| Author |
Message |
|
|
Just one question concerning the dump...
If the dump is meant to evolve, wouldn't it be easier if it used some XML format, that is both more readable and backward-compatible ?
This is purely a suggestion, the most important thing is that we have the data in whatever format
Sami Dalouche
|
 |
|
|
Hi,
there's one of these weird escape characters(ESC) in alternate name id #1210278
1210278 2578103 AïneESC Djahfa
thanks
Sami
|
 |
|
|
|
A 4 character-long code can also be "post", for postal codes
|
 |
|
|
Hmm.. actually, I was wrong.. the file does contain everything...
So, please excuse my mistake
Sami
|
 |
|
|
Hi,
I am wondering whether the alternate names dump is working correctly :
http://download.geonames.org/export/dump/alternateNames.zip
Indeed, the dump does not seem to contain :
- the "isOfficial" column
- the misc. additional information, such as the postcodes, that we can find in the incremental updates (http://download.geonames.org/export/dump/alternateNamesModifications-2007-04-21.txt)
Is there something I am missing ? or is it really the dump that is broken ?
Thanks,
Sami Dalouche
|
 |
|
|
marc wrote:
Hi Sami
I have added a comment header to the file countryInfo.txt : http://download.geonames.org/export/dump/countryInfo.txt
A header with explanation is required anyhow and it could be sufficient. There are only a handful of countries with special currencies. I don't know how much sense it makes to automatically parse them as every application has to decide what to do with these special cases.
Marc
I agree this is currently the best thing to do.
Just one question : what is the difference between http://geonames.cvs.sourceforge.net/geonames/data/countries.txt?revision=1.7&view=markup
and http://download.geonames.org/export/dump/countryInfo.txt ?
The file on CVS seems to have more fields....
|
 |
|
|
marc wrote:
Hi Isa
The number of alternate names per language is unlimited. For Paris we have alone four alternate names in French : Paris, Lutece, Pantruche and Paname. Only the number of 'isOfficial/isPreferred' is limited to one.
For the postal codes in Paris it is a little bit more complicated, because we have an entry in the geonames database for each of the 20 Arrondissements. The postal codes 75001-75020 are assigned to the corresponding Arrondissement.
Regards
Marc
Hi,
is it possible to easily query for all arrondissements of Paris, for instance ?
Thanks,
Sami
|
 |
|
|
the advantage I see for the additional column is that it can be automatically processed, whereas the comments need to be processed by someone, etc...
On the other hand, I agree that these columns (equivalentFipsCode and this one) are not generic enough to be there....
An example of where comments can be problematic is for your incremental updates, for instance... for deletions, comments are written in english (duplicate of #..), which means that is is impossible to take any automatic action regarding the deletion. I would tend to think we should not have that kind of stuff everywhere...
So, something different that could be possible would be to prefix the iso codes by something that would describe if it is ISO or not :
iso:GGP
noniso:JEP
What do you think ?
|
 |
|
|
You mean "currency code" instead of "country code", right ?
I'm not really sure of the best approach... because saying that guernsay uses GBP is not 100% true since they issue their own notes, and banks seem to consider GGP (and JEP, since jersey is in the same case) as different currencies... But at the same time, entering a non ISO 4217 currency code is kind of awkward and error-prone...
BTW, in the file, it looks like Jersey is associated to "GPD", which doesn't seem to exist.. the british pound is GBP, and Jersey Pound non-official code is JEP..
So, basically, we have 6 countries that are in the same case, and whatever we choose, we must be consistent... So we can :
1] put the official iso code for these 6 territories,
2] set no currency for these countries
3] add the non ISO code as a separate column.
I would tend to think solution 3) would the best, but what do you think ?
----------------
http://en.wikipedia.org/wiki/ISO_4217 :
A number of territories are not included in ISO 4217, because their currencies are not per se an independent currency, but a variant of another currency. These currencies are:
1. Faroese króna (1:1 pegged to the Danish krone)
2. Guernsey pound (1:1 pegged to the pound sterling)
3. Jersey pound (1:1 pegged to the pound sterling)
4. Isle of Man pound (1:1 pegged to the pound sterling)
5. Tuvaluan dollar (1:1 pegged to the Australian dollar).
6. Cook Islands dollar (1:1 pegged to the New Zealand dollar).
The following non-ISO codes are, however, sometimes used: "GGP" for the Guernsey pound", "JEP" for the Jersey pound and "IMP" for the Isle of Man pound [1].
|
 |
|
|
And yet another correction :
http://www.iso.org/iso/en/prods-services/popstds/currencycodeslist.html
Mozambique's currency is MZN, not MZM.
corrected
|
 |
|
|
I also forgot :
Serbia => RSD and not CSD
Serbia and Montenegro => RSD
Montenegro => EUR
(http://en.wikipedia.org/wiki/Serbia_and_Montenegro ,
http://www.iso.org/iso/en/prods-services/popstds/currencycodeslist.html
)
These are also corrected in the file
|
 |
|
|
Oh, and another correction I added in the file is for Slovenia.
Slovenia doesn't use the Tholar anymore, since they adopted the Euro since 2007.. (http://en.wikipedia.org/wiki/Slovenia)
|
 |
|
|
Here is a modified version of your countries files, with a few corrections :
http://www.sirika.com/data/geonames/geonamesCountries.20070407.txt
Corrections :
- French Southern Territories
http://fr.wikipedia.org/wiki/Terres_australes_et_antarctiques_fran%C3%A7aises
> La monnaie officielle des TAAF est l'euro (et non le franc Pacifique utilisé dans > les collectivités d'outre-mer du Pacifique, bien que la Terre Adélie soit plus
> proche de l'Australie et la Nouvelle Calédonie que de la Réunion) ;
> c'était le franc français avant 1999.
The currency was not set, so it's now set to EUR
Guernsey currency :
http://en.wikipedia.org/wiki/Guernsey_pound
GPD => GGP
Regards,
Sami Dalouche
|
 |
|
|
Ok, another problem.. this time, it's the alternate Name #1210278 : AïneESC Djahfa that has this weird ESC character...
Hopefully, this is the last one I'll discover
Regards,
Sami Dalouche
|
 |
|
|
Oh, Another feature that has a weird character : 4818789
this one contains the Ctrl-C sequence
Regards,
Sami Dalouche
|
 |
|
|