Thunderstone Support Forums

We have several aliases for each mission on our site, to avoid the problem of people using caps. all-caps, or lower-case. We also have (naturally) direct paths to the files. Unfortunately, successive web designers have chosen different ways of referring to a mission so :-

http://www.rssd.esa.int/Gaia/index.html
http://www.rssd.esa.int/GAIA/index.html
http://www.rssd.esa.int/gaia/index.html
http://www.rssd.esa.int/SA-general/Proj ... index.html

All refer to the same page ... and I want to do a project-based categorisation ... so ... my question ...

I can see that using URL/sub-dir/* works when defining categories, but is it possible to use a construction like */sub-dir/* ??

Hey ... no need to answer - I tried it and it works .. it also works with sub-strings like *Compu* ... I love it !!

Now can I refine it any further using regexp expressions like *[Cc]ompu* perhaps ???

You can't use regular expressions or REX, but you can give multiple wildcard expressions per category, separated by spaces. Eg.:

*compu* *Compu*

Thanks Kai and John ... I think I'll take a look at Vortex scripting and see if I can get my head around using a regexp ... I need to know how the cosding works anyway, so now's the time to get stuck in ...

OTOH - I also discovered the "ignore case in URLs" option. Setting that to "yes" gives me the equivalent of a capitalisation regexp ... i.e. *xeus* matches *Xeus* and *XEUS* as well ...

Next step .. designing treed categories

xeus -- xmm -- rosita >> Xray-Spectrum >>--->>
GAIA --hipparcos -- hubble >> Visual Spectrum >>--->> Astrophysics Division
Planck -- Herschel >> Microwave Spectrum >>--->>

Permitting me to offer -

Search this Mission
Search this Frequency Band
Search the Division
Search Whole Site

Are there any other expressions that can be used in the Category field that aren't in the documentation like
*/?ompu*. It is a great option to be able to use a question mark in this example are there others?

Sorry, no. We might normally use a rex expression for things like that but it was simplified for the masses. An adventurous person could modify the dowalk script functions findcat and recat to use rex.

Hmmm ... I return ...

I'm reworking our site as a database-backend, PHP-driven system, so my urls are all turning into things like http://www.rssd.esa.int/index.php?ptoject=TOP&page=pubs ... and now I have a new version of the old problem.

It would seem that the search is not correctly categorising urls with a query string. It will handle things like *compu* for all urls which contain "computing" **in the main url string** but not where it occurs in the query string.

Is there a trick, or do I need to REALLY learn Vortex scripting this time ???

Thanks in advance ...

The entire url is considered when looking for category matches. Make sure "Strip queries" is off.