pdf

Files (1)

• SKILL.md

SKILL.md (first lines)

# PDF Processing Guide

## IMPORTANT: Do NOT Use the Read Tool for PDFs

**The Read tool cannot properly extract tabular data from PDFs.** It will only show you a limited preview of the first page's text content, missing most of the data.

For PDF files, especially multi-page PDFs with tables:
- **DO**: Use `pdfplumber` as shown in this skill
- **DO NOT**: Use the Read tool to view PDF contents

If you need to extract data from a PDF, write Python code using pdfplumber. This is the only reliable way to get complete table data from all pages.

## Overview

Extract text and tables from PDF documents using Python libraries.

## Table Extraction with pdfplumber

pdfplumber is the recommended library for extracting tables from PDFs.

### Basic Table Extraction

```python
import pdfplumber

with pdfplumber.open("document.pdf") as pdf:
    for page in pdf.pages:
        tables = page.extract_tables()
        for table in tables:
            print(table)  # List of lists
```

### Convert to pandas DataFrame

```python
import pdfplumber
import pandas as pd

with pdfplumber.open("document.pdf") as pdf:
    page = pdf.pages[0]  # First page
    tables = page.extract_tables()

    if tables:
        # First row is usually headers
        table = tables[0]
        df = pd.DataFrame(table[1:], columns=table[0])
        print(df)
```

### Extract All Tables from Multi-Page PDF

```python
import pdfplumber
import pandas as pd

all_tables = []

with pdfplumber.open("document.pdf") as pdf:
    for page in pdf.pages:
        tables = page.extract_tables()
        for table in tables:
            if table and len(table) > 1:  # Has data
                df = pd.DataFrame(table[1:], columns=table[0])
                all_tables.append(df)

# Combine if same structure
if all_tables:
    combined = pd.concat(all_tables, ignore_index=True)
```

## Text Extraction

### Full Page Text

```python
import pdfplumber

with pdfplumber.open("document.pdf") as pdf:
    text = ""
    for page in pdf.pages:
        text += page.extract_text() + "\n"
```

### Text with Layout Preserved

```python
with pdfplumber.open("document.pdf") as pdf:
    page = pdf.pages[0]
    text = page.extract_text(layout=True)
```

## Common Patterns

### Extract Multiple Tables and Create Mappings

When a PDF contains multiple related tables (possibly spanning multiple pages), extract from ALL pages and build lookup dictionaries:

```python
import pdfplumber
import pandas as pd

category_map = {}      # CategoryID -> CategoryName
product_map = {}       # ProductID -> (ProductName, CategoryID)

with pdfplumber.open("catalog.pdf") as pdf:
    for page_num, page in enumerate(pdf.pages):
        tables = page.extract_tables()

        for table in tables:
            if not table or len(table) < 2:
                continue

            # Check first row to identify table type
            header = [str(cell).strip() if cell else '' for cell in table[0]]

            # Determine if first row is header or data (continuation page)
            if 'CategoryID' in header and 'CategoryName' in header:
                # Categories table with header
                for row in table[1:]:

From benchflow-ai/skillsbench, Apache-2.0.

More from benchflow-ai/skillsbench