Because it ends up treating it as a local variable, and `a.x`
is not a valid local variable name.
I'm not big on pattern matching, but conceptually it makes sense to me
to treat anything inside ^() to not be
pattern matching syntax?
80dbd85c45
`StringNode` and `SymbolNode` don't have the same shape
(`content` vs `value`) and that wasn't handled.
I believe the logic for the common case can be reused.
I simply left the special handling for implicit nodes in pattern matching
and fall through otherwise.
NOTE: patterns.txt is not actually tested at the moment,
because it contains syntax that `parser` mistakenly rejects.
But I checked manually that this doesn't introduce other failures.
https://github.com/whitequark/parser/pull/106055adfaa895
Mostly around newlines and line continuation.
* percent arrays need special backslash handling in the ast
* Fix offset issue for heredocs with many line continuations (used wrong variable as index access)
* More refined rules on when to simplify string tokens
* Handle line continuations in squiggly heredocs
* Correctly dedent squiggly heredocs with interpolation
* Consider `':foo:` and `%s[foo]` to not be interpolation
4edfe9d981
I see `Array.include?` as 2.4% runtime. Probably because of `LPAREN_CONVERSION_TOKEN_TYPES` but
the others will be faster as well.
Also remove some inline array checks. They are specifically optimized in Ruby since 3.4, but for now prism is for >= 2.7
ca9500a3fc
`Integer#chr` performs some validation that we don't want/need. Octal escapes can go above 255, where it will then raise trying to convert.
`append_as_bytes` actually allows to pass a number, so we can just skip that call.
Although, on older rubies of course we still need to handle this in the polyfill.
I don't really like using `pack` but don't know of another way to do so.
For the utf-8 escapes, this is not an issue. Invalid utf-8 in these is simply a syntax error.
161c606b1f
Mostly around newlines and line continuation.
* percent arrays need special backslash handling in the ast
* Fix offset issue for heredocs with many line continuations (used wrong variable as index access)
* More refined rules on when to simplify string tokens
* Handle line continuations in squiggly heredocs
* Correctly dedent squiggly heredocs with interpolation
* Consider `':foo:` and `%s[foo]` to not be interpolation
4edfe9d981
Turns out, it was already almost correct. If you disregard \c and \M style escapes, only a single character is allowed to be escaped in a regex so most tests passed already.
There was also a mistake where the wrong value was constructed for the ast, this is now fixed.
One test fails because of this, but I'm fairly sure it is because of a parser bug. For `/\“/`, the backslash is supposed to be removed because it is a multibyte character. But tbh,
I don't entirely understand all the rules.
Fixes more than half of the remaining ast differences for rubocop tests
e1c75f304b
The offset cache contains an entry for each byte so it can't be accessed via the string length.
Adds tests for all variants except for this:
```
"fo
o" "ba
’"
```
For some reason, this still has the wrong offset.
a651126458
There are a few other locations that should be included in that check.
I think the end location must always be present but I left it in to be safe (maybe implicit begin somehow?)
545d07ddc3
Mostly around newlines and line continuation.
* percent arrays need special backslash handling in the ast
* Fix offset issue for heredocs with many line continuations (used wrong variable as index access)
* More refined rules on when to simplify string tokens
* Handle line continuations in squiggly heredocs
* Correctly dedent squiggly heredocs with interpolation
* Consider `':foo:` and `%s[foo]` to not be interpolation
4edfe9d981
1. The string starts out as binary
2. `ち` is appended, forcing it back into utf-8
3. Some invalid byte sequences are tried to append
> incompatible character encodings: UTF-8 and BINARY (ASCII-8BIT)
This makes use of my wish to use `append_as_bytes`. Unfortunatly that method is rather new
so it needs a fallback
e31e94a775
I see `Array.include?` as 2.4% runtime. Probably because of `LPAREN_CONVERSION_TOKEN_TYPES` but
the others will be faster as well.
Also remove some inline array checks. They are specifically optimized in Ruby since 3.4, but for now prism is for >= 2.7
ca9500a3fc
Temoprary backwards-compat code so that current users
don't break.
Eventually the Translation::Parser initializer should asser that the correct class is passed in.
66b0162b35
## Summary
`itblock` node is added to support the `it` block parameter syntax introduced in Ruby 3.4.
```console
$ ruby -Ilib -rprism -rprism/translation/parser34 -e 'buffer = Parser::Source::Buffer.new("path"); buffer.source = "proc { it }"; \
p Prism::Translation::Parser34.new.tokenize(buffer)[0]'
s(:itblock,
s(:send, nil, :proc), :it,
s(:lvar, :it))
```
This node design is similar to the `numblock` node, which was introduced for the numbered parameter syntax in Ruby 2.7.
```
$ ruby -Ilib -rprism -rprism/translation/parser34 -e 'buffer = Parser::Source::Buffer.new("path"); buffer.source = "proc { _1 }"; \
p Prism::Translation::Parser34.new.tokenize(buffer)[0]'
s(:numblock,
s(:send, nil, :proc), 1,
s(:lvar, :_1))
```
The difference is that while numbered parameters can have multiple parameters, the `it` block parameter syntax allows only a single parameter.
In Ruby 3.3, the conventional node prior to the `it` block parameter syntax is returned.
```console
$ ruby -Ilib -rprism -rprism/translation/parser33 -e 'buffer = Parser::Source::Buffer.new("path"); buffer.source = "proc { it }"; \
p Prism::Translation::Parser33.new.tokenize(buffer)[0]'
s(:block,
s(:send, nil, :proc),
s(:args),
s(:send, nil, :it))
```
## Development Note
The Parser gem does not yet support the `it` block parameter syntax. This is the first case where Prism's node design precedes that of the Parser gem.
When implementing https://github.com/whitequark/parser/issues/962, this node design will need to be taken into consideration.
c141e1420a
I want to add new node types to the parser translator, for example `itblock`. The bulk of the work is already done by prism itself. In the `parser`
builder, this would be a 5-line change at most but we don't control that here.
Instead, we can add our own builder and either overwrite the few methods we need,
or just inline the complete builder. I'm not sure yet which would be better.
`rubocop-ast` uses its own builder for `parser`. For this to correctly work, it must explicitly choose to extend the
prism builder and use it, same as it currently chooses to use a different parser when prism is used.
I'd like to enforce that the builder for prism extends its custom one since it will lead to
some pretty weird issues otherwise. But first, I'd like to change `rubocop-ast` to make use of this.
b080e608a8
`builder.pair_label` is no good since it makes use of variables that the parser gem encountered.
Since the prism translator doesn't keep proper track of that information, the following code interprets
the implicit value as a local variable, even though it is not in scope:
```rb
def foo
bar = 123
end
{ bar: }
```
bbeb5b083a
Turns out, it was already almost correct. If you disregard \c and \M style escapes, only a single character is allowed to be escaped in a regex so most tests passed already.
There was also a mistake where the wrong value was constructed for the ast, this is now fixed.
One test fails because of this, but I'm fairly sure it is because of a parser bug. For `/\“/`, the backslash is supposed to be removed because it is a multibyte character. But tbh,
I don't entirely understand all the rules.
Fixes more than half of the remaining ast differences for rubocop tests
Tests worked around this but the incompatibility is not hard to fix.
This fixes 17 token incompatibilies in tests here that were previously passing
101962526d
In https://github.com/ruby/prism/pull/3393 I made a mistake.
When there is no previous token, it wraps around to -1. Oops
Additionally, if a comment has no newline then the offset should be kept as is
3c266f1de4
There appear to be a bunch of rules, changing behaviour for
inline comments, multiple comments after another, etc.
This seems to line up with reality pretty closely, token differences for RuboCop tests go from 1129 to 619 which seems pretty impressive
2e1b92670c
These are not line continuations. They either should be taken literally,
or allow the word array to contain the following whitespace (newlines in this case)
Before:
```
0...1: tSTRING_BEG => "'"
1...12: tSTRING_CONTENT => "foobar\\\n"
12...16: tSTRING_CONTENT => "baz\n"
16...17: tSTRING_END => "'"
17...18: tNL => nil
```
After:
```
0...1: tSTRING_BEG => "'"
1...6: tSTRING_CONTENT => "foo\\\n"
6...12: tSTRING_CONTENT => "bar\\\n"
12...16: tSTRING_CONTENT => "baz\n"
16...17: tSTRING_END => "'"
17...18: tNL => nil
```
b6554ad64e
Much of this logic should be shared between interpolated symbols and regexps.
It's also incorrect when the node contains a literal `\\n` (same as for plain string nodes at the moment)
561914f99b
Turns out, the vast majority of work was already done with handling the same for heredocs
I'm confident this should also apply to actual string nodes (there's even a todo for it) but
no tests change if I apply it there too, so I can't say for sure if the logic would be correct.
The individual test files are a bit too large, maybe something else would break that currently passes.
Leaving it for later to look more closely into that.
6bba1c54e1
Blocks and lambdas inherit anonymous arguments from the method they are a part of.
They themselves don't allow to introduce new anonymous arguments.
While you can write this:
```rb
def foo(*)
bar { |**| }
end
```
referecing the new parameter inside of the block will always be a syntax error.
2cbd27e134
The offset cache contains an entry for each byte so it can't be accessed via the string length.
Adds tests for all variants except for this:
```
"fo
o" "ba
’"
```
For some reason, this still has the wrong offset.
a651126458
