Fixes #152.

One of our production services uses re2j to match several hundred mostly
case-insensitive patterns of varying complexity against text.
We observed that approximately 12% of CPU time was being spent in
toLowerCase() as called from simpleFold(), due to the necessity of doing
at least one character data lookup per Inst.Rune in the common case that
the input rune being examined did not match the instruction.

As a fix, implement a method equalsIgnoreCase() that performs
Unicode-aware case-insensitive comparison between two runes, with a fast
path for the common case where both input runes are ASCII, and use it in
Inst for single-rune case-insensitive comparison. This takes character
data lookups out of the hot path.

The existing re2j benchmarks did not exercise case-insensitive patterns,
so add a new benchmark that executes a mostly ASCII regex pattern on a
text containing a mix of ASCII and Unicode characters (generated using
a Hungarian "lorem ipsum" text generator).

Also add unit tests for the new equality comparison logic.

Signed-off-by: Máté Szabó <[email protected]>

This follows google/re2@6148386 (and
golang/go@ee61186) to some extent.

dc7d6e5 unfortunately increases the
incidence of infinite loops during case folding if re2j is running on a
JVM newer than the version used to generate the bundled
UnicodeTables.java and the input contains a rune that would require
special case folding rules to form a closed fold loop. \u1C80 (Cyrillic
Small Letter Rounded Ve) is an example of such a rune.

Workaround the issue by inverting the order of parameters passed to
equalsIgnoreCase() so that the rune from the pattern being matched,
rather than the input content, undergoes case folding instead. This does
not fully eliminate the possibility of an infinite loop in this
scenario, since the pattern may well contain one of the problematic
runes, but it effectively restores the situation as it was pre
dc7d6e5, since the previous logic also
performed case folding on the rune from the pattern and not on the
content.

Signed-off-by: Máté Szabó <[email protected]>

PiperOrigin-RevId: 574899560

…from bytes. The platform default character set is guaranteed to be UTF-8.

PiperOrigin-RevId: 576577338