Re: NULL Coercion Consistency

On 28 May 2022, at 03:36, Craig Francis <craig@craigfrancis.co.uk> wrote: > On 8 Apr 2022, at 18:34, Craig Francis <craig@craigfrancis.co.uk> wrote: >> I've written a new draft RFC to address the NULL coercion problems: >> https://wiki.php.net/rfc/null_coercion_consistency > > > I give up. For everyone affected by this... Rector has a "solution" via NullToStrictStringFuncCallArgRector. It has a list of 275 functions (442 parameters), and every time these functions are used, and Rector isn't sure a variable cannot be NULL, it will add a string type cast, e.g. - <?= htmlspecialchars($var) ?> + <?= htmlspecialchars((string) $var) ?> So yeah, it's very messy, and while I think it makes the code worse, at least it's one way to avoid the fatal errors (coming in 9.0). Oh, and developers can easily re-introduce more "problems" (because they might not be aware which variables can be NULL), so Rector will need to re-run. I'm also tempted to create a composer library to provide a second solution, by using a namespace to re-define these functions, e.g. <?php namespace null_coercion_fixer; function strlen(?string $string): int { return \strlen(\strval($string)); } ?> Have fun, Craig https://github.com/rectorphp/rector-src/blob/main/rules/Php81/Rector/FuncCall/NullToStrictStringFuncCallArgRector.php#L37