This is the message I get when trying
https://api.tfl.gov.uk/Stoppoint/490001944E1
Similar for 490001255E1, 490016430E1, 490000116E1, 490000839E1, 490006300E1, 490012561E3, 490014772E1, 490080161E1, 490032130E1, 490032290E1
These are all ATCO codes that feature in the current Datastore Journey Planner files and I have no reason to suppose that they are incorrect.
The eagle eyed observer will note that they all have something in common - they end in E1 or E3, which immediately screams “it’s been interpreted as scientific notation” at me but there are plenty of other codes with similar endings which do not return an error.
I should also say that I have only identified these because I was searching for SMS codes that were missing from my records (of which more later on the complete bus stops list thread) so there may be others, unless I’d searched for them before unsuccessfully without noticing the pattern.
hi @mjcarchive - do you know which routes contain these?
Thanks,
James
490001255E1 - 427 =(first stop at Southall)
490016430E1 - 350 etc (first stop at Terminal 5)
490000116E1 - 120 H20 at Hounslow Central; the JP files show the same stop each way which must be wrong
490000839E1 - last stop on 237 at Hounslow Heath
490001944E1 - 196 P5 near Stockwell
490006300E1 - 65 where it crosses A4
490012561E3 - S4 at Stanley Park Road (not even sure this exists as I can’t find it on map!)
490014772E1 - E2 H91 at Windmill Road
490080161E1 - 111 etc (first stop at Heathrow Central)
490032130E1 - 488 at Downs Road
490032290E1 - W5 near Crouch End
1 Like
Thanks - I’ll take a look and let you know what I find
@jamesevans
Oops - I have a horrible feeling that the E suffices are being generated by my translation processes, as some of them (at least) are not present in the original XML, where the codes are entirely numeric. That’s going to take me a bit of time tracking down Sorry but it simply had not occurred to me.
No problem - I did a similar thing when doing some analysis a few months ago!
@jamesevans
To clean up my own mess…
The issue arises with the handful of TfL ATCO codes (starting 49…) which are not in the standard form 49nnnnnnna[n] where a is a letter. My coding for stopping cases like …E14 being interpreted as scientific notation was also catching these. As there are so few I had not noticed the problem but equally that means hard coding the examples is the easiest way of dealing with them. I have also put in an error trap in case I have missed any.
For the record…
“490001255E1” is 4900012554 or 4900012555
“490016430E1” is 4900164301 or 4900164307
“490000116E1” is 4900001161 or 4900001162
“490000839E1” is 4900008396
“490001944E1” is 4900019445
“490006300E1” is 4900063001
“490012561E3” is 490012561265
“490014772E1” is 4900147722
“490080161E1” is 4900801619
“490032130E1” is 4900321300
“490032290E1” is 4900322900
One might ask why these entirely numeric codes exist at all but, once established, a code is a code is a code so it doesn’t really matter. 490012561265 is particularly odd as the format is much more like an out-county code (BTW on the S4, the file it is on, some stops are shown as OTH and PTP on different journey patterns; they should be consistent).
I’ll put it in the “you leant something every day” pile…