Re: tv_extractinfo_en problem again!

>>> These are the last few lines of the trace before the crash...
>>> ==================================================
>>> testing word 2.75 at position 37
>>> word is '2.75'
>>> trying 2.75 as 24 hour time
>>> found a time that interprets: 2:75
>> ...

>> It's testing if it should split the programme at 2.75. The test
>> returns 2:75 as a valid time (which it is not)

> I have to say that that still looks like a bogus path. If the
> distance mentioned in the description were 2.39 miles instead of
> 2.75 miles, it would still be trying to pull out a time (39<  75),
> which is clearly not the right thing to be doing here.

Why is that bogus? The programme stopped when it tried to handle 2:75
as a time. The check avoids just that.

quoting from the man page:
 > This tool also attempts to split multipart programmes into their
 > constituents, by looking for a description that seems to contain lots
 > of times and titles.  But this depends on the description following
 > one particular style and is useful only for some listings sources
 > (Ananova).

After digging into the code it's considering to split programmes that
have at least 3 valid times in their description that are between start
and end time and where there is no better match from another extraction
function.

Parsing machine readable data from raw text is black art after all :)

Regards,
Karl

------------------------------------------------------------------------------
WhatsUp Gold - Download Free Network Management Software
The most intuitive, comprehensive, and cost-effective network 
management toolset available today.  Delivers lowest initial 
acquisition cost and overall TCO of any competing solution.
http://p.sf.net/sfu/whatsupgold-sd