On 4/08/2013 10:04 AM, Ze'ev Atlas wrote:
I just built pcre-8.33 and was impressed by how easy it was build. I ran
./configure --enable-ebcdic && make and it built clean straight off the
bat. I ran the test suite and had a quick look at pcregrep which is
significantly more powerful
than POSIX grep. It's my first look a pcre and it's great package.
OK, this proves to me that I was wrong all along and should have looked that direction
from the beginning. In my defense, I am a "classic" z/OS programmer in the
core and all these Unix tools confuse me.
Would you be so kind and send me (maybe off the list) a full script of what you are doing
including an explanation "for Dummies" assuming I know nothing!
Thanks
ZA
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
Here's what I did. It's a order of magnitude simpler than rubbing two
sticks together with JCL and PDS data sets.
Run the following commands from a z/OS UNIX shell (I use putty). You
will need to download the UNIX tools that I mentioned before for gzip
etc and have ported tools
installed if you want to use curl.
1. download PCRE source code tarball
curl -O
ftp://ftp.csx.cam.ac.uk/pub/software/programming/pcre/pcre-8.33.tar.gz
2. decompress the package
gzip -d pcre-8.33.tar.gz
3. extract the archive converting ASCII to EBCDIC
pax -o from=iso8859-1 -rvf pcre-8.33.tar.gz
4. build the software
cd pcre-8.33
./configure --enable-ascii && gmake
5. run the test suite
RunTest
6. simple pcregrep test
pcregrep '#includ' *.cpp
Simple as that. Took less than 10 minutes end-to-end.
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
