askai-python 0.2.0__tar.gz → 0.2.1__tar.gz

{askai_python-0.2.0/askai_python.egg-info → askai_python-0.2.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: askai-python
-Version: 0.2.0
+Version: 0.2.1
 Summary: AI Made Stupid Simple. Unified API for OpenAI, Groq, Google, and more.
 Author-email: Hossein Ghorbani <hosseingh1068@gmail.com>
 License: MIT
@@ -53,16 +53,6 @@ print(Groq().ask("Explain black holes like I'm 5").text)
 - **Zero config**: Keys are pulled from the environment automatically
 - **SDK-first, not a framework**: It stays out of your way.
 
-## ⚖️ How it compares
-
-| Feature        | ask-ai | LangChain |
-| -------------- | ------ | --------- |
-| Setup time     | 30 sec | 1 hour    |
-| Learning curve | ⭐      | ⭐⭐⭐⭐⭐     |
-| Async support  | ⏳ *(Coming soon)* | ⚠️ Complex |
-| Retry/Timeout  | ✅ Built-in | ❌ Manual |
-| Gateway needed | ❌ No      | ❌ No      |
-
 ## 🚫 What this project is NOT
 
 > ❌ Not an AI framework  
@@ -73,6 +63,33 @@ It does one thing perfectly: **Simplifying the API call to LLMs.**
 
 ---
 
+## 🛠️ Advanced Usage
+
+### 🧰 Developer QoL Utilities (Auto-Parsing)
+Stop writing Regex to clean up model outputs! `ask-ai` comes with built-in text processing flags:
+
+```python
+from ask_ai import OpenAI
+ai = OpenAI()
+
+# 1. Clean Markdown (Removes ```json and ``` tags)
+# Perfect for extracting raw data from models that wrap everything in markdown
+clean_text = ai.ask("Write JSON", clean=True).text
+
+# 2. Extract Code (Returns ONLY the code block, ignores conversational filler)
+# Great for automation pipelines
+code = ai.ask("Write a python ping script", code=True).text
+
+# 3. Strip Tags (Removes <think> blocks and HTML)
+# Essential for reasoning models like DeepSeek-R1
+answer_only = ai.ask("What is 1+1?", strip=True).text
+
+# 4. Enforce & Parse JSON (Directly returns a Parsed Python Dictionary)
+# Adds JSON instructions to the prompt and safely runs json.loads()
+data_dict = ai.ask("Extract user info", json=True).json
+print(data_dict['name'])
+```
+
 ## 🚀 Built-in Resiliency (Retries & Timeouts)
 
 Build reliable apps without writing your own loops. `askai-python` handles rate limits (`429`) and network drops via an internal exponential backoff.

{askai_python-0.2.0 → askai_python-0.2.1}/README.md

@@ -48,17 +48,6 @@ Groq().ask("Explain black holes like I'm 5").text
 - **Zero config**: Keys are pulled from the environment automatically
 - **SDK-first, not a framework**: It stays out of your way.
 
-## ⚖️ How it compares
-
-| Feature        | ask-ai | LangChain |
-| -------------- | ------ | --------- |
-| Setup time     | 30 sec | 1 hour    |
-| Learning curve | ⭐      | ⭐⭐⭐⭐⭐     |
-| Async support  | ⏳ *(Coming soon)* | ⚠️ Complex |
-| Retry/Timeout  | ✅ Built-in | ❌ Manual |
-| Gateway needed | ❌ No      | ❌ No      |
-| Lines to switch| **1**      | 20+       |
-
 ## 🚫 What this project is NOT
 
 > ❌ Not an AI framework  
@@ -71,7 +60,32 @@ It does one thing perfectly: **Simplifying the API call to LLMs.**
 
 ## 🛠️ Advanced Usage
 
-### Built-in Retries & Timeout (Resiliency)
+### 🧰 Developer QoL Utilities (Auto-Parsing)
+Stop writing Regex to clean up model outputs! `ask-ai` comes with built-in text processing flags:
+
+```python
+from ask_ai import OpenAI
+ai = OpenAI()
+
+# 1. Clean Markdown (Removes ```json and ``` tags)
+# Perfect for extracting raw data from models that wrap everything in markdown
+clean_text = ai.ask("Write JSON", clean=True).text
+
+# 2. Extract Code (Returns ONLY the code block, ignores conversational filler)
+# Great for automation pipelines
+code = ai.ask("Write a python ping script", code=True).text
+
+# 3. Strip Tags (Removes <think> blocks and HTML)
+# Essential for reasoning models like DeepSeek-R1
+answer_only = ai.ask("What is 1+1?", strip=True).text
+
+# 4. Enforce & Parse JSON (Directly returns a Parsed Python Dictionary)
+# Adds JSON instructions to the prompt and safely runs json.loads()
+data_dict = ai.ask("Extract user info", json=True).json
+print(data_dict['name'])
+```
+
+### 🔄 Built-in Retries & Timeout (Resiliency)
 Handle rate limits (`429`) and network drops automatically:
 
 ```python

{askai_python-0.2.0 → askai_python-0.2.1}/ask_ai/base.py

@@ -9,9 +9,30 @@ class Response:
     """
     Unified response object for all easyai requests.
     """
-    def __init__(self, text: str = "", media: Union[ImageObject, AudioObject, None] = None):
-        self.text = text
+    def __init__(self, text: str = "", media: Union[ImageObject, AudioObject, None] = None, **kwargs):
+        self._raw_text = text
         self.media = media
+        self._kwargs = kwargs
+
+    @property
+    def text(self) -> str:
+        text = self._raw_text
+        if self._kwargs.get('strip'):
+            from .utils import strip_tags
+            text = strip_tags(text)
+        if self._kwargs.get('clean'):
+            from .utils import clean_markdown
+            text = clean_markdown(text)
+        if self._kwargs.get('code'):
+            from .utils import extract_code
+            text = extract_code(text)
+        return text
+
+    @property
+    def json(self) -> Union[Dict[str, Any], list]:
+        """Returns parsed JSON. If 'json=True' was passed, this is highly reliable."""
+        from .utils import parse_json
+        return parse_json(self._raw_text)
 
     def __str__(self):
         return self.text
@@ -96,7 +117,10 @@ class BaseProvider:
         last_error = None
         while attempts <= retry:
             try:
-                return self._send_request(messages, final_config, output_type)
+                response = self._send_request(messages, final_config, output_type)
+                # Attach parsing arguments to the response object dynamically
+                response._kwargs.update(kwargs)
+                return response
             except Exception as e:
                 last_error = e
                 if self._is_retryable_error(e) and attempts < retry:
@@ -133,10 +157,16 @@ class BaseProvider:
 
     def _prepare_messages(self, query: str, config: AdvancedConfig) -> List[Dict[str, str]]:
         messages = []
-        if config.system_message:
-            messages.append({"role": "system", "content": config.system_message})
-        elif self.persona:
-            messages.append({"role": "system", "content": self.persona})
+        
+        system_msg = config.system_message or self.persona or ""
+        
+        # Smart capability: if json=True, enforce JSON system prompt
+        if config.extra.get('json'):
+            json_instruction = "You must respond with valid JSON only, without any markdown formatting or tags."
+            system_msg = f"{system_msg}\n{json_instruction}".strip()
+
+        if system_msg:
+            messages.append({"role": "system", "content": system_msg})
         
         messages.append({"role": "user", "content": query})
         return messages

{askai_python-0.2.0 → askai_python-0.2.1}/ask_ai/exceptions.py

@@ -22,5 +22,9 @@ class MediaTypeNotSupportedError(AskAIError):
     """Raised when a provider can't handle the requested media type."""
     pass
 
+class AskAIParsingError(AskAIError):
+    """Raised when expect_json is True but the response cannot be parsed."""
+    pass
+
 # Alias for backwards compatibility internally
 APIKeyError = AskAIConfigError

askai_python-0.2.1/ask_ai/utils.py

@@ -0,0 +1,43 @@
+import re
+import json
+from typing import Dict, Any, Union
+
+def clean_markdown(text: str) -> str:
+    """Removes markdown code blocks like ```json ... ```"""
+    text = text.strip()
+    if text.startswith("```"):
+        # find the first newline
+        first_newline = text.find("\n")
+        if first_newline != -1:
+            text = text[first_newline+1:]
+        if text.endswith("```"):
+            text = text[:-3]
+    return text.strip()
+
+def extract_code(text: str) -> str:
+    """Extracts code blocks from text, ignoring conversational filler."""
+    text = str(text)
+    matches = re.findall(r'```(?:[\w]*\n)?(.*?)```', text, re.DOTALL)
+    if matches:
+        return "\n\n".join(m.strip() for m in matches)
+    return text
+
+def strip_tags(text: str) -> str:
+    """Removes XML/HTML tags, including <think>...</think> blocks."""
+    # First remove <think>...</think> entirely if we don't want the thought process
+    text = re.sub(r'<think>.*?</think>', '', text, flags=re.DOTALL)
+    # Remove any other HTML tags
+    text = re.sub(r'<[^>]+>', '', text)
+    return text.strip()
+
+def parse_json(text: str) -> Union[Dict[str, Any], list]:
+    """Tries to parse JSON from text, employing cleaning if necessary."""
+    try:
+        return json.loads(text)
+    except json.JSONDecodeError:
+        cleaned = clean_markdown(text)
+        try:
+            return json.loads(cleaned)
+        except json.JSONDecodeError as e:
+            from .exceptions import AskAIParsingError
+            raise AskAIParsingError(f"Failed to parse JSON: {e}\nContent snippet: {text[:100]}...")

{askai_python-0.2.0 → askai_python-0.2.1/askai_python.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: askai-python
-Version: 0.2.0
+Version: 0.2.1
 Summary: AI Made Stupid Simple. Unified API for OpenAI, Groq, Google, and more.
 Author-email: Hossein Ghorbani <hosseingh1068@gmail.com>
 License: MIT
@@ -53,16 +53,6 @@ print(Groq().ask("Explain black holes like I'm 5").text)
 - **Zero config**: Keys are pulled from the environment automatically
 - **SDK-first, not a framework**: It stays out of your way.
 
-## ⚖️ How it compares
-
-| Feature        | ask-ai | LangChain |
-| -------------- | ------ | --------- |
-| Setup time     | 30 sec | 1 hour    |
-| Learning curve | ⭐      | ⭐⭐⭐⭐⭐     |
-| Async support  | ⏳ *(Coming soon)* | ⚠️ Complex |
-| Retry/Timeout  | ✅ Built-in | ❌ Manual |
-| Gateway needed | ❌ No      | ❌ No      |
-
 ## 🚫 What this project is NOT
 
 > ❌ Not an AI framework  
@@ -73,6 +63,33 @@ It does one thing perfectly: **Simplifying the API call to LLMs.**
 
 ---
 
+## 🛠️ Advanced Usage
+
+### 🧰 Developer QoL Utilities (Auto-Parsing)
+Stop writing Regex to clean up model outputs! `ask-ai` comes with built-in text processing flags:
+
+```python
+from ask_ai import OpenAI
+ai = OpenAI()
+
+# 1. Clean Markdown (Removes ```json and ``` tags)
+# Perfect for extracting raw data from models that wrap everything in markdown
+clean_text = ai.ask("Write JSON", clean=True).text
+
+# 2. Extract Code (Returns ONLY the code block, ignores conversational filler)
+# Great for automation pipelines
+code = ai.ask("Write a python ping script", code=True).text
+
+# 3. Strip Tags (Removes <think> blocks and HTML)
+# Essential for reasoning models like DeepSeek-R1
+answer_only = ai.ask("What is 1+1?", strip=True).text
+
+# 4. Enforce & Parse JSON (Directly returns a Parsed Python Dictionary)
+# Adds JSON instructions to the prompt and safely runs json.loads()
+data_dict = ai.ask("Extract user info", json=True).json
+print(data_dict['name'])
+```
+
 ## 🚀 Built-in Resiliency (Retries & Timeouts)
 
 Build reliable apps without writing your own loops. `askai-python` handles rate limits (`429`) and network drops via an internal exponential backoff.

{askai_python-0.2.0 → askai_python-0.2.1}/askai_python.egg-info/SOURCES.txt

@@ -8,6 +8,7 @@ ask_ai/cli.py
 ask_ai/config.py
 ask_ai/exceptions.py
 ask_ai/media.py
+ask_ai/utils.py
 ask_ai/providers/__init__.py
 ask_ai/providers/anthropic_provider.py
 ask_ai/providers/azure_provider.py

{askai_python-0.2.0 → askai_python-0.2.1}/docs/PYPI_README.md

@@ -34,16 +34,6 @@ print(Groq().ask("Explain black holes like I'm 5").text)
 - **Zero config**: Keys are pulled from the environment automatically
 - **SDK-first, not a framework**: It stays out of your way.
 
-## ⚖️ How it compares
-
-| Feature        | ask-ai | LangChain |
-| -------------- | ------ | --------- |
-| Setup time     | 30 sec | 1 hour    |
-| Learning curve | ⭐      | ⭐⭐⭐⭐⭐     |
-| Async support  | ⏳ *(Coming soon)* | ⚠️ Complex |
-| Retry/Timeout  | ✅ Built-in | ❌ Manual |
-| Gateway needed | ❌ No      | ❌ No      |
-
 ## 🚫 What this project is NOT
 
 > ❌ Not an AI framework  
@@ -54,6 +44,33 @@ It does one thing perfectly: **Simplifying the API call to LLMs.**
 
 ---
 
+## 🛠️ Advanced Usage
+
+### 🧰 Developer QoL Utilities (Auto-Parsing)
+Stop writing Regex to clean up model outputs! `ask-ai` comes with built-in text processing flags:
+
+```python
+from ask_ai import OpenAI
+ai = OpenAI()
+
+# 1. Clean Markdown (Removes ```json and ``` tags)
+# Perfect for extracting raw data from models that wrap everything in markdown
+clean_text = ai.ask("Write JSON", clean=True).text
+
+# 2. Extract Code (Returns ONLY the code block, ignores conversational filler)
+# Great for automation pipelines
+code = ai.ask("Write a python ping script", code=True).text
+
+# 3. Strip Tags (Removes <think> blocks and HTML)
+# Essential for reasoning models like DeepSeek-R1
+answer_only = ai.ask("What is 1+1?", strip=True).text
+
+# 4. Enforce & Parse JSON (Directly returns a Parsed Python Dictionary)
+# Adds JSON instructions to the prompt and safely runs json.loads()
+data_dict = ai.ask("Extract user info", json=True).json
+print(data_dict['name'])
+```
+
 ## 🚀 Built-in Resiliency (Retries & Timeouts)
 
 Build reliable apps without writing your own loops. `askai-python` handles rate limits (`429`) and network drops via an internal exponential backoff.

{askai_python-0.2.0 → askai_python-0.2.1}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "askai-python"
-version = "0.2.0"
+version = "0.2.1"
 description = "AI Made Stupid Simple. Unified API for OpenAI, Groq, Google, and more."
 readme = "docs/PYPI_README.md"
 requires-python = ">=3.8"