Home

Awesome

asciicast2gif

asciicast2gif converts asciicast files to animated GIF files

News

2017-05-06: Marcin Kulik, the developer of asciinema, is working on a similar tool, also named asciicast2gif (originally named a2gif), which offers improvements over my tool and which is likely to be better supported in the future. Thus, you might want to use Marcin's asciicast2gif instead of mine.

Usage

Usage asciicast GIF

Generating the output GIF file might take several seconds/minutes, depending on the duration of the asciicast.

Synopsis

Run without arguments to get the following help message:

$ ./asciicast2gif

NAME
    asciicast2gif - convert an asciicast (asciinema JSON file) to an animated GIF file

SYNOPSIS
    ./asciicast2gif ASCIICAST.json [GIF_FILE_NAME] [OPTIONS]

OPTIONS
    --fps=NUM : generate gif with NUM frames per second (default: 10)
    --head=NUM : discard first NUM screenshots (default: 0)
    --keep : do not delete the temporary directory (default: delete)
    --nogif : do not generate GIF file (default: generate)
    --port=NUM : use port NUM for the local HTTP server (default: 8000)
    --size=[small|medium|big] : set player size (default: small)
    --speed=VAL : set player speed to VAL (default: 1)
    --tail=NUM : discard last NUM screenshots (default: 0)
    --theme=[asciinema|monokai|tango|solarized-dark|solarized-light] : use player theme (default: asciinema)
    --server=[py|py2|py3|py2.7|py3.5|py3.6|ruby|none] : use the given local HTTP server (default: py)

EXAMPLES
    $ ./asciicast2gif your.json               => generate your.json.gif
    $ ./asciicast2gif your.json foo.gif       => generate foo.gif
    $ ./asciicast2gif your.json --fps=20      => generate your.json.gif at 20 frames/s
    $ ./asciicast2gif your.json --speed=1.5   => play asciicast at 1.5x speed
    $ ./asciicast2gif your.json --head=3      => discard first 3 screenshots
    $ ./asciicast2gif your.json --tail=5      => discard last 5 screenshots
    $ ./asciicast2gif your.json --server=ruby => use Ruby to start local HTTP server
    $ ./asciicast2gif your.json --server=none => local HTTP server is already running on port 8000
    $ ./asciicast2gif your.json --size=big    => set the player size to big
    $ ./asciicast2gif your.json --port=5000   => use local HTTP server port 5000
    $ ./asciicast2gif your.json --theme=tango => set the player theme to tango

Note: it is strongly suggested to copy your asciicast JSON file into the git-cloned project directory, cd into the latter, and run asciicast2gif from there.

Requirements

  1. a POSIX shell, e.g. the Bash shell)
  2. phantomjs, i.e. PhantomJS (>=2.0.0)
  3. convert, i.e. ImageMagick
  4. gifsicle, i.e. Gifsicle
  5. python (Python) or ruby (Ruby) or a running HTTP server of your choice

How It Works

asciicast2gif is a POSIX shell script that reads an asciicast JSON file created by asciinema, patches a template HTML page generating a page.patched.html file, and runs a local HTTP server to serve it in background.

(If you have python or ruby, asciicast2gif starts and terminates the HTTP server for you, attempting to determine the correct command line parameters. Alternatively, you can manually run any other HTTP server of your choice, passing --server=none to asciicast2gif.)

The screenshot.js script, run inside phantomjs, periodically takes screenshots of the asciinema-player while it reproduces the asciicast, saving them as PNG files.

Finally, convert and gifsicle generate the output GIF file.

Note: currently you need a local HTTP server because it seems that phantomjs and/or asciinema-player.js do not work when fed a static local HTML file with the asciicast JSON file embedded. Removing this limitation would help many users. Please contribute if you figure it out.

License

asciicast2gif is released under the terms of the MIT License.

The page.template.html file in this repository contains an inlined version of asciinema-player v2.0.0 (GPLv3 License).

Changelog

Limitations and Missing Features

Contribution Policy

Contributions are welcome!

Please follow the usual GitHub fork/new branch/pull-request flow.

Acknowledgments