diff options
author | Arnold D. Robbins <arnold@skeeve.com> | 2020-11-06 14:17:43 +0200 |
---|---|---|
committer | Arnold D. Robbins <arnold@skeeve.com> | 2020-11-06 14:17:43 +0200 |
commit | 793fc4309e655d68ba2972fb0e7c4be4346424ea (patch) | |
tree | 04f6abe151269fd023d86cd4b85c699068e51099 /support/regex_internal.c | |
parent | 22ca86022481cbdfbdc3a2a1f0977735067b2447 (diff) | |
parent | 1dbaf410e0b3ec754ee4ab78f861d4ef034e7962 (diff) | |
download | egawk-793fc4309e655d68ba2972fb0e7c4be4346424ea.tar.gz egawk-793fc4309e655d68ba2972fb0e7c4be4346424ea.tar.bz2 egawk-793fc4309e655d68ba2972fb0e7c4be4346424ea.zip |
Merge branch 'master' into feature/cmp_scalars
Diffstat (limited to 'support/regex_internal.c')
-rw-r--r-- | support/regex_internal.c | 19 |
1 files changed, 10 insertions, 9 deletions
diff --git a/support/regex_internal.c b/support/regex_internal.c index e1b6b4d5..ed0a1346 100644 --- a/support/regex_internal.c +++ b/support/regex_internal.c @@ -300,18 +300,20 @@ build_wcs_upper_buffer (re_string_t *pstr) while (byte_idx < end_idx) { wchar_t wc; + unsigned char ch = pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx]; - if (isascii (pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx]) - && mbsinit (&pstr->cur_state)) + if (isascii (ch) && mbsinit (&pstr->cur_state)) { - /* In case of a singlebyte character. */ - pstr->mbs[byte_idx] - = toupper (pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx]); /* The next step uses the assumption that wchar_t is encoded ASCII-safe: all ASCII values can be converted like this. */ - pstr->wcs[byte_idx] = (wchar_t) pstr->mbs[byte_idx]; - ++byte_idx; - continue; + wchar_t wcu = __towupper (ch); + if (isascii (wcu)) + { + pstr->mbs[byte_idx] = wcu; + pstr->wcs[byte_idx] = wcu; + byte_idx++; + continue; + } } remain_len = end_idx - byte_idx; @@ -348,7 +350,6 @@ build_wcs_upper_buffer (re_string_t *pstr) { /* It is an invalid character, an incomplete character at the end of the string, or '\0'. Just use the byte. */ - int ch = pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx]; pstr->mbs[byte_idx] = ch; /* And also cast it to wide char. */ pstr->wcs[byte_idx++] = (wchar_t) ch; |