Robert Bernier <robert@pg-live.info> wrote:
> Still it would be nice if people could mention here on the list
> real world use-cases using text-search.
The Wisconsin Supreme Court and Court of Appeals software uses
tsearch2 against text extracted from PDF documents to allow
justices, appeals court judges, and their clerks to search filings,
motions, opinions, etc. Custom dictionaries are added to the
dictionary chain to allow accurate recognition of legal phrases
like "amicus curiae" or "power of attorney", and the lexemes
generated by the dictionaries are combined with the results of
passing the text with a regular expression search to find dates,
case numbers, and statute cites (which get higher priority in the
searches). If any portion of the search string entered by the user
has quotes around it, the normal tsearch2 results are further
filtered by looking for an exact match on the quoted string.
-Kevin
