convert

Convert.py

The convert.py module is part of the osm_fieldwork package and provides functionality for converting ODK forms to OSM XML using a YAML configuration file.

Even if an XLSForm is carefully designed to have a one to one match between ODK and OSM, this is not always possible, as not all survey questions are for OSM.

The Config File Sections

There are several sections the config file. The default one is called xforms.yaml, and is included in the sources and the python package. It is possible to use a different config file.

convert

This section supports one to one conversion of tags, as well as one to many. This example shows all poossible conversion types. The simple ones like altitude just change the tag, and the value is used unchanged. A more complicated conversion is changing the value in addition to the tag. Anything with an equals sign is split into the appropriate tag and value for OSM. The final one is where a singe survey question creates multiple tahg and value pairs, deliminated by a comma. Each of the pairs is handled as a separate tag and value in OSM. convert: - latitude: lat - longitude: lon - altitude: ele - cemetery_services: - cemetery: amenity=grave_yard - cremation: amenity=crematorium - amenity: - coffee: amenity=cafe,cuisine=coffee_shop ...

private

Not all collected data is suitable for OSM. This may include data that has no equivalant tag in OSM, or personal data.

private:
 - income
 - age
 - gender
 - education

ignore

ODK supports many tags useful only internally. These go into the ignore section of the config file. Any tag in this section gets removed from from all output files. An example would be this:

ignore:
 - attachmentsexpected
 - attachmentspresent
 - reviewstate
 - edits
 - gps_type
 - accuracy
 - deviceid
...

multiple

Not all survey questions have a single answer. Anything using select_multiple may have more than one value. As the default assumes one answer per question, this specifies the questions with multiple answers since they have to be processed seperately. The normal conversion process is applied to these too.

multiple:
 - healthcare
 - amenity_type
 - specialty