On Tue, 26 Aug 2025 17:40:12 GMT, Daniel Gredler <[email protected]> wrote:
> `TextLayout` should deal more gracefully with zero length strings. Currently
> the exception listed below is the one that is thrown.
>
> `new TextLayout("", f, new FontRenderContext(null, false, false));`
>
>> Exception in thread "main" java.lang.IllegalArgumentException: Zero length
>> string passed to TextLayout constructor.
>> at java.lang.Throwable.<init>(Compiled Code)
>> at java.lang.Exception.<init>(Compiled Code)
>> at java.lang.RuntimeException.<init>(Compiled Code)
>> at java.lang.IllegalArgumentException.<init>(Compiled Code)
>> at java.awt.font.TextLayout.<init>(Compiled Code)
>> at test.main(Compiled Code)
>
> **REVIEWER NOTE:** Please check the empty-string `TextLayout` behavior
> documented in `TextLayoutConstructorTest` carefully; a badly-behaving empty
> `TextLayout` is probably worse than a `TextLayout` which doesn't allow empty
> strings...
Several JCK tests for TextLayout constructor fail (surprise !)
They all fail because they use a null FRC.
We need a CSR for this anyway but I'll need to co-ordinate problem listing
those tests as part of this bug so we don't see failures as soon as it is
pushed.
src/java.desktop/share/classes/java/awt/font/TextLine.java line 839:
> 837:
> 838: if (result == null) {
> 839: result = new Rectangle2D.Float(Float.MAX_VALUE,
> Float.MAX_VALUE, Float.MIN_VALUE, Float.MIN_VALUE);
That's interesting. I'd have assumed what you changed it to would be the
answer in such. a case ? Why was it like this ? Perhaps it was never possible
before and this was intended to 'flag a problem' ??
src/java.desktop/share/classes/java/awt/font/TextLine.java line 934:
> 932:
> 933: if (!requiresBidi) {
> 934: requiresBidi = Bidi.requiresBidi(chars, 0, characterCount);
This doesn't change anything, does it ? So why is the change made ? This - and
white space changes - make it look like there are more changes than there
really are.
src/java.desktop/share/classes/sun/font/TextLabelFactory.java line 122:
> 120: int limit) {
> 121:
> 122: if (start > limit || start < lineStart || limit > lineLimit) {
Why is this change needed ? I'm supposing that for zero length start==limit==0
but we get here because TextLine calls it in a do { ... } block so will call it
at least once, and I wonder if that should change ? So why did you do it this
way ?
-------------
PR Review: https://git.openjdk.org/jdk/pull/26947#pullrequestreview-3166753411
PR Review Comment: https://git.openjdk.org/jdk/pull/26947#discussion_r2308701015
PR Review Comment: https://git.openjdk.org/jdk/pull/26947#discussion_r2308698669
PR Review Comment: https://git.openjdk.org/jdk/pull/26947#discussion_r2308697022
