From: Ben Faga <faga@cshl.edu>
Date: Mon, 28 Jul 2008 12:36:07 +0000 (+0000)
Subject: Applied patches written by Nigel Metheringham.  His notes follow.
X-Git-Tag: v0.11008~308
X-Git-Url: http://git.shadowcat.co.uk/gitweb/gitweb.cgi?a=commitdiff_plain;h=999859d79388a00c262c32c08d1439578e925eb9;p=dbsrgits%2FSQL-Translator.git

Applied patches written by Nigel Metheringham.  His notes follow.
---

The postgres producer uses a hash as both a single level and a 2 level
symbol table for ensuring names are unique which causes some interesting
failures when a database has constraint names which match table names.

I've replaced the code with effectively the same code as the mysql
producer uses, massaged very slightly so as in the most cases to
produce the same name remapping as the old code but without the
extra explosions when it attempts to deref a scalar as a hash.

---

When a table has more than one index, all but the last CREATE INDEX
commands produced have a doubled semi-colon at the end of them.

This causes the parser for postgresql to explode :-)

One line bug fix appears to sort it.
---

diff --git a/lib/SQL/Translator/Producer/PostgreSQL.pm b/lib/SQL/Translator/Producer/PostgreSQL.pm
index afea2b8..a7abf10 100644
--- a/lib/SQL/Translator/Producer/PostgreSQL.pm
+++ b/lib/SQL/Translator/Producer/PostgreSQL.pm
@@ -38,7 +38,7 @@ producer.
 
 use strict;
 use warnings;
-use vars qw[ $DEBUG $WARN $VERSION ];
+use vars qw[ $DEBUG $WARN $VERSION %used_names ];
 $VERSION = sprintf "%d.%02d", q$Revision: 1.29 $ =~ /(\d+)\.(\d+)/;
 $DEBUG = 1 unless defined $DEBUG;
 
@@ -182,6 +182,7 @@ sub produce {
     my $add_drop_table = $translator->add_drop_table;
     my $schema         = $translator->schema;
     my $pargs          = $translator->producer_args;
+    local %used_names  = ();
 
     my $postgres_version = $pargs->{postgres_version} || 0;
 
@@ -192,7 +193,6 @@ sub produce {
     
     my $output;
     $output .= header_comment unless ($no_comments);
-#    my %used_index_names;
 
     my (@table_defs, @fks);
     for my $table ( $schema->get_tables ) {
@@ -292,18 +292,14 @@ sub unreserve {
 
 # -------------------------------------------------------------------
 sub next_unused_name {
-    my $name       = shift || '';
-    my $used_names = shift || '';
-
-    my %used_names = %$used_names;
-
-    if ( !defined($used_names{$name}) ) {
+    my $name = shift || '';
+    if ( !defined( $used_names{$name} ) ) {
         $used_names{$name} = $name;
         return $name;
     }
-    
+
     my $i = 2;
-    while ( defined($used_names{$name . $i}) ) {
+    while ( defined( $used_names{ $name . $i } ) ) {
         ++$i;
     }
     $name .= $i;
@@ -311,6 +307,7 @@ sub next_unused_name {
     return $name;
 }
 
+
 sub create_table 
 {
     my ($table, $options) = @_;
@@ -401,7 +398,7 @@ sub create_table
                             "\n);"
                             ;
 
-    $create_statement .= "\n" . join(";\n", @index_defs) . "\n";
+    $create_statement .= "\n" . join("\n", @index_defs) . "\n";
     
     return $create_statement, \@fks;
 }
@@ -480,9 +477,6 @@ sub create_table
     }
 }
 
-{
-    my %used_index_names;
-
     sub create_index
     {
         my ($index, $options) = @_;
@@ -494,11 +488,9 @@ sub create_table
 
         my ($index_def, @constraint_defs);
 
-        $used_index_names{$table_name} ||= {};
         my $name = $index->name || '';
         if ( $name ) {
-            $name = next_unused_name($name, $used_index_names{$table_name});
-            $used_index_names{$name} = $name;
+            $name = next_unused_name($name);
         }
 
         my $type = $index->type || NORMAL;
@@ -543,8 +535,7 @@ sub create_table
 
         my $name = $c->name || '';
         if ( $name ) {
-            $name = next_unused_name($name, \%used_index_names);
-            $used_index_names{$name} = $name;
+            $name = next_unused_name($name);
         }
 
         my @fields     = 
@@ -564,8 +555,7 @@ sub create_table
                 '('.$qf . join( $qf.', '.$qf, @fields ) . $qf.')';
         }
         elsif ( $c->type eq UNIQUE ) {
-            $name = next_unused_name($name, \%used_index_names);
-            $used_index_names{$name} = $name;
+            $name = next_unused_name($name);
             push @constraint_defs, "${def_start}UNIQUE " .
                 '('.$qf . join( $qf.', '.$qf, @fields ) . $qf.')';
         }
@@ -604,7 +594,6 @@ sub create_table
 
         return \@constraint_defs, \@fks;
     }
-}
 
 sub convert_datatype
 {