robust-pdf-extraction

Robust PDF Extraction Workflow

This skill provides a systematic approach to extracting text from PDF files, handling both text-based and scanned/image-based documents through progressive fallback methods.

When to Use

• Processing PDFs of unknown or mixed types (text vs. scanned images)
• Critical document processing where extraction failure is not acceptable
• Batch processing multiple PDFs with varying formats

Workflow Steps

Step 1: Verify File Accessibility

Before attempting extraction, confirm the PDF exists and is readable:

# Check file exists and get basic info
ls -la /path/to/document.pdf

# Or search for files if location uncertain
find /path -name "*.pdf" -type f 2>/dev/null

Step 2: Attempt Primary Extraction (pdfplumber)

Start with pdfplumber for best text structure preservation:

import pdfplumber

def extract_with_pdfplumber(pdf_path):
    text = ""
    with pdfplumber.open(pdf_path) as pdf:
        for page in pdf.pages:
            page_text = page.extract_text()
            if page_text:
                text += page_text + "\n"
    return text.strip()

Step 3: Fallback to Secondary Method (pypdfium2)

If pdfplumber returns empty or incomplete text:

import pdfium2

def extract_with_pypdfium2(pdf_path):
    pdf = pdfium2.PdfDocument(pdf_path)
    text = ""
    for page in pdf:
        text_page = page.get_textpage()
        page_text = text_page.get_text_bounded()
        if page_text:
            text += page_text + "\n"
    return text.strip()

Step 4: Fallback to Tertiary Method (pdftotext)

If pypdfium2 also fails, use command-line pdftotext:

pdftotext /path/to/document.pdf - 2>/dev/null

Or in Python:

import subprocess

def extract_with_pdftotext(pdf_path):
    result = subprocess.run(
        ['pdftotext', pdf_path, '-'],
        capture_output=True,
        text=True
    )
    return result.stdout.strip()

Step 5: Detect Scanned/Image-Based PDFs