Changes

Jump to: navigation, search

Addon:PlaceCleanupGramplet

436 bytes added, 13:23, 3 October 2018
Faq
== Faq ==
* GeoNames data is primarily intended for present day usage. It contains little historical information. I have heard that it is pretty complete for United States, Canada, and less so for Europe, although it seems to have all the European cities I have tested so far. If you find an error or omission in the GeoNames data, you may want to tell GeoNames about it [http://www.geonames.org/manual.html| http://www.geonames.org/manual.html].
* GeoNames finds multiple places: Example, 'Santa Rosa, CA'. GeoNames finds both 'Santa Rosa, Sonoma, California, United States' and 'City or of Santa Rosa, Sonoma, California, United States'. The way the GeoNames database is structured, they have a 'Populated Place', which generally has lots of Alternate names and postal codes, and an 'Administrative subdivision' for the same place. The 'City of' version is the Administrative subdivision, which generally doesn't have postal codes and alternate names. I personally prefer the first version, which should appear first in the 'Choices' list.
* Timeouts: When using the GeoNames web database with a free account, you will sometimes get a timeout. The Timeout is set to 20 seconds for each request, unfortunately some actions take several requests, so the user may see unresponsive Gramps for up to 40 seconds. You can always try the request again by pressing the last used button. According to the GeoNames web site, if you pay for an account, you may get better performance, as the paid accounts use more lightly loaded servers.
* Not Found: While the GeoNames database contains some historical names, it mostly tries to be up to date with the current times. Many places in a typical family tree are described as of the time the event occurred. So it is entirely possible that the described place no longer exists. It may now belong to a different country, or other administrative subdivision. Or it may have subsumed with another larger place. Initially you should try removing the intermediate place data, leaving only the initial segment and the country. You may get a lot of matches, but that may give you a better clue as to the current situation. Or you may have to do some research outside the Gramplet to figure out what it might be called now.
: If the structure is not always known or used, then the algorithm attempts to assign place types by looking at the place name. For example if the place name includes a word that matches a place type, such as "Harris County" or "Smith Township", then the place type is set accordingly. To do this, all the various place alternate names are scanned, in case the primary name does not include a suitable word.
: If you know that a particular country has a rigid (or even mostly rigid) hierarchical structure, and this addon seems to be getting it wrong a lot, please let the author know at paulr2787 at gmail.com. I can modify the code to make it work better if I understand the correct country structure.
* IDs: When GeoNames is the source of place data, the normal place ID ('P0001') is replaced by a GeoNames specific ID ('GEO12345'). The number portion of this ID is the GeoNames identifier for the place. By using this type of ID, the addon can easily identify when enclosing places are already present in your database, which means that the user doesn't have to do a local search for and selection of each level of enclosing place.
 
== Notes ==
This is a new tool and fairly complex. Please report any issues or desired enhancements to paulr2787 at gmail.com
I have included a French translation of the tool, mostly as a test of the internationalization code. Since I don't speak French (USA Texas version of English only), please excuse my French.
[[Category:Addons]]
266
edits

Navigation menu