dsoprea / Geonamesrdf

Introduction

This is a client library for the Geonames <http://www.geonames.org>_ RDF-compatible web-services <http://www.geonames.org/export/ws-overview.html>. At this time, only the Search API <http://www.geonames.org/export/geonames-search.html> provides RDF responses and is, therefore, the only adapter available.

RDF is often used in association with the Semantic Web <http://www.w3.org/standards/semanticweb>_ and/or the IoT <https://en.wikipedia.org/wiki/Internet_of_Things>_.

Requirements

• requests
• Python 2.x or 3.x

Installation

Via PyPI::

$ sudo pip install geonames_rdf

Usage

As a library::

import geonames.adapters.search

_USERNAME = 'your_username'
sa = geonames.adapters.search.Search(_USERNAME)

result = sa.query('detroit').country('us').max_rows(5).execute()
for id_, name in result.get_flat_results():
    # make_unicode() is only used here for Python version-compatibility.
    print(geonames.compat.make_unicode("[{0}]: [{1}]").format(id_, name))

A result object will be returned from the execute() method. The following are the ways to read the results (get_flat_results is used above):

• The raw lxml object is exposed through the xml property.
• The get_xml_nodes method will return a list of Feature lxml.etree objects from the resulting document, representing each of the results.
• The get_flat_results method will simply yields a list of (id, name) 2-tuples if you're not interested in anything else. This is also what's returned, by default, from the command-line utility.

As a command-line utility::

$ gn_search dsoprea -p query detroit -p country us -p max_rows 5
[http://sws.geonames.org/4990729/]: [Detroit]
[http://sws.geonames.org/6955112/]: [Detroit-Warren-Livonia]
[http://sws.geonames.org/5024238/]: [Detroit Lakes Airport]
[http://sws.geonames.org/4990733/]: [Detroit City Airport]
[http://sws.geonames.org/4990742/]: [Detroit Metropolitan Wayne County Airport]

You may only provide string (non-boolean) parameters to the command-line utility.

NOTE: You can also get the complete RDF response by passing the "-x" parameter.

Whether you're using the library or the tool, the parameters that are available are:

========================= ===================== Library Parameter Name API Parameter Name ========================= ===================== query q place_name_like name place_name_equals name_equals place_name_starts_with name_startsWith max_rows maxRows start_row startRow country country show_country_first countryBias continent continentCode admin_code1 adminCode1 admin_code2 adminCode2 admin_code3 adminCode3 feature_class featureClass feature_code featureCode population_class cities language lang verbosity style force_matching_name isNameRequired tag tag operator operator charset charset fuzzy fuzzy east_west_north_south_box east,west,north,south place_name_language searchlang order_by orderby ========================= =====================

Some of the parameters can be specified more than once. See the API documentation for more information.

Design Notes

• Note that most of the parameters in the library are named differently from the API. Normally this goes against my policy, but I found so many of the parameters to be ambiguous or cryptic that I renamed those parameters to be clearer and just renamed the rest to underscore-separated naming while I was at it.

• We mostly rely on the API to correctly validate against bad parameters. We don't like arbitrarily validating parameters in the client since the rules may change in the server and it's one more thing that might interfere with your implementation. However, if there is a type of validation that, for one reason or another, makes sense and/or would make things easier to develop against, let me know.