Suspicious duplicate condition in java.util.regex.Grapheme#isExcludedSpacingMark
Naoto Sato
naoto.sato at oracle.com
Tue Sep 7 16:15:01 UTC 2021
It does look incorrect. I will take a look.
Naoto
On 9/6/21 12:16 AM, Andrey Turbanov wrote:
> Hello.
> I found suspicious condition in the method
> java.util.regex.Grapheme#isExcludedSpacingMark
> It's detected by IntelliJ IDEA inspection 'Condition is covered by
> further condition'
> https://github.com/openjdk/jdk/blob/master/src/java.base/share/classes/java/util/regex/Grapheme.java#L157
>
> ```
> private static boolean isExcludedSpacingMark(int cp) {
> return cp == 0x102B || cp == 0x102C || cp == 0x1038 ||
> cp >= 0x1062 && cp <= 0x1064 ||
> cp >= 0x1062 && cp <= 0x106D || // <== here is the warning
> cp == 0x1083 ||
> cp >= 0x1087 && cp <= 0x108C ||
> cp == 0x108F ||
> cp >= 0x109A && cp <= 0x109C ||
> cp == 0x1A61 || cp == 0x1A63 || cp == 0x1A64 ||
> cp == 0xAA7B || cp == 0xAA7D;
> }
> ```
> There are 2 sub-conditions in this complex condition:
> cp >= 0x1062 && cp <= 0x1064 ||
> cp >= 0x1062 && cp <= 0x106D ||
>
> The second condition is _wider_ than the first one.
> I believe it's a bug. The second condition (according to
> https://www.compart.com/en/unicode/category/Mc) should look like this:
>
> cp >= 0x1067 && cp <= 0x106D ||
>
> 0x1065, 0x1066 are not from the Spacing_Mark category.
>
>
> Andrey Turbanov
>
More information about the core-libs-dev
mailing list