avoid skipping ways if mixing encoders in OSMRoadClassParser #1845
Conversation
| public IntsRef handleWayTags(IntsRef edgeFlags, ReaderWay readerWay, EncodingManager.Access access, IntsRef relationFlags) { | ||
| if (!access.isWay()) | ||
| public IntsRef handleWayTags(IntsRef edgeFlags, ReaderWay readerWay, boolean ferry, IntsRef relationFlags) { | ||
| if (ferry) |
There was a problem hiding this comment.
This is the actual fix. I.e. it skips only for ferries (not really necessary but also does not hurt - at least now ;))
There was a problem hiding this comment.
Ah I see, that explains the issue 👍.
There was a problem hiding this comment.
BTW: do we need to pass the ferry boolean? We pass the that should contain the ferry information? Or is this for some ferry relations or something like this?
There was a problem hiding this comment.
Thanks this looks good. I gave it a quick test with #1844 and it seems to work fine now 👍. One thing though, should we remove the boolean ferry completely from the method? We already pass the ReaderWay which should contain this information anyway. On the other hand, the import could get slightly slower if we have to perform the ferry check multiple times.
Yes, true. I would do this later when we remove the FlagEncoders completely. E.g. I'm unsure if we need the separate Another solution would be to create artificial tags like we already do with |
fixes #1843
The reason why there is a difference for bike vs. mtb: in AcceptWay we use HashMap and getAccess uses the first entry which changes for different String keys.