[Ekiga-list] Implementing ekiga-callto (plain number management)




This is a discussion started in http://bugzilla.gnome.org/show_bug.cgi?id=576486. Making a try to move discussion to the mailing list. Please read the bug to get context, specifically the "stories" attachment.

The basic idea is that users should be able to use ordinary phone numbers when using Ekiga.

> Snark skrev:
There are several parts in all of this :
- what can the user enter ?
User can enter anything which would be handled by a normal phone. User can also enter a number with formatting characters like in '+46 (0)70-532 11 22', typically when pasting from documents. User can also enter a foreign number, meant to be evaluated outside user's country, but in this case user might have to edit the result before using it.

"User" could also mean other programs (e. g., Firefox) invoking ekiga through the CLI interface. This situation is more or less the same, besides that there will no be formatting characters in this case.

Another "user" is the address book, where users store numbers in all sorts of ways, often without countrycode similar to how numbers are entered interactive.
- what can the user modify after entering ?
I don't really understand what you mean...
- what does ekiga understand.
Well, in the end Ekiga only understands a complete url on the form sip:+1234567899 sip provider net My code makes the transformations from the forms mentioned above to this form.
I'm not sure your code handles everything... isn't it only for '
US-based phone numbers? Different countries have different schemes,
> have a look there :
http://en.wikipedia.org/wiki/Local_conventions_for_writing_telephone_numbers
No, my code is designed to handle anything, and it seems like it does as well. My code can handle even odd things like local prefixes of more than one digit etc. There is a table in the very heart, a text file, which miss some entries for small islands/countries/territories but should be complete otherwise. You can try the ekiga-callto package, it's python but uses the same logic. Or you can compile my package, it compiles to a test binary which makes CLI conversions, see the README file.

Alas, I'm not handling missing area codes. I don't think this is a real problem, though.

The URL you mention is a little bit off-topic. The rewriting rules are more strict, the way people actually write their numbers is much more relaxed - and the list is far from complete. Fortunately, it doesn't really matter. To get a feeling what the rewriting rules are, you might be interested in http://www.kropla.com/dialcode.htm.

I'd rather have a discussion on the ekiga-devel mailing list, since I'm not that familiar with what opal can handle : it would be better if others could step in!

Now, we are here! But this is *not* about changing the interface to OPAL. It's about preprocessing numbers given by users to the form we already use. OPAL should not see any difference.

> Snark

--alec


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]