[OSM-dev] extract_polygon_0.5.pl bugs

Frederik Ramm frederik at remote.org
Tue Oct 9 00:45:30 BST 2007


> 1. The "sort" option ("-s") isn't recognized and doesn't combine the 
> three output files into a single one

True. The comment "# fixme sort" somehow doesn't do the job ;)

> 2. Show stopper: Does NOT recognize OSM tags (i.e. '<tag ...>') and XML 
> closing tags (i.e. '</...>') => produces useless, invalid XML (that's 
> how I found out)!

I think you're being a bit too harsh with the script. It does not
contain explicit regular expressions for the tags you mention but it
doesn't need them. When it encounters, say, an opening <way> tag, it
starts buffering the raw lines, and if, before reaching another <way>,
a <nd ref=...> is found that makes it want to include this way in the
output, then the buffer is simply written out, and any further lines
are also written, until the next opening <way> is reached. This
includes any <tag> elements and the closing </way>, without having to
explicitly parse them.

There may be a bug in the script but the bug surely isn't that the
programmer was too stupid to think about closing tags ;-) can you give
more details about the problem?

> Will try planetosm-excerpt-area.pl now.

I think that one hasn't been made 0.5 compliant but you can of course
do that if you prefer. I had originally thought that the polygon
script could supersede that one.


Frederik Ramm  ##  eMail frederik at remote.org  ##  N49°00.09' E008°23.33'

More information about the dev mailing list