Fix a severe GPX parse slowdown and add route (<rte>) support
CI / check (push) Successful in 43s
CI / deploy (push) Successful in 16s

GPX tracks were silently failing to render for large real-world files:
point_extensions used REXML's "extensions//*" XPath, and the // descendant
axis costs low-single-digit milliseconds per call regardless of subtree
size. Run once per point, that's negligible on small files but ~30s on a
real 8000-point track — long enough to hit a proxy timeout in production,
so the file just never appeared with no error. Replaced with plain
recursive child traversal (~40x faster on the same file, confirmed
29.9s -> 3.16s); test_parsing_many_points_with_extensions_stays_fast
guards against reintroducing it.

Also adds support for <rte> routes (not just <trk> tracks and <wpt>
waypoints), sharing the track/route conversion logic via a new
line_feature helper. Routes get their own "route" kind and a dashed
ROUTE_ICON in the file tree so the type is visible at a glance.
This commit is contained in:
randy committed 2026-09-07 14:36:12 +02:00
1 parent 0394befdec
commit 708c08259b
6 files changed
+149 -19

No files matched your search

+62 -15
View File
@@ -4,9 +4,9 @@ require_relative "track_stats"
module Maptool
# Converts a GPX file to a GeoJSON FeatureCollection: waypoints (<wpt>)
# become Point features, tracks (<trk>) become MultiLineString features
# (one line per <trkseg>). Routes (<rte>) are not handled — GPX in
# practice is either points or recorded tracks for this app.
# become Point features, tracks (<trk>) and routes (<rte>) both become
# MultiLineString features — "kind" (=> "track"/"route") is the only
# difference in shape, so the file tree can show which one a file is.
module Gpx
class ParseError < StandardError; end
@@ -20,6 +20,7 @@ module Maptool
kinds = []
kinds << "waypoint" if content.include?("<wpt")
kinds << "track" if content.include?("<trk")
kinds << "route" if content.include?("<rte")
kinds
end
@@ -30,6 +31,7 @@ module Maptool
features = []
root.each_element("wpt") { |wpt| features << waypoint_feature(wpt) }
root.each_element("trk") { |trk| features << track_feature(trk) }
root.each_element("rte") { |rte| features << route_feature(rte) }
{ "type" => "FeatureCollection", "features" => features }
rescue REXML::ParseException => e
@@ -58,43 +60,88 @@ module Maptool
trk.each_element("trkseg") do |seg|
points = []
seg.each_element("trkpt") do |pt|
point = coordinate(pt)
point, time, extra = process_point(pt)
points << point
all_points << point
times << parse_time(child_text(pt, "time"))
extras << point_extensions(pt)
times << time
extras << extra
end
lines << points unless points.empty?
end
line_feature(kind: "track", name: child_text(trk, "name"), activity_type: child_text(trk, "type"),
lines: lines, all_points: all_points, times: times, extras: extras)
end
private_class_method :track_feature
# Structurally simpler than a track: <rte> holds <rtept> directly, no
# <trkseg>-style grouping — a route is one continuous line by
# definition, planned rather than recorded.
def self.route_feature(rte)
points = []
times = []
extras = []
rte.each_element("rtept") do |pt|
point, time, extra = process_point(pt)
points << point
times << time
extras << extra
end
line_feature(kind: "route", name: child_text(rte, "name"), activity_type: nil,
lines: points.empty? ? [] : [points], all_points: points, times: times, extras: extras)
end
private_class_method :route_feature
def self.line_feature(kind:, name:, activity_type:, lines:, all_points:, times:, extras:)
{
"type" => "Feature",
"geometry" => { "type" => "MultiLineString", "coordinates" => lines },
"properties" => { "kind" => "track", "name" => child_text(trk, "name"), "activity_type" => child_text(trk, "type") }
"properties" => { "kind" => kind, "name" => name, "activity_type" => activity_type }
.merge(TrackStats.compute(all_points, times: times))
.merge("profile" => TrackStats.build_profile(all_points, times: times, extras: extras)),
}
end
private_class_method :track_feature
private_class_method :line_feature
def self.process_point(pt)
[coordinate(pt), parse_time(child_text(pt, "time")), point_extensions(pt)]
end
private_class_method :process_point
# Heart rate/cadence/speed live in a device-specific <extensions> block
# (usually Garmin's own TrackPointExtension) — matched by local element
# name only (`Element#name` already strips the namespace prefix), so
# this doesn't care which extension schema a given file actually
# declares, as long as it names its elements the same way Garmin's does.
#
# Walked by hand (elements.each with no argument — plain child
# iteration) rather than the `"extensions//*"` XPath this used
# originally: REXML's `//` (descendant) axis turned out to cost
# milliseconds *per call* regardless of how small the subtree actually
# is, and this runs once per point — fine on a short test file, ~30s
# for an 8000-point real one. Plain traversal is ~40x faster on that
# same file.
def self.point_extensions(point_el)
values = {}
point_el.elements.each("extensions//*") do |el|
case el.name
when "hr" then values[:heart_rate_bpm] = el.text&.strip&.to_i
when "cad" then values[:cadence_rpm] = el.text&.strip&.to_i
when "speed" then values[:speed_mps] = el.text&.strip&.to_f
end
end
extensions = point_el.elements["extensions"]
collect_extension_values(extensions, values) if extensions
values
end
private_class_method :point_extensions
def self.collect_extension_values(el, values)
el.elements.each do |child|
case child.name
when "hr" then values[:heart_rate_bpm] = child.text&.strip&.to_i
when "cad" then values[:cadence_rpm] = child.text&.strip&.to_i
when "speed" then values[:speed_mps] = child.text&.strip&.to_f
end
collect_extension_values(child, values)
end
end
private_class_method :collect_extension_values
def self.parse_time(text)
text && Time.parse(text)
rescue ArgumentError