transitime / core Goto Github PK

Hi,
The pom.xml file in transitime contains a reference to RmiQuery but this class is in transitimeAPI. So you get a class not found when you try to run RMIQuery.jar.
Cheers,
Sean.

This is the stack trace produced when first run with empty agency table.

Exception in thread "main" java.lang.IndexOutOfBoundsException: Index: 0, Size: 0
at java.util.ArrayList.rangeCheck(ArrayList.java:635)
at java.util.ArrayList.get(ArrayList.java:411)
at org.transitime.gtfs.DbConfig.getFirstAgency(DbConfig.java:705)
at org.transitime.applications.Core.(Core.java:130)
at org.transitime.applications.Core.createCore(Core.java:184)
at org.transitime.applications.Core.getInstance(Core.java:197)
at org.transitime.db.structs.Trip.getRoute(Trip.java:832)
at org.transitime.db.structs.Trip.getRouteName(Trip.java:848)
at org.transitime.db.structs.Trip.getRouteShortName(Trip.java:819)
at org.transitime.gtfs.ScheduleBasedTravelTimesProcessor.determineTravelTimesBasedOnSchedule(ScheduleBasedTravelTimesProcessor.java:270)
at org.transitime.gtfs.ScheduleBasedTravelTimesProcessor.processTrips(ScheduleBasedTravelTimesProcessor.java:474)
at org.transitime.gtfs.ScheduleBasedTravelTimesProcessor.process(ScheduleBasedTravelTimesProcessor.java:568)
at org.transitime.gtfs.GtfsData.processData(GtfsData.java:2344)
at org.transitime.applications.GtfsFileProcessor.process(GtfsFileProcessor.java:254)
at org.transitime.applications.GtfsFileProcessor.main(GtfsFileProcessor.java:604)

A null direction stops the handleArrivalDeparture method of PredictionAccuracyModule matching an arrivalDeparture event to a PredAccuracyPrediction in memory. The map key includes direction.

PredictionKey key = new PredictionKey(arrivalDeparture.getVehicleId(), 
                arrivalDeparture.getDirectionId(), arrivalDeparture.getStopId());
        List<PredAccuracyPrediction> predsList = predictionMap.get(key);

Not sure on the solution but for me I did this in the new GTFSRealtimePredictionAccuracyModule I am working on. This is similar to what is done in NextBusPredictionAccuracyModule.

String direction=null;

                         if(update.getTrip().hasDirectionId())
                                direction=""+update.getTrip().getDirectionId();

                            if (update.getTrip() != null) {
                                Trip trip = dbConfig.getTrip(update.getTrip().getTripId());
                                if (trip != null) {
                                    direction = trip.getDirectionId();
                                } else {
                                    logger.error("Got tripTag={} but no such trip in "
                                            + "the configuration.", update.getTrip().getTripId());
                                }
                            }

Hi,

The build command should be

mvn install -DskipTests.

Cheers,

Sean.

See log

core.log:15:03:46.734 INFO  thread=main [o.t.a.Core:404] Starting up optional module org.transitime.core.predAccuracy.PredictionAccuracyModule
core.log:15:03:46.756 ERROR thread=main [o.t.m.Module:120] Could not run module org.transitime.core.predAccuracy.PredictionAccuracyModule
core.log:java.lang.NoSuchMethodException: org.transitime.core.predAccuracy.PredictionAccuracyModule.<init>(java.lang.String)

Perhaps this constructor should be public instead of protected?

protected PredictionAccuracyModule(String agencyId)

The constructors

public ServiceUtils(DbConfig dbConfig)
public Time(DbConfig dbConfig)

call

this.calendar = new GregorianCalendar(dbConfig.getFirstAgency().getTimeZone());

Both throw exceptions if no agency in database.

Could catch exception and use GregorianCalendar() for first run?

@skibu Will I implement this fix in a branch?

Reports use dates in ISO, MDY format for the date fields. If you set up a postgres database it uses lc_time locale behaviours as the default which in Ireland expects ISO, DMY.

From the postgres docs

The built-in default is ISO, MDY, but initdb will initialize the configuration file with a setting that corresponds to the behavior of the chosen lc_time locale.

The GTFS-rt spec states the time stamps for vehicle positions are to be in seconds from epoch not milliseconds.

            if (vehicle.hasTimestamp())
                gpsTime = vehicle.getTimestamp();
            else
                gpsTime = System.currentTimeMillis();

I expect this should be

        if (vehicle.hasTimestamp())
                gpsTime = vehicle.getTimestamp()*1000;
            else
                gpsTime = System.currentTimeMillis();

I'm testing Transitime for the transport system of Santiago of Chile.

I downloaded the latest version of Transitime, December 23, commit 464a055. But this aborts loading the GTFS file
(http://datos.gob.cl/dataset/8583ae05-d62f-42d7-a56d-4f865711b65a/resource/dc31a5c7-e717-4b9d-a9a6-93b4d4d2aaca/download/gtfsv17.zip).

The error says:

14:02:14.139 WARN - Modifying headsign "(M) La Cisterna" to "(M) La Cisterna to PG1583 - Parada / Est. Intermodal La Cisterna" since it has a different last stop. TripPattern H
erna" direction 1 from stop PF152 to stop PG1583. Other TripPattern Headsign "(M) La Cisterna" direction 1 from stop PF152 to stop PG1573
Exception in thread "main" java.lang.NullPointerException
at org.transitime.gtfs.GtfsData.makeHeadsignsUniqueIfDifferentLastStop(GtfsData.java:1501)
at org.transitime.gtfs.GtfsData.createTripsAndTripPatterns(GtfsData.java:1469)
at org.transitime.gtfs.GtfsData.processStopTimesData(GtfsData.java:1118)
at org.transitime.gtfs.GtfsData.processData(GtfsData.java:2537)
at org.transitime.applications.GtfsFileProcessor.process(GtfsFileProcessor.java:294)
at org.transitime.applications.GtfsFileProcessor.main(GtfsFileProcessor.java:642)

On the Active Blocks page, when calculating the assigned percentage schedules based vehicles (created by the SchedBasedPredsModule) are taken into account but the number of assigned vehicles does not include these.

I am not sure if the intention is to include them or not in the totals/percentage but I expect it should be consistent?

"The goal is to use real-time GPS data to make useful public transportation information readily available for. "

The sentence is incomplete, it does not say say "for" who?

In TravelTimesProcessor.java on line 636 it has

                logger.error("For segmentIdx={} segment speed of {}m/s is "
                        + "above the limit of maxSegmentSpeedMps={}m/s. "
                        + "Therefore it is being reset to min segment speed. "
                        + "arrDep1={} arrDep2={}",
                        i, StringUtils.twoDigitFormat(segmentSpeedMps), 
                        maxSegmentSpeedMps.getValue(), arrDep1, arrDep2);

This should say max not min in the log message.

Possible better solution.

http://stackoverflow.com/questions/520328/can-you-find-all-classes-in-a-package-using-reflection

This would require a change to getClasses method in SchemaGenerator.

When trying to run GtfsFileProcessor.java, the program would terminate with an error when trying to read into the database the values in the column containsid from fare_rules.txt.

G07:47:30.912 WARN [o.h.e.j.s.SqlExceptionHelper:144] SQL Error: 0, SQLState: 23502
07:47:30.912 ERROR [o.h.e.j.s.SqlExceptionHelper:146] ERROR: null value in column "containsid" violates not-null constraint
Detail: Failing row contains (233, , a, , null, 1).
07:47:30.919 ERROR [o.t.g.DbWriter:204] Error writing GTFS configuration data to db.
org.hibernate.exception.ConstraintViolationException: could not execute statement
at org.hibernate.exception.internal.SQLStateConversionDelegate.convert(SQLStateConversionDelegate.java:129) ~[hibernate-core-4.3.9.Final.jar:4.3.9.Final]
at org.hibernate.exception.internal.StandardSQLExceptionConverter.convert(StandardSQLExceptionConverter.java:49)
.
.
.

... 15 common frames omitted

07:47:30.920 ERROR [o.t.a.GtfsFileProcessor:644] could not execute statement

i temporally resolved the error by filling in numbers for the containsid.
but containsid is an optional data entry and shouldn't cause an error when empty

orginal file which caused error
fare_rules.txt

changed file that fixed error
fare_rules.txt

while setting up the transitime test server we are facing such issue while calling RMI server for API call to fetch some data.
we believe apart from some minor bugs, the major part is configuration and setting properties for agency database and core database is challenge as structure of how application works is not clear. so please help us with the configuration and setting up properties file and connect with database and please give us idea of the architecture and application flow too.

Thank you.

You can view unassigned vehicles using the showUnassignedVehicles=true parameter in the map URL.

ogcrudden.koding.io:8080/web/map/map.jsp?a=02&showUnassignedVehicles=true&r=DART

The vehicle background appears yellow when unassigned. When they become assigned the background remains yellow which I believe is incorrect.

I will fix this.

When run for the first time on an empty database GTFSFileProcessor complains

" ActiveRevisions in database is not valid. The configuration revs must be set to proper values. ActiveRevisions [configRev=-1, travelTimesRev=-1]"

The configRev option needs to be added to the processCommandLineOptions method as if the database does not have any values in the ActiveRevisions table it could use the one from the command line.

I think that is what was intended.

Whenever I get a database jdbc error/exception all subsequent calls to the database result in a

"ERROR: current transaction is aborted, commands ignored until end of transaction block"

For example I got an exception on writing a MonitoringEvent where the message length was greater than 512 characters, which then results in the above error when writing subsequent AVL reports to the database.

I think this is specific to Postgres.

$(#routeId-99-11-r11-1 #blockId-68.TA.99-11-r11-1.2.O #block) does not evaluate properly. The . means something to jQuery.

Hi Mike,

I have two sources of AVL so an ability to filter by source for an AVL Map Report would be very useful. I have done something locally to just show the AVL source when all are displayed on the map.

Will I extend this so you can have an AVL Map Report for each individual AVL source, and send on a pull request?

Cheers,

Sean.