From b24cbbe22c6e1a4cf0d9d95cf4c69ee796665975 Mon Sep 17 00:00:00 2001
From: Holger Bruch <hb@mfdz.de>
Date: Tue, 30 Jan 2024 10:21:32 +0100
Subject: [PATCH] work around synthetic parent station ids

see https://github.com/mfdz/GTFS-Issues/issues/141
---
 makefile                            |  2 +
 scripts/patch_nvbw_station_stops.py | 66 +++++++++++++++++++++++++++++
 2 files changed, 68 insertions(+)
 create mode 100644 scripts/patch_nvbw_station_stops.py

diff --git a/makefile b/makefile
index 98f0388..f3d10fb 100644
--- a/makefile
+++ b/makefile
@@ -102,6 +102,8 @@ data/gtfs/%.merged.with_flex.gtfs: data/gtfs/%.merged.gtfs.zip
 	$(info unzipping $* GTFS feed)
 	rm -rf $@
 	unzip -d $@ $<
+	$(info patching station stop_ids referenced in stoptimes)
+	python3 scripts/patch_nvbw_station_stops.py $@  
 	$(info patching GTFS-Flex data into the GTFS feed)
 	# todo: pick flex rules file based on GTFS feed
 	docker run -i --rm -v $(HOST_MOUNT)/data/gtfs/$(@F):/gtfs derhuerst/generate-gtfs-flex:4 stadtnavi-herrenberg-flex-rules.js
diff --git a/scripts/patch_nvbw_station_stops.py b/scripts/patch_nvbw_station_stops.py
new file mode 100644
index 0000000..7a819b0
--- /dev/null
+++ b/scripts/patch_nvbw_station_stops.py
@@ -0,0 +1,66 @@
+# This script patches stations referenced as stops by 
+# adding a new stop with `:na:na` appended to the id and
+# replacing the original station's ID in stop_times.txt
+# by this new stop ID.
+# Note: This script is a quick and dirty hack and no
+# general purpose patch script. I.e. GTFS feeds with transfers
+# might need additional changes.
+
+# Background: this fix is necssary
+
+
+import csv
+import os
+from os.path import join
+import sys
+
+gtfs_dir = sys.argv[1] if len(sys.argv) >= 2 else ''
+
+original_stops_file_path = join(gtfs_dir, 'stops.txt')
+new_stops_file_path = join(gtfs_dir, 'new_stops.txt')
+original_stop_times_file_path = join(gtfs_dir, 'stop_times.txt')
+new_stop_times_file_path= join(gtfs_dir, 'new_stop_times.txt')
+
+with open(original_stops_file_path) as csvfile:
+
+	reader = csv.DictReader(csvfile)
+	stations = {}
+	for row in reader:
+		if row['location_type'] == '1':
+			stations[row['stop_id']] = row
+
+renamed_stop_ids = {}
+# check if they are references by stop_times
+with open(original_stop_times_file_path) as csvfile:
+	with open(new_stop_times_file_path, 'w') as new_stops_file:
+
+		reader = csv.DictReader(csvfile)
+		writer = csv.DictWriter(new_stops_file, reader.fieldnames)
+		writer.writeheader()
+		for row in reader:
+			if row['stop_id'] in stations:
+				print(row['stop_id'],row['trip_id'],row['departure_time'])
+				new_stop_id = row['stop_id']+':na:na'
+				renamed_stop_ids[row['stop_id']] = new_stop_id
+				row['stop_id']= new_stop_id
+
+			writer.writerow(row)
+
+with open(original_stops_file_path) as csvfile:
+	with open(new_stops_file_path, 'w') as new_stops_file:
+		reader = csv.DictReader(csvfile)
+		writer = csv.DictWriter(new_stops_file, reader.fieldnames)
+		writer.writeheader()
+
+		for row in reader:
+			writer.writerow(row)
+			if row['stop_id'] in renamed_stop_ids:
+				row['stop_id'] = renamed_stop_ids[row['stop_id'] ]
+				row['location_type'] = '0'
+				writer.writerow(row)
+
+os.remove(original_stops_file_path)
+os.rename(new_stops_file_path, original_stops_file_path)
+
+os.remove(original_stop_times_file_path)
+os.rename(new_stop_times_file_path, original_stop_times_file_path)