(MySQL v5 knows how to do index merges) in _type_by_external_id():

  # decrease in time by 35%
  $dbea->list_transcript_ids_by_extids('BRCA2');

  # decrease in time by 28%
  $dbea->list_translation_ids_by_extids('GO:0004835');

  # decrease in time by 24%
  $dbea->list_gene_ids_by_extids('ARSE');


Also some formatting.

got rid of a bad line

extras for tetraodon

in human is number 118 in length, no gene, %gc, or SNP density features
were being calculated for it.

added support for different cache methods (currently implemented: build_cache_all() and build_cache_by_seq_region()

methods for initial db checks; two variants of build_cache(): build_cache_by_seq_region() and build_cache_all()

1)  Indicate truncated Xref descriptions by appending ' /.../' to
    the end of them.  The maximum length of the descriptions is 255
    characters (including this truncation marker).

2)  Properly check for successful running of the 'xref_config2sql.pl'
    script and be verbose about reporting failures (from the
    documentation of the 'system' command).

multiple spaces after the words 'source' or 'species' are flagged as
errors.

Reworked recalc_display_xrefs functionality so that it provides a more complete of SQL files (e.g. unset display_xrefs and descriptions, and also gene status updates).