[OSM-talk] planet files appear truncated to osmosis, was: old planet files with 'nolt' in the name
Toby Murray
toby.murray at gmail.com
Sun Jan 15 20:10:56 GMT 2012
On Sun, Jan 15, 2012 at 10:31 AM, Martijn van Exel <m at rtijn.org> wrote:
> Hi,
>
> I had the following issue with processing a planet file:
>
> On Sat, Jan 14, 2012 at 11:32 AM, Martijn van Exel <m at rtijn.org> wrote:
> [..]>
>> Also, I'm running into problems with the Jan 6, 2010 planet:
>>
>> org.openstreetmap.osmosis.core.OsmosisRuntimeException: Unable to
>> parse xml file /osm/planet/historical/planet-100106.osm.bz2.
>> publicId=(null), syst
>> emId=(null), lineNumber=4312, columnNumber=75.
>>
>> The MD5 hash check was OK.
>> Anyone had trouble with this particular file before?
>
> It seems that the problem is not with the particular file. I tried the
> same process with a few other files, all give the same error around
> line 4000. I'm officially confused. I used osmosis a lot before on the
> same machine.
>
> This is what I am doing:
>
> mvexel at lima$ /osm/software/osmosis-0.39/bin/osmosis --rx
> /osm/planet/historical/planet-090701.osm.bz2 --bp
> file=/osm/poly/us.poly --wb
> /osm/planet/historical/us/planet-090701-us.osm.pbf
>
> I tried osmosis 0.39 and 0.40.1 with the same result.
I think you're running into the problem in java's internal bzip
implementation. It doesn't deal with multiple streams. This presents
as a malformed XML problem because it stops parsing at the end of the
first stream. You should unzip the file in a separate process and pipe
the result into osmosis like so:
bzcat planet-090701.osm.bz2 | osmosis --rx -
(or maybe --rx /dev/stdin)
Toby
More information about the talk
mailing list