summary refs log tree commit diff
path: root/.buildkite/scripts (follow)
Commit message (Collapse)AuthorAgeFilesLines
* Remove buildkite from portdb CI testsDavid Robertson2021-08-102-88/+0
|
* Move test_old_deps.sh to new ci dirDavid Robertson2021-08-101-16/+0
|
* Fix off-by-one-error in synapse_port_db (#9991)Richard van der Hoff2021-05-241-0/+4
| | | fixes #9979
* Fix port_db on empty db (#9930)Richard van der Hoff2021-05-103-45/+57
| | | | ... and test it.
* Remove redundant "coding: utf-8" lines (#9786)Jonathan de Jong2021-04-141-1/+0
| | | | | | | Part of #9744 Removes all redundant `# -*- coding: utf-8 -*-` lines from files, as python 3 automatically reads source code as utf-8 now. `Signed-off-by: Jonathan de Jong <jonathan@automatia.nl>`
* update test_old_deps scriptRichard van der Hoff2021-04-081-1/+1
|
* unpin olddeps build from py36Richard van der Hoff2021-04-081-3/+3
|
* Use interpreter from $PATH instead of absolute paths in various scripts ↵Quentin Gliech2021-03-252-2/+2
| | | | | | | | | using /usr/bin/env (#9689) On NixOS, `bash` isn't under `/bin/bash` but rather in some directory in `$PATH`. Locally, I've been patching those scripts to make them work. `/usr/bin/env` seems to be the only [portable way](https://unix.stackexchange.com/questions/29608/why-is-it-better-to-use-usr-bin-env-name-instead-of-path-to-name-as-my) to use binaries from the PATH as interpreters. Signed-off-by: Quentin Gliech <quentingliech@gmail.com>
* Fix Python 3.5 old deps build by using a compatible pip version. (#9217)Patrick Cloke2021-01-251-0/+3
| | | | | Co-authored-by: Dan Callahan <danc@element.io> pip 21.0 stopped supporting Python 3.5.
* Run the linters on a consistent list of files (#9038)Richard van der Hoff2021-01-081-0/+1
| | | | | We were running some linters on some files and some on others. Extract a common setting and use it everywhere.
* Add basic SAML tests for mapping users. (#8800)Patrick Cloke2020-12-021-1/+1
|
* Fix py35-old CI by using native tox. (#7018)Erik Johnston2020-03-021-6/+1
| | | | | I'm not really sure how this was going wrong, but this seems like the right approach anyway.
* Update setuptools for python 3.5 tests (#6880)Richard van der Hoff2020-02-101-0/+18
| | | Workaround for jaraco/zipp#40
* Fix CI for synapse_port_db (#6276)Brendan Abolivier2019-10-292-4/+47
| | | | | | | | | | | | * Don't use a virtualenv * Generate the server's signing key to allow it to start * Add signing key paths to CI configuration files * Use a Python script to create the postgresql database * Improve logging
* Add CI for synapse_port_db (#6140)Brendan Abolivier2019-10-281-0/+29
This adds: * a test sqlite database * a configuration file for the sqlite database * a configuration file for a postgresql database (using the credentials in `.buildkite/docker-compose.pyXX.pgXX.yaml`) as well as a new script named `.buildkite/scripts/test_synapse_port_db.sh` that: 1. installs Synapse 2. updates the test sqlite database to the latest schema and runs background updates on it 3. creates an empty postgresql database 4. run the `synapse_port_db` script to migrate the test sqlite database to the empty postgresql database (with coverage) Step `2` is done via a new script located at `scripts-dev/update_database`. The test sqlite database is extracted from a SyTest run, so that it can be considered as an actual homeserver's database with actual data in it.