| [railML3] identifierRange in PatternTrain [message #4057] |
Mon, 22 June 2026 09:53  |
Milan Wölke
Messages: 233
Registered: April 2007
|
Senior Member |
|
|
Hi all,
in railML 3.4 we will introduce the pattern train to timetable in order to describe service concepts. We have had this in railML 2.5 already and added this to 3.4 as part of the efforts to fulfill the requirements of the use cases LTST (Long Term Strategic Timetabling) and SCOT (Service COncept for a Tender).
Part of this new PatternTrain is the option to describe the range of identifiers to be used by the actual commercial and operational trains that may result from such PatternTrain. As per the time of this writing the idea here is to select the type of identifier using the same enumeration that is being used when specifying identifiers for commercial and operational trains, plus a lower bound and an upper bound.
This implies that all identifiers to be used are numeric. My question is, if this is sufficient. How do you deal with train numbers that are not strictly numerical, but for example contain prefixes? Would you need or want to encode these with the train number pattern? One could imagine some kind of regular expression based approach, however, I would be reluctant to add this if there is no need from the community for this.
Please share your thoughts on this, so I can adapt the modelling if needed. If lower bound and upper bound is fine for you, please also leave a note. Thanks in advance.
Best regards, Milan
Milan Hoffmann – Timetable schema coordinator
railML.org (Registry of Associations: VR 5750)
Altplauen 19h; 01187 Dresden; Germany www.railML.org
|
|
|
|
| Re: [railML3] identifierRange in PatternTrain [message #4104 is a reply to message #4057] |
Tue, 18 August 2026 08:47  |
David Lichti
Messages: 59
Registered: December 2020
|
Member |
|
|
I currently don't see us (Hacon, TPS.plan) using pattern trains.
But from our experience, train numbers often do not fit in a purely numeric domain.
We are often dealing with alphanumerical train numbers. And letters are not necessarily prefixes; they may appear anywhere in the train number. Though that should not prevent the use of ranges. As long as we can agree on a common alphabet and its ordering, it would still be possible to have consistent range definitions across an alphanumerical domain. (Numbers before letters, or vice-versa? While the German alphabet is from A to Z, the regular Swedish alphabet is from A to Ö.)
We also see numbering schemes where trains in opposite directions have interleaving train numbers. The services from A to B would have odd train numbers, while the services from B to A would have the even train numbers in between. Since pattern trains would usually only represent services going in one direction, a train number range 101..121 should expand to 101, 103, 105, ..., not to 101, 102, 103, ... So, there should probably also be some attribute to indicate the stepping pattern.
|
|
|
|
Members
Search
Help
Login
Home
