PDF Reader
A hybrid PDF indexing and search system designed for AI assistant integration via MCP (Model Context Protocol). Rather than dumping entire PDF content into context windows, this tool provides intelligent document processing, full-text search, and optimized image handling.
Features
- Hybrid Processing: Local (tesserocr) and remote (Modal cloud) OCR processing
- Smart Search: Tantivy-based full-text search with BM25 ranking and query escaping
- AI Enhancement: LLM-generated document titles and page summaries
- Provider Optimization: Automatic image resizing for OpenAI/Anthropic API limits
- MCP Integration: RESTful API for seamless AI assistant integration
Quick Start
# Activate virtual environment
source .venv/bin/activate
# Example usage (if you have an existing index)
python -c "
from pdf_reader.index import PDFIndex
index = PDFIndex.load('indexes/index2')
results = index.search_pages(['your search term'])
print(f'Found {len(results)} results')
"
Architecture
The system uses a hybrid approach:
- Tantivy: Fast full-text search across OCR text and AI summaries
- SQLite: Blob storage for complete page data (text + images)
- AI Enhancement: Document titles and page summaries via LLM
- Provider-Aware Images: Optimized for different LLM API requirements
MCP Tools
list_pdfs: List available PDF documentssearch_pdfs: Keyword search across all documentsread_pdf: Read specific pages from a documentread_pdfs: Bulk read across multiple documents
Development
See CLAUDE.md for detailed architecture documentation, development guidelines, and troubleshooting information.
Metadata
Release files for pdf-reader-0jsfjf 0.1.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| pdf_reader_0jsfjf-0.1.0.tar.gz | 35.5 kB | Without provenance | Details for pdf_reader_0jsfjf-0.1.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| pdf_reader_0jsfjf-0.1.0-py3-none-any.whl | Python 3 | none | any | Details for pdf_reader_0jsfjf-0.1.0-py3-none-any.whl |
Total release size: 76.1 kB
Release files / pdf_reader_0jsfjf-0.1.0.tar.gz
| Download URL | pdf_reader_0jsfjf-0.1.0.tar.gz |
|---|---|
| Size | 35.5 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
df7f1c4a0b1867c14259c9b441bfcd2a066aac11864c6a7f0490587e78f48625
|
|
BLAKE2b-256 checksum How to use checksums |
e7141ab472918acbb47c96233b265ab3e618daba3fb4fa7db60b816ec75b488b
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.12.14
|
Release files / pdf_reader_0jsfjf-0.1.0-py3-none-any.whl
| Download URL | pdf_reader_0jsfjf-0.1.0-py3-none-any.whl |
|---|---|
| Size | 40.6 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
d0710ee74a9ad722dca34ed45d91f3384d6cb09e5e70f0eeba678094031efa57
|
|
BLAKE2b-256 checksum How to use checksums |
f0ac76f215a0d6cd08570528a1f6c8be169f4977f2bdd465d71e95914a05cbf5
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.12.14
|
Security for pdf-reader-0jsfjf 0.1.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.