Leezer: Another crash, on NYCT-1/NTT/1 Line - (To Van Courtlandt Park - 242nd Street)

Hi Leezer,

I got this crash on the route shown.  See attached.

Everything is freshly downloaded from the openbveconnect.com website.

Hmmm... I do not see how to add an attachment.   Here is the crash log. I'll see if I can add an attachment with the other log, in a reply.

Code:

6/7/2017 4:31:24 PM
OpenBVE 1.5.464.795-travis Crash Log
Program is running on the Linux backend
Current screen resolution is: Windowed 1276px X 976px
Current routefile is: /home/edg/openbve/UserData/Railway/Route/NYCT-1/NTT/1 Line - (To Van Courtlandt Park - 242nd Street) - (7 Avenue LCL - Broadway LCL )  (v1.0) - BETA.csv
Current train is: /home/edg/openbve/UserData/Train/R46 2014/R46 2014 (8 Car)
The current routefile caused the following exception:
System.ArgumentException: The relative path contains invalid characters.
  at OpenBveApi.Path.CombineFile (System.String absolute, System.String relative) <0x41315dc0 + 0x00513> in <filename unknown>:0
  at OpenBve.CsvRwRouteParser.ParseRouteForData (System.String FileName, Boolean IsRW, System.Text.Encoding Encoding, Expression[] Expressions, System.String TrainPath, System.String ObjectPath, System.String SoundPath, System.Double[] UnitOfLength, OpenBve.RouteData& Data, Boolean PreviewOnly) <0x4051c000 + 0x2206f> in <filename unknown>:0
  at OpenBve.CsvRwRouteParser.ParseRouteForData (System.String FileName, Boolean IsRW, System.Text.Encoding Encoding, System.String TrainPath, System.String ObjectPath, System.String SoundPath, OpenBve.RouteData& Data, Boolean PreviewOnly) <0x40506fc0 + 0x00127> in <filename unknown>:0
  at OpenBve.CsvRwRouteParser.ParseRoute (System.String FileName, Boolean IsRW, System.Text.Encoding Encoding, System.String TrainPath, System.String ObjectPath, System.String SoundPath, Boolean PreviewOnly) <0x405028c0 + 0x03f67> in <filename unknown>:0
  at OpenBve.Loading.LoadEverythingThreaded () <0x4060b000 + 0x001ff> in <filename unknown>:0
  at OpenBve.Loading.LoadThreaded () <0x40608730 + 0x0001b> in <filename unknown>:0  at System.Environment.get_StackTrace () [0x00000] in <filename unknown>:0
  at OpenBve.Loading.LoadThreaded () [0x00000] in <filename unknown>:0
  at System.Threading.ThreadHelper.ThreadStart_Context (System.Object state) [0x00000] in <filename unknown>:0
  at System.Threading.ExecutionContext.RunInternal (System.Threading.ExecutionContext executionContext, System.Threading.ContextCallback callback, System.Object state, Boolean preserveSyncCtx) [0x00000] in <filename unknown>:0
  at System.Threading.ExecutionContext.Run (System.Threading.ExecutionContext executionContext, System.Threading.ContextCallback callback, System.Object state, Boolean preserveSyncCtx) [0x00000] in <filename unknown>:0
  at System.Threading.ExecutionContext.Run (System.Threading.ExecutionContext executionContext, System.Threading.ContextCallback callback, System.Object state) [0x00000] in <filename unknown>:0
  at System.Threading.ThreadHelper.ThreadStart () [0x00000] in <filename unknown>:0
Current program memory usage: 184.16mb
System memory free: 0mb

Thanks,
Ed G

Unable to find how to upload an attachment. I'm using a real "reply" not a quick reply.

Pls advise.

And fixed as soon as the current nightly builds
(I'd actually been away for a few days, so you're quite lucky with the timing!)

This is caused by a comment on Line 3047:

Code:

; SIGNAL ON TRACK 2: 06 F X 316,

As the comment starts with the word SIGNAL and then a space, the parser misinterprets this as a new signal declaration, and attempts to parse the following text as a relative path, which then crashes due to the colon......

Shouldn't the parser ignore anything between a semicolon and \n?

Quork wrote:Shouldn't the parser ignore anything between a semicolon and \n?

No

Remember that the semicolon is used as an argument separator too, and so it's impossible (At least without any massive amount of hacking) to determine whether whether the semicolon is deliminating a comment, an argument, or is simply misplaced.....
It's a horrid method of deliminating a comment, but it's what we're stuck with

So it basically boils down to determining, whether there is a command accepting semicolon-separated arguments before the semicolon... Ugh. Even more arguments for the new format project...

Quork wrote:So it basically boils down to determining, whether there is a command accepting semicolon-separated arguments before the semicolon... Ugh. Even more arguments for the new format project...

Oh, I know

I'm working towards this, but it's an incredibly slow job, as Michelle left hacks *everywhere* in the source code, and 90% of the time, it's only designed to work in the one specific way, and changing it is a massive job.
What looks simple to change, 95% of the time isn't, and sometimes we get knockon effects on other stuff......

I guess it's a very extreme example of the old 80-20-rule (80% of the work gets done in 20% of the time)...

I moved all those ; SIGNAL comments to their own lines in the route csv file, and was able to run the Express route just fine.   I'm still working on getting the proper objects for the local route.