re2c-general Mailing List for re2c scanner generator

Has anyone been able to successfully build and run this example lately?

http://fasterparser.blogspot.com/2010/11/re2c-lexer-lemon-parser-calculator.html

I took a shot at it, but I get the following when I try to run it:

Starting program: /home/user/calc-example/build/calc
warning: no loadable sections found in added symbol-file
system-supplied DSO at 0x2aaaaaaab000

Program received signal SIGSEGV, Segmentation fault.
0x0000000000401577 in scan (s=0x603660, token=0x603680) at
/home/user/calc-example/build/scanner.c:34
34                              yych = *YYCURSOR;
(gdb) bt
#0  0x0000000000401577 in scan (s=0x603660, token=0x603680) at
/home/user/calc-example/build/scanner.c:34
#1  0x0000000000401506 in parse_math_expr (s=0x0) at parser.c:966
#2  0x00000000004017a7 in main (argc=1, argv=0x7fffffffe578) at
/home/user/calc-example/main.c:12

On Tue, Jun 21, 2011 at 10:27 AM, Rui Maciel <rui...@gm...> wrote:
>
> If that "official"build means that re2c abandons autotools and ceases to
> be an autotools project then what you perceive as being a convenience
> for yourself constitutes a serious inconvenience for others such as
> myself.  Therefore, with this in mind, I don't have any interest in
> seeing re2c migrate to cmake, and therefore I'm opposed to it.

It doesn't have to be an "either/or" situation - the CMake files can
co-exist with the autotools files.  The main reason I thought it might
be of interest is the multiple sln/vcproj files currently in
Subversion trunk - CMake generates sln/vcproj files for a given
version of Microsoft Visual Studio.  I can maintain the CMake files
separately - just thought it might be of interest to a wider audience,
given one of re2c's main advantages seems to be it's cross-platform
nature.

Cheers,
CY

On 06/21/2011 02:22 PM, Clifford Yapp wrote:
> Hi!  I've been looking around for cross platform, C based/C generating
> alternatives to flex/bison.  I have been seeing some references to the
> combination of re2c/lemon as a viable alternative.
>
> I was wondering - would there be interest in a CMake based build of
> re2c?  My primary project makes use of CMake to ensure one unified
> cross platform build logic, and if re2c were to prove viable I would
> need to write a CMake build anyway (we have to ensure the tool is
> available if no system installed version can be found, which means we
> need to build it).  Would there be any interest in making such a build
> "official"?  It really is helpful to have one build system cover a
> wide variety of build tools.
If that "official"build means that re2c abandons autotools and ceases to 
be an autotools project then what you perceive as being a convenience 
for yourself constitutes a serious inconvenience for others such as 
myself.  Therefore, with this in mind, I don't have any interest in 
seeing re2c migrate to cmake, and therefore I'm opposed to it.


Rui Maciel

Hi!  I've been looking around for cross platform, C based/C generating
alternatives to flex/bison.  I have been seeing some references to the
combination of re2c/lemon as a viable alternative.

I was wondering - would there be interest in a CMake based build of
re2c?  My primary project makes use of CMake to ensure one unified
cross platform build logic, and if re2c were to prove viable I would
need to write a CMake build anyway (we have to ensure the tool is
available if no system installed version can be found, which means we
need to build it).  Would there be any interest in making such a build
"official"?  It really is helpful to have one build system cover a
wide variety of build tools.

Also, can anyone recommend a good introductory document for
re2c/lemon?  I've found a couple examples but a really good
step-by-step walkthrough would be awesome.

Cheers,
CY