webscout 7.3__py3-none-any.whl → 7.5__py3-none-any.whl

webscout/Provider/AISEARCH/__init__.py

@@ -1,3 +1,4 @@
-from .felo_search import Felo
-from .DeepFind import DeepFind
-from .ISou import Isou
+from .felo_search import *
+from .DeepFind import *
+from .ISou import *
+from .genspark_search import *

webscout/Provider/AISEARCH/genspark_search.py

@@ -0,0 +1,208 @@
+import cloudscraper
+from uuid import uuid4
+import json
+import re
+from typing import Dict, Optional, Generator, Union, Any
+
+from webscout.AIbase import AISearch
+from webscout import exceptions
+from webscout import LitAgent
+
+
+class Response:
+    """A wrapper class for Genspark API responses.
+    
+    This class automatically converts response objects to their text representation
+    when printed or converted to string.
+    
+    Attributes:
+        text (str): The text content of the response
+        
+    Example:
+        >>> response = Response("Hello, world!")
+        >>> print(response)
+        Hello, world!
+        >>> str(response)
+        'Hello, world!'
+    """
+    def __init__(self, text: str):
+        self.text = text
+    
+    def __str__(self):
+        return self.text
+    
+    def __repr__(self):
+        return self.text
+
+
+class Genspark(AISearch):
+    """A class to interact with the Genspark AI search API.
+    
+    Genspark provides a powerful search interface that returns AI-generated responses
+    based on web content. It supports both streaming and non-streaming responses.
+    
+    Basic Usage:
+        >>> from webscout import Genspark
+        >>> ai = Genspark()
+        >>> # Non-streaming example
+        >>> response = ai.search("What is Python?")
+        >>> print(response)
+        Python is a high-level programming language...
+        
+        >>> # Streaming example
+        >>> for chunk in ai.search("Tell me about AI", stream=True):
+        ...     print(chunk, end="", flush=True)
+        Artificial Intelligence is...
+        
+        >>> # Raw response format
+        >>> for chunk in ai.search("Hello", stream=True, raw=True):
+        ...     print(chunk)
+        {'text': 'Hello'}
+        {'text': ' there!'}
+    
+    Args:
+        timeout (int, optional): Request timeout in seconds. Defaults to 30.
+        proxies (dict, optional): Proxy configuration for requests. Defaults to None.
+        max_tokens (int, optional): Maximum tokens to generate. Defaults to 600.
+    """
+
+    def __init__(
+        self,
+        timeout: int = 30,
+        proxies: Optional[dict] = None,
+        max_tokens: int = 600,
+    ):
+        """Initialize the Genspark API client.
+        
+        Args:
+            timeout (int, optional): Request timeout in seconds. Defaults to 30.
+            proxies (dict, optional): Proxy configuration for requests. Defaults to None.
+            max_tokens (int, optional): Maximum tokens to generate. Defaults to 600.
+        """
+        self.session = cloudscraper.create_scraper()
+        self.max_tokens = max_tokens
+        self.chat_endpoint = "https://www.genspark.ai/api/search/stream"
+        self.stream_chunk_size = 64
+        self.timeout = timeout
+        self.last_response = {}
+        
+        self.headers = {
+            "Accept": "*/*",
+            "Accept-Encoding": "gzip, deflate, br, zstd",
+            "Accept-Language": "en-US,en;q=0.9,en-IN;q=0.8",
+            "Content-Type": "application/json",
+            "DNT": "1",
+            "Origin": "https://www.genspark.ai",
+            "Priority": "u=1, i",
+            "Sec-CH-UA": '"Chromium";v="128", "Not;A=Brand";v="24", "Microsoft Edge";v="128"',
+            "Sec-CH-UA-Mobile": "?0",
+            "Sec-CH-UA-Platform": '"Windows"',
+            "Sec-Fetch-Dest": "empty",
+            "Sec-Fetch-Mode": "cors",
+            "Sec-Fetch-Site": "same-origin",
+            "User-Agent": LitAgent().random(),
+        }
+        
+        self.cookies = {
+            "i18n_redirected": "en-US",
+            "agree_terms": "0",
+            "session_id": uuid4().hex,
+        }
+
+        self.session.headers.update(self.headers)
+        self.session.proxies = proxies or {}
+
+    def search(
+        self,
+        prompt: str,
+        stream: bool = False,
+        raw: bool = False,
+    ) -> Union[Dict[str, Any], Generator[Union[Dict[str, Any], str], None, None]]:
+        """Search using the Genspark API and get AI-generated responses.
+        
+        Args:
+            prompt (str): The search query or prompt to send to the API.
+            stream (bool, optional): If True, yields response chunks as they arrive.
+                                   If False, returns complete response. Defaults to False.
+            raw (bool, optional): If True, returns raw response dictionaries with 'text' key.
+                                If False, returns Response objects that convert to text automatically.
+                                Defaults to False.
+        
+        Returns:
+            Union[Dict[str, Any], Generator[Union[Dict[str, Any], str], None, None]]: 
+                - If stream=False: Returns complete response
+                - If stream=True: Yields response chunks as they arrive
+        
+        Raises:
+            APIConnectionError: If the API request fails
+        """
+        url = f"https://www.genspark.ai/api/search/stream?query={prompt.replace(' ', '%20')}"
+        
+        def for_stream():
+            try:
+                with self.session.post(
+                    url,
+                    headers=self.headers,
+                    cookies=self.cookies,
+                    json={},  # Empty payload as query is in URL
+                    stream=True,
+                    timeout=self.timeout,
+                ) as response:
+                    if not response.ok:
+                        raise exceptions.APIConnectionError(
+                            f"Failed to generate response - ({response.status_code}, {response.reason}) - {response.text}"
+                        )
+                    
+                    for line in response.iter_lines(decode_unicode=True):
+                        if line and line.startswith("data: "):
+                            try:
+                                data = json.loads(line[6:])
+                                if "field_name" in data and "delta" in data:
+                                    if data["field_name"].startswith("streaming_detail_answer"):
+                                        delta_text = data.get("delta", "")
+                                        
+                                        # Clean up markdown links in text
+                                        delta_text = re.sub(r"\[.*?\]\(.*?\)", "", delta_text)
+                                        
+                                        if raw:
+                                            yield {"text": delta_text}
+                                        else:
+                                            yield Response(delta_text)
+                            except json.JSONDecodeError:
+                                continue
+                                
+            except cloudscraper.exceptions as e:
+                raise exceptions.APIConnectionError(f"Request failed: {e}")
+                
+        def for_non_stream():
+            full_response = ""
+            for chunk in for_stream():
+                if raw:
+                    yield chunk
+                else:
+                    full_response += str(chunk)
+            
+            if not raw:
+                # Process the full response to clean up any JSON structures
+                try:
+                    text_json = json.loads(full_response)
+                    if isinstance(text_json, dict) and "detailAnswer" in text_json:
+                        full_response = text_json.get("detailAnswer", full_response)
+                except (json.JSONDecodeError, TypeError):
+                    # Not valid JSON or not a dictionary, keep as is
+                    pass
+                    
+                self.last_response = Response(full_response)
+                return self.last_response
+
+        return for_stream() if stream else for_non_stream()
+
+
+if __name__ == "__main__":
+
+    from rich import print
+    
+    ai = Genspark()
+    response = ai.search(input(">>> "), stream=True, raw=False)
+    for chunk in response:
+        print(chunk, end="", flush=True)

webscout/Provider/AllenAI.py

@@ -0,0 +1,282 @@
+import requests
+import json
+import os
+from uuid import uuid4
+from typing import Any, Dict, Optional, Generator, Union
+
+from webscout.AIutel import Optimizers
+from webscout.AIutel import Conversation
+from webscout.AIutel import AwesomePrompts, sanitize_stream
+from webscout.AIbase import Provider, AsyncProvider
+from webscout import exceptions
+from webscout import LitAgent
+
+class AllenAI(Provider):
+    """
+    A class to interact with the AllenAI (Ai2 Playground) API.
+    """
+
+    AVAILABLE_MODELS = [
+        'tulu3-405b',
+        'OLMo-2-1124-13B-Instruct',
+        'tulu-3-1-8b',
+        'Llama-3-1-Tulu-3-70B',
+        'olmoe-0125'
+    ]
+
+
+    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 = "tulu3-405b",
+        system_prompt: str = "You are a helpful AI assistant.",
+    ):
+        """Initializes the AllenAI API client."""
+        if model not in self.AVAILABLE_MODELS:
+            raise ValueError(f"Invalid model: {model}. Choose from: {self.AVAILABLE_MODELS}")
+            
+        self.url = "https://playground.allenai.org"
+        self.api_endpoint = "https://olmo-api.allen.ai/v4/message/stream"
+        
+        # Use LitAgent for user-agent
+        self.headers = {
+            'User-Agent': LitAgent().random(),
+            'Accept': '*/*',
+            'Accept-Language': 'en-US,en;q=0.9',
+            'Origin': self.url,
+            'Referer': f"{self.url}/",
+            'Connection': 'keep-alive',
+            'Cache-Control': 'no-cache',
+            'Pragma': 'no-cache',
+            'Sec-Fetch-Dest': 'empty',
+            'Sec-Fetch-Mode': 'cors',
+            'Sec-Fetch-Site': 'same-site',
+            'sec-ch-ua': '"Google Chrome";v="119", "Chromium";v="119", "Not?A_Brand";v="24"',
+            'sec-ch-ua-mobile': '?0',
+            'sec-ch-ua-platform': '"Windows"'
+        }
+        
+        self.session = requests.Session()
+        self.session.headers.update(self.headers)
+        self.session.proxies.update(proxies)
+        self.model = model
+        self.system_prompt = system_prompt
+        self.is_conversation = is_conversation
+        self.max_tokens_to_sample = max_tokens
+        self.timeout = timeout
+        self.last_response = {}       
+        # Generate user ID if needed
+        self.x_anonymous_user_id = str(uuid4())
+        self.parent = None
+
+        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 format_prompt(self, messages):
+        """Format messages into a prompt string"""
+        formatted = []
+        for msg in messages:
+            role = msg.get("role", "")
+            content = msg.get("content", "")
+            if role == "system":
+                formatted.append(f"System: {content}")
+            elif role == "user":
+                formatted.append(f"User: {content}")
+            elif role == "assistant":
+                formatted.append(f"Assistant: {content}")
+        return "\n".join(formatted)
+
+    def ask(
+        self,
+        prompt: str,
+        stream: bool = False,
+        raw: bool = False,
+        optimizer: str = None,
+        conversationally: bool = False,
+        host: str = "inferd",
+        private: bool = True,
+        top_p: float = None,
+        temperature: float = 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}")
+
+        # Generate boundary for multipart form
+        boundary = f"----WebKitFormBoundary{uuid4().hex}"
+        
+        # Set content-type header for this specific request
+        self.session.headers.update({
+            "content-type": f"multipart/form-data; boundary={boundary}",
+            "x-anonymous-user-id": self.x_anonymous_user_id
+        })
+        
+        # Format messages for AllenAI
+        messages = [
+            {"role": "system", "content": self.system_prompt},
+            {"role": "user", "content": conversation_prompt}
+        ]
+        
+        # Build multipart form data
+        form_data = [
+            f'--{boundary}\r\n'
+            f'Content-Disposition: form-data; name="model"\r\n\r\n{self.model}\r\n',
+            
+            f'--{boundary}\r\n'
+            f'Content-Disposition: form-data; name="host"\r\n\r\n{host}\r\n',
+            
+            f'--{boundary}\r\n'
+            f'Content-Disposition: form-data; name="content"\r\n\r\n{self.format_prompt(messages)}\r\n',
+            
+            f'--{boundary}\r\n'
+            f'Content-Disposition: form-data; name="private"\r\n\r\n{str(private).lower()}\r\n'
+        ]
+        
+        # Add parent if exists
+        if self.parent:
+            form_data.append(
+                f'--{boundary}\r\n'
+                f'Content-Disposition: form-data; name="parent"\r\n\r\n{self.parent}\r\n'
+            )
+        
+        # Add optional parameters
+        if temperature is not None:
+            form_data.append(
+                f'--{boundary}\r\n'
+                f'Content-Disposition: form-data; name="temperature"\r\n\r\n{temperature}\r\n'
+            )
+        
+        if top_p is not None:
+            form_data.append(
+                f'--{boundary}\r\n'
+                f'Content-Disposition: form-data; name="top_p"\r\n\r\n{top_p}\r\n'
+            )
+        
+        form_data.append(f'--{boundary}--\r\n')
+        data = "".join(form_data).encode()
+
+        def for_stream():
+            nonlocal data  # Explicitly capture the data variable from outer scope
+            try:
+                response = self.session.post(
+                    self.api_endpoint,
+                    data=data,
+                    stream=True,
+                    timeout=self.timeout
+                )
+                
+                if response.status_code != 200:
+                    raise exceptions.FailedToGenerateResponseError(
+                        f"Request failed with status code {response.status_code}: {response.text}"
+                    )
+                
+                streaming_text = ""
+                current_parent = None
+                
+                for chunk in response.iter_content(chunk_size=1024, decode_unicode=False):
+                    if not chunk:
+                        continue
+                        
+                    decoded = chunk.decode(errors="ignore")
+                    for line in decoded.splitlines():
+                        line = line.strip()
+                        if not line:
+                            continue
+                        
+                        try:
+                            data = json.loads(line)
+                        except json.JSONDecodeError:
+                            continue
+                        
+                        if isinstance(data, dict):
+                            # Update the parent ID
+                            if data.get("children"):
+                                for child in data["children"]:
+                                    if child.get("role") == "assistant":
+                                        current_parent = child.get("id")
+                                        break
+                            
+                            # Process content
+                            if "message" in data and data.get("content"):
+                                content = data["content"]
+                                if content.strip():
+                                    streaming_text += content
+                                    resp = dict(text=content)
+                                    yield resp if raw else resp
+                            
+                            # Handle completion
+                            if data.get("final") or data.get("finish_reason") == "stop":
+                                if current_parent:
+                                    self.parent = current_parent
+                                
+                                # Update conversation history
+                                self.conversation.update_chat_history(prompt, streaming_text)
+                                self.last_response = {"text": streaming_text}
+                                return
+                
+            except requests.RequestException as e:
+                raise exceptions.FailedToGenerateResponseError(f"Request failed: {str(e)}")
+
+        def for_non_stream():
+            streaming_text = ""
+            for resp in for_stream():
+                streaming_text += resp["text"]
+            self.last_response = {"text": streaming_text}
+            return self.last_response
+
+        return for_stream() if stream else for_non_stream()
+
+    def chat(
+        self,
+        prompt: str,
+        stream: bool = False,
+        optimizer: str = None,
+        conversationally: bool = False,
+    ) -> str:
+        def for_stream():
+            for response in self.ask(prompt, True, optimizer=optimizer, conversationally=conversationally):
+                yield self.get_message(response)
+        def for_non_stream():
+            return self.get_message(
+                self.ask(prompt, False, optimizer=optimizer, conversationally=conversationally)
+            )
+        return for_stream() if stream else for_non_stream()
+
+    def get_message(self, response: dict) -> str:
+        assert isinstance(response, dict), "Response should be of dict data-type only"
+        return response["text"]
+
+if __name__ == "__main__":
+    from rich import print
+    ai = AllenAI(timeout=5000)
+    response = ai.chat("write a poem about AI", stream=True)
+    for chunk in response:
+        print(chunk, end="", flush=True)