Community
Participate
Working Groups
Build Identifier: 201106081058 Parser is unable to correctly process and index following statements: static int i = #include "one.txt" ; where one.txt contains one char ('1' char) Syntax error is incorrectly reported, probably include isn't substituted before feeding source to parser leading to syntax error. Reproducible: Always
Generalization of bug 315964
Will this ever be supported? If so, is there some kind of estimate as to when?
(In reply to Nicklas Hansson from comment #2) > Will this ever be supported? If so, is there some kind of estimate as to > when? I plan to work on this at some point, following Markus' suggestion in https://bugs.eclipse.org/bugs/show_bug.cgi?id=315964#c7. It's hard to give a time estimate. There are many open bugs in CDT, and not that many people fixing them; I'm working through the indexer-related ones in my spare time. Hopefully sometime in 2015. If you'd like it fixed sooner, you're welcome to write a patch yourself. I can try to give you guidance (here in the bug, or on the mailing list) and review your patch.
(In reply to Nathan Ridge from comment #3) > (In reply to Nicklas Hansson from comment #2) > > Will this ever be supported? If so, is there some kind of estimate as to > > when? > > I plan to work on this at some point, following Markus' suggestion in > https://bugs.eclipse.org/bugs/show_bug.cgi?id=315964#c7. > > It's hard to give a time estimate. There are many open bugs in CDT, and not > that many people fixing them; I'm working through the indexer-related ones > in my spare time. Hopefully sometime in 2015. > > If you'd like it fixed sooner, you're welcome to write a patch yourself. I > can try to give you guidance (here in the bug, or on the mailing list) and > review your patch. I tried to get the environment going because I'd like to contribute to CDT. I ran into some issues with the test environment but it seems fine now. Anyway, is there some type of architectural description available somewhere or is it more deep diving into the code that's the strategy for getting started? I suppose this discussion would be more suitable for the mailing list, but with some basic pointers regarding architecture above would be good and any further questions can be taken through the mailing list (or forums?).
(In reply to Nicklas Hansson from comment #4) > I tried to get the environment going because I'd like to contribute to CDT. > I ran into some issues with the test environment but it seems fine now. > Anyway, is there some type of architectural description available somewhere > or is it more deep diving into the code that's the strategy for getting > started? > > I suppose this discussion would be more suitable for the mailing list, but > with some basic pointers regarding architecture above would be good and any > further questions can be taken through the mailing list (or forums?). Unfortunately I'm not aware of much in the way of design documentation. My suggestion for getting a feel for the architecture is to have a look at the patches for some bugs that are already fixed. I think it's particularly helpful to look at a patch that implements a new language feature, albeit a small one so it's not overwhelming. For example, you could have a look at my patch in bug 411196 which adds support for the '__underlying_type(T)' compiler builtin that's commonly used to implement std::underlying_type in standard libraries. You can see the various parts of the code that are touched: adding a new token, adding a new AST node, modifying the parser to create the new AST node, adding new semantic entities (in this case, a new kind of type). For this bug, most of the work involved will touch CPreprocessor and GNUCPPSourceParser, so it will probably help to familiarize yourself with these classes (obviously do not read all of GNUCPPSourceParser, just enough to get an idea for how it works).
(In reply to Nicklas Hansson from comment #4) > any further questions can be taken through the mailing list (or forums?). Please ask any further questions either on the mailing list or in this bug. I don't read the forums.
*** Bug 445183 has been marked as a duplicate of this bug. ***
