How to Check Installed Extractors and Get Install Commands in book-to-skill

Use the --check CLI flag to run a pre-flight dependency scan that reports which optional extractors (PDF, EPUB, DOCX, HTML, RTF, MOBI/AZW) are available and prints the exact pip or system commands needed to install any missing components.

The book-to-skill repository provides a built-in diagnostic command for verifying extractor availability. This guide explains how to run the check, interpret the output, and automate missing dependency installation based on the source code in virgiliojr94/book-to-skill.

Running the Dependency Check

Basic Check Command

The fastest way to verify installed extractors is through the CLI entry point:

book-to-skill --check

Alternatively, use the repository wrapper script when working from source:

python3 scripts/extract.py --check

How the Check Is Processed

The --check flag is intercepted early in book_to_skill/utils.py at lines 73-75. When detected, it immediately calls run_dependency_check() from dependencies.py and exits without processing any files.

This short-circuit design ensures you get instant feedback without triggering the full extraction pipeline.

What the Check Detects

The run_dependency_check() function (lines 52-58 in book_to_skill/dependencies.py) iterates over the DEPENDENCY_GROUPS definition to evaluate each extractor category.

Detection Categories

For every format group, the check evaluates:

• Python modules — tests importability via python_module_available
• External binaries — verifies presence on PATH
• Fallback status — indicates whether the format can use lower-quality parsing

Report Output Structure

The diagnostic prints a ✔︎/✘ status line for each dependency. Missing Python packages are aggregated into a single install command, while missing system binaries receive commented hints.

Example output:

book-to-skill — dependency check

  PDF (smart inspection / native Markdown)
    ✗ python: pdf_inspector
      ↳ a `pdf_inspector` command exists at /usr/local/bin/pdf_inspector...

  PDF (text-heavy)
    ✗ python: pypdf
    ✗ python: pdfminer
    ✗ system: pdftotext (poppler-utils)  # sudo apt install poppler-utils

  → fallback available (install for best quality) — any one of pdftotext / pypdf / pdfminer is enough

...

To enable the best extractor for every format, install the missing pieces:

  /usr/bin/python3 -m pip install pypdf pdfminer
  # poppler-utils: sudo apt install poppler-utils

The "→" indicator reveals three states per format:

Indicator	Meaning
✔︎ all green	Best-quality extractor ready
→ fallback available	Functional but lower quality
✗ required missing	Cannot process this format (e.g., Calibre for MOBI)

Automating Missing Dependency Installation

Automatic Python Package Installation

Enable auto-installation with the --install-missing flag:

book-to-skill --install-missing=yes --check

This triggers offer_dependency_install() in dependencies.py, which attempts to pip install missing Python wheels before falling back to stdlib parsers.

You can also set the environment variable:

export BOOK_SKILL_INSTALL_MISSING=yes
book-to-skill --check

Programmatic Access

Capture the dependency report in Python scripts:

import subprocess
import sys

result = subprocess.run(
    [sys.executable, "-m", "book_to_skill", "--check"],
    capture_output=True,
    text=True,
)

print("Dependency report:")
print(result.stdout)

This approach is useful for CI/CD pipelines or environment validation scripts.

Core Implementation Details

File	Key Function	Purpose
book_to_skill/dependencies.py	run_dependency_check()	Defines DEPENDENCY_GROUPS, probes modules, builds install commands (lines 18-27, 60-84)
book_to_skill/utils.py	CLI argument parsing	Detects --check flag and routes to dependency check (lines 73-75)
book_to_skill/cli.py	main() entry point	UTF-8 stream setup and PDF-inspector hook loading
scripts/extract.py	Wrapper script	Alternative invocation path for repository users

The install-command generation logic (lines 18-27 in dependencies.py) aggregates all missing Python packages into one pip install line and provides shell-commented hints for system packages like poppler-utils or calibre.

Summary

• Run book-to-skill --check to scan all optional extractors and see exact install commands
• Check output shows ✔︎/✘ status per dependency with fallback indicators
• Copy-paste the final command block to install missing Python packages; follow commented hints for system binaries
• Use --install-missing=yes for automatic Python package installation
• Key source files: dependencies.py for detection logic, utils.py for CLI flag handling

Run the check after any environment change to ensure optimal extraction quality without manually consulting documentation.

Frequently Asked Questions

How do I check if PDF extraction will work in book-to-skill?

Run book-to-skill --check and examine the "PDF" section. The tool reports three detection paths: smart inspection via pdf_inspector, text-heavy parsing via pypdf/pdfminer/pdftotext, and fallback availability. Any one working dependency enables PDF processing, though quality varies by source document structure.

What if the check reports "required missing" for a format?

This indicates no fallback parser exists for that format—install the specified dependency before proceeding. For MOBI/AZW files, the check requires Calibre's ebook-convert binary and will not proceed without it, unlike PDF or EPUB which offer stdlib fallbacks.

Can I suppress the install hints and just see status?

The --check flag always produces the full report including install suggestions. To parse status programmatically, capture stdout and filter lines containing "✔︎" or "✗" indicators. The structured output format remains consistent across versions for reliable parsing.

Where does book-to-skill look for external binaries like pdftotext?

The python_module_available helper and binary detection routines scan the current PATH environment variable. Binaries installed via package managers (apt, brew, choco) are detected automatically when their installation directory appears in PATH before running the check.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}