Valid python files, that fails to parse in ruff

[qarmin]

Ruff 0.0.291
Cpython 3.11.4

Tested with

ruff format . --check

and cpython part

python -m compileall .

RuffBroken.zip

Example of file

from ..const import d#ta_source_const

def validate_incming_payload(baseballdc_request):

    requested_data_source_trimmed = baseballdc_request['data_source'].strip().upper()

    if(requested_data_source_trimmed not in data_source_const.DATA_SOURCE_LIST):
        raise ValueError(generate_data_source_error_message(baseballdc_request))

def generate_data_source_error_message(baseballdc_request): 

    requested_data_source = baseballdc_request['data_source']
    valid_data_sources = data_source_cons.DATA_SOURCE_LIST

    error_message_text = 'Value error in the incoming request payload:'
    data_sourcÛ_error_text = f'The data source requested ("{requested_data_source}") was not valid.'
    valid_data_source_text = f'The current list of valid datasources to choose from for basblldc is: {valid_data_sources}'

    return f'{error_message_text}\n\n{data_source_erro%_text}\{valid_data_source_text}\n';

error

error: Failed to format 48_IDX_0_RAND_14765242697460627399.py: source contains syntax errors: ParseError { error: Lexical(FStringError(SingleRbrace)), offset: 908, source_path: "<filename>" }

[dhruvmanila]

The following files contains f-string errors which will be fixed by #7376. I've verified that those files are parsed without any errors on that branch:

• 335_IDX_0_RAND_13905148532772824398.py
• 48_IDX_0_RAND_14765242697460627399.py
• 586_IDX_0_RAND_36061242038301280.py
• 855_IDX_0_RAND_15689123331827052431.py

Other files contains identifier starting with unicode character which we're unable to identify which then produces a syntax error.

[qarmin]

Looks that #7376 introduced regression, and now some valid python files shows errors:

error: Failed to format 549_IDX_0_RAND_5726799750657298695.py: source contains syntax errors: LexicalError { error: FStringError(UnterminatedString), location: 3282 }
error: Failed to format 230_IDX_0_RAND_14394415519784108053.py: source contains syntax errors: ParseError { error: UnrecognizedToken(For, None), offset: 1799, source_path: "<filename>" }

RuffBroken.zip

[MichaReiser]

Looks that #7376 introduced regression, and now some valid python files shows errors:

error: Failed to format 549_IDX_0_RAND_5726799750657298695.py: source contains syntax errors: LexicalError { error: FStringError(UnterminatedString), location: 3282 }
error: Failed to format 230_IDX_0_RAND_14394415519784108053.py: source contains syntax errors: ParseError { error: UnrecognizedToken(For, None), offset: 1799, source_path: "<filename>" }

RuffBroken.zip

CC: @dhruvmanila

[charliermarsh]

@dhruvmanila - We should probably try to fix these regressions before the next release. I can also help out if needed.

[dhruvmanila]

230_IDX_0_RAND_14394415519784108053.py: This fails on 3.12.0rc3 as well because the generator is missing the parentheses. I'm surprised that it parses on 3.11. Maybe it's because the parentheses were added manually? Atleast in our case we did the expression parsing by adding the surrounding parentheses manually. So,

- f"', '.join({ckt.name for ckt in self._ckts_with_errors})"
+ f"', '.join({(ckt.name for ckt in self._ckts_with_errors)})"

It fails in the same way the below code fails:

foo = ckt.name for ckt in self._ckts_with_errors

---

549_IDX_0_RAND_5726799750657298695.py: This is failing because of Windows newline. A minimal repro:

f'text \\r\n
more text'