Merge pull request #45957 from lupoDharkael/natural-comp

[3.2] Add natural string comparison
This commit is contained in:
Rémi Verschelde 2021-02-22 11:07:54 +01:00 committed by GitHub
commit 0abf702d4b
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
2 changed files with 15 additions and 0 deletions

View file

@ -237,6 +237,7 @@ struct _VariantCall {
VCALL_LOCALMEM1R(String, casecmp_to);
VCALL_LOCALMEM1R(String, nocasecmp_to);
VCALL_LOCALMEM1R(String, naturalnocasecmp_to);
VCALL_LOCALMEM0R(String, length);
VCALL_LOCALMEM3R(String, count);
VCALL_LOCALMEM3R(String, countn);
@ -1558,6 +1559,7 @@ void register_variant_methods() {
/* STRING */
ADDFUNC1R(STRING, INT, String, casecmp_to, STRING, "to", varray());
ADDFUNC1R(STRING, INT, String, nocasecmp_to, STRING, "to", varray());
ADDFUNC1R(STRING, INT, String, naturalnocasecmp_to, STRING, "to", varray());
ADDFUNC0R(STRING, INT, String, length, varray());
ADDFUNC2R(STRING, STRING, String, substr, INT, "from", INT, "len", varray(-1));

View file

@ -621,6 +621,19 @@
Returns the MD5 hash of the string as a string.
</description>
</method>
<method name="naturalnocasecmp_to">
<return type="int">
</return>
<argument index="0" name="to" type="String">
</argument>
<description>
Performs a case-insensitive [i]natural order[/i] comparison to another string. Returns [code]-1[/code] if less than, [code]1[/code] if greater than, or [code]0[/code] if equal. "less than" or "greater than" are determined by the [url=https://en.wikipedia.org/wiki/List_of_Unicode_characters]Unicode code points[/code] of each string, which roughly matches the alphabetical order. Internally, lowercase characters will be converted to uppercase during the comparison.
When used for sorting, natural order comparison will order suites of numbers as expected by most people. If you sort the numbers from 1 to 10 using natural order, you will get [code][1, 2, 3, ...][/code] instead of [code][1, 10, 2, 3, ...][/code].
[b]Behavior with different string lengths:[/b] Returns [code]1[/code] if the "base" string is longer than the [code]to[/code] string or [code]-1[/code] if the "base" string is shorter than the [code]to[/code] string. Keep in mind this length is determined by the number of Unicode codepoints, [i]not[/i] the actual visible characters.
[b]Behavior with empty strings:[/b] Returns [code]-1[/code] if the "base" string is empty, [code]1[/code] if the [code]to[/code] string is empty or [code]0[/code] if both strings are empty.
To get a boolean result from a string comparison, use the [code]==[/code] operator instead. See also [method nocasecmp_to] and [method casecmp_to].
</description>
</method>
<method name="nocasecmp_to">
<return type="int">
</return>