In order to better accommodate an expectation of more general state for prior line break context text and to facilitate use of prior context in non-ASCII line break path, change to a fixed UChar array for storage of prior context and change accessors/methods to use better names.
This change is to simplify (reduce the size of) the patch proposed for bug 105692 in order to ease the review and landing process.
No behavioral change is introduced, so no new tests are required.
Note that it is expected that this prior context state may be further generalized in the future in order to accommodate greater than the last two context characters. This is expected to be required to support Thai (and similar) dictionary based line breaking, where the prior context may consist of an entire word.
Created attachment 196227 [details]
Comment on attachment 196227 [details]
Clearing flags on attachment: 196227
Committed r147506: <http://trac.webkit.org/changeset/147506>
All reviewed patches have been landed. Closing bug.