Re: Optimization for lower(), upper(), casefold() functions.

12.03.2025 22:39, Jeff Davis wrote:

[...]

>> 2. Added a fast path for codepoint < 0x80.
>>
>> v3j-0002:
>> In the fast path for codepoints < 0x80, I added a premature return.
>> This avoided additional insertions, which increased performance.
> 
> What do you mean "additional insertions"?

Sorry for my English. I mean, we immediately do a return in the
if () condition. To avoid further branching/checking.

> Also, should we just compute the results in the fast path? We don't
> even need a table. Rough patch attached to go on top of v4-0001.
> 
> Should we properly return CASEMAP_SELF when *simple == u1, or is it ok
> to return CASEMAP_SIMPLE? It probably doesn't matter performance-wise,
> but it feels more correct to return CASEMAP_SELF.

It seems to disrupt the overall "beauty" of the approach. Thus, we will
copy code (bloat code), make optimizations that do not improve
performance but bloat code. I would refrain from such practices.
Especially since we'll be changing all that in the next patch (v4-0002).

>>
>> Perhaps for general
>> beauty it should be made static inline, I don't have a rigid position
>> here.
> 
> We ordinarily use "static inline" if it's in a header file, and
> "static" if it's in a .c file, so I'll do it that way.

Great, I've changed this place. Performance has not changed in any way.

>> I was purely based on existing approaches in Postgres, the
>> Normalization Forms have them separated into different headers. Just
>> trying to be consistent with existing approaches.
> 
> I think that was done for normalization primarily because it's not used
> #ifndef FRONTEND (see unicode_norm.c), and perhaps also because it's
> just a more complex function worthy of its own file.
> 
> I looked into the history, and commit 783f0cc64d explains why perfect
> hashing is not used in the frontend:
> 
> "The decomposition table remains the same, getting used for the binary
> search in the frontend code, where we care more about the size of the
> libraries like libpq over performance..."

I removed the extra file (unicode_case_func.h). You are right, we should
not create unnecessary clutter.

v5 attached.

Regards,
Alexander Borisov