Possible dead or incorrect code in Lexer.create_token() related to TRANSLATOR_COMMENT_MARK

[Chris Jerdonek]

While working on ​this PR for ticket #32919, I noticed that Lexer.create_token() has a branch of code that looks like it might be wrong or not needed.

Specifically, ​this line seems always to be true. It's true both when TRANSLATOR_COMMENT_MARK is found and not found. It's only false when token_string starts with TRANSLATOR_COMMENT_MARK since str.find() is used (as it returns the index). However, when I tried modifying the code in ​this test PR to raise an exception in the false code path, the test suite still passed. So it's not clear to me if the str.find() call is needed, or if it's just incorrect.

[Mariusz Felisiak]

content for non-translator comments is unnecessary, so I'd change this to token_string.find(TRANSLATOR_COMMENT_MARK) > -1.

Incorrect since its introduction in 17b329ae08f9e3886da2baafc9e53949000480f9.

[Chris Jerdonek]

In that case, you'd want TRANSLATOR_COMMENT_MARK in token_string, right? (See note at ​https://docs.python.org/3/library/stdtypes.html#str.find )

[Mariusz Felisiak]

Replying to Chris Jerdonek:

In that case, you'd want TRANSLATOR_COMMENT_MARK in token_string, right? (See note at ​https://docs.python.org/3/library/stdtypes.html#str.find )

TBH, checking TRANSLATOR_COMMENT_MARK is probably completely unnecessary here since 47ddd6a4082d55d8856b7e6beac553485dd627f7. We can remove find().

[Chris Jerdonek]

Okay, sounds good.

[Chris Jerdonek]

PR: ​https://github.com/django/django/pull/14743

[Mariusz Felisiak <felisiak.mariusz@…>]

In e79ae5c:

Fixed #32986 -- Removed unneeded str.find() call in Lexer.create_token().

Unnecessary since 47ddd6a4082d55d8856b7e6beac553485dd627f7.

[Mariusz Felisiak <felisiak.mariusz@…>]

In 4fe3774:

Refs #32986 -- Moved TRANSLATOR_COMMENT_MARK to django.utils.translation.template.