Bug 177607 – Extract logic to compute text to render into common function

• Component: Layout and Rendering
• Status: RESOLVED FIXED

[Daniel Bates]

Extract logic to compute text to render into common function.

[Daniel Bates]

Created attachment 322091 [details]
Patch

[Daniel Bates]

Comment on attachment 322091 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=322091&action=review

> Source/WebCore/rendering/InlineTextBox.cpp:1100
> +auto InlineTextBox::text(bool ignoreCombinedText, bool ignoreHyphen) const -> StringAndLengthIgnoringTruncation

Maybe it would be preferred to use enum instead of booleans? Ideally, with the exception of the offset-to-position and position-to-offset functions all of the code would always respect combined text and hyphens (i.e. call text(), which is equal to calling text(false, false)).

[Daniel Bates]

(In reply to Daniel Bates from comment #2)
> Comment on attachment 322091 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=322091&action=review
> 
> > Source/WebCore/rendering/InlineTextBox.cpp:1100
> > +auto InlineTextBox::text(bool ignoreCombinedText, bool ignoreHyphen) const -> StringAndLengthIgnoringTruncation
> 
> Maybe it would be preferred to use enum...

*enums (though we could define one enum with two enumerators and have this function take an OptionSet if that is preferred).

[Daniel Bates]

As far as I can tell if a text box is combined text or has a hyphen then it cannot be truncated. I make use of this assumption in the implementation of InlineTextBox::text() in attachment 322091 [details] to always return the substring from the start offset associated wit the text box (InlineTextBox::m_start) to the end of the text box (InlineTextBox::m_len) when ignoring combined text or ignoring the hyphen. Let me know if this assumption is wrong.

[alan]

Comment on attachment 322091 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=322091&action=review

> Source/WebCore/rendering/InlineTextBox.cpp:725
> +    auto text = this->text(ignoreCombinedText, ignoreHyphen);
> +    TextRun textRun = createTextRun(text);

At this point StringAndLengthIgnoringTruncation is only used for creating a run. Is this just an intermediate step/change?

[Daniel Bates]

(In reply to zalan from comment #5)
> Comment on attachment 322091 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=322091&action=review
> 
> > Source/WebCore/rendering/InlineTextBox.cpp:725
> > +    auto text = this->text(ignoreCombinedText, ignoreHyphen);
> > +    TextRun textRun = createTextRun(text);
> 
> At this point StringAndLengthIgnoringTruncation is only used for creating a
> run. Is this just an intermediate step/change?

Actually, we do not need StringAndLengthIgnoringTruncation because we do not need to track the length of the text ignoring truncation. See bug #177620.

[Daniel Bates]

Created attachment 322135 [details]
Patch

[Daniel Bates]

Comment on attachment 322135 [details]
Patch

Clearing flags on attachment: 322135

Committed r222670: <http://trac.webkit.org/changeset/222670>

[Daniel Bates]

All reviewed patches have been landed.  Closing bug.

[Radar WebKit Bug Importer]

<rdar://problem/34750483>