On 10/31/2012 12:03 PM, Alek Paunov wrote:
On 31.10.2012 20:46, Klaus Rudolph wrote:
...
File "/home/zfrdh/work/gcc-python-plugin/configbuilder.py", line 126
with open(self.filename, 'w') as f:
Unfortunately, your python is ancient too :-).
Believe me, to change the trend for your day so far :-), just install
a Fedora 17 VM (Fedora is very similar to a current OpenSUSE - just
replace your usual zypper invocations with yum) and try several of the
Python plugin examples (GCC there is 4.7.2 as in your /opt build).
And also thanks for the lua environment. My intention was to do some
things with c++, first try was a jave script engine, next goes python
and now we have a lua plugin. Great, I have to learn some more languages
to walk through my c++ code :-)
Taras note was about that the Mozilla team has been switched to Clang
for this sort of things, and not going to support the GCC plugin
anymore. Actually, it should not be too much of work to bring Dehydra
back to life. After/If you manage to start successfully your project
with Pyhton or Lua, you always can step-in and start to maintain the
Spidermonkey plugin.
Actually one big reason that dehydra failed is that spidermonkey became
a lot harder to embed due to increased churn(new jits landing, etc). So
I strongly advice against wasting time on it. Certain other JS JIT
engines have much more stable APIs(but much crappier JS, eg lack of
generators makes dehydra less fun). Python has the benefit of maturity +
shipping with generators, comprehensions, lambdas which make it much
better suited than various JS implementations for this.
The other thing to note is that the best way to do something like
dehydra is to build it as a highlevel wrapper on top of a lower level
api like python-gcc. The only reason dehydra was implemented the way it
was is because it came first before lower-level AST compiler bindings
were shown to be practical.
Taras
Cheers,
Alek
_______________________________________________
dev-static-analysis mailing list
[email protected]
https://lists.mozilla.org/listinfo/dev-static-analysis
_______________________________________________
dev-static-analysis mailing list
[email protected]
https://lists.mozilla.org/listinfo/dev-static-analysis
