RE: [edict-jmdict] Re: creating jmdict database in postgresql

["Stuart McGraw" <smcg4191@********>]

jasonmcgregorshepherd wrote:
[...]
> Does this python script have the same inherent problems with xrefs as
> the make file does? I guess it's not too hard to connect to Postgres
> with Java so either is ok really.

Where did this script come from?  Was it in the snapshot tar file?

> As long as the lookup's are fast and
> accurate. Which is what xrefs help with I suppose.

The xrefs I'm referring to are the database representation of
the <xref> and <ant> elements in the JMdict file.  They hold 
the "see also" and "antonym" cross references (in the form 
of the id number of the entry that is being referred to.)  They 
have no effect on accuracy or speed of lookups (searches).  
They actually slow down retrieval of data after one has found 
an entry looked for, because more data has to be fetched if 
the found entry contains any xrefs.  However, there are only
about 2800 of them.