aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
Diffstat
-rw-r--r--src/nvim/regexp.c57
-rw-r--r--src/nvim/regexp_nfa.c29
2 files changed, 37 insertions, 49 deletions
diff --git a/src/nvim/regexp.c b/src/nvim/regexp.c
index 6316129c6a..1c88bd4ba4 100644
--- a/src/nvim/regexp.c
+++ b/src/nvim/regexp.c
@@ -797,8 +797,7 @@ static int get_equi_class(char_u **pp)
*/
static void reg_equi_class(int c)
{
- if (enc_utf8 || STRCMP(p_enc, "latin1") == 0
- || STRCMP(p_enc, "iso-8859-15") == 0) {
+ {
switch (c) {
// Do not use '\300' style, it results in a negative number.
case 'A': case 0xc0: case 0xc1: case 0xc2:
@@ -1141,7 +1140,7 @@ static char_u *skip_anyof(char_u *p)
if (*p == ']' || *p == '-')
++p;
while (*p != NUL && *p != ']') {
- if (has_mbyte && (l = (*mb_ptr2len)(p)) > 1) {
+ if ((l = (*mb_ptr2len)(p)) > 1) {
p += l;
} else if (*p == '-') {
p++;
@@ -1876,7 +1875,7 @@ static char_u *regatom(int *flagp)
EMSG_RET_NULL(_("E63: invalid use of \\_"));
/* When '.' is followed by a composing char ignore the dot, so that
* the composing char is matched here. */
- if (enc_utf8 && c == Magic('.') && utf_iscomposing(peekchr())) {
+ if (c == Magic('.') && utf_iscomposing(peekchr())) {
c = getchr();
goto do_multibyte;
}
@@ -2242,11 +2241,7 @@ collection:
if (*regparse == '[')
endc = get_coll_element(&regparse);
if (endc == 0) {
- if (has_mbyte) {
- endc = mb_ptr2char_adv((const char_u **)&regparse);
- } else {
- endc = *regparse++;
- }
+ endc = mb_ptr2char_adv((const char_u **)&regparse);
}
/* Handle \o40, \x20 and \u20AC style sequences */
@@ -2256,8 +2251,8 @@ collection:
if (startc > endc) {
EMSG_RET_NULL(_(e_reverse_range));
}
- if (has_mbyte && ((*mb_char2len)(startc) > 1
- || (*mb_char2len)(endc) > 1)) {
+ if ((*mb_char2len)(startc) > 1
+ || (*mb_char2len)(endc) > 1) {
// Limit to a range of 256 chars
if (endc > startc + 256) {
EMSG_RET_NULL(_(e_large_class));
@@ -2502,9 +2497,9 @@ do_multibyte:
&& !one_exactly
&& !is_Magic(c))); ++len) {
c = no_Magic(c);
- if (has_mbyte) {
+ {
regmbc(c);
- if (enc_utf8) {
+ {
int l;
/* Need to get composing character too. */
@@ -2516,8 +2511,7 @@ do_multibyte:
skipchr();
}
}
- } else
- regc(c);
+ }
c = getchr();
}
ungetchr();
@@ -4248,15 +4242,13 @@ static bool regmatch(
opnd = OPERAND(scan);
// Inline the first byte, for speed.
if (*opnd != *rex.input
- && (!rex.reg_ic
- || (!enc_utf8
- && mb_tolower(*opnd) != mb_tolower(*rex.input)))) {
+ && (!rex.reg_ic)) {
status = RA_NOMATCH;
} else if (*opnd == NUL) {
// match empty string always works; happens when "~" is
// empty.
} else {
- if (opnd[1] == NUL && !(enc_utf8 && rex.reg_ic)) {
+ if (opnd[1] == NUL && !rex.reg_ic) {
len = 1; // matched a single byte above
} else {
// Need to match first byte again for multi-byte.
@@ -4267,7 +4259,7 @@ static bool regmatch(
}
// Check for following composing character, unless %C
// follows (skips over all composing chars).
- if (status != RA_NOMATCH && enc_utf8
+ if (status != RA_NOMATCH
&& UTF_COMPOSINGLIKE(rex.input, rex.input + len)
&& !rex.reg_icombine
&& OP(next) != RE_COMPOSING) {
@@ -4336,7 +4328,7 @@ static bool regmatch(
break;
case RE_COMPOSING:
- if (enc_utf8) {
+ {
// Skip composing characters.
while (utf_iscomposing(utf_ptr2char(rex.input))) {
MB_CPTR_ADV(rex.input);
@@ -5366,9 +5358,10 @@ do_class:
if (got_int) {
break;
}
- } else if (has_mbyte && (l = (*mb_ptr2len)(scan)) > 1) {
- if (testval != 0)
+ } else if ((l = (*mb_ptr2len)(scan)) > 1) {
+ if (testval != 0) {
break;
+ }
scan += l;
} else if ((class_tab[*scan] & mask) == testval) {
scan++;
@@ -5481,7 +5474,7 @@ do_class:
/* Safety check (just in case 'encoding' was changed since
* compiling the program). */
if ((len = (*mb_ptr2len)(opnd)) > 1) {
- if (rex.reg_ic && enc_utf8) {
+ if (rex.reg_ic) {
cf = utf_fold(utf_ptr2char(opnd));
}
while (count < maxcount && (*mb_ptr2len)(scan) >= len) {
@@ -5490,7 +5483,7 @@ do_class:
break;
}
}
- if (i < len && (!rex.reg_ic || !enc_utf8
+ if (i < len && (!rex.reg_ic
|| utf_fold(utf_ptr2char(scan)) != cf)) {
break;
}
@@ -6383,7 +6376,7 @@ static int cstrncmp(char_u *s1, char_u *s2, int *n)
}
// if it failed and it's utf8 and we want to combineignore:
- if (result != 0 && enc_utf8 && rex.reg_icombine) {
+ if (result != 0 && rex.reg_icombine) {
char_u *str1, *str2;
int c1, c2, c11, c12;
int junk;
@@ -6501,10 +6494,10 @@ char_u *regtilde(char_u *source, int magic)
STRMOVE(p, p + 2); /* remove '\~' */
--p;
} else {
- if (*p == '\\' && p[1]) /* skip escaped characters */
- ++p;
- if (has_mbyte)
- p += (*mb_ptr2len)(p) - 1;
+ if (*p == '\\' && p[1]) { // skip escaped characters
+ p++;
+ }
+ p += (*mb_ptr2len)(p) - 1;
}
}
@@ -6940,7 +6933,7 @@ static int vim_regsub_both(char_u *source, typval_T *expr, char_u *dest,
else /* just copy */
cc = c;
- if (has_mbyte) {
+ {
int l;
// Copy composing characters separately, one
@@ -6953,8 +6946,6 @@ static int vim_regsub_both(char_u *source, typval_T *expr, char_u *dest,
utf_char2bytes(cc, dst);
}
dst += utf_char2len(cc) - 1;
- } else if (copy) {
- *dst = cc;
}
dst++;
}
diff --git a/src/nvim/regexp_nfa.c b/src/nvim/regexp_nfa.c
index 7cd1ae93d2..7dfd16fb4f 100644
--- a/src/nvim/regexp_nfa.c
+++ b/src/nvim/regexp_nfa.c
@@ -704,8 +704,7 @@ static void nfa_emit_equi_class(int c)
#define EMIT2(c) EMIT(c); EMIT(NFA_CONCAT);
#define EMITMBC(c) EMIT(c); EMIT(NFA_CONCAT);
- if (enc_utf8 || STRCMP(p_enc, "latin1") == 0
- || STRCMP(p_enc, "iso-8859-15") == 0) {
+ {
#define A_grave 0xc0
#define A_acute 0xc1
#define A_circumflex 0xc2
@@ -1246,7 +1245,7 @@ static int nfa_regatom(void)
}
// When '.' is followed by a composing char ignore the dot, so that
// the composing char is matched here.
- if (enc_utf8 && c == Magic('.') && utf_iscomposing(peekchr())) {
+ if (c == Magic('.') && utf_iscomposing(peekchr())) {
old_regparse = regparse;
c = getchr();
goto nfa_do_multibyte;
@@ -1737,11 +1736,10 @@ collection:
EMIT(endc);
EMIT(NFA_RANGE);
EMIT(NFA_CONCAT);
- } else if (has_mbyte && ((*mb_char2len)(startc) > 1
- || (*mb_char2len)(endc) > 1)) {
- /* Emit the characters in the range.
- * "startc" was already emitted, so skip it.
- * */
+ } else if ((*mb_char2len)(startc) > 1
+ || (*mb_char2len)(endc) > 1) {
+ // Emit the characters in the range.
+ // "startc" was already emitted, so skip it.
for (c = startc + 1; c <= endc; c++) {
EMIT(c);
EMIT(NFA_CONCAT);
@@ -1819,9 +1817,8 @@ collection:
nfa_do_multibyte:
// plen is length of current char with composing chars
- if (enc_utf8 && ((*mb_char2len)(c)
- != (plen = utfc_ptr2len(old_regparse))
- || utf_iscomposing(c))) {
+ if ((*mb_char2len)(c) != (plen = utfc_ptr2len(old_regparse))
+ || utf_iscomposing(c)) {
int i = 0;
/* A base character plus composing characters, or just one
@@ -4995,7 +4992,7 @@ static long find_match_text(colnr_T startcol, int regstart, char_u *match_text)
}
if (match
// check that no composing char follows
- && !(enc_utf8 && utf_iscomposing(PTR2CHAR(s2)))) {
+ && !utf_iscomposing(PTR2CHAR(s2))) {
cleanup_subexpr();
if (REG_MULTI) {
rex.reg_startpos[0].lnum = rex.lnum;
@@ -5248,7 +5245,7 @@ static int nfa_regmatch(nfa_regprog_T *prog, nfa_state_T *start,
{
// If the match ends before a composing characters and
// rex.reg_icombine is not set, that is not really a match.
- if (enc_utf8 && !rex.reg_icombine && utf_iscomposing(curc)) {
+ if (!rex.reg_icombine && utf_iscomposing(curc)) {
break;
}
nfa_match = true;
@@ -5747,7 +5744,7 @@ static int nfa_regmatch(nfa_regprog_T *prog, nfa_state_T *start,
case NFA_ANY_COMPOSING:
// On a composing character skip over it. Otherwise do
// nothing. Always matches.
- if (enc_utf8 && utf_iscomposing(curc)) {
+ if (utf_iscomposing(curc)) {
add_off = clen;
} else {
add_here = true;
@@ -6019,7 +6016,7 @@ static int nfa_regmatch(nfa_regprog_T *prog, nfa_state_T *start,
// Bail out quickly when there can't be a match, avoid the overhead of
// win_linetabsize() on long lines.
- if (op != 1 && col > t->state->val * (has_mbyte ? MB_MAXBYTES : 1)) {
+ if (op != 1 && col > t->state->val * MB_MAXBYTES) {
break;
}
@@ -6132,7 +6129,7 @@ static int nfa_regmatch(nfa_regprog_T *prog, nfa_state_T *start,
// If rex.reg_icombine is not set only skip over the character
// itself. When it is set skip over composing characters.
- if (result && enc_utf8 && !rex.reg_icombine) {
+ if (result && !rex.reg_icombine) {
clen = utf_ptr2len(rex.input);
}
generated by cgit (git 2.34.1) at 2025-04-29 21:53:33 +0000