gitbook-convert

CLI to convert an existing document to a GitBook.

Install

Install this globally and you'll have access to the gitbook-convert command anywhere on your system.

$ npm install gitbook-convert -g

Use

$ gitbook-convert [options] <file> [export-directory]

Options

Short	Long	Description	Type	Default
-t	--document-title	Name used for the main document title	string	null
-a	--assets-dir	Name of the document's assets export directory	string	assets
-m	--max-depth	Maximum title depth to use to split your original document into sub-chapters	integer	2
-p	--prefix	Prefix filenames by an incremental counter	flag	false
-d	--debug	Log stack trace when an error occurs	flag	false

After converting your document, the corresponding GitBook files will be placed in the provided export-directory folder. The folder is created during conversion.

If export-directory is not provided, a new export folder is created in the current working directory. The GitBook files are then placed here.

If the --document-title argument is not passed, the filename without the file extension will be used as the main document title.

Currently accepted formats

Type	Extension
Microsoft Office Open XML Document	.docx
OpenOffice / Open Document Format	.odt
Docbook Markup Language	.xml
HyperText Markup Language	.html

Output

This version of gitbook-convert generates markdown files only. Support for asciidoc might be added later.

Document processing

gitbook-convert divides your original document into chapters and sub-chapters, if any, one per output file. To do this, gitbook-convert automatically detects the headers in your document and uses the -m flag to split it into sub-chapters.

When converting a Docbook file though, the depth is always detected automatically.

Thus, converting the following document named History of modern computers.docx with the default --max-depth flag:

Chapter 1

What the world used to be.

The beginning

At the beginning was the big bang...

The following

Strange creatures called “humans” had trouble living in peace...

Chapter 2

What the world is now.

The awakening

Computers came to rule the world...

The end

The power supply went disconnected.

will produce the following output:

user @ cwd/export/history_of_modern_computers
  README.md
  SUMMARY.md
  assets/
  chapter_1/
    README.md
    the_beginning.md
    the_following.md
  chapter_2/
    README.md
    the_awakening.md
    the_end.md

While using 1 for --max-depth would produce:

user @ cwd/export/history_of_modern_computers
  chapter_1.md
  chapter_2.md
  README.md
  SUMMARY.md
  assets/

Summary

The SUMMARY.md file is created automatically.

For our first example:

# Summary

* [Introduction](README.md)
* [Chapter 1](chapter_1/README.md)
  * [The beginning](chapter_1/the_beginning.md)
  * [The following](chapter_1/the_following.md)
* [Chapter 2](chapter_2/README.md)
  * [The awakening](chapter_2/the_awakening.md)
  * [The end](chapter_2/the_end.md)

With --max-depth set to 1:

# Summary

* [Introduction](README.md)
* [Chapter 1](chapter_1.md)
* [Chapter 2](chapter_2.md)

README

The content of the README.md file depends on your document structure. Anyways, the filename of your original document will be used as the main title here.

Original document starts with a main header

gitbook-convert creates the default GitBook README.md file:

# History of modern computers

This file serves as your book's preface, a great place to describe your book's content and ideas.

Original document has an introduction

Otherwise, everything before the first main header is used as the README.md content. If we modify our example to be:

A short history of modern computers.

Chapter 1

The beginning

At the beginning was the big bang...

The following

...

The content of the README.md file will be:

# History of modern computers

A short history of modern computers.

The behavior is the same when --max-depth is set to higher levels. Each README.md in the sub-chapters folders will contain the preface for the current chapter.

Converters

The appropriate converter for a document type is deduced from its extension.

For now, the converters should:

• be placed in lib/converters,
• with its filename being the document-type extension, for example /lib/converters/docx.js,
• added to the lib/converters/index.js file for reference and use.

docx

The .docx converter uses mwilliamson's mammoth.js to convert your document to HTML before generating the output.

gitbook-convert will try to export your inline images in the /assets folder, using the image title as the image filename if provided.

odt

The .odt converter uses odt2html to convert your document to HTML before generating the output. Because there was no node module out there to convert OpenOffice documents to HTML, we built our own.

gitbook-convert will try to export your inline images in the /assets folder, using the image name in the document as the image filename if provided.

docbook

gitbook-convert requires xsltproc to be installed to process a Docbook. If you are using MacOS or a Linux distribution, it should be installed by default.

You can test that xsltproc is installed using:

$ which xsltproc

xsltproc uses the last version of docbook.xsl to convert your Docbook to HTML first. Since the Docbook XML markup is very large, gitbook-convert will try to convert the meta-data as well as possible. Extended conversion might be added to the tool based on user requests.

When you install gitbook-convert using npm, the docbook.xsl stylesheets are downloaded and installed along with the app.

We recommend using the tool with Docbook version 5. Here is a walk-through for converting an existing Docbook in version 4 to version 5.