Incorrect passing of the DBConnection object to the execute update closure which meant we attempted to call out to attributes which did not exist on that object.

Cannot use the begin_work() command if we have turned off AutoCommit manually. We were doing the same thing

Many improvements to reduce memory usage. Now uses registry, so don't need to specify databases explicitly, just use e.g. -species human

specified.  Patch from Dan S.

not needed and move declaration of a statment handle to aviod having to
declare another one.

Slice::dbID().

attached slice.

allow user to set a list of sources that will not have the [Source:...] bit at the end of the gene description

MGI have added -ps to psuedo gene names but havana have not doen this yet so check for this when getting descriptions etc

push_master_tables.pl.

Making sure if AutoCommit was off we issue a rollback to signal the start of a fresh block. This seems to be easier than using DBI's begin_work() since that will error if we have already begun a transaction

for specifying inheritance.

line switches.

This program takes the master tables from the production database
and compares it to the corresponding tables on the given servers (by
default, the staging servers).

The program will display any discrepancies on the display
while writing SQL to files in the current directory that will
correct the discrepancies.

Each SQL patch file will have the generic name "fix-DBNAME.sql"
where "DBNAME" is the name of the database, e.g.,
"fix-oryctolagus_cuniculus_otherfeatures_60_3.sql".

A discrepancy is patched by

1)  Insertion into the master table in the production database in the
    case where a new entry has been added to a database without being
    added to the master table.

2)  Insertion into the database table in the case where a new master
    entry is missing in the database.

3)  Updating the database entry in the case where an entry (identified by
    its primary key only) differs in any of its fields.

The SQL patch files may then be used to patch the databases:

  $ mysql -h server -u user -ppass < fix-DBNAME.sql




                    BE SURE TO REVIEW THESE SQL PATCH FILES
                        (along with the program output)

                            WITH YOUR EYE AND BRAIN

                              BEFORE APPLYING THEM

the wrong directory.

# This little script will bootstrap the master tables in the production
# database (and nothing else).  This means copying the relevant tables
# from a known correct database (the "template database") into the
# master_% tables in the production database.
#
# The template database needs to live on the same server as the
# production database.
#
# The otuput of this script is SQL written to standard output.  The SQL
# needs to be run against the MySQL server manually.
#