A document may contain date in the traditional format. For example it
may contain '11/1/2007'. It will be useful if we can directly search for
year in a document. However, the 'default' tsearch2 parser does not
break down integers separated by '/'. So I my search for '2007' will not
match tsvector for '11/1/2007'. Here is an example
cmsdb=# select to_tsvector('default', '11/1/2007');
to_tsvector
----------------
'11/1/2007':1
I think this can be easily fixed if we use '/' as a word separator. Is
there an way to specify word separators in tsearch2 module?
Thank you,
-Sushant.