Epstein Files – Complete Downloader & Text Extractor

A Python tool that downloads all publicly available Jeffrey Epstein–related document collections from multiple sources and extracts searchable text from them.

This project aggregates court records, document dumps, and OCR archives into a single local dataset, then extracts raw text layers from PDFs to enable full-text search and analysis.

---

Features

• Downloads all known public Epstein document collections from:
  • Internet Archive (court filings, black books, phone books, oversight releases)
  • GitHub OCR archive (8,186 structured documents with metadata & entities)
• Extracts raw text layers from PDFs
  • Uses PyMuPDF when available
  • Falls back to pdftotext if needed
• Preserves document structure and metadata
• Generates searchable .txt files
• Detects potentially interesting patterns (emails, phone numbers)
• Unified search across all downloaded material
• Automatically installs missing Python dependencies

---

Data Sources

Internet Archive

Text-searchable PDF collections, including:

• Giuffre v. Maxwell court documents
• House Oversight Committee releases
• Epstein Black Book (multiple versions)
• Epstein phone books

---

Requirements

• Python 3.8+
• Internet connection
• Disk space:
  • ~5–7 GB for PDFs
  • Additional space for extracted text

Optional (Recommended)

• PyMuPDF (auto-installed)
• pdftotext (system fallback)

---

Installation

git clone https://github.com/lanefiedler731-gif/Epstein-Downloader.git
cd Epstein-Downloader
python3 epstein_downloader.py --all

Dependencies are installed automatically if missing.

---

Usage

Download everything:

python3 epstein_downloader.py --all

Show status:

python3 epstein_downloader.py --status

Search all documents:

python3 epstein_downloader.py --search "keyword"

Only Internet Archive:

python3 epstein_downloader.py --ia-only

Only GitHub OCR:

python3 epstein_downloader.py --github-only

Extract PDF text only:

python3 epstein_downloader.py --extract

---

Output Structure

documents/
├── internet_archive/
├── github_ocr/
│   └── analyses.json
extracted_text/
├── pdf_text/
└── interesting_finds.json

---

Text Extraction Notes

• Extracts raw PDF text layers
• Improper redactions may still expose underlying text
• Blank pages are flagged
• No OCR is performed on image-only PDFs

---

Legal & Ethical Notice

• Only publicly available documents are accessed
• No private or restricted material is obtained
• Users are responsible for interpretation and use
• Intended for research, journalism, and archival transparency

---

Disclaimer

This software makes no claims regarding accuracy, completeness, or interpretation of the documents.
It is a data aggregation and text extraction utility only.

---

License

Without a license, all rights are reserved.