update(self.headers)
+        if proxies:
+            self.session.proxies = proxies
+
+    def _make_payload(self, prompt: str) -> bytes:        # This is a static payload for the demo; in production, generate dynamically as needed
+        boundary = "----WebKitFormBoundaryQsWD5Qs3QqDkNBPH"
+        parts = [
+            f"--{boundary}\r\nContent-Disposition: form-data; name=\"1\"\r\n\r\n{{\"id\":\"71bb616ba5b7cbcac2308fe0c249a9f2d51825b7\",\"bound\":null}}\r\n",
+            f"--{boundary}\r\nContent-Disposition: form-data; name=\"2\"\r\n\r\n{{\"id\":\"8bcca1d0cb933b14fefde88dacb2865be3d1d525\",\"bound\":null}}\r\n",
+            f"--{boundary}\r\nContent-Disposition: form-data; name=\"3_input\"\r\n\r\n{prompt}\r\n",
+            f"--{boundary}\r\nContent-Disposition: form-data; name=\"3_id\"\r\n\r\nx1GUVzl\r\n",
+            f"--{boundary}\r\nContent-Disposition: form-data; name=\"3_userId\"\r\n\r\nnull\r\n",
+            f"--{boundary}\r\nContent-Disposition: form-data; name=\"0\"\r\n\r\n[{{\"action\":\"$F1\",\"options\":{{\"onSetAIState\":\"$F2\"}}}},{{\"messages\":[],\"chatId\":\"\"}},\"$K3\"]\r\n",
+            f"--{boundary}--\r\n"
+        ]
+        return "".join(parts).encode("utf-8")
+
+    @staticmethod
+    def _stellar_extractor(chunk: Union[str, bytes, Dict[str, Any]]) -> Optional[str]:
+        """Extracts content from the Stellar stream format with hex keys and diff arrays. Handles both str and bytes input."""
+        if isinstance(chunk, bytes):
+            try:
+                chunk = chunk.decode('utf-8', errors='replace')
+            except Exception:
+                return None
+        if not isinstance(chunk, str):
+            return None
+        # Match patterns like 6e:{"diff":[0," empathy"],"next":"$@6f"}
+        pattern = r'[0-9a-f]+:\{"diff":\[0,"([^"\\]*)"\]'
+        matches = re.findall(pattern, chunk)
+        if matches:
+            extracted_text = ''.join(matches)
+            # Fix escaped newlines
+            extracted_text = extracted_text.replace('\\n', '\n').replace('\\n\\n', '\n\n')
+            return extracted_text if extracted_text.strip() else None
+        return None
+
+    def search(self, prompt: str, stream: bool = False, raw: bool = False) -> Union[SearchResponse, Generator[Union[Dict[str, str], SearchResponse, str], None, None]]:
+        payload = self._make_payload(prompt)
+        try:
+            response = self.session.post(
+                self.api_endpoint,
+                data=payload,
+                timeout=self.timeout,
+                proxies=self.proxies,
+                stream=stream,
+            )
+            if not response.ok:
+                raise exceptions.APIConnectionError(f"Failed to get response: {response.status_code} {response.text}")
+
+            def _yield_stream():
+                # Use sanitize_stream for real-time extraction from the response iterator
+                processed_stream = sanitize_stream(
+                    data=response.iter_lines(decode_unicode=True),
+                    intro_value=None,
+                    to_json=False,
+                    content_extractor=self._stellar_extractor
+                )
+                full_response = ""
+                for content in processed_stream:
+                    if content and isinstance(content, str):
+                        full_response += content
+                        if raw:
+                            yield {"text": content}
+                        else:
+                            yield content
+                # Do NOT yield SearchResponse(full_response) in streaming mode to avoid duplicate output
+
+            if stream:
+                return _yield_stream()
+            else:
+                # Use sanitize_stream for the full response text
+                processed_stream = sanitize_stream(
+                    data=response.text.splitlines(),
+                    intro_value=None,
+                    to_json=False,
+                    content_extractor=self._stellar_extractor
+                )
+                full_response = ""
+                for content in processed_stream:
+                    if content and isinstance(content, str):
+                        full_response += content
+                if raw:
+                    return {"text": full_response}
+                else:
+                    return SearchResponse(full_response)
+        except requests.RequestException as e:
+            raise exceptions.APIConnectionError(f"Request failed: {e}")
+
+if __name__ == "__main__":
+    from rich import print
+    ai = Stellar()
+    user_query = input(">>> ")
+    response = ai.search(user_query, stream=True, raw=False)
+    for chunk in response:
+        print(chunk, end="", flush=True)

webscout/Provider/ExaChat.py

@@ -2,7 +2,7 @@ from curl_cffi import CurlError
 from curl_cffi.requests import Session, Response # Import Response
 import json
 import uuid
-from typing import Any, Dict, Union, Optional, List
+from typing import Any, Dict, Union, Optional, List, Generator
 from datetime import datetime
 from webscout.AIutel import Optimizers, Conversation, AwesomePrompts, sanitize_stream # Import sanitize_stream
 from webscout.AIbase import Provider 
@@ -264,10 +264,11 @@ class ExaChat(Provider):
     def ask(
         self,
         prompt: str,
+        stream: bool = False,
         raw: bool = False,
         optimizer: str = None,
         conversationally: bool = False,
-    ) -> Dict[str, Any]:
+    ) -> Union[Dict[str, Any], Generator[Any, None, None]]:
         """Sends a prompt to the API and returns the response."""
         conversation_prompt = self.conversation.gen_complete_prompt(prompt)
         if optimizer:
@@ -281,78 +282,103 @@ class ExaChat(Provider):
 
         payload = self._build_payload(conversation_prompt)
         response = self._make_request(payload)
-        
-        try:
+        processed_stream = sanitize_stream(
+            data=response.iter_content(chunk_size=None),
+            intro_value=None,
+            to_json=True,
+            content_extractor=self._exachat_extractor,
+            yield_raw_on_error=False,
+            raw=raw
+        )
+        if stream:
+            streaming_text = ""
+            for content_chunk in processed_stream:
+                if content_chunk and isinstance(content_chunk, str):
+                    content_chunk = content_chunk.replace('\\\\', '\\').replace('\\"', '"')
+                if raw:
+                    if content_chunk and isinstance(content_chunk, str):
+                        streaming_text += content_chunk
+                    yield content_chunk
+                else:
+                    if content_chunk and isinstance(content_chunk, str):
+                        streaming_text += content_chunk
+                        yield dict(text=content_chunk)
+            self.last_response = {"text": streaming_text}
+            self.conversation.update_chat_history(prompt, streaming_text)
+        else:
             full_response = ""
-            # Use sanitize_stream to process the response
-            processed_stream = sanitize_stream(
-                data=response.iter_content(chunk_size=None), # Pass byte iterator
-                intro_value=None, # API doesn't seem to use 'data:' prefix
-                to_json=True,     # Stream sends JSON lines
-                content_extractor=self._exachat_extractor, # Use the specific extractor
-                yield_raw_on_error=False # Skip non-JSON lines or lines where extractor fails
-            )
-
             for content_chunk in processed_stream:
-                # content_chunk is the string extracted by _exachat_extractor
                 if content_chunk and isinstance(content_chunk, str):
-                    full_response += content_chunk
-                            
+                    content_chunk = content_chunk.replace('\\\\', '\\').replace('\\"', '"')
+                if raw:
+                    if content_chunk and isinstance(content_chunk, str):
+                        full_response += content_chunk
+                else:
+                    if content_chunk and isinstance(content_chunk, str):
+                        full_response += content_chunk
             self.last_response = {"text": full_response}
             self.conversation.update_chat_history(prompt, full_response)
-            return self.last_response if not raw else full_response # Return dict or raw string
-            
-        except json.JSONDecodeError as e:
-            raise exceptions.FailedToGenerateResponseError(f"Invalid JSON response: {e}") from e
+            return self.last_response if not raw else full_response
 
     def chat(
         self,
         prompt: str,
+        stream: bool = False,
         optimizer: str = None,
         conversationally: bool = False,
-    ) -> str:
-        """Generate response."""
-        response = self.ask(
-            prompt, optimizer=optimizer, conversationally=conversationally
-        )
-        return self.get_message(response)
+        raw: bool = False,
+    ) -> Union[str, Generator[str, None, None]]:
+        def for_stream():
+            for response in self.ask(
+                prompt, stream=True, raw=raw, optimizer=optimizer, conversationally=conversationally
+            ):
+                if raw:
+                    yield response
+                else:
+                    yield self.get_message(response)
+        def for_non_stream():
+            result = self.ask(
+                prompt, stream=False, raw=raw, optimizer=optimizer, conversationally=conversationally
+            )
+            if raw:
+                return result if isinstance(result, str) else str(result)
+            return self.get_message(result)
+        return for_stream() if stream else for_non_stream()
 
     def get_message(self, response: Union[Dict[str, Any], str]) -> str:
-        """
-        Retrieves message from response.
-        
-        Args:
-            response (Union[Dict[str, Any], str]): The response to extract the message from
-            
-        Returns:
-            str: The extracted message text
-        """
         if isinstance(response, dict):
-            return response.get("text", "")
-        return str(response)
+            text = response.get("text", "")
+        else:
+            text = str(response)
+        return text.replace('\\\\', '\\').replace('\\"', '"')
 
 if __name__ == "__main__":
-    print("-" * 80)
-    print(f"{'Model':<50} {'Status':<10} {'Response'}")
-    print("-" * 80)
+    # print("-" * 80)
+    # print(f"{'Model':<50} {'Status':<10} {'Response'}")
+    # print("-" * 80)
     
-    # Test all available models
-    working = 0
-    total = len(ExaChat.AVAILABLE_MODELS)
+    # # Test all available models
+    # working = 0
+    # total = len(ExaChat.AVAILABLE_MODELS)
     
-    for model in ExaChat.AVAILABLE_MODELS:
-        try:
-            test_ai = ExaChat(model=model, timeout=60)
-            response = test_ai.chat("Say 'Hello' in one word")
-            response_text = response
+    # for model in ExaChat.AVAILABLE_MODELS:
+    #     try:
+    #         test_ai = ExaChat(model=model, timeout=60)
+    #         response = test_ai.chat("Say 'Hello' in one word")
+    #         response_text = response
             
-            if response_text and len(response_text.strip()) > 0:
-                status = "✓"
-                # Truncate response if too long
-                display_text = response_text.strip()[:50] + "..." if len(response_text.strip()) > 50 else response_text.strip()
-            else:
-                status = "✗"
-                display_text = "Empty or invalid response"
-            print(f"{model:<50} {status:<10} {display_text}")
-        except Exception as e:
-            print(f"{model:<50} {'✗':<10} {str(e)}")
+    #         if response_text and len(response_text.strip()) > 0:
+    #             status = "✓"
+    #             # Truncate response if too long
+    #             display_text = response_text.strip()[:50] + "..." if len(response_text.strip()) > 50 else response_text.strip()
+    #         else:
+    #             status = "✗"
+    #             display_text = "Empty or invalid response"
+    #         print(f"{model:<50} {status:<10} {display_text}")
+    #     except Exception as e:
+    #         print(f"{model:<50} {'✗':<10} {str(e)}")
+    from rich import print
+    ai = ExaChat(model="gemini-2.0-flash")
+    response = ai.chat("tell me a joke", stream=True, raw=False)
+    for chunk in response:
+        print(chunk, end='', flush=True)

webscout/Provider/GeminiProxy.py

@@ -0,0 +1,140 @@
+from typing import Any, Dict, Optional, Union, Generator
+import requests
+import base64
+from webscout.litagent import LitAgent
+from webscout.AIutel import Optimizers, AwesomePrompts
+from webscout.AIutel import Conversation
+from webscout.AIbase import Provider
+from webscout import exceptions
+
+class GeminiProxy(Provider):
+    """
+    GeminiProxy is a provider class for interacting with the Gemini API via a proxy endpoint.
+    """
+    AVAILABLE_MODELS = [
+        "gemini-2.0-flash-lite",
+        "gemini-2.0-flash",
+        "gemini-2.5-pro-preview-06-05",
+        "gemini-2.5-pro-preview-05-06",
+        "gemini-2.5-flash-preview-04-17",
+        "gemini-2.5-flash-preview-05-20",
+
+    ]
+
+    def __init__(
+        self,
+        is_conversation: bool = True,
+        max_tokens: int = 2048,
+        timeout: int = 30,
+        intro: str = None,
+        filepath: str = None,
+        update_file: bool = True,
+        proxies: dict = {},
+        history_offset: int = 10250,
+        act: str = None,
+        model: str = "gemini-2.0-flash-lite",
+        system_prompt: str = "You are a helpful assistant.",
+        browser: str = "chrome"
+    ):
+        if model not in self.AVAILABLE_MODELS:
+            raise ValueError(f"Invalid model: {model}. Choose from: {self.AVAILABLE_MODELS}")
+        self.base_url = "https://us-central1-infinite-chain-295909.cloudfunctions.net/gemini-proxy-staging-v1"
+        self.agent = LitAgent()
+        self.fingerprint = self.agent.generate_fingerprint(browser)
+        self.headers = self.fingerprint.copy()
+        self.session = requests.Session()
+        self.session.headers.update(self.headers)
+        self.session.proxies.update(proxies)
+        self.is_conversation = is_conversation
+        self.max_tokens_to_sample = max_tokens
+        self.timeout = timeout
+        self.last_response = {}
+        self.model = model
+        self.system_prompt = system_prompt
+        self.__available_optimizers = (
+            method
+            for method in dir(Optimizers)
+            if callable(getattr(Optimizers, method)) and not method.startswith("__")
+        )
+        Conversation.intro = (
+            AwesomePrompts().get_act(
+                act, raise_not_found=True, default=None, case_insensitive=True
+            )
+            if act
+            else intro or Conversation.intro
+        )
+        self.conversation = Conversation(
+            is_conversation, self.max_tokens_to_sample, filepath, update_file
+        )
+        self.conversation.history_offset = history_offset
+
+    def get_image(self, img_url):
+        try:
+            response = self.session.get(img_url, stream=True, timeout=self.timeout)
+            response.raise_for_status()
+            mime_type = response.headers.get("content-type", "application/octet-stream")
+            data = base64.b64encode(response.content).decode("utf-8")
+            return {"mime_type": mime_type, "data": data}
+        except Exception as e:
+            raise exceptions.FailedToGenerateResponseError(f"Error fetching image: {e}")
+
+    def ask(
+        self,
+        prompt: str,
+        stream: bool = False,
+        raw: bool = False,
+        optimizer: str = None,
+        conversationally: bool = False,
+        img_url: Optional[str] = None,
+    ) -> Union[Dict[str, Any], Generator]:
+        conversation_prompt = self.conversation.gen_complete_prompt(prompt)
+        if optimizer:
+            if optimizer in self.__available_optimizers:
+                conversation_prompt = getattr(Optimizers, optimizer)(
+                    conversation_prompt if conversationally else prompt
+                )
+            else:
+                raise Exception(f"Optimizer is not one of {self.__available_optimizers}")
+        parts = []
+        if img_url:
+            parts.append({"inline_data": self.get_image(img_url)})
+        parts.append({"text": conversation_prompt})
+        request_data = {
+            "model": self.model,
+            "contents": [{"parts": parts}]
+        }
+        def for_non_stream():
+            try:
+                response = self.session.post(self.base_url, json=request_data, headers=self.headers, timeout=self.timeout)
+                response.raise_for_status()
+                data = response.json()
+                self.last_response = data
+                self.conversation.update_chat_history(prompt, self.get_message(data))
+                return data
+            except Exception as e:
+                raise exceptions.FailedToGenerateResponseError(f"Error during chat request: {e}")
+        # Gemini proxy does not support streaming, so only non-stream
+        return for_non_stream()
+
+    def chat(
+        self,
+        prompt: str,
+        stream: bool = False,
+        optimizer: str = None,
+        conversationally: bool = False,
+        img_url: Optional[str] = None,
+    ) -> str:
+        data = self.ask(prompt, stream=stream, optimizer=optimizer, conversationally=conversationally, img_url=img_url)
+        return self.get_message(data)
+
+    def get_message(self, response: dict) -> str:
+        assert isinstance(response, dict), "Response should be of dict data-type only"
+        try:
+            return response['candidates'][0]['content']['parts'][0]['text']
+        except Exception:
+            return str(response)
+
+if __name__ == "__main__":
+    ai = GeminiProxy(timeout=30, model="gemini-2.5-flash-preview-05-20")
+    response = ai.chat("write a poem about AI")
+    print(response)