1) If gen.sh fails the build stops.
2) checks if ant is available before attempting to launch ant check & ant
no-build.
3) passes ${GUMP_WORKSPACE}.xml to the ant scripts
Somebody mind checking these in?
Here is a cvs diff patch:
Here is the new file (I tested it, as far as failing after gen.sh failure)
Questions:
1) Since this is intended to be "nightly builds" (not manual testing) ought
we always clean the cache? rm -rf it?
2) Ought we similarly wipe the gump word area?
3) What about grepping for "Dropping project" on the output of gen.sh and
mailing it to some predetermined alias?
regards
Adam
--
<http://www.try.sybase.com>
Experience Sybase Technology ...
Index: gump.sh
===================================================================
RCS file: /home/cvspublic/jakarta-gump/gump.sh,v
retrieving revision 1.9
diff -u -r1.9 gump.sh
--- gump.sh 19 Jul 2003 21:55:19 -0000 1.9
+++ gump.sh 7 Aug 2003 21:30:33 -0000
@@ -93,19 +93,25 @@
cvs -q update -dP >> $GUMP_LOG 2>&1
rm -f .timestamp
-#
-# generate nobuild projects
-#
-cd $GUMP
-echo $SEPARATOR >> $GUMP_LOG
-ant nobuild-projects >> $GUMP_LOG 2>&1
+export ANT=`which ant 2>/dev/null`
+if [ -n "$ANT" ] ; then
+ #
+ # generate nobuild projects
+ #
+ cd $GUMP
+ echo $SEPARATOR >> $GUMP_LOG
+ $ANT -Dworkspace=${GUMP_WORKSPACE}.xml nobuild-projects >> $GUMP_LOG 2>&1
+
+ #
+ # Check the projects
+ #
+ cd $GUMP
+ echo $SEPARATOR >> $GUMP_LOG
+ $ANT -Dworkspace=${GUMP_WORKSPACE}.xml check >> $GUMP_LOG 2>&1
+else
+ echo "Ant was not found in the environment, steps skipped." >> $GUMP_LOG
+fi
-#
-# Check the projects
-#
-cd $GUMP
-echo $SEPARATOR >> $GUMP_LOG
-ant check >> $GUMP_LOG 2>&1
#
# Do a gen
@@ -113,7 +119,12 @@
cd $GUMP
echo $SEPARATOR >> $GUMP_LOG
bash gen.sh ${GUMP_WORKSPACE}.xml >> $GUMP_LOG 2>&1
+if [ $? -ge 1 ] ; then
+ echo "Gump failed in generation, can't continue."
+ exit 1
+fi
echo >> $GUMP_LOG
+
#
# Do a clean
#!/bin/bash
#
# $Header: /home/cvspublic/jakarta-gump/gump.sh,v 1.9 2003/07/19 21:55:19 nickchalko
Exp $
if [ -e local-env.sh ] ; then
. local-env.sh
fi
if [ ! $GUMP ] ; then
echo "Set the \$GUMP variable to your gump install"
exit 1
fi
if [ ! $GUMP_WS ] ; then
echo "Set the \$GUMP_WS variable to your gump install"
exit 1
fi
if [ ! $GUMP_LOG_DIR ] ; then
echo "Set the \$GUMP_LOG_DIR variable to your gump install"
exit 1
fi
if [ -n "$1" ] ; then
export GUMP_TARGET=$1
else
export GUMP_TARGET=all
fi
#
# Calculated
#
export GUMP_HOST=`hostname -s`
export GUMP_DATE=`date`
export GUMP_LOG=$GUMP_LOG_DIR/gump.html
export GUMP_PROFILE_LOG_DIR=$GUMP_LOG_DIR/myprofile
if [ -z "$GUMP_WORKSPACE" ] ; then
export GUMP_WORKSPACE=${GUMP_HOST}
fi
export SEPARATOR='------------------------------------------------------- G U M P'
#
# Generate gump.html from this (into the WWW site)
#
umask 002
echo \<XMP\> > $GUMP_LOG
echo $SEPARATOR >> $GUMP_LOG
echo $SEPARATOR >> $GUMP_LOG
echo "Gump run on $GUMP_HOST at $GUMP_DATE" >> $GUMP_LOG
echo $SEPARATOR >> $GUMP_LOG
echo $SEPARATOR >> $GUMP_LOG
echo "Gump Target : $GUMP_TARGET" >> $GUMP_LOG
echo "Gump Location : $GUMP" >> $GUMP_LOG
echo "Gump Workspace : $GUMP_WS" >> $GUMP_LOG
echo "Gump Log : $GUMP_LOG" >> $GUMP_LOG
echo $SEPARATOR >> $GUMP_LOG
echo $SEPARATOR >> $GUMP_LOG
echo >> $GUMP_LOG
#
# Capture environment
#
echo $SEPARATOR >> $GUMP_LOG
export >> $GUMP_LOG
#
# Store the profile (into a myprofile dir)
#
cd $GUMP
if [ ! -d $GUMP_PROFILE_LOG_DIR ] ; then
mkdir $GUMP_PROFILE_LOG_DIR;
fi
cp $GUMP/gump.sh $GUMP_PROFILE_LOG_DIR
cp ${GUMP_WORKSPACE}.xml $GUMP_PROFILE_LOG_DIR
cp -R `grep profile ${GUMP_WORKSPACE}.xml | cut -d\" -f2` $GUMP_PROFILE_LOG_DIR
#
# Do a CVS update
#
echo $SEPARATOR >> $GUMP_LOG
cd $GUMP
cvs -q update -dP >> $GUMP_LOG 2>&1
rm -f .timestamp
export ANT=`which ant 2>/dev/null`
if [ -n "$ANT" ] ; then
#
# generate nobuild projects
#
cd $GUMP
echo $SEPARATOR >> $GUMP_LOG
$ANT -Dworkspace=${GUMP_WORKSPACE}.xml nobuild-projects >> $GUMP_LOG 2>&1
#
# Check the projects
#
cd $GUMP
echo $SEPARATOR >> $GUMP_LOG
$ANT -Dworkspace=${GUMP_WORKSPACE}.xml check >> $GUMP_LOG 2>&1
else
echo "Ant was not found in the environment, steps skipped." >> $GUMP_LOG
fi
#
# Do a gen
#
cd $GUMP
echo $SEPARATOR >> $GUMP_LOG
bash gen.sh ${GUMP_WORKSPACE}.xml >> $GUMP_LOG 2>&1
if [ $? -ge 1 ] ; then
echo "Gump failed in generation, can't continue."
exit 1
fi
echo >> $GUMP_LOG
#
# Do a clean
#
echo $SEPARATOR >> $GUMP_LOG
bash $GUMP_WS/build.sh clean >> $GUMP_LOG
echo >> $GUMP_LOG
#
# Do an update (from CVS)
#
cd $GUMP_WS
echo $SEPARATOR >> $GUMP_LOG
bash update.sh $GUMP_TARGET >> $GUMP_LOG 2>&1
#
# Do the build
#
cd $GUMP_WS
echo $SEPARATOR >> $GUMP_LOG
bash build.sh $GUMP_TARGET >> $GUMP_LOG 2>&1
cd $GUMP
echo $SEPARATOR >> $GUMP_LOG
if [ -n "$STARTED_FROM_CRON" ] ; then
perl nag.pl work/naglist >> $GUMP_LOG 2>&1
fi;
echo \</XMP\> >> $GUMP_LOG
pkill -P $$
# $Log: gump.sh,v $
# Revision 1.9 2003/07/19 21:55:19 nickchalko
# Added GUMP_WORKSPACE variable
# PR:
# Obtained from:
# Submitted by: Adam Jack [EMAIL PROTECTED]
# Reviewed by:
# CVS: ----------------------------------------------------------------------
# CVS: PR:
# CVS: If this change addresses a PR in the problem report tracking
# CVS: database, then enter the PR number(s) here.
# CVS: Obtained from:
# CVS: If this change has been taken from another system, such as NCSA,
# CVS: then name the system in this line, otherwise delete it.
# CVS: Submitted by:
# CVS: If this code has been contributed to Apache by someone else; i.e.,
# CVS: they sent us a patch or a new module, then include their name/email
# CVS: address here. If this is your work then delete this line.
# CVS: Reviewed by:
# CVS: If we are doing pre-commit code reviews and someone else has
# CVS: reviewed your changes, include their name(s) here.
# CVS: If you have not had it reviewed then delete this line.
#
# Revision 1.8 2003/06/26 06:39:55 nickchalko
# Add ant nobuild-project to the gump sequence.
# PR:
# Obtained from:
# Submitted by:
# Reviewed by:
# CVS: ----------------------------------------------------------------------
# CVS: PR:
# CVS: If this change addresses a PR in the problem report tracking
# CVS: database, then enter the PR number(s) here.
# CVS: Obtained from:
# CVS: If this change has been taken from another system, such as NCSA,
# CVS: then name the system in this line, otherwise delete it.
# CVS: Submitted by:
# CVS: If this code has been contributed to Apache by someone else; i.e.,
# CVS: they sent us a patch or a new module, then include their name/email
# CVS: address here. If this is your work then delete this line.
# CVS: Reviewed by:
# CVS: If we are doing pre-commit code reviews and someone else has
# CVS: reviewed your changes, include their name(s) here.
# CVS: If you have not had it reviewed then delete this line.
#
# Revision 1.7 2003/06/02 18:18:44 nickchalko
# Fixed typo. $ should be inside the ""
# PR:
# Obtained from: Adam Jack [EMAIL PROTECTED]
# Submitted by:
# Reviewed by:
#
# Revision 1.6 2003/05/30 22:02:56 nickchalko
# Fixing incomplete update from Adam
# PR:
# Obtained from:
# Submitted by:Adam Jack [EMAIL PROTECTED]
# Reviewed by:
#
# Revision 1.4 2003/04/09 07:31:42 nickchalko
# -mRemoved reference to tsbuild1
#
# Revision 1.3 2003/04/09 07:27:27 nickchalko
# Moved user specific vars to local-env.sh which is not checked in.
#
# Revision 1.2 2003/04/08 19:18:58 nickchalko
# Here is adams rework. I moved the cvs log to the bottom, and added a comment about
using a local_env.sh file.
#
#
# Commit notes from Adam Jack:
#
# I've taken your enhancements and tried to go one step further.
#
# I have this accepting three environment variables GUMP/GUMP_WS/GUMP_LOG and
# try to do everything off those. It exits (with 1 -- that right?) if they are
# not set. I set those three and it seems to run in my environment. Normally
# I'd expect folks to hack these entries.
#
# 1) I have it copy `hostname -s`.xml (likely workspace) and 'likely profile'
# and the gump.sh to myprofile:
#
# cp $GUMP/gump.sh $GUMP_PROFILE_LOG_DIR
# cp $GUMP_HOST.xml $GUMP_PROFILE_LOG_DIR
# cp `grep profile $GUMP_HOST.xml | cut -d\" -f2` $GUMP_PROFILE_LOG_DIR
#
# 2) I added some CVS headers, but am not sure I got those right.
#
# 3) I set it as bash not sh -- that ok?
#
# 4) I made it echo some separators, but it could use more verbose information
# about each section.
#
# 5) I made the output be gump.html (better name than gen.html).
#
# 6) I don't like the way cron is detected (for running nag.pl) something
# better is needed.
#
# 7) I made it do the CVS update each time, and a build clean each time. Maybe
# these are overkill.
#
# This is a major work in progress (it'd be nice if it output HTML as pretty
# as build.sh or update.sh) and it needs a few switches (so folks don't have
# to do everything every time.)
#
# This is is "Linux" format.
#
#
# Submitted by: Adam Jack [EMAIL PROTECTED]
#
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
