| Age | Commit message (Collapse) | Author |
|---|
|
https://github.com/ruby/prism/commit/6f886be0a4
|
|
https://github.com/ruby/prism/commit/4d8929ff6a
|
|
https://github.com/ruby/prism/commit/2752f0b8df
|
|
https://github.com/ruby/prism/commit/9e200dd1c1
|
|
https://github.com/ruby/prism/commit/a0b978d25b
|
|
After finding the "if if" typo, some additional typos identified by running `codespell` are also being corrected:
https://github.com/codespell-project/codespell
https://github.com/ruby/prism/commit/e6a34cfeeb
|
|
https://github.com/ruby/prism/commit/592128de4d
|
|
"unparser/corpus/literal/def.txt"
See the discussion on https://github.com/ruby/ruby/pull/9923
|
|
The previous commit changes Ripper lex state behavior of `tLABEL`.
For example, Ripper lex state for the second `bar` was changed
by the previous commit.
```
def foo(bar: bar())
end
```
Before the commit, Ripper didn’t add label `bar` id to `vtable`
because `formal_argument` function for Ripper didn’t return id
but returned `VALUE lhs`.
Therefore Ripper didn’t `SET_LEX_STATE(EXPR_END|EXPR_LABEL)` for following `bar`
in `parse_ident` because `lvar_defined` for following `bar` was not true.
Currently Ripper does `SET_LEX_STATE(EXPR_END|EXPR_LABEL)` then
the result of this test case is changed like below.
```
Prism::ParseTest#test_filepath_unparser/corpus/literal/def.txt [ruby/test/prism/parse_test.rb:280]:
<[[80, 13], :on_ident, “bar”, END|LABEL]> expected but was
<[[80, 13], :on_ident, “bar”, ARG]>.
```
Parser sets lex state to `END|LABEL` on master branch.
Therefore previous commit makes Ripper’s behavior aligned with parser’s behavior.
```
$ ruby -v -y -e "def foo(bar: bar())" -e "end"
ruby 3.4.0dev (2024-02-11T23:52:05Z master 697ade7bda) [arm64-darwin21]
...
Reading a token
parser_dispatch_scan_event:11210 (1: 12|1|7)
lex_state: ARG|LABELED -> ARG at line 11113
lex_state: ARG -> END|LABEL at line 11128
parser_dispatch_scan_event:11877 (1: 13|3|4)
Next token is token “local variable or method” (1.13-1.16: bar)
Shifting token “local variable or method” (1.13-1.16: bar)
```
|
|
* Faster that way:
$ ruby -Ilib -rprism -rbenchmark/ips -e 'Benchmark.ips { |x| x.report("parse") { Prism.parse("1 + 2") } }'
195.722k (± 0.5%) i/s
rb_iv_set():
179.609k (± 0.5%) i/s
rb_funcall():
190.030k (± 0.3%) i/s
before this PR:
183.319k (± 0.4%) i/s
|
|
This PR implements proper file parsing error handling. Previously
`file_options` would call `pm_string_mapped_init` which would print an
error from `perror`. However this wouldn't raise a proper Ruby error so
it was just a string output. I've done the following:
- Raise an error from `rb_syserr_fail` with the filepath in
`file_options`.
- No longer return `Qnil` if `file_options` returns false (because now
it will raise)
- Update `file_options` to return `static void` instead of `static
bool`.
- Update `file_options` and `profile_file` to check the type so when
passing `nil` we see a `TypeError`.
- Delete `perror` from `pm_string_mapped_init`
- Update `FFI` backend to raise appropriate errors when calling
`pm_string_mapped_init`.
- Add tests for `dump_file`, `lex_file`, `parse_file`,
`parse_file_comments`, `parse_lex_file`, and `parse_file_success?`
when a file doesn't exist and for `nil`.
- Updates the `bin/parse` script to no longer raise it's own
`ArgumentError` now that we raise a proper error.
Fixes: ruby/prism#2207
https://github.com/ruby/prism/commit/b2f7494ff5
|
|
https://github.com/ruby/prism/commit/12649ee74c
|
|
https://github.com/ruby/prism/commit/e190308845
|
|
https://github.com/ruby/prism/commit/aa8c702271
|
|
Ruby allows for 0 or negative line start, this is often used
with `eval` calls to get a correct offset when prefixing a snippet.
e.g.
```ruby
caller = caller_locations(1, 1).first
class_eval <<~RUBY, caller.path, caller.line - 2
# frozen_string_literal: true
def some_method
#{caller_provided_code_snippet}
end
RUBY
```
https://github.com/ruby/prism/commit/0d14ed1452
|
|
https://github.com/ruby/prism/commit/8582d372a3
|
|
https://github.com/ruby/prism/commit/bf9bdb9d82
|
|
|
|
|
|
|
|
|
|
|
