On 21.07.2024 at 13:28, Tim Düsterhus wrote:
> On 7/20/24 20:31, Christoph M. Becker wrote:
>
>> If the tokenizer would tokenize a whole file as a single token, would
>> that also be correct? Of course, I'm exaggerating, but
>
> The function is documented as:
>
>> token_get_all() parses the given code string into PHP language tokens
>> using the Zend engine's lexical scanner.
>
> So if Zend engine's lexical scanner would tokenize a complete PHP file
> as a single token, then the output would be correct, yes.
Apparently we cannot even agree that there is a bug, and as such we
cannot fix it in PHP 8.3 according to the release process documentation.
>> <https://3v4l.org/qIf2c> doesn't look
>> correct to me – "yield /* comment
>> */ from" shouldn't be a single token.
>
> Looking at this from a user expectation perspective and not a technical
> perspective, neither should yield from be a single token in:
>
> var_dump(token_get_all('<?php yield from $foo;'));
ACK
Cheers,
Christoph