Log Message:
-----------
Added script to extract the user schema (without slony cruft)
from an existing node. The node must be the origin.

Jan

Modified Files:
--------------
    slony1-engine/tools:
        Makefile (r1.2 -> r1.3)

Added Files:
-----------
    slony1-engine/tools:
        slony1_extract_schema.sh (r1.1)

-------------- next part --------------
--- /dev/null
+++ tools/slony1_extract_schema.sh
@@ -0,0 +1,80 @@
+#!/bin/sh
+# ----------
+# slony1_extract_schema.sh
+#
+#	Script to extract the user schema of a slony node in the original
+#	state with all Slony related cruft removed.
+# ----------
+
+# ----
+# Check for correct usage
+# ----
+if test $# -ne 3 ; then
+	echo "usage: $0 dbname clustername tempdbname" >&2
+	exit 1
+fi
+
+# ----
+# Remember call arguments and get the nodeId of the DB specified
+# ----
+dbname=$1
+cluster=$2
+tmpdb=$3
+nodeid=`psql -q -At -c "select \"_$cluster\".getLocalNodeId('_$cluster')" $dbname`
+
+TMP=tmp.$$
+
+# ----
+# Print a warning for sets originating remotely that their
+# triggers and constraints will not be included in the dump.
+# ----
+psql -q -At -c "select 'Warning: Set ' || set_id || ' does not origin on node $nodeid - original triggers and constraints will not be included in the dump' from \"_$cluster\".sl_set where set_origin <> $nodeid" $dbname >&2
+
+# ----
+# Step 1.
+#
+# Dump the current schema and the Slony-I configuration data. Add
+# UPDATE commands that correct the table oid's in the config data
+# when this dump is restored.
+# ----
+pg_dump -s $dbname >$TMP.sql
+pg_dump -a -n _$cluster $dbname >>$TMP.sql
+
+psql -q -t -c "select 'update \"_$cluster\".sl_table set tab_reloid=(select C.oid from \"pg_catalog\".pg_class C, \"pg_catalog\".pg_namespace N where C.relnamespace = N.oid and C.relname = ''' || C2.relname || ''' and N.nspname = ''' || N2.nspname || ''') where tab_id = ''' || tab_id || ''';' from \"_$cluster\".sl_table T, \"pg_catalog\".pg_class C2, \"pg_catalog\".pg_namespace N2 where T.tab_reloid = C2.oid and C2.relnamespace = N2.oid" $dbname >>$TMP.sql
+
+psql -q -t -c "select 'update \"_$cluster\".sl_sequence set seq_reloid=(select C.oid from \"pg_catalog\".pg_class C, \"pg_catalog\".pg_namespace N where C.relnamespace = N.oid and C.relname = ''' || C2.relname || ''' and N.nspname = ''' || N2.nspname || ''') where seq_id = ''' || seq_id || ''';' from \"_$cluster\".sl_sequence T, \"pg_catalog\".pg_class C2, \"pg_catalog\".pg_namespace N2 where T.seq_reloid = C2.oid and C2.relnamespace = N2.oid" $dbname >>$TMP.sql
+
+# ----
+# Step 2.
+#
+# Create a temporary database and restore the schema including all
+# Slony related cruft.
+# ----
+createdb $tmpdb >/dev/null 
+psql $tmpdb <$TMP.sql >/dev/null 2>&1
+
+# ----
+# Step 3.
+#
+# Use the slonik "uninstall node" command to restore the original schema.
+# ----
+slonik >/dev/null 2>&1 <<_EOF_
+cluster name = $cluster;
+node $nodeid admin conninfo = 'dbname=$tmpdb';
+uninstall node (id = $nodeid);
+_EOF_
+
+# ----
+# Step 4.
+#
+# Use pg_dump on the temporary database to dump the user schema
+# to stdout.
+# ----
+pg_dump -s $tmpdb
+
+# ----
+# Remove temporary files and the database
+# ----
+sleep 2
+dropdb $tmpdb >/dev/null
+rm $TMP.*
Index: Makefile
===================================================================
RCS file: /usr/local/cvsroot/slony1/slony1-engine/tools/Makefile,v
retrieving revision 1.2
retrieving revision 1.3
diff -Ltools/Makefile -Ltools/Makefile -u -w -r1.2 -r1.3
--- tools/Makefile
+++ tools/Makefile
@@ -13,7 +13,7 @@
 
 SUBDIRS = altperl
 
-DISTFILES = Makefile slony_setup.pl generate_syncs.sh
+DISTFILES = Makefile slony_setup.pl generate_syncs.sh slony1_extract_schema.sh
 
 all install installdirs: