Feat/text rewriter

[Sarang-Nambiar]

Description

This PR introduces the Text-Rewriter Tool, a feature that enables users to rewrite text with specific instructions or by uploading files. Below are the main functionalities implemented in this PR:

Key Features

• Accept input text directly or via file uploads in the following formats:
  • Supported Formats: CSV, PDF, DOCX, PPT, Plain Text.
  • URLs: YouTube videos, websites, Google Sheets.
• Process the uploaded content based on the provided rewriting instructions (e.g., simplify, summarize, rephrase).
• Generate rewritten text tailored to user instructions.
• Export rewritten text in formats such as Plain Text, DOCX, or PDF.

---

Related Issue

This PR addresses the following issue:
#128

---

Type of Change

Please select the type(s) of change that apply:

• New feature: A non-breaking change that adds functionality.

---

Proposed Solution

The implementation follows a modular structure and integrates seamlessly with the existing codebase. Key details include:

• Structure:

  • The tool is implemented within app/features/text-rewriter.
  • Core functionalities such as data validation and document retrieval are defined in core.py.
  • The main business logic resides in tools.py.

• Rewriting Logic:

  • Input text and rewriting instructions are sent to the LLM for processing.
  • The LLM response is parsed using JsonOutputParser from LangChain.
  • The processed output is returned as a response to the user.

---

How to Test

Testing Instructions

1. Start the FastAPI server from the root directory:

   ./local-start.sh

2. Use Postman or a similar tool to send a POST request to:
   http://localhost:8000/submit-tool

   Headers:

   • api-key: dev

   JSON Body:

   {
     "user": {
       "id": "string",
       "fullName": "string",
       "email": "string"
     },
     "type": "tool",
     "tool_data": {
       "tool_id": "text-rewriter",
       "inputs": [
         {
           "name": "instructions",
           "value": "Rephrase it in a formal tone"
         },
         {
           "name": "raw_text",
           "value": "The quick brown fox jumped over the fence."
         },
         {
           "name": "file_url",
           "value": ""
         },
         {
           "name": "file_type",
           "value": ""
         }
       ]
     }
   }

3. Expected Output:

   {
     "data": {
       "rewritten_text": "The agile brown fox successfully cleared the fence."
     }
   }

---

Unit Tests

Tests Added:

• Valid Inputs: Ensures successful processing for all supported file types.
• Invalid Inputs: Validates error handling for incorrect formats or missing data.

---

Documentation Updates

Does this PR require documentation updates?

• Yes
• No

If yes, provide the necessary details and link the relevant documentation.

---

Checklist

• Performed a self-review of the code.
• Added comments to clarify complex areas.
• Updated documentation where necessary.
• Verified all unit tests pass locally.
• Added unit tests to verify functionality.
• No new warnings are introduced with these changes.
• Ensured dependent changes are merged and published downstream.

---

Additional Information

Feel free to reach out for further clarification or if additional changes are required.

---

[stevenrayhinojosa-gmail-com]

please include Langchain

[Sarang-Nambiar]

I have made the requested changes

[stevenrayhinojosa-gmail-com]

Awesome, I will run this on my machine and I will get back to you, great job !