Hi Steve:
This version should fix you; in my little test it was about 20x faster doing repeated calculations. You still take a hit if skipping holidays (as defibed in /etc/acct/holidays) or if inputting/outputting month or week day names because those still rely upon awk but for the usual applications this version is much faster.
Again, you can simply yank out the cal_jdate, jdate_cal, and wkday functions and use them inside your own scripts.
Use it like this:
#!/usr/bin/sh
INFILE="./mydates"
caljd.sh -S "/" -n 30 -x 0 -x 6 -c -f | caljd.sh -S "/" -f |&
# The above is all one line
cat ${INFILE} | while read DT
do
print -p ${DT}
read -p S
echo "${DT} --> ${S}"
done
Note the new "-f" arguments. That instructs caljd.sh to do read stdin and write on stdout.
The command line args are parsed only once.
Pay careful attention to the "|&" invocation. That starts caljd.sh as a co-process. The -p arguments for the read and print statements direct the i/o from/to the co-process.
Now, I've got to do the same thing to caljd.pl but it should be much easier.
If it ain't broke, I can fix that.
