Re: Fix for fragile code in ltree/crc32.c

On 25/08/2026 11:13, Heikki Linnakangas wrote:
>> /*
>>  * Expansion factor of string length, not including terminating NUL.
>> That is,
>>  * the upper bound of the number of multibyte characters in the result
>> string
>>  * per multibyte character in the input string.
>>  *
>>  * NB: assumes no provider exceeds the Unicode-defined maximum.
>>  */
>> #define PG_MAX_CASEMAP_MBCHARS        UNICODE_MAX_CASEMAP_CODEPOINTS
>
> Is it a number of codepoints or bytes? The comment says "multibyte
> characters" which I think means codepoints in unicode, but the ltree
> code does this:
>
>>         /* max space required to map single codepoint, including NUL */
>>         char        foldstr[PG_CASEMAP_BUFSZ];
>
> Which sure looks like # of bytes.

Never mind. Just after sending this I saw that PG_CASEMAP_BUFSZ is
different from PG_MAX_CASEMAP_MBCHARS. Sorry for the noise.

- Heikki