vim-patch:8.2.2607: strcharpart() cannot include composing characters

Problem:    strcharpart() cannot include composing characters.
Solution:   Add the {skipcc} argument.

02b4d9b18a

Co-authored-by: Bram Moolenaar <Bram@vim.org>
This commit is contained in:
zeertzjq
2023-05-04 17:42:34 +08:00
parent b441dafdf5
commit 3724e65c30
4 changed files with 41 additions and 8 deletions

View File

@@ -486,7 +486,7 @@ str2list({expr} [, {utf8}]) List convert each character of {expr} to
str2nr({expr} [, {base} [, {quoted}]]) str2nr({expr} [, {base} [, {quoted}]])
Number convert String to Number Number convert String to Number
strcharlen({expr}) Number character length of the String {expr} strcharlen({expr}) Number character length of the String {expr}
strcharpart({str}, {start} [, {len}]) strcharpart({str}, {start} [, {len} [, {skipcc}]])
String {len} characters of {str} at String {len} characters of {str} at
character {start} character {start}
strchars({expr} [, {skipcc}]) Number character count of the String {expr} strchars({expr} [, {skipcc}]) Number character count of the String {expr}
@@ -7779,6 +7779,7 @@ slice({expr}, {start} [, {end}]) *slice()*
Similar to using a |slice| "expr[start : end]", but "end" is Similar to using a |slice| "expr[start : end]", but "end" is
used exclusive. And for a string the indexes are used as used exclusive. And for a string the indexes are used as
character indexes instead of byte indexes. character indexes instead of byte indexes.
Also, composing characters are not counted.
When {end} is omitted the slice continues to the last item. When {end} is omitted the slice continues to the last item.
When {end} is -1 the last item is omitted. When {end} is -1 the last item is omitted.
Returns an empty value if {start} or {end} are invalid. Returns an empty value if {start} or {end} are invalid.
@@ -8129,12 +8130,16 @@ strcharlen({string}) *strcharlen()*
GetText()->strcharlen() GetText()->strcharlen()
strcharpart({src}, {start} [, {len}]) *strcharpart()* strcharpart({src}, {start} [, {len} [, {skipcc}]]) *strcharpart()*
Like |strpart()| but using character index and length instead Like |strpart()| but using character index and length instead
of byte index and length. Composing characters are counted of byte index and length.
separately. When {skipcc} is omitted or zero, composing characters are
counted separately.
When {skipcc} set to 1, Composing characters are ignored,
similar to |slice()|.
When a character index is used where a character does not When a character index is used where a character does not
exist it is assumed to be one character. For example: > exist it is omitted and counted as one character. For
example: >
strcharpart('abc', -1, 2) strcharpart('abc', -1, 2)
< results in 'a'. < results in 'a'.

View File

@@ -386,7 +386,7 @@ return {
str2list={args={1, 2}, base=1}, str2list={args={1, 2}, base=1},
str2nr={args={1, 3}, base=1}, str2nr={args={1, 3}, base=1},
strcharlen={args=1, base=1}, strcharlen={args=1, base=1},
strcharpart={args={2, 3}, base=1, fast=true}, strcharpart={args={2, 4}, base=1, fast=true},
strchars={args={1, 2}, base=1}, strchars={args={1, 2}, base=1},
strdisplaywidth={args={1, 2}, base=1}, strdisplaywidth={args={1, 2}, base=1},
strftime={args={1, 2}, base=1}, strftime={args={1, 2}, base=1},

View File

@@ -1834,12 +1834,26 @@ void f_strcharpart(typval_T *argvars, typval_T *rettv, EvalFuncData fptr)
const size_t slen = strlen(p); const size_t slen = strlen(p);
int nbyte = 0; int nbyte = 0;
varnumber_T skipcc = false;
bool error = false; bool error = false;
varnumber_T nchar = tv_get_number_chk(&argvars[1], &error); varnumber_T nchar = tv_get_number_chk(&argvars[1], &error);
if (!error) { if (!error) {
if (argvars[2].v_type != VAR_UNKNOWN
&& argvars[3].v_type != VAR_UNKNOWN) {
skipcc = tv_get_bool(&argvars[3]);
if (skipcc < 0 || skipcc > 1) {
semsg(_(e_using_number_as_bool_nr), skipcc);
return;
}
}
if (nchar > 0) { if (nchar > 0) {
while (nchar > 0 && (size_t)nbyte < slen) { while (nchar > 0 && (size_t)nbyte < slen) {
if (skipcc) {
nbyte += utfc_ptr2len(p + nbyte);
} else {
nbyte += utf_ptr2len(p + nbyte); nbyte += utf_ptr2len(p + nbyte);
}
nchar--; nchar--;
} }
} else { } else {
@@ -1854,9 +1868,13 @@ void f_strcharpart(typval_T *argvars, typval_T *rettv, EvalFuncData fptr)
if (off < 0) { if (off < 0) {
len += 1; len += 1;
} else {
if (skipcc) {
len += utfc_ptr2len(p + off);
} else { } else {
len += utf_ptr2len(p + off); len += utf_ptr2len(p + off);
} }
}
charlen--; charlen--;
} }
} else { } else {

View File

@@ -31,4 +31,14 @@ func Test_strcharpart()
call assert_equal('a', strcharpart('àxb', 0, 1)) call assert_equal('a', strcharpart('àxb', 0, 1))
call assert_equal('̀', strcharpart('àxb', 1, 1)) call assert_equal('̀', strcharpart('àxb', 1, 1))
call assert_equal('x', strcharpart('àxb', 2, 1)) call assert_equal('x', strcharpart('àxb', 2, 1))
call assert_equal('a', strcharpart('àxb', 0, 1, 0))
call assert_equal('à', strcharpart('àxb', 0, 1, 1))
call assert_equal('x', strcharpart('àxb', 1, 1, 1))
call assert_fails("let v = strcharpart('abc', 0, 0, [])", 'E745:')
call assert_fails("let v = strcharpart('abc', 0, 0, 2)", 'E1023:')
endfunc endfunc
" vim: shiftwidth=2 sts=2 expandtab