[antlr-interest] Verifying bug reports in C runtime, beta testers, unit tests

[Jim Idle]

I am just about finished with the updates for the C runtime for the
impending ANTLR 3.1 release, but I need to verify a few things and look
for a little help.

 

Firstly, a number of you sent me bug reports, which I believe I kept
track of correctly, but I think I missed putting some of them in Jira.
Then my new mail server trashed all my archives and an eagle swooped
down from the sky and stole my homework, and… The long and short of it
is that in order that I can make sure I really have covered all the
reported bugs, if you have reported a bug or asked for a feature from me
since the last release, please send me a repeat email with the bug and I
will endeavor to ensure that the bugs are fixed. Sorry for the extra
work, but in future I will be putting things straight into Jira as I get
them rather than fooling myself that I will have the time to verify the
bugs first and so on. I am fine with getting repeat reports and won’t  b
upset or offended, so just email me anything you have in case I have
missed it. Thanks in advance for that.

 

 

Secondly, once the code is testable I would hope that some of you might
wish to try it out for m and tell me of any bugs that you find. I have
not had a lot of testing time for this release so if a few of us get
together then the testing will take no time at all. I don’t need anyone
to find the actual reason for the bugs, just give me grammar snippets or
code snippets and tell me what goes wrong. Let m know if you have time
to do that.

 

Thirdly, the other targets have unit tests that can be run for
regressions and so on. Apart from hating producing unit test, I always
think it is a good idea for someone other than the developer to write
such things. If anyone feels like they are pedantic enough to write good
unit tests for the C runtime (I don’t care how they are written etc, so
long as it does not involve installing 60 Java jars to do it ;-), then
please let me know and I will arrange some access to perforce for
writing these with Ter.

 

Finally, I intend that this release will allow the generated code to be
compiled as C++. This does not mean that the libraries will compile as
C++, they will still be an external C library, but it should mean that
you can use your own C++ classes in your grammar and they will compile
just fine. The output file will still be a .c, so it is up to you to
switch on whatever flags you need to compile it as C++. I would like to
know if anyone is willing to test this out as soon as I say go. There is
nothing special to do but try it out, but I would just like to know that
one or two people HAVE tried it out, rather than just hope they have.

 

Finally finally, this release will come with a huge improvement in API
and usage documentation, which will be placed on the web site from the
usual API documentation links on the front page.  I am hoping that a few
people might run through it looking for errors or places where my
Yorkshire/Cockney/American grammar combination doesn’t make sense.

 

Thanks for you help,

 

Jim


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.antlr.org/pipermail/antlr-interest/attachments/20080216/505d0195/attachment.html