Bugzilla – Full Text Bug Listing
| Summary: | Javadoc inside DOM AST does not support starting characters in unicode | ||
|---|---|---|---|
| Product: | [Eclipse Project] JDT | Reporter: | David Audel <david_audel> |
| Component: | Core | Assignee: | Frederic Fusier <frederic_fusier> |
| Status: | VERIFIED FIXED | QA Contact: | |
| Severity: | normal | ||
| Priority: | P3 | ||
| Version: | 3.0 | ||
| Target Milestone: | 3.0 M7 | ||
| Hardware: | PC | ||
| OS: | Windows 2000 | ||
| Whiteboard: | |||
In fact this is only when unicode is on one of the three starting character (e.g. '/', first '*' or second '*'). Otherwise, unicode is correctly handled inside javadoc comment... Fixed. Now javadoc comments which have unicode in starting (or ending) characters) are correctly parsed and DOM AST nodes hierarchy matches content. [jdt-dev internal] Changes applied in parse(int,int) method of AbstractCommentParser. In initial implementation, we started parse at javadoc start position + 3. It was enough before as we didn't store text. But because now we store all characters, it's really important to start _precisely_ after the starting characters '/**'. Use readChar() three times to be sure of parse starting point... Test cases added in jdt.core.dom.tests.ASTConverterJavadocTest Verified for 3.0M7 |
build 20040229 ------------------- /*\u002A * Test */ public class X { } ------------------- \u002A is * character The DOM Javadoc of this test case seems not correct Javadoc +TagElement +TextElement '02A' +TextElement 'Test'