Comments (8)
Another idea here might be to break the slow tests out into their own CI job. That would mean that things like coverage, linting & the main suite of tests could give their feedback sooner, while still running the slower tests.
from asttokens.
Because CI has ASTTOKENS_SLOW_TESTS=1
.
from asttokens.
It looks like #128 does speed it up, which makes sense because the main slow test runs twice, once for ast
and once for astroid
. So getting that working would help. If it's possible to effectively use more than 2 workers then parametrising that test could help further.
It might also help to run the slow tests without coverage, and if that causes coverage to drop then we should be writing normal deterministic tests to fill the gap.
from asttokens.
It looks like the initial speedup from #128 was in part due to it running without coverage. That was actually what was failing. Adding in pytest-cov
fixes that, but the tests are slow again. I like the idea of the slow tests running without coverage.
Separately it looks like the slow tests are in fact all in a single test case, which means that parallelisation can only help so far with that. Adding in parameterized
kinda helps with that, however the ordering of the sys.modules
isn't stable and the tests can flake out. Applying sorting helps, but I'm not sure if we'd see divergences in the members of sys.modules
between workers too. Needs more investigation.
from asttokens.
It's a single test method, but it lives in TestMarkTokens
which is subclassed by TestAstroid
so there are two versions of the test that run. But I don't know how the division of tests works, so maybe a single worker was allocated both of them.
from asttokens.
pytest-xdist
has a number of strategies for worker allocation. My understanding of the default is that workers for tests are allocated in groups on the fly rather than upfront, so I think in general the two long tests should end up on separate workers. I think we could also explicitly configure those tests to end up on separate workers with pytest marks. Docs for the strategies are at https://pytest-xdist.readthedocs.io/en/stable/distribution.html
from asttokens.
Ah, we might want to be using --dist worksteal
if we see highly variable times.
from asttokens.
Is the --junitxml=./rspec.xml
used for anything? I can't see a use for it, but I don't know if it's somehow implicitly needed by coveralls or something.
from asttokens.
Related Issues (20)
- DeprecationWarning: The 'astroid.node_classes' module is deprecated HOT 5
- ASTText astroid support HOT 1
- IndexError in asttokens.ASTTokens HOT 2
- Timelines around support for legacy Python HOT 2
- Replace deprecated imp module with importlib HOT 2
- Compatibility with Python 3.12 and unreleased astroid 3.0.0 HOT 1
- Drop Python 2.7? HOT 4
- ModuleNotFoundError: No module named 'six.moves' with python 3.12 HOT 5
- RFE: is it possible to start making github releases?🤔
- ASTText: Module found error HOT 3
- 2.4.1 + master (ab2c6289): pytest fails in tests/test_astroid.py and tests/test_mark_tokens.py units HOT 3
- 2.4.1: sphinx warnings `reference target not found`
- 2.0.5: pytest is failing in few `astroid` units HOT 15
- `asttokens` 2.0.6 has undeclared dependency on `astroid` HOT 8
- Followup to astroid import fix
- mypy reports `error: Name "asttokens.ASTTokens" is not defined` HOT 4
- Support for constructing around a list of existing tokens HOT 6
- Docs outdated HOT 3
- Consider a clearer optional-functionality API HOT 15
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from asttokens.