Validation - Ignore duplicate warnings if services are distinct

[GoogleCodeExporter]

Ignore duplicate warnings if services are distinct. For example, agency may
provide route A twice in routes.txt under different routes ids

route_id, route_short_name
1,A
2,A

but 1 and 2 are on distinct services

Original issue reported on code.google.com by fredf...@gmail.com on 25 Jul 2008 at 10:41

[GoogleCodeExporter]

How does one know if the services are distinct? The [union of stops visited by 
the
trips] of each route are disjoint?

Original comment by tom.brow...@gmail.com on 25 Jul 2008 at 11:05

• Added labels: App-FeedValidator, Language-Python

[GoogleCodeExporter]

to expand on my example,

trips.txt
route_id,service_id,trip_id
1,W1,T1
1,W1,T2
...
2,W2,TX
2,W2,TY
...

calendar.txt
W1,....,20080101,20080501
W2,....,20080502,20080901

W1 and W2 are service distinct.

Original comment by fredf...@gmail.com on 25 Jul 2008 at 11:09

[GoogleCodeExporter]

I'm not sure why trips that are in disjoint time periods should be different 
routes.
For example, if you open a station info window in google maps near the cutoff 
between
two periods you probably don't care that A for 20080101,20080501 is different 
from A
for 20080502,20080901 unless something big changed, such as them going to 
different
stops. I don't think feed parsers should need to cluster routes by 
route_short_name.

Original comment by tom.brow...@gmail.com on 25 Jul 2008 at 11:55

[GoogleCodeExporter]

I agree ideally it should be one route. Some agencies however need to export 
data
like this and the currently receive errors. I've been told that this is fine so 
why
warn? It creates another communication.

Original comment by fredf...@gmail.com on 26 Jul 2008 at 12:00

[GoogleCodeExporter]

I think it is correct for duplicate routes to cause a warning, even when they 
have 
disjoint service periods. It is true that for internally at Google we handle 
them, but 
it makes for cleaner GTFS.
An error would be inappropriate. I just checked that it is a warning.

Original comment by tom.brow...@gmail.com on 21 Sep 2009 at 7:19

• Changed state: WontFix

[GoogleCodeExporter]

Issue 94 has been merged into this issue.

Original comment by tom.brow...@gmail.com on 21 Sep 2009 at 9:36