General

Profile

make_now_just (Hiroya Fujinami)

Issues

open closed Total
Assigned issues 1 28 29
Reported issues 2 8 10

Projects

Project Roles Registered on
Ruby Committer 12/13/2022

Activity

11/18/2026

10:18 PM Ruby Revision 9a6ccf70 (git): Check the end of the subject in string_cmp_ic
The case-insensitive backreference comparison folded the character at
the end of the subject when the captured text and the subject differ in
the byte length of their case folds, and could report a match ending
beyond the subject.
make_now_just (Hiroya Fujinami)
09:57 PM Ruby Revision b469749b (git): Check the end of the subject in string_cmp_ic
The case-insensitive backreference comparison folded the character at
the end of the subject when the captured text and the subject differ in
the byte length of their case folds, and could report a match ending
beyond the subject.
make_now_just (Hiroya Fujinami)
09:39 PM Ruby Revision 1184052e (git): Check the end of the subject in string_cmp_ic
The case-insensitive backreference comparison folded the character at
the end of the subject when the captured text and the subject differ in
the byte length of their case folds, and could report a match ending
beyond the subject.
make_now_just (Hiroya Fujinami)
08:52 AM Ruby Revision f4b895ce (git): Fix memoization for the `/(...){0}/` case (#13169)
In this case, the previous implementation counted an extra number of
opcodes to cache and the matching was unstable on memoization.
This patch is to fix that problem by not counting an number of opcodes
to cache in the parentheses of `(...
make_now_just (Hiroya Fujinami)

11/13/2026

03:35 AM Ruby Bug #22495 (Closed): UTF-32 strings with a byte 0x80 or more in the most significant position are reported as valid
Applied in changeset commit:git|e1a9429918d5ead25a00f79a92f750abbeaa519a.
----------
Compute UTF-32 code points in unsigned arithmetic
utf32be_mbc_to_code and utf32le_mbc_to_code computed the code point in
int, which overflowed when th...
make_now_just (Hiroya Fujinami)
03:34 AM Ruby Revision e1a94299 (git): Compute UTF-32 code points in unsigned arithmetic
utf32be_mbc_to_code and utf32le_mbc_to_code computed the code point in
int, which overflowed when the most significant byte was 0x80 or more.
With gcc -O2 such sequences were accepted as valid UTF-32.
[Bug #22495]
make_now_just (Hiroya Fujinami)

11/12/2026

06:58 PM Ruby Revision ebf8b9e8 (git): Keep the groups of a regexp compiled again for a string
When a regexp is compiled again for the encoding of a string, pass its
options to the preprocessor so that the comments of extended mode are
dropped as in the first compilation. Also make is_code_ctype of
Emacs-Mule return false for non...
make_now_just (Hiroya Fujinami)
06:58 PM Ruby Revision 6e7a8c5d (git): Do not use the leading .* shortcut in a backward search
A match at an earlier position does not imply one at the start position, so String#rindex with /.*.../m missed matches.
[Bug #22492]
(cherry picked from commit d55b67dfa3c1f94a4855381b400da7aa987bb5be)
make_now_just (Hiroya Fujinami)
06:58 PM Ruby Revision 85e6cd39 (git): Do not treat a leading .* after an assertion as an implicit anchor
When an assertion such as ^, \b or a lookahead precedes the leading .*, it may fail at the start position and hold later, so the search must try later positions.
[Bug #22494]
(cherry picked from commit dbfe6bac2285ea61a76a2dec970cf859f...
make_now_just (Hiroya Fujinami)

11/09/2026

12:00 PM Ruby Bug #22494 (Closed): Regexp search misses a match when an assertion precedes a leading .*
Applied in changeset commit:git|dbfe6bac2285ea61a76a2dec970cf859f5a8fef4.
----------
Do not treat a leading .* after an assertion as an implicit anchor
When an assertion such as ^, \b or a lookahead precedes the leading .*, it may fail...
make_now_just (Hiroya Fujinami)

Also available in: Atom