crazypdf

A generic, extensible PDF processing library for Go.

Features

• Text Extraction — Extract text from PDFs with three layout modes:
  • Simple — Plain text, words joined by spaces, rows by newlines
  • Raw — Content stream order, preserving the internal PDF text order
  • Physical — Spatial layout preservation using x,y coordinates
• Per-Page Access — Access individual pages by index
• CLI Tool — Command-line utility with subcommand architecture
• Pure Go — No CGo dependencies, built on ledongthuc/pdf

Installation

As a library

go get github.com/ayushanand18/crazypdf

CLI tool

go install github.com/ayushanand18/crazypdf/cmd/crazypdf@latest

Quick Start

Library Usage

package main

import (
    "fmt"
    "log"

    "github.com/ayushanand18/crazypdf/pkg/crazypdf"
    "github.com/ayushanand18/crazypdf/pkg/extract"
)

func main() {
    // Open a PDF document
    doc, err := crazypdf.Open("document.pdf")
    if err != nil {
        log.Fatal(err)
    }
    defer doc.Close()

    // Extract all text (simple mode)
    text, err := extract.Text(doc)
    if err != nil {
        log.Fatal(err)
    }
    fmt.Println(text)
}

Open from Bytes

// Read PDF into memory
data, err := os.ReadFile("document.pdf")
if err != nil {
    log.Fatal(err)
}

// Open from byte slice
doc, err := crazypdf.OpenBytes(data)
if err != nil {
    log.Fatal(err)
}
defer doc.Close()

Per-Page Extraction

// Iterate over pages
for _, page := range doc.Pages() {
    text, err := extract.PageText(page)
    if err != nil {
        log.Printf("Page %d error: %v", page.Number, err)
        continue
    }
    fmt.Printf("--- Page %d ---\n%s\n", page.Number, text)
}

Layout Modes

// Physical layout — preserves spatial positioning
text, _ := extract.Text(doc, extract.WithLayout(extract.LayoutPhysical))

// Raw — content stream order
text, _ = extract.Text(doc, extract.WithLayout(extract.LayoutRaw))

// Custom page separator
text, _ = extract.Text(doc, extract.WithPageSeparator("\n---\n"))

Encrypted PDFs

doc, err := crazypdf.Open("encrypted.pdf", crazypdf.WithPassword("secret"))

CLI Usage

# Extract text from a PDF (prints to stdout)
crazypdf text document.pdf

# Save to file
crazypdf text document.pdf output.txt

# Preserve physical layout
crazypdf text -layout document.pdf

# Raw content stream order
crazypdf text -raw document.pdf

# Specific pages
crazypdf text -pages 1-3 document.pdf
crazypdf text -pages 1,3,5 document.pdf

# Encrypted PDF
crazypdf text -password secret encrypted.pdf

Architecture

crazypdf/
├── pkg/
│   ├── crazypdf/            # Core library package
│   │   ├── doc.go           # Package documentation
│   │   ├── document.go      # Document struct, Open/Close
│   │   ├── page.go          # Page struct, text accessors
│   │   ├── options.go       # Config, functional options
│   │   └── errors.go        # Shared error types
│   │
│   └── extract/             # Feature: Text Extraction
│       ├── text.go          # Text, PageText, AllPages
│       └── options.go       # LayoutMode, extraction options
│
├── internal/pdf/            # Internal PDF reader wrapper
│   └── reader.go            # Wraps ledongthuc/pdf
│
├── cmd/crazypdf/            # CLI tool
│   └── main.go              # Subcommand-based CLI
│
└── testdata/                # Test fixtures
    └── sample.pdf

Contributing

Happy to seek new contributions!. To add a new feature:

1. Create a new directory under pkg/ (e.g., pkg/structurize/, pkg/metadata/)
2. Accept *crazypdf.Document or *crazypdf.Page as input
3. Use public accessor methods (PlainText(), TextByRow(), StyledTexts(), ContentStream())
4. Add a new subcommand to cmd/crazypdf/main.go

Planned Features

• structurize — Convert PDF structure into machine-readable format (headings, paragraphs, lists)
• metadata — Extract document metadata (title, author, keywords, dates)
• tables — Detect and extract tabular data

API Reference

Core Package (pkg/crazypdf)

Type/Function	Description
Open(path, ...Option) (*Document, error)	Open a PDF file
WithPassword(string) Option	Set password for encrypted PDFs
Document.NumPages() int	Get page count
Document.Page(index int) (*Page, error)	Get page by 0-based index
Document.Pages() []*Page	Get all pages
Document.Close() error	Release resources
Page.PlainText() (string, error)	Get plain text from page
Page.TextByRow() ([]TextRow, error)	Get text organized by rows
Page.StyledTexts() ([]StyledText, error)	Get styled text with positions
Page.ContentStream() ([]byte, error)	Get raw content stream

Extract Package (pkg/extract)

Type/Function	Description
Text(doc, ...Option) (string, error)	Extract text from entire document
PageText(page, ...Option) (string, error)	Extract text from single page
AllPages(doc, ...Option) ([]string, error)	Extract text from all pages
WithLayout(LayoutMode) Option	Set layout mode
WithPageSeparator(string) Option	Set page separator
LayoutSimple	Plain text extraction
LayoutRaw	Content stream order
LayoutPhysical	Spatial layout preservation

License

See LICENSE for details.

Thanks!