Add developers help file.

This commit is contained in:
Bruce Momjian 1997-08-06 03:37:34 +00:00
parent 1ebc1280e8
commit 3bea7b138b
1 changed files with 66 additions and 0 deletions

66
src/DEV_TIPS Normal file
View File

@ -0,0 +1,66 @@
Bruce Momjian <maillist@candle.pha.pa.us>
Here are some of the scripts I use to make development easier.
First, I use 'cpdir' on every file I am about to change. This makes a
copy with the extension .orig. If an .orig already exists, I am warned.
:
# cporig
for FILE
do
if [ ! -f "$FILE.orig" ]
then cp $FILE $FILE.orig
else echo "$FILE.orig exists" 1>&2
fi
done
I can get really fancy with this. I can do 'cporig *' and make a .orig
for every file in the current directory. I can:
cporig `grep -l HeapTuple *`
If I use mkid (from ftp.postgreSQL.org), I can do:
cporig `lid -kn 'fsyncOff'`
and get a copy of every file containing that word. I can then do:
vi `find . -name '*.orig'`
or even better (using mkid):
eid fsyncOff
to edit all those files.
When I am ready to generate a patch, I run this command from the top of
the source tree:
:
#difforig
if [ "$#" -eq 0 ]
then APATH="."
else APATH="$1"
fi
find $APATH -name '*.orig' -print | sort | while read FILE
do
NEW="`dirname $FILE`/`basename $FILE .orig`"
echo "$NEW" 1>&2
diff -c $FILE $NEW
done
I pipe the output of this to a file to hold my patch, and the file names
it processes appear on my screen. It creates a nice patch for me of all
the files I used with cporig.
Finally, I remove my old copies with:
:
# rmorig
if [ "$#" -eq 0 ]
then APATH="."
else APATH="$1"
fi
find $APATH -name '*.orig' -exec rm {} \;