generate_images: add puppeteer(pptr) backend #1268
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
- jsdom: run Bravura and other font to ensure backcompat filename.
(for recent flow.html dom tree).
… in parallel. - Moved the implementation of child process management from each backend to generate_images.js.
- jsdom: fix old build support.
ex: npm run get:releases 3.0.9 then load flow.html?module=Beam&ver=releases/3.0.9
|
This is still a draft. |
h-sug1no
changed the title
|
ready to review. |
|
And puppeteer is back! This is great @h-sug1no. :-) Just a minor comment. |
0xfe
reviewed
Dec 23, 2021
tools/generate_images.js
Outdated
|
|
||
| const usage = () => { | ||
| log('Usage:'); | ||
| log(' node generate_images.js ver imageDir [--backends=<backends>] [--parallel=<jobs>]'); |
There was a problem hiding this comment.
Can we make this look like:
Usage: node generate_images.js <version> /path/to/image/dir [options...]
Options:
--backends=(all|jsdom|pptr) : Specify which backend(s) to run.
--parallel=<num> : Number of parallel processes. Defaults to the number of CPUs.
There was a problem hiding this comment.
like this? 5d26dff
$ node ./tools/generate_images.js
Usage: node generate_images.js <version> /path/to/image/dir [options...]
Options:
--backends=(all|jsdom|pptr) : Specify which backend(s) to run.
--parallel=<num> : Number of parallel processes. Defaults to the number of CPUs.
|
Hi, thanks for this PR! I am excited to try it out when I have some time. Unfortunately due to the holidays, I won't be able to test & review this until January 2nd or 3rd. Maybe someone else will review it before I get around to it, but I just wanted to let you know my availability. :-) |
|
I am also wishing to see if it detects the changes in #1262 that the current visual regression system does not detect. :) |
# Conflicts: # package.json # tests/flow.html
|
use ver/cjs if available as follows:
svg diff is detected as follows ( #1241 ) : OK
|
|
ready for review now. |
0xfe
reviewed
Dec 25, 2021
| @@ -90,6 +90,11 @@ if (!global.$) { | |||
|
|
|||
| export type TestFunction = (options: TestOptions, contextBuilder: ContextBuilder) => void; | |||
|
|
|||
| export type RunOptions = { | |||
| jobs: number; | |||
| job: number; | |||
There was a problem hiding this comment.
This is a bit confusing, maybe numJobs and jobIndex? (sorry, I know it's all over the code, but it makes it hard to tell what it's for.)
There was a problem hiding this comment.
- numJobs is length of VexFlowTests.tests.
- to map tests to available cpus split tests into jobs.
- Since flow.html contains multiple canvases and svgs, and since the number of tests is likely to increase in the future, we will divide it into 10 tests each for fear of exhausting resources. : https://github.com/h-sug1no/vexflow/blob/0a8eec238857cf52d82cc711a9fb61aa230cf756/tools/generate_images.js#L120
- If split the test too much, the initialization cost of pptr/chrome will increase and the total performance will be worse. 10 is not such a bad value.
- Since flow.html contains multiple canvases and svgs, and since the number of tests is likely to increase in the future, we will divide it into 10 tests each for fear of exhausting resources. : https://github.com/h-sug1no/vexflow/blob/0a8eec238857cf52d82cc711a9fb61aa230cf756/tools/generate_images.js#L120
- to map tests to available cpus split tests into jobs.
- VexFlowTests.run() interprets jobs and job arguments as follows
- https://github.com/h-sug1no/vexflow/blob/0a8eec238857cf52d82cc711a9fb61aa230cf756/tests/vexflow_test_helpers.ts#L160
- This is the implementation because we believe that tests should not depend on the order of execution.
- https://github.com/h-sug1no/vexflow/blob/0a8eec238857cf52d82cc711a9fb61aa230cf756/tests/vexflow_test_helpers.ts#L160
I hope that explains it...
|
Looks good, thanks @h-sug1no! |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR add puppeteer(pptr) backend:
msys file systemsomthing is slow, so it is difficult to run tools/visual_regression.sh...Options
VF_GENERATE_OPTIONS:
--backends=<backends>: Which backend should be enabled by default?--parallel=<jobs>:Examples
VF_GENERATE_OPTIONS is applied to commands that invoke
generate:hoge(e.g.VF_GENERATE_OPTIONS='--backends=jsdom' npm run test:reference).VF_GENERATE_OPTIONS='--backends=jsdom' npm run generate:hogeVF_GENERATE_OPTIONS='--backends=pptr' npm run generate:hogeVF_GENERATE_OPTIONS='--parallel=2' npm run generate:hogeVF_GENERATE_OPTIONS='--backends=jsdom --fonts=petaluma' npm run generate:hogeVF_GENERATE_OPTIONS='--backends=pptr --module=Beam --parallel=1' npm run generate:hogenpm run get:releases 3.0.9node tools/generate_images.js releases/3.0.9 build/images/referenceSpeed measurement
Core(TM) i7-4790S CPU @ 3.20GHz
ubuntu(18.04.6 LTS (Bionic Beaver)) VMware on window10
memory: 13G
Limitations
Please review it if it seems to work for this project!