Digest a tarfile full of GPX files into a more manageable CSV file

targpxtocsv.py

import lzma
from xml.dom.minidom import parseString, parse
import dateutil.parser
import calendar
from pyproj import Geod
import tarfile
from xml.parsers import expat

def parse_trkpt( node ):
  lat = float(node.getAttribute("lat"))
	lon = float(node.getAttribute("lon"))
	strtime = node.getElementsByTagName("time")[0].firstChild.data
	dt = dateutil.parser.parse(strtime)
	tt = calendar.timegm( dt.timetuple() )

	return (lat,lon,tt)

gg = Geod(ellps='clrk66')

fn = "gpx-planet-2013-04-09.tar.xz"
lzfp = lzma.LZMAFile( fn )

tf = tarfile.open( "gpx-planet-2013-04-09.tar" )
counts = {'i':0}
fpout = open( "points.csv", "w" )

class GPXParser(object):
	def __init__(self, counts):
		self._ontime = False
		self._curpt = None
		self._curtimechars = None
		parser = expat.ParserCreate()
		def onStartElement(name, attributes):
			if name=="trk":
				counts['i'] += 1
				print counts
			if name=="trkpt":
				self._curpt = attributes
			if name=="time":
				self._ontime = True
				self._curtimechars = ""
		def onEndElement(name):
			if name=="time":
				self._ontime = False
			elif name=="trkpt":
				fpout.write( "%s,%s,%s,%s\n"%(counts['i'],self._curtimechars,self._curpt['lat'],self._curpt['lon']) )

				self._curpt = None
				self._curtimechars = None
		def onCharacterData(data):
			if self._ontime:
				self._curtimechars += data
		parser.StartElementHandler = onStartElement
		parser.EndElementHandler = onEndElement
		parser.CharacterDataHandler = onCharacterData

		self._parser = parser

	def parseFile(self, fp):
		self._parser.ParseFile( fp )

members = tf.getmembers()

for i, info in enumerate( members ):
	print "%s/%s"%(i, len(members))	
	if not info.isfile():
		continue
	if info.name.split("/")[-1]=="metadata.xml":
		continue

	fp = tf.extractfile( info )
	parser = GPXParser(counts)
	parser.parseFile( fp )