[TYPO3-50-general] Evaluating FLOW3 for a new project

[Ries van Twisk]

On Apr 9, 2009, at 10:46 AM, Martin Kutschker wrote:

> Ries van Twisk schrieb:
>>
>>> but PostgreSQL has some features to make a good language dependent  
>>> full
>>> text search (using dictionaries, synonym lists, a theraus and
>>> ispell/myspell/hunspell stemming dictionaries).
>>>
>>> From SQL it's quite awkward to use ([1]), but it's tokenizer [2] and
>>> dictionaries [3] are quite impressive. Much better then what Mysql  
>>> has
>>> to offer (at least in the stable versions).
>>
>> I am using this myself in a multi-lingual application
>> where the database stored content coming in form different languages.
>>
>> Searches for 'wheels' results in records that also contain wheel,
>> something that is very nice to have.
>
> For me this is a must-have for a modern text search.

Fully agreed. It's often humans that do the searches in the first place,
and it's hard to explain to them that they have to search for the  
exact word,
or else the system might not find it (wheels vs wheel).

For us this was also a requirement, and it took a couple of hours to  
get it implemented
from start to finish.

>
>> It's not really awkward once you know the implementation
>> and why it was created in such a way.
>
> I only read the docs and it wasn't an easy read for me ;)

We put indexes on the tables much like this:

CREATE INDEX tbl_copy_translation_idx_ts_2
   ON mkt.tbl_copy_translation
   USING gin
   (to_tsvector('english'::regconfig, copy))
   WHERE lang_id = 2;

The table has a record for each language in the same table. We

lang_id is in this case english, this speeds up the system enormously.



Ries




>
> Masi