Skip to content

Latest commit

 

History

History
93 lines (68 loc) · 2.72 KB

README.md

File metadata and controls

93 lines (68 loc) · 2.72 KB

Anyparser CrewAI

https://anyparser.com

Integrate Anyparser's powerful content extraction capabilities with CrewAI for enhanced AI workflows. This integration package enables seamless use of Anyparser's document processing and data extraction features within your CrewAI applications, making it easier than ever to build sophisticated AI pipelines.

Installation

pip install anyparser-crewai

Setup

Before running the examples, make sure to set your Anyparser API credentials as environment variables:

export ANYPARSER_API_KEY="your-api-key"
export ANYPARSER_API_URL="https://anyparserapi.com"

Anyparser LangChain Examples

This examples directory contains examples demonstrating different ways to use the Anyparser LangChain integration.

python examples/01_single_file_markdown.py
python examples/02_single_file_json.py
python examples/03_multiple_files_markdown.py
python examples/04_multiple_files_json.py
python examples/05_directory_read.py
python examples/06_ocr_markdown.py
python examples/07_ocr_json.py
python examples/08_web_crawler.py
python examples/09_web_crawler_json.py

Examples

1. Single File Processing

  • 01_single_file_markdown.py: Process a single file with markdown output
  • 02_single_file_json.py: Process a single file with JSON output

2. Multiple File Processing

  • 03_multiple_files_markdown.py: Process multiple files with markdown output
  • 04_multiple_files_json.py: Process multiple files with JSON output
  • 05_directory_read.py: Load and process all files from a folder (max 5 files)

3. OCR Processing

  • 06_ocr_markdown.py: Process images/scans with OCR (markdown output)
  • 07_ocr_json.py: Process images/scans with OCR (JSON output)

4. Web Crawling

  • 08_web_crawler.py: Basic web crawling with essential settings (markdown)
  • 09_web_crawler_json.py: Web crawling (JSON)

Features Demonstrated

Document Processing

  • Different output formats (markdown, JSON)
  • Multiple file handling
  • Folder processing
  • Metadata handling

Web Crawling

  • Basic crawling with depth and scope control
  • Advanced URL and content filtering
  • Crawling strategies (BFS, LIFO)
  • Rate limiting and robots.txt respect

Notes

  • All examples use async/await for better performance
  • Error handling is included in all examples
  • Each example includes detailed comments explaining the options used
  • OCR examples support multiple languages
  • Crawler examples demonstrate various filtering and control options

Features Demonstrated

  • Different output formats (markdown, JSON)
  • OCR capabilities with language support
  • OCR performance presets
  • Image extraction
  • Table extraction
  • Metadata handling
  • Error handling
  • Async/await usage

License

Apache-2.0