Sorry I guess I lost you on this one. Where can I change this variable?<div><br></div><div>I am currently setting this on my ~/.profile</div><div><br></div><div><div>PYTHONPATH="/Users/carlosandrade/Desktop/llvm/tools/clang/bindings/python:$PYTHONPATH"</div>

<div>export PYTHONPATH</div><div><br></div><div>I only found on read me file after running grep this:</div><div><br></div><div><div>$ env PYTHONPATH=$(echo ~/llvm/tools/clang/bindings/python/) \</div><div>      LD_LIBRARY_PATH=$(llvm-config --libdir) \</div>

<div>  nosetests -v</div></div><div><br></div><div><br></div><div>But I am a little clueless on how to go about this!</div><div><br></div><div>Thanks,</div><div><br></div><br clear="all">Carlos Andrade<br><div><a href="http://carlosandrade.co" target="_blank">http://carlosandrade.co</a></div>

<br>
<br><br><div class="gmail_quote">2012/7/12 Gregory Szorc <span dir="ltr"><<a href="mailto:gregory.szorc@gmail.com" target="_blank">gregory.szorc@gmail.com</a>></span><br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">


  
    
  
  <div bgcolor="#FFFFFF" text="#000000">
    <div>I just ran the most recent Python
      bindings on my OS X build and it works fine.<br>
      <br>
      I suspect an older libclang is being loaded by the Python
      bindings. Try:<br>
      <br>
      LD_LIBRARY_PATH=/path/to/llvm/lib python<br>
      <br>
      You may also want to debug the library loading code and ensure the
      proper path is being picked up. You can always through a print
      statement in there or something.<span class="HOEnZb"><font color="#888888"><br>
      <br>
      Gregory</font></span><div><div class="h5"><br>
      <br>
      On 7/12/12 5:25 PM, Carlos Andrade wrote:<br>
    </div></div></div><div><div class="h5">
    <blockquote type="cite">Hi,
      <div><br>
      </div>
      <div>I was using the old version of lib clang python interface and
        have been following up on this. I checked llvm, clang and the
        compiler as recommended on the get started of the clang webpage
        (<a href="http://clang.llvm.org/get_started.html" target="_blank">http://clang.llvm.org/get_started.html</a>).
        After changing the python path to the new build I am now getting
        an error while trying to run the code. </div>
      <div><br>
      </div>
      <div>
        <div>Traceback (most recent call last):</div>
        <div>  File "aura.py", line 7, in <module></div>
        <div>    import clang.cindex</div>
        <div>  File
          "/Users/carlosandrade/Desktop/llvm/tools/clang/bindings/python/clang/cindex.py",
          line 2816, in <module></div>
        <div>    register_functions(lib)</div>
        <div>  File
          "/Users/carlosandrade/Desktop/llvm/tools/clang/bindings/python/clang/cindex.py",
          line 2407, in register_functions</div>
        <div>    lib.clang_CompilationDatabase_dispose.argtypes =
          [c_object_p]</div>
        <div>  File
          "/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/ctypes/__init__.py",
          line 378, in __getattr__</div>
        <div>    func = self.__getitem__(name)</div>
        <div>  File
          "/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/ctypes/__init__.py",
          line 383, in __getitem__</div>
        <div>    func = self._FuncPtr((name_or_ordinal, self))</div>
        <div>AttributeError: dlsym(0x1003b21b0,
          clang_CompilationDatabase_dispose): symbol not found</div>
      </div>
      <div><br>
      </div>
      <div>
        <div><br>
        </div>
        <div>Does anyone know what is that?</div>
        <div><br>
        </div>
        <div>Thanks,</div>
        <div><br>
        </div>
        <br clear="all">
        Carlos Andrade<br>
        <div><a href="http://carlosandrade.co" target="_blank">http://carlosandrade.co</a></div>
        <br>
        <br>
        <br>
        <div class="gmail_quote">2012/7/11 Manuel Klimek <span dir="ltr"><<a href="mailto:klimek@google.com" target="_blank">klimek@google.com</a>></span><br>
          <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
            <div>
              <div>On Tue, Jul 10, 2012 at 3:27 PM, Manuel Klimek <<a href="mailto:klimek@google.com" target="_blank">klimek@google.com</a>> wrote:<br>
                > On Mon, Jul 9, 2012 at 8:13 AM, Gregory Szorc <<a href="mailto:gregory.szorc@gmail.com" target="_blank">gregory.szorc@gmail.com</a>>
                wrote:<br>
                >> On 7/1/12 11:48 PM, Manuel Klimek wrote:<br>
                >><br>
                >> +        lib.clang_tokenize(tu, extent,
                byref(tokens_memory),<br>
                >> +                byref(tokens_count))<br>
                >><br>
                >> Strange indent. Hm, after reading further I see
                that that kind of indent is<br>
                >> quite common. Is there a guideline for how to
                indent this?<br>
                >><br>
                >> It appears my editor is indenting 8 spaces when
                wrapping lines. I'm pretty<br>
                >> sure that's provided by a popular "macro"
                settings file to handle Python.<br>
                >> I'll change it to 4 spaces, as that seems more
                conventional.<br>
                >><br>
                >><br>
                >><br>
                >> --- /dev/null<br>
                >> +++ b/bindings/python/clang/enumerations.py<br>
                >><br>
                >> +TokenKinds = [<br>
                >> +    ('PUNCTUATION', 0),<br>
                >> +    ('KEYWORD', 1),<br>
                >> +    ('IDENTIFIER', 2),<br>
                >> +    ('LITERAL', 3),<br>
                >> +    ('COMMENT', 4),<br>
                >> +]<br>
                >><br>
                >> This doesn't seem to match the comment. Those
                "TokenKinds" are different<br>
                >> from what I see in clang/Basic/TokenKinds.def.
                Am I missing something?<br>
                >><br>
                >> What doesn't match the comment?<br>
                >><br>
                >> Also, these come from<br>
                >> <a href="http://clang.llvm.org/doxygen/group__CINDEX__LEX.html#gaf63e37eee4280e2c039829af24bbc201" target="_blank">http://clang.llvm.org/doxygen/group__CINDEX__LEX.html#gaf63e37eee4280e2c039829af24bbc201</a>,<br>


                >> not the low-level tokens from TokenKinds.def.<br>
                ><br>
                > Oh, wow. Can you add a comment on where these
                TokenKinds come from.<br>
                > With the overloaded meaning it's really confusing.<br>
                <br>
              </div>
            </div>
            And apart from that lgtm.<br>
            <div>
              <div><br>
                ><br>
                > Thanks!<br>
                > /Manuel<br>
                ><br>
                >><br>
                >><br>
                >> On Sat, Jun 30, 2012 at 4:16 AM, Gregory Szorc
                <<a href="mailto:gregory.szorc@gmail.com" target="_blank">gregory.szorc@gmail.com</a>><br>
                >> wrote:<br>
                >>><br>
                >>> Updated patch attached.<br>
                >>><br>
                >>> * Added Cursor.get_tokens()<br>
                >>> * Refactored TokenGroup out of
                TranslationUnit to support above<br>
                >>> * Refactored with TranslationUnit.get_* API
                changes from last patch.<br>
                >>><br>
                >>> On Fri, Jun 29, 2012 at 12:18 AM, Gregory
                Szorc <<a href="mailto:gregory.szorc@gmail.com" target="_blank">gregory.szorc@gmail.com</a>><br>
                >>> wrote:<br>
                >>> > This exposes all of libclang's token
                functions in the Python bindings.<br>
                >>> ><br>
                >>> > I'm putting the TokenKind enumerations
                in a new module,<br>
                >>> > clang.enumerations. I plan to
                eventually move all existing<br>
                >>> > enumerations there so they are all
                consolidated. And, maybe one day,<br>
                >>> > we can even generate that file
                automatically by parsing the libclang<br>
                >>> > header files. I know Anders has code
                around somewhere that does<br>
                >>> > this...<br>
                >>> ><br>
                >>> > ---<br>
                >>> >  bindings/python/clang/cindex.py      
                             | 177<br>
                >>> > +++++++++++++++++++--<br>
                >>> >  bindings/python/clang/enumerations.py
                             |  32 ++++<br>
                >>> >
                 bindings/python/tests/cindex/test_token_kind.py    |
                 43 +++++<br>
                >>> >
                 bindings/python/tests/cindex/test_tokens.py        |
                 52 ++++++<br>
                >>> >
                 .../python/tests/cindex/test_translation_unit.py   |
                 24 ++-<br>
                >>> >  5 files changed, 312 insertions(+),
                16 deletions(-)<br>
                >>> >  create mode 100644
                bindings/python/clang/enumerations.py<br>
                >>> >  create mode 100644
                bindings/python/tests/cindex/test_token_kind.py<br>
                >>> >  create mode 100644
                bindings/python/tests/cindex/test_tokens.py<br>
                >>><br>
                >>>
                _______________________________________________<br>
                >>> cfe-commits mailing list<br>
                >>> <a href="mailto:cfe-commits@cs.uiuc.edu" target="_blank">cfe-commits@cs.uiuc.edu</a><br>
                >>> <a href="http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits" target="_blank">http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits</a><br>
                >>><br>
                >><br>
                >><br>
              </div>
            </div>
          </blockquote>
        </div>
        <br>
      </div>
    </blockquote>
    <br>
  </div></div></div>

</blockquote></div><br></div>