Transform scheduled stop point IDs before importing into OTP (#231)

* Implement XSLT for transforming ScheduledStopPoints

* Add documentation

* Add reuse headers

* Run XLST during build

* Use specific Saxon directory

* Update input file name

* use correct file name

* Exclude Amarillo feed

* Add artifact

* Move transformed file into data

* Add SIRI-Lite feeds

* Add test for EntityResolver

* Remove separate script

* Use un-released version to use scheduled stop points

* Re-enable Amarillo

* Bump actions version

.github/workflows/main.yml

@@ -140,6 +140,13 @@ jobs:
         # the script uses error handling with +e, so please don't run it with bash prepended in this line
         run: ./build-graph.sh
 
+
+      - name: Store artifact for sta.netex.correct-ssids.xml
+        uses: actions/upload-artifact@v4
+        with:
+          name: sta-netex-correct-ssids
+          path: data/sta.netex.correct-ssids.xml
+
       - name: Build and push
         if: github.repository_owner == 'noi-techpark'
         uses: noi-techpark/github-actions/docker-build-and-push@v2

.gitignore

@@ -11,3 +11,7 @@ journey/tmp
 geocoder/.env
 
 graph.obj
+
+*.xml
+*.zip
+saxon

.otp-version

@@ -1,4 +1,4 @@
 # SPDX-FileCopyrightText: NOI Techpark <[email protected]>
 #
 # SPDX-License-Identifier: CC0-1.0
-export OTP_IMAGE="opentripplanner/opentripplanner:2.7.0_2024-11-19T11-28"
+export OTP_IMAGE="lehrenfried/opentripplanner:siri-scheduled-stop-point"

build-graph.sh

@@ -22,6 +22,15 @@ TRANSIT_NETEX_URL="https://rapuser:[email protected]/netex/api/v4/download
 TRANSIT_NETEX_XML=data/sta-netex.xml
 TRANSIT_NETEX_GZ=${TRANSIT_NETEX_XML}.gz
 TRANSIT_NETEX_ZIP=${TRANSIT_NETEX_XML}.zip
+
+# config for transforming the ids of scheduled stop points
+SAXON_URL="https://github.com/Saxonica/Saxon-HE/releases/download/SaxonHE12-5/SaxonHE12-5J.zip"
+SAXON_ZIP="saxon.zip"
+SAXON_JAR="saxon/saxon-he-12.5.jar"
+XSL_FILE="transform-scheduled-stop-point-ids.xsl"
+
+SSIDS_TRANSFORMED_XML="data/sta.netex.correct-ssids.xml"
+
 # parking
 PARKING_NETEX_URL=https://transmodel.api.opendatahub.com/netex/parking
 PARKING_NETEX_XML=data/shared-data.xml
@@ -55,7 +64,24 @@ rm -f ${TRANSIT_NETEX_ZIP}
 echo "Downloading NeTEx transit data from ${TRANSIT_NETEX_URL}"
 ${CURL} "${TRANSIT_NETEX_URL}" -o ${TRANSIT_NETEX_GZ}
 gunzip --force ${TRANSIT_NETEX_GZ}
-zip ${TRANSIT_NETEX_ZIP} ${TRANSIT_NETEX_XML}
+
+# Configuration
+
+if [ ! -f "${SAXON_JAR}" ]; then
+  $CURL $SAXON_URL -o $SAXON_ZIP
+  unzip $SAXON_ZIP -d saxon
+fi
+
+# the scheduled stop point ids and the SIRI StopPointRefs do not match, so we have to transform
+# the NeTEx feed so that they do: https://github.com/noi-techpark/odh-mentor-otp/issues/215
+echo "Running Saxon transformation..."
+java -jar "$SAXON_JAR" -s:"${TRANSIT_NETEX_XML}" -xsl:"$XSL_FILE" -o:"$SSIDS_TRANSFORMED_XML"
+
+
+zip ${TRANSIT_NETEX_ZIP} ${SSIDS_TRANSFORMED_XML}
+
+ls -lah
+ls -lah data
 
 # download parking data and put it into a zip
 rm -f ${PARKING_NETEX_XML} ${PARKING_NETEX_ZIP}

infrastructure/docker/otp/Dockerfile

@@ -1,5 +1,5 @@
 # Simon, do you know how we could use the value from .otp-version here?
-FROM opentripplanner/opentripplanner:2.7.0_2024-11-19T11-28
+FROM lehrenfried/opentripplanner:siri-scheduled-stop-point
 
 WORKDIR /var/otp
 

router-config.json

@@ -78,9 +78,15 @@
   },
   "updaters": [
     {
-      "type": "stop-time-updater",
-      "url": "https://mobility.api.opendatahub.testingmachine.eu/v2/GTFS-RT",
-      "feedId": "sta"
+      "type": "siri-et-lite",
+      "feedId": "sta",
+      "url": "https://efa.sta.bz.it/siri-lite/estimated-timetable/xml",
+      "fuzzyTripMatching": true
+    },
+    {
+      "type": "siri-sx-lite",
+      "feedId": "sta",
+      "url": "https://efa.sta.bz.it/siri-lite/situation-exchange/xml"
     },
     {
       "type" : "vehicle-rental",

transform-scheduled-stop-point-ids.xsl

@@ -0,0 +1,47 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!--
+SPDX-FileCopyrightText: NOI Techpark <[email protected]>
+
+SPDX-License-Identifier: CC0-1.0
+-->
+<xsl:stylesheet version="3.0"
+  xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
+  xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+  xmlns:n="http://www.netex.org.uk/netex"
+  xsi:schemaLocation="http://www.netex.org.uk/netex https://raw.githubusercontent.com/NeTEx-CEN/NeTEx/master/xsd/NeTEx_publication.xsd"
+  exclude-result-prefixes="n"  >
+
+  <!-- Identity template to copy every node and attribute by default -->
+  <xsl:mode on-no-match="shallow-copy"/>
+
+  <!--
+    Replace the IDs of ScheduledStopPoints from this format
+
+      it:apb:ScheduledStopPoint:it-22101-7010-51-32073:
+
+    to this
+
+      it:22101:7010:51:32073
+
+    .
+
+    This because the SIRI feeds use the latter format, and we need to match up the two sources.
+  -->
+  <xsl:template match="//n:ScheduledStopPoint/@id">
+    <xsl:attribute name="id">
+      <xsl:value-of select="replace(replace(., '.*ScheduledStopPoint:(.+?):$', '$1'), '-', ':')"/>
+    </xsl:attribute>
+  </xsl:template>
+
+  <xsl:template match="//n:RoutePointRef/@ref">
+    <xsl:attribute name="ref">
+      <xsl:value-of select="replace(replace(., '.*ScheduledStopPoint:(.+?):$', '$1'), '-', ':')"/>
+    </xsl:attribute>
+  </xsl:template>
+
+  <xsl:template match="//n:ScheduledStopPointRef/@ref">
+    <xsl:attribute name="ref">
+      <xsl:value-of select="replace(replace(., '.*ScheduledStopPoint:(.+?):$', '$1'), '-', ':')"/>
+    </xsl:attribute>
+  </xsl:template>
+</xsl:stylesheet>