| commit | b43935f58581968f4b3597923f6b5d00baaf618c | [log] [tgz] |
|---|---|---|
| author | Jason Robbins <jrobbins@chromium.org> | Mon Oct 06 15:38:34 2014 |
| committer | Jason Robbins <jrobbins@chromium.org> | Mon Oct 06 15:38:34 2014 |
| tree | cbe89df840aec6c495eff01444700f7db5df024c | |
| parent | 9bbca54eada88d8219d01c4d017bd73b8509ecc2 [diff] |
Make MIME boundary harder to accidentally match. BUG=419156 R=iannucci@chromium.org Review URL: https://codereview.chromium.org/624513002
This document explains how to contribute to the Chrome infrastructure codebase. If you want to contribute to the Chromium browser, you're in the wrong place. See http://dev.chromium.org/getting-involved instead. You can find more information on the Chrome infrastructure here.
If you‘re reading this file, you’re probably involved in the Chromium project already. If this is not the case, you might want to read this page to get some background information. In particular, depot_tools needs to be installed by following instructions here.
The proper way to check out this repository is (assuming you have depot_tools somewhere in your path) to run:
mkdir chrome_infra # or whatever name you please cd chrome_infra gclient config https://chromium.googlesource.com/infra/infra.git gclient sync
This will check out the base repository (infra/) and its dependencies.
(See bootstrap/README.md for more details).
Manually create a bootstrap virtualenv environment by running:
./bootstrap/bootstrap.py --deps_file bootstrap/deps.pyl
This is done for you automatically by gclient sync (or gclient runhooks).
Mixing modules and scripts in the same hierarchy can sometimes be a pain in Python, because it usually requires updating the Python path. The goal for infra/ is to be able to check out the repository and be able to run code right away, without setting up anything. The adopted solution is to use main.py files everywhere.
Example: python -m infra.services.lkgr_finder will run the lkgr_finder script.
To make things easier, a convenience script is located at root level. This will do the same thing: run.py infra.services.lkgr_finder. It also provides some additional goodness, like listing all available tools (when invoked without any arguments), and allowing for autocompletion.
If you want run.py to auto-complete, just run:
# BEGIN = ZSH ONLY autoload -Uz bashcompinit bashcompinit # END = ZSH ONLY eval "$(/path/to/infra/ENV/bin/register-python-argcomplete run.py)" eval "$(/path/to/infra/ENV/bin/register-python-argcomplete test.py)"
And that's it. You may want to put that in your .bashrc somewhere.
Infra.git hosts several google appengine projects. In order to support ease of testing and pylint all of the python code for these projects is stored in one shared python package “appengine_module”.
In order to interface well with dev_appserver.py and appcfg.py individual directories exist for the appengine projects under “appengine_apps”. Symlinks exist from those directories back into appengine_module to expose the necessary parts of the appengine_module package to run the app in question.
The ‘appengine’ directory holds the as-of-yet fully converted Appengine apps. All of those should be split into appengine_apps and appengine_module pieces and appengine_module should be renamed to ‘appengine’. See crbug.com/407734.