aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
Diffstat
-rw-r--r--runtime/doc/pattern.txt3
-rw-r--r--src/nvim/regexp.c34
-rw-r--r--src/nvim/regexp_nfa.c32
-rw-r--r--src/nvim/testdir/test_regexp_utf8.vim55
4 files changed, 123 insertions, 1 deletions
diff --git a/runtime/doc/pattern.txt b/runtime/doc/pattern.txt
index adfab07758..7129c6cd58 100644
--- a/runtime/doc/pattern.txt
+++ b/runtime/doc/pattern.txt
@@ -1111,6 +1111,9 @@ x A single character, with no special meaning, matches itself
*[:tab:]* [:tab:] the <Tab> character
*[:escape:]* [:escape:] the <Esc> character
*[:backspace:]* [:backspace:] the <BS> character
+*[:ident:]* [:ident:] identifier character (same as "\i")
+*[:keyword:]* [:keyword:] keyword character (same as "\k")
+*[:fname:]* [:fname:] file name character (same as "\f")
The brackets in character class expressions are additional to the
brackets delimiting a collection. For example, the following is a
plausible pattern for a Unix filename: "[-./[:alnum:]_~]\+" That is,
diff --git a/src/nvim/regexp.c b/src/nvim/regexp.c
index b4e2c7d766..9705896e9b 100644
--- a/src/nvim/regexp.c
+++ b/src/nvim/regexp.c
@@ -602,6 +602,12 @@ static int get_char_class(char_u **pp)
#define CLASS_BACKSPACE 14
"escape:]",
#define CLASS_ESCAPE 15
+ "ident:]",
+#define CLASS_IDENT 16
+ "keyword:]",
+#define CLASS_KEYWORD 17
+ "fname:]",
+#define CLASS_FNAME 18
};
#define CLASS_NONE 99
int i;
@@ -2417,6 +2423,27 @@ collection:
case CLASS_ESCAPE:
regc(ESC);
break;
+ case CLASS_IDENT:
+ for (cu = 1; cu <= 255; cu++) {
+ if (vim_isIDc(cu)) {
+ regmbc(cu);
+ }
+ }
+ break;
+ case CLASS_KEYWORD:
+ for (cu = 1; cu <= 255; cu++) {
+ if (reg_iswordc(cu)) {
+ regmbc(cu);
+ }
+ }
+ break;
+ case CLASS_FNAME:
+ for (cu = 1; cu <= 255; cu++) {
+ if (vim_isfilec(cu)) {
+ regmbc(cu);
+ }
+ }
+ break;
}
} else {
// produce a multibyte character, including any
@@ -3276,6 +3303,13 @@ void free_regexp_stuff(void)
#endif
+// Return true if character 'c' is included in 'iskeyword' option for
+// "reg_buf" buffer.
+static bool reg_iswordc(int c)
+{
+ return vim_iswordc_buf(c, rex.reg_buf);
+}
+
/*
* Get pointer to the line "lnum", which is relative to "reg_firstlnum".
*/
diff --git a/src/nvim/regexp_nfa.c b/src/nvim/regexp_nfa.c
index a744071a6a..506c4e87db 100644
--- a/src/nvim/regexp_nfa.c
+++ b/src/nvim/regexp_nfa.c
@@ -230,7 +230,10 @@ enum {
NFA_CLASS_TAB,
NFA_CLASS_RETURN,
NFA_CLASS_BACKSPACE,
- NFA_CLASS_ESCAPE
+ NFA_CLASS_ESCAPE,
+ NFA_CLASS_IDENT,
+ NFA_CLASS_KEYWORD,
+ NFA_CLASS_FNAME,
};
/* Keep in sync with classchars. */
@@ -1643,6 +1646,15 @@ collection:
case CLASS_ESCAPE:
EMIT(NFA_CLASS_ESCAPE);
break;
+ case CLASS_IDENT:
+ EMIT(NFA_CLASS_IDENT);
+ break;
+ case CLASS_KEYWORD:
+ EMIT(NFA_CLASS_KEYWORD);
+ break;
+ case CLASS_FNAME:
+ EMIT(NFA_CLASS_FNAME);
+ break;
}
EMIT(NFA_CONCAT);
continue;
@@ -2407,6 +2419,9 @@ static void nfa_set_code(int c)
case NFA_CLASS_RETURN: STRCPY(code, "NFA_CLASS_RETURN"); break;
case NFA_CLASS_BACKSPACE: STRCPY(code, "NFA_CLASS_BACKSPACE"); break;
case NFA_CLASS_ESCAPE: STRCPY(code, "NFA_CLASS_ESCAPE"); break;
+ case NFA_CLASS_IDENT: STRCPY(code, "NFA_CLASS_IDENT"); break;
+ case NFA_CLASS_KEYWORD: STRCPY(code, "NFA_CLASS_KEYWORD"); break;
+ case NFA_CLASS_FNAME: STRCPY(code, "NFA_CLASS_FNAME"); break;
case NFA_ANY: STRCPY(code, "NFA_ANY"); break;
case NFA_IDENT: STRCPY(code, "NFA_IDENT"); break;
@@ -4474,6 +4489,21 @@ static int check_char_class(int class, int c)
return OK;
}
break;
+ case NFA_CLASS_IDENT:
+ if (vim_isIDc(c)) {
+ return OK;
+ }
+ break;
+ case NFA_CLASS_KEYWORD:
+ if (reg_iswordc(c)) {
+ return OK;
+ }
+ break;
+ case NFA_CLASS_FNAME:
+ if (vim_isfilec(c)) {
+ return OK;
+ }
+ break;
default:
// should not be here :P
diff --git a/src/nvim/testdir/test_regexp_utf8.vim b/src/nvim/testdir/test_regexp_utf8.vim
index 209de2c35e..4466ad436a 100644
--- a/src/nvim/testdir/test_regexp_utf8.vim
+++ b/src/nvim/testdir/test_regexp_utf8.vim
@@ -54,6 +54,12 @@ func s:classes_test()
let tabchar = ''
let upperchars = ''
let xdigitchars = ''
+ let identchars = ''
+ let identchars1 = ''
+ let kwordchars = ''
+ let kwordchars1 = ''
+ let fnamechars = ''
+ let fnamechars1 = ''
let i = 1
while i <= 255
let c = nr2char(i)
@@ -105,6 +111,24 @@ func s:classes_test()
if c =~ '[[:xdigit:]]'
let xdigitchars .= c
endif
+ if c =~ '[[:ident:]]'
+ let identchars .= c
+ endif
+ if c =~ '\i'
+ let identchars1 .= c
+ endif
+ if c =~ '[[:keyword:]]'
+ let kwordchars .= c
+ endif
+ if c =~ '\k'
+ let kwordchars1 .= c
+ endif
+ if c =~ '[[:fname:]]'
+ let fnamechars .= c
+ endif
+ if c =~ '\f'
+ let fnamechars1 .= c
+ endif
let i += 1
endwhile
@@ -124,6 +148,37 @@ func s:classes_test()
call assert_equal("\t\n\x0b\f\r ", spacechars)
call assert_equal("\t", tabchar)
call assert_equal('0123456789ABCDEFabcdef', xdigitchars)
+
+ if has('win32')
+ let identchars_ok = '0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz€‚ƒ„…†‡ˆ‰Š‹ŒŽ‘’“”•–—˜™š›œžŸ ¡¢£¤¥¦§µÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõöøùúûüýþÿ'
+ let kwordchars_ok = '0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyzµÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ'
+ elseif has('ebcdic')
+ let identchars_ok = 'ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz€ŒŽœž¬®µº¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ'
+ let kwordchars_ok = 'ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz€ŒŽœž¬®µº¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ'
+ else
+ let identchars_ok = '0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyzµÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ'
+ let kwordchars_ok = '0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyzµÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ'
+ endif
+
+ if has('win32')
+ let fnamechars_ok = '!#$%+,-./0123456789:=@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]_abcdefghijklmnopqrstuvwxyz{}~ ¡¢£¤¥¦§¨©ª«¬­®¯°±²³´µ¶·¸¹º»¼½¾¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ'
+ elseif has('amiga')
+ let fnamechars_ok = '$+,-./0123456789:ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz~ ¡¢£¤¥¦§¨©ª«¬­®¯°±²³´µ¶·¸¹º»¼½¾¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ'
+ elseif has('vms')
+ let fnamechars_ok = '#$%+,-./0123456789:;<>ABCDEFGHIJKLMNOPQRSTUVWXYZ[]_abcdefghijklmnopqrstuvwxyz~ ¡¢£¤¥¦§¨©ª«¬­®¯°±²³´µ¶·¸¹º»¼½¾¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ'
+ elseif has('ebcdic')
+ let fnamechars_ok = '#$%+,-./=ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz~ ¡¢£¤¥¦§¨©ª«¬­®¯°±²³´µ¶·¸¹º»¼½¾¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ'
+ else
+ let fnamechars_ok = '#$%+,-./0123456789=ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz~ ¡¢£¤¥¦§¨©ª«¬­®¯°±²³´µ¶·¸¹º»¼½¾¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ'
+ endif
+
+ call assert_equal(identchars_ok, identchars)
+ call assert_equal(kwordchars_ok, kwordchars)
+ call assert_equal(fnamechars_ok, fnamechars)
+
+ call assert_equal(identchars1, identchars)
+ call assert_equal(kwordchars1, kwordchars)
+ call assert_equal(fnamechars1, fnamechars)
endfunc
func Test_classes_re1()
generated by cgit (git 2.34.1) at 2025-05-08 12:50:32 +0000