| Commit message (Collapse) | Author | Age | Lines |
... | |
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|\ |
|
| | |
|
| | |
|
| | |
|
|/ |
|
|\ |
|
| | |
|
|/ |
|
|\
| |
| | |
Add python-levenshtein and sort Pipfile (dev)packages
|
| | |
|
|/ |
|
|\
| |
| | |
add --dev flag to readme
|
|/ |
|
|\
| |
| | |
Fix venv being created in Dockerfile
|
| |
| |
| |
| |
| | |
`pipenv run` creates a venv because it cannot detect that a --system
install was done. The solution is to invoke gunicorn directly.
|
| |
| |
| |
| |
| | |
This will make it easy to maintain a consistent config without relying
on invocation via pipenv.
|
|/
|
|
|
| |
There will be more config files to come so it's cleaner to have them
together than littering the root directory with more files.
|
|\
| |
| | |
Document a simpler way to run snekbox
|
| | |
|
| |
| |
| |
| |
| |
| | |
As convenient as it may be, it is redundant to list out the config in
the docs. It also may fall out of sync with the actual config should
someone forget to update the docs.
|
| |
| |
| |
| | |
Makes the Markdown less cluttered when editing it.
|
|/
|
|
|
|
|
|
| |
The current run instructions are geared towards developers. A simpler
way to run snekbox is to start a container with `docker run` via the
image published on Docker Hub.
Resolves #57
|
|\
| |
| | |
Update contributor doc
|
|/ |
|
|\
| |
| | |
Add sentry integration
|
| | |
|
| |
| |
| | |
Co-Authored-By: Mark <[email protected]>
|
|/ |
|
|\
| |
| | |
Use port mapping, change container name to snekbox.
|
| | |
|
| |\
| |/
|/| |
|
| |
| |
| | |
This is to attempt to avoid matching the bot's typing event being resent at the 5 second mark, which seems to be causing it to hang around after the timeout message.
|
|\ \
| | |
| | | |
CI Improvements
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Error handling is performed inside can_pull so the callers of the
function don't always check its exit code. Because set -e present, if
can_pull returns 1, bash would consider that function a failed call and
thus exit the entire script with code 1. That, in turn, would cause the
CI job to fail.
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The Azure website has proven to not be reliable for displaying the
JUnit output. Furthermore, some may simply prefer the format of the
output in the terminal over the JUnit representation on the Azure site.
Nevertheless, the JUnit output isn't that bad (when there's actually
a lint error) so it will remain for now. It also provides historical
statistics on occurrences of errors, which is kind of cool, I guess...
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Since #62, the dependencies have been installed to the system
interpreter. Therefore, it's not necessary to run through pipenv to
ensure the commands run through an activated virtual environment.
This should also fix pipenv run creating a virtual environment. It seems
it cannot tell when it should be using the system interpreter. It
probably wasn't designed for that anyway i.e. the intent was to run
commands directly in such case, which is what this PR will do.
|
| | | |
|
| | |
| | |
| | |
| | | |
The array shouldn't be expanded when testing with -v.
|
| | |
| | |
| | |
| | |
| | |
| | | |
The script may need to use the master commit several times. The easiest
way to implement the cache was to just cache the response rather than
the commit hash.
|
| | |
| | |
| | |
| | | |
Hyphens are disallowed so they are replaced with underscores.
|
| | |
| | |
| | |
| | |
| | | |
* Remove duplicate build steps
* Move push steps to a separate YAML file
|
| | | |
|
| | | |
|
| | | |
|