Phabritex

Phabritex is my rough effort for adding LaTeX support into the Phabricator system.

Install

First clone the render2katex utility here. Install as guided to somewhere outside the Phabricator directory.

Copy all files into the phabricator directory and manually patch two files (to add the KaTeX CSS).

Patch <phabricator>/webroot/rsrc/css/core/remarkup.css, change the header into:

/**
 * @provides phabricator-remarkup-css
 * @requires katex-css
 */

Patch <phabricator>/webroot/rsrc/css/phui/phui-document.css, change the header into:

/**
 * @provides phui-document-view-css
 * @requires katex-css
**/

Replace the path of index.js into the path you installed render2katex in the two files in src/extensions

Then run <phabricator>/bin/celerity map and restart Apache or php5-fpm (for nginx).

Upgrading Phabricator with Local Changes

It is recommended here that you manage your own local fork of the stable branch.

After you first patched the two css files, commit them (DO NOT commit celerity.map!) to your local stable branch. After that, you can upgrade as usual using git pull to the upstream.

Troubleshooting

First check that you have already restarted everything(nginx, Apache, php daemon, etc.). If it still does not work, try the following (as suggested by @followyourheart):

bin/cache purge --purge-all
Disable SELinux or add exceptions

Compatibility

Tested with upstream until 2019/09. Please report compatibility as you update from upstream so I can make the corresponding changes.

Usage

You can use it everywhere with remarkup support :)

In Phriction wiki:

# This is currently a test

{nav Home}

The statistical model estimated the probability, {tex \pi_{i}}, of capturing dolphins on a tow, {tex i}. A year effect, {tex logit(\pi_i) = \lambda_{j[i]} + \sum_c \beta_c x_{ic}.} was estimated for each year, {tex j}, allowing for annual variation in the capture event rates that was unrelated to the covariates, {tex x}. The contribution of each covariate, indexed by {tex c}, was governed by a regression coefficient, {tex \beta_c}, that was estimated by the model. The logit transform of the capture event probability was defined as the sum of the year effect, {tex \lambda}, and the covariates:

katex {{{
logit(\pi_i) = \lambda_{j[i]} + \sum_c \beta_c x_{ic}.
}}}
determines the result.

# The following is more tests

katex {{{
f(x) = \int_{-\infty}^\infty
    \hat f(\xi)\,e^{2 \pi i \xi x}
    \,d\xi 
}}}

katex {{{
\begin{bmatrix}
    a & b \\
    c & a 
    \end{bmatrix}
}}}

katex {{{
\begin{bmatrix}
    a & c \\
    c & a 
    \end{bmatrix}
}}}

renders into

and in comments

Development

Please see this post for a through introduction to the codebase.

License

MIT

Zarrathustra/phabritex