Bug 367827

Description Hyukmin Kwon 2012-01-04 05:59:57 EST

Build Identifier: Version: 3.7.1 Build id: M20110909-1335

Get tu (an IASTTranslationUnit) by parsing following code with CDT C parser.

#define NULL_STATEMENT_MACRO ;
void macro_test()
{
	int x = 1;

	NULL_STATEMENT_MACRO //comment1

	x; //comment2
}

The obtained position of 'comment1' by using tu.getComments() is incorrect.
The character at the position is not '/' but ' ' (space before '/').
For the 'comment2' the obtained position is correct.
Only comments after the NULL_STATEMENT_MACRO at the same line give incorrect positions.

Same phenomenon is occurred for "#define NULL_BLOCK_MACRO {}".

Reproducible: Always

Steps to Reproduce:
1.Parse following code with CDT C parser and get an IASTTranslationUnit 'tu'.
#define NULL_STATEMENT_MACRO ;
void macro_test()
{
	int x = 1;

	NULL_STATEMENT_MACRO //comment1

	x; //comment2
}

2.Get comment list from tu.getComments()
IASTComment[] cl = tu.getComments();

3.Get node offset of comments by folowing code
commentOffset[i] = cl[i].getFileLocation().getNodeOffset();

4.If we get the character at commentOffset[i] by using IDocument.getChar() method, then commentOffset[1] is '/' but commentOffset[0] is ' ' (space) (i.e. incorrect offset). The difference with correct offset is 1.