Re: New member

--Boundary-00=_pX8jQv6BmWwHNjf
Content-Type: Text/Plain;
  charset="utf-8"
Content-Transfer-Encoding: 8bit

On Monday 29 October 2012 08:56:45 Alfonso Muñoz-Pomer Fuentes wrote:
> (...)
> I wanted to know what are the best way to post, suggestions and possible 
> patches. Should I use the SF mailing list, GitHub issues, SF forum or SF 
> bug tracker?

SF forum & bugtracker are dead, as far as I know.

This mailing list and github are active.

Patches are always welcome, either on the list, or in a github.
My personnal preference would be github.

I did some automatic mdb -> postgres migration in the past.
I would recommand exportting the schema without the foreign keys
Then use psql COPY to get the csv file
Then add the foreign keys.
Attached is a an exemple script. "Works for me"© ;)

Cheers

--Boundary-00=_pX8jQv6BmWwHNjf
Content-Type: application/x-shellscript;
  name="mdb2pgsql"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
	filename="mdb2pgsql"

#!/usr/bin/env bash
#
# This file generates to stdout sql for postgres
# foreign keys are printed for standard output 3
#
# Parameter:
# The mdb file name to process (ex: /home/nirgal/kod/friends/friends-migration/db.in/india/eft/ImpexBibitData.mdb )
#
# Environment variables:
# TMP: (ex: /home/nirgal/tmp/china/ImpExpBitBit )
#
# Author: Nirgal Vourgère <[email protected]>
# Permission is granted to copy, modify, resell that script, no guaranties.

set -e
#set -x

if [ ${#@} != 1 ]; then
	echo You must give exactly one parameter: the MDB file name to process. >&2
	exit 1
fi
MDB="$1"
if [ ! -r "$MDB" ]; then
	echo "Can't read file $MDB">&2
	exit 1
fi

SCHEMA=`basename "$MDB" .mdb`

TMP="$TMP/$SCHEMA"
mkdir -p "$TMP"

echo "-- Fragment generated by $0 for $1" > "$TMP/import.sql"
echo "--" `date --rfc-3339=seconds` >> "$TMP/import.sql"
echo "-- Fragment generated by $0 for $1" >&3
echo "--" `date --rfc-3339=seconds` >&3

# Create schema
echo "Processing SCHEMA $SCHEMA">&2
echo "CREATE SCHEMA \"$SCHEMA\";" >> "$TMP/import.sql"
echo "SET search_path='$SCHEMA';" >> "$TMP/import.sql"
echo "SET search_path='$SCHEMA';" >&3


# Get the schema, without the foreign keys nor indexes
mdb-schema --no-drop-table --not-null --default-values --not-empty --indexes --no-relations "$MDB" postgres  | grep -v '^CREATE INDEX' | grep -v '^CREATE UNIQUE INDEX' >> "$TMP/import.sql" # key the primary keys
# Print indexes on standard output 3
mdb-schema --no-drop-table --no-not-null --no-default-values --no-not-empty --indexes --no-relations "$MDB" postgres | grep '^CREATE INDEX' >&3 || true
mdb-schema --no-drop-table --no-not-null --no-default-values --no-not-empty --indexes --no-relations "$MDB" postgres | grep '^CREATE UNIQUE INDEX' >&3 || true
# Print foreign keys on standard output 3
mdb-schema --no-drop-table --no-not-null --no-default-values --no-not-empty --no-indexes --relations "$MDB" postgres | grep REFERENCES >&3 || true


# Generate CSV files and matching COPY sql statements
mdb-tables -1 "$MDB" | while read table; do
	cleantable=`echo $table|sed -e "s/\///g"` # remove / from table names
	echo generating $cleantable.csv>&2
	# Add leading 0 to dates before 1000AD so that years always are 4 digits long
	mdb-export -X \\ -D "{postgresdate}%Y-%m-%d %H:%M:%S" "$MDB" "$table" | sed -re 's/\{postgresdate\}([0-9]{2})-/00\1-/g' | sed -re 's/\{postgresdate\}([0-9]{3})-/0\1-/g' | sed -re 's/\{postgresdate\}//g' > "$TMP/$cleantable.csv"
	squotedcleantable=`echo $cleantable|sed -e "s/'/''/g"`
	echo "COPY \"$table\" FROM '$TMP/$squotedcleantable.csv' WITH DELIMITER ',' CSV HEADER QUOTE AS '\"' ESCAPE AS E'\\\\' ;" >> "$TMP/import.sql"
done

# Update the sequences
echo "-- Generated sequence next val updates from default column values"  >> "$TMP/import.sql"
cat "$TMP/import.sql" | while read line
do
	if [[ "$line" =~ ^CREATE\ TABLE\ \"(.*)\" ]]
	then
		#echo $line
		current_table=${BASH_REMATCH[1]}
		squoted_current_table=`echo $current_table|sed -e "s/'/''/g"`
	fi
	if [[ $line =~ [\ \t]*\"([^\"]*)\".*[\ \t]*SERIAL ]]
	then
		#echo $line
		current_field=${BASH_REMATCH[1]}
		squoted_current_field=`echo $current_field|sed -e "s/'/''/g"`
		echo "SELECT pg_catalog.setval('\"${squoted_current_table}_${squoted_current_field}_seq\"', (SELECT MAX(\"$current_field\") FROM \"$current_table\"), true);" >> "$TMP/import.sql"
	fi
done

# Add OIDS for tables with a primary key over several columns
echo "-- Adding OIDS for tables with a primary key over several columns" >> "$TMP/import.sql"
for table in `grep "PRIMARY KEY .*," "$TMP/import.sql" | cut -d '"' -f 2`; do
	echo "ALTER TABLE \"$table\" SET WITH OIDS;" >> "$TMP/import.sql"
done

# Add OIDS for tables that have no primary key
echo "-- Adding OIDS for tables with no primary keys" >> "$TMP/import.sql"
mdb-tables -1 "$MDB" | while read table; do
	if ! grep -q "ALTER TABLE \"$table\" ADD CONSTRAINT .* PRIMARY KEY" "$TMP/import.sql"; then
		echo "ALTER TABLE \"$table\" SET WITH OIDS;" >> "$TMP/import.sql"
	fi
done

# print the SQL result
cat "$TMP/import.sql"

--Boundary-00=_pX8jQv6BmWwHNjf
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

------------------------------------------------------------------------------
Everyone hates slow websites. So do we.
Make your web apps faster with AppDynamics
Download AppDynamics Lite for free today:
http://p.sf.net/sfu/appdyn_sfd2d_oct
--Boundary-00=_pX8jQv6BmWwHNjf
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
mdbtools-dev mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mdbtools-dev

--Boundary-00=_pX8jQv6BmWwHNjf--