Sanitizing GPX information for public sharing – Ole Begemann

August 27, 2022

1

GPX is a well-liked XML format for working or biking tracks with geocoordinates. It is a how-to for cleansing up a GPX file by eradicating undesirable or privacy-sensitive data.

Many apps that report exercise routes and might export them as GPX information embrace extra knowledge than the plain GPS coordinates. As an example, a GPX file from my favourite recording app, Guru Maps, appears like this:

<?xml model="1.0" encoding="utf-8"?>
<gpx model="1.1" creator="Guru Maps/4.5.2" 
  xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" 
  xmlns="http://www.topografix.com/GPX/1/1" 
  xmlns:gom="https://gurumaps.app/gpx/v2" 
  xsi:schemaLocation="http://www.topografix.com/GPX/1/1 http://www.topografix.com/GPX/1/1/gpx.xsd https://gurumaps.app/gpx/v2 https://gurumaps.app/gpx/v2/schema.xsd">
  <trk>
    <title>Barnimer Dörferweg</title>
    <kind>TrackStyle_FF7F00C8</kind>
    <trkseg>
      <trkpt lat="52.6254614634" lon="13.4092010169">
        <ele>54.238586451</ele>
        <time>2020-05-10T05:30:38.997Z</time>
        <hdop>4.6875</hdop>
        <vdop>3.375</vdop>
        <extensions>
          <gom:pace>5.5661926282</gom:pace>
          <gom:course>329.1938658731</gom:course>
        </extensions>
      </trkpt>
      …
      <!-- 1000's of observe factors -->

This observe contains the next properties for every observe level:

Geocoordinates (latitude and longitude)
Elevation
Timestamp
Horizontal and vertical dilution of precision (hdop/vdop)
Present pace
Present course/heading

Plus, Guru Maps makes use of the observe’s <kind> attribute to encode the colour of the observe as displayed within the app in a non-standardized format (TrackStyle_FF7F00C8).

Some apps additionally embrace coronary heart charge or different health measurements.

All this knowledge is helpful for archiving tracks or importing them into one other app. However earlier than sharing this observe publicly, I’d need to clear the information up first:

The one actually necessary items of knowledge are the coordinates and presumably the elevation.
Timestamps are non-public knowledge. I don’t need to share these.
The opposite measurements are largely irrelevant.

GPX information can develop into fairly massive (1000’s of observe factors is widespread), so lowering the quantity of information can be good for file sizes and parsing efficiency.

Necessities

XmlStarlet

I exploit Xml to do many of the XML processing. On macOS, you possibly can set up XMLStarlet through Homebrew:
xmllint

One optionally available processing step makes use of xmllint, which comes preinstalled on macOS.
XSLT file for eradicating unused namespaces

Lastly, obtain this XSLT file remove-unused-namespaces.xslt, both from this Gist or from my server. We’re gonna use it in a single processing step to strip unused namespaces from the GPX file.

Authentic supply: Dimitre Novatchev on Stack Overflow.

Operating the command

Assuming your supply file is known as enter.gpx and the XSLT file you downloaded above is within the present listing, that is the total command to course of the GPX file and save the end result to output.gpx:

xmlstarlet ed 
  -d "//_:extensions" 
  -d "/_:gpx/_:metadata/_:time" 
  -d "/_:gpx/_:trk/_:kind" 
  -d "//_:trkpt/_:time" 
  -d "//_:trkpt/_:hdop" 
  -d "//_:trkpt/_:vdop" 
  -d "//_:trkpt/_:pdop" 
  -u "/_:gpx/@creator" -v "Shell script" 
  enter.gpx 
  | xmlstarlet tr remove-unused-namespaces.xslt - 
  | xmlstarlet ed -u "/_:gpx/@xsi:schemaLocation" -v "http://www.topografix.com/GPX/1/1 http://www.topografix.com/GPX/1/1/gpx.xsd" 
  | xmllint --c14n11 --pretty 2 - 
  > output.gpx

This sequence performs the next steps:

Delete all <extensions> components.
Delete the timestamp from the file’s <metadata> part if current.
Delete the <trk><kind> component.
Delete the <time>, <hdop>, <vdop>, and <pdop> components from all observe factors.
Set the file’s creator attribute.
Now that extension fields are gone, take away all unused XML namespaces from the file header.
Delete all xsi:schemaLocation entries besides the one for the GPX schema.
Run the file by xmllint for formatting. The --c14n11 choice performs XML Canonicalization (C14N). Amongst many different issues, canonicalization replaces numeric character entities within the XML with their regular Unicode characters, which is necessary for my use case.

For instance, the textual content “Dörferweg” within the supply would develop into “Dörferweg”. I discovered that a number of the instruments I exploit insert non-ASCII characters as numeric codes and different instruments don’t show these accurately.

The processed GPX file appears like this:

<gpx xmlns="http://www.topografix.com/GPX/1/1" 
  xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" 
  creator="Shell script" model="1.1" 
  xsi:schemaLocation="http://www.topografix.com/GPX/1/1 http://www.topografix.com/GPX/1/1/gpx.xsd">
  <trk>
    <title>Barnimer Dörferweg</title>
    <trkseg>
      <trkpt lat="52.6254614634" lon="13.4092010169">
        <ele>54.238586451</ele>
      </trkpt>
      <trkpt lat="52.6255090307" lon="13.4091548326">
        <ele>53.9600219977</ele>
      </trkpt>
      …

The processing steps above are those that work for me given the apps I exploit. Your mileage might differ in case your instruments add different knowledge to your GPX information. Be happy to edit the command accordingly. XmlStarlet makes use of XPath syntax to pick out which components to function on. The xmlstarlet sel command is helpful for inspecting a supply file and making an attempt out the required XPath incantations.

Validation

Lastly, it’s a good suggestion to validate the processed GPX file towards the official GPX schema:

xmlstarlet val --quiet --err --xsd 
  http://www.topografix.com/GPX/1/1/gpx.xsd 
  output.gpx

Comfortable processing!

PS: If you happen to’re ever in Berlin, this can be a good lengthy bike route (55 km) with minimal automotive visitors. Begins and ends at Hauptbahnhof. Obtain the (sanitized) GPX file.