[cfe-dev] 回复:Re: Public the interface for cxa_demangler.cpp

via cfe-dev cfe-dev at lists.llvm.org
Sun Apr 15 20:35:50 PDT 2018


Thank you Erik, I think, the new demangler might meet all the requirement I needed. I will try with it. BTW, does your code will sync with the cxa_demangler.cpp in the llvm runtime ? I notice that, the demangler in the llvm and llvm runtime look like the same for mostly,
----- 原始邮件 -----发件人:Erik Pilkington <erik.pilkington at gmail.com>
收件人:bluedream_zqs at sina.com, cfe-dev <cfe-dev at lists.llvm.org>
主题:Re: [cfe-dev] Public the interface for cxa_demangler.cpp
日期:2018年04月13日 20点54分


  
  
    No plans as of yet! I never planned to export the AST in this
      way, doing so would probably require some surgery (ie, make all
      the identifiers it uses reserved). I very recently landed r329951,
      which provides a simple opaque API to the AST for the copy of the
      demangler in llvm. Maybe this is enough for your use case? What
      kind of analysis are you planning on performing?
    Thanks,

    Erik

    

    On 2018-04-13 4:22 AM, via cfe-dev
      wrote:

    
     The
      file "llvm.stl/libcxxabi/src/cxa_demangler.cpp" implement the
      interface "__cxa_demangle()" defined in the C++ABI. It decoded the
      mangled string into the demangled one, that is, "string -->
      string".  Inside it, it is implemented by "string --> AST
      --> string". However, someone might need the internal AST
      instead of the demangled string to do the further analysis for the
      mangled string. Do we have any plan not to hide the AST inside the
      cxa_demangler.cpp, but public it into the cxa_demangle.h ? Thank
      you.  

      

      
      

      _______________________________________________
cfe-dev mailing list
cfe-dev at lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev

    
    

  


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-dev/attachments/20180416/83e369eb/attachment-0001.html>


More information about the cfe-dev mailing list