ruby.git/test/prism, branch master

[ruby/prism] Fix parser translation when escaped newline with trailing content

2026-05-26T15:48:53+00:00

https://github.com/ruby/prism/commit/81e07f3cdb

[ruby/prism] Don't replicate unary method bug in parser translator

2026-05-26T14:29:38+00:00

Closes https://github.com/ruby/prism/issues/4112

Also see https://github.com/ruby/prism/issues/2501, for which this was done.
The expectation in rubocop is incorrect (produces code with semantic difference), so the test should be updated instead.

In any way, it should also have applied to `+`, where the same happens

https://github.com/ruby/prism/commit/1f8cae2a75

Reject `END { next } ` as well

2026-05-15T10:58:27+00:00

Followup for https://bugs.ruby-lang.org/issues/20409

[ruby/prism] Take the strings encoding as the initial encoding in the ripper translator

2026-05-08T20:12:59+00:00

When no magic encoding comment is present, it does not default to utf-8,
and takes the encoding of the string that contains the source code instead.
Most of the time that will be utf-8, but not always.

https://github.com/ruby/prism/commit/1a273db780

[ruby/prism] Respect `encoding` option in `Prism.lex` and friends

2026-05-08T20:12:59+00:00

utf-8 is the default for source files but can be overwritten via options

https://github.com/ruby/prism/commit/355f451528

[ruby/prism] Fix UTF-8 code units to match the number of bytes

2026-04-30T14:27:16+00:00

(https://github.com/ruby/prism/pull/4098)

https://github.com/ruby/prism/commit/442bd904ed

[ruby/prism] Reject `END { break }` for Ruby 4.0

2026-04-27T14:01:26+00:00

For [Bug #20409]

https://github.com/ruby/prism/commit/4848eb344e

[ruby/prism] Implement `on_label_end` for ripper

2026-04-24T13:35:04+00:00

It is emitted for string-like symbols only

https://github.com/ruby/prism/commit/6d63302c7b

[ruby/prism] Fix incompatibilities for `on_word_sep` in ripper

2026-04-23T23:49:06+00:00

It was mostly good, just a few edgecases:
* word_sep always only contains a single line. if there are multiple lines, it is also multiple word_sep
* Handle trailing whitespace

The excluded testcase is interpolation mixed with heredoc.
That is not currently handled properly and word_sep contains
the content of the heredoc

I also switched start/end_offset names in the method, they seem to be backwards

https://github.com/ruby/prism/commit/2e151ad41b

[ruby/prism] Implement bracket/braces events for ripper

2026-04-14T13:26:49+00:00

Drops the check against order. Very often ripper emits events in a order
that is not easy to mimic. It's only getting worse now that most events are implemented.

Perhaps the test can be brought back at a later time. For now, I used it while making this commit
but adding all these exceptions makes not much sense.
It already was a pretty long list.

https://github.com/ruby/prism/commit/f43299a18b