Conversation
8b757f6 to
b0f4531
Compare
a2c8d68 to
17f9e9d
Compare
runtime-testsuite/resources/descriptors/ParserExec/ListLabelForClosureContext.txt
Outdated
Show resolved
Hide resolved
|
A huge amount of work here. thank you! |
Thank you for your reviews and PR accepting :) I suggest changing other runtimes (moving |
|
Are we sure the others have deadlock issues? Not sure I have the energy to update. Maybe we can ping those target folks. |
Not sure (because they can use another initialization mechanizm) but it's not fully excluded (C# is very similar to Java). But I also like the idea of synchonized API across runtimes. I can do that. |
Signed-off-by: Ivan Kochurkin <kvanttt@gmail.com>
Signed-off-by: Ivan Kochurkin <kvanttt@gmail.com>
Signed-off-by: Ivan Kochurkin <kvanttt@gmail.com>
38e7869 to
fb6fee8
Compare
|
I had a random thought: Is any user code relying on EMPTY? I don't think so from a quick check but we want to be careful we don't break any user code as a result of this. |
Yes, we should be careful, but removing deadlocks is much more important than insignificant changing of API. And since we change Java code, code of other runtimes can be changed as well. |
|
I've almost completed everything, but I also want to update documentation. |
Referencing subclass ParserRuleContext from superclass RuleContext initializer might lead to class loading deadlock Signed-off-by: Ivan Kochurkin <kvanttt@gmail.com>
Signed-off-by: Ivan Kochurkin <kvanttt@gmail.com>
…odeCount Signed-off-by: Ivan Kochurkin <kvanttt@gmail.com>
Signed-off-by: Ivan Kochurkin <kvanttt@gmail.com>
Remove useless loadTemplates overloads Signed-off-by: Ivan Kochurkin <kvanttt@gmail.com>
Signed-off-by: Ivan Kochurkin <kvanttt@gmail.com>
Signed-off-by: Ivan Kochurkin <kvanttt@gmail.com>
Signed-off-by: Ivan Kochurkin <kvanttt@gmail.com>
Signed-off-by: Ivan Kochurkin <kvanttt@gmail.com>
…ent navigation to tests data in IntelliJ Signed-off-by: Ivan Kochurkin <kvanttt@gmail.com>
Signed-off-by: Ivan Kochurkin <kvanttt@gmail.com>
Signed-off-by: Ivan Kochurkin <kvanttt@gmail.com>
Signed-off-by: Ivan Kochurkin <kvanttt@gmail.com>
fb6fee8 to
ff655e3
Compare
|
I think now it's ready for review and merge. |
Signed-off-by: Ivan Kochurkin <kvanttt@gmail.com>
Signed-off-by: Ivan Kochurkin <kvanttt@gmail.com>
ff655e3 to
f235d33
Compare
|
In the middle of the review but wanted to comment that this is very well structured...each commit does a logical unit of work. thanks! And amazing piece of refactoring overall 💯 |
runtime-testsuite/test/org/antlr/v4/test/runtime/RuntimeTests.java
Outdated
Show resolved
Hide resolved
doc/antlr-project-testing.md
Outdated
|
|
||
| All test are run in parallel both via maven and via IDE. | ||
|
|
||
| In IntelliJ, it's very easy to go to source by right-click on any test and pressing `Jump to source` (F4). |
runtime-testsuite/resources/org/antlr/v4/test/runtime/helpers/RuntimeTestParser.java
Show resolved
Hide resolved
I'd like to thank my @JetBrains team. They taught me to keep structured git history and write very conscious code. |
Signed-off-by: Ivan Kochurkin <kvanttt@gmail.com>
Fix misprint Signed-off-by: Ivan Kochurkin <kvanttt@gmail.com>
|
An amazing PR if I ever saw one thanks!!! Merging... |
See antlr/antlr4#3752 Commit 875911c2b2e6e50317e75ff2b5e80f4014a75954
@parrt I've fixed Java and C# runtimes. If it's ok, I'll unify other runtimes as well (where multithreading is actual)
Now Java runtime tests are superfast since they are inprocess (30s instead of 1.5 minutes)!
Also, now it's very easy to navigate to test data via tests in IntelliJ! See the video:
idea64_dBV9JpR0fg.mp4