On Thu, Nov 06, 2008 at 03:00:23PM +0100, Sascha Silbe wrote:
Wenn wir eine XML-Datei (so wie vom Server geliefert) nehmen (und daraus dann die CSV generieren), kann ich ein Tool schreiben, da[s] die Werte in der XML-Datei an Hand der Daten aus dem CSV anpasst.Ist fertig, hängt an. Ausgabeformat "new" ist die entsprechend geänderte XML-Datei, "oldNew" enthält sowohl alte als auch neue Daten für jeden einzelnen veränderten Eintrag.
Felder für die CSV-Dateien: type ("node" oder "way"), id (ID aus der OSM-API), lat/lon Koordinaten (wird ignoriert), alle weiteren Feldnamen sind die Namen von Tags.
Beispieldateien hängen ebenfalls an. CU Sascha -- http://sascha.silbe.org/ http://www.infra-silbe.de/
#!/usr/bin/env python # arch-tag: 43b684ec-eb85-41d9-9b9a-c879f5b8b732 # Author: Sascha Silbe -- http://sascha.silbe.org/ # "License": Public Domain import csv import xml.dom.minidom import xml.xpath import sys generatorName="osmcsvdiff" elemTypes=['node', 'way', 'relation'] nonTagAttrs=['type', 'id', 'lat', 'lon'] def extractTags(entry) : tags = dict(entry) for name in nonTagAttrs : if name in tags : del tags[name] return tags def readCSVs(fNameOld, fNameNew) : oldLines = csv.DictReader(file(fNameOld, "rb")) newLines = csv.DictReader(file(fNameNew, "rb")) oldByType = dict([(type_,{}) for type_ in elemTypes]) newByType = dict([(type_,{}) for type_ in elemTypes]) for entry in oldLines : oldByType[entry['type']][int(entry['id'])] = {'tags': extractTags(entry)} for entry in newLines : newByType[entry['type']][int(entry['id'])] = {'tags': extractTags(entry)} return (oldByType, newByType) def uniq(l) : return dict.fromkeys(l).keys() def genChangesType(oldVals, newVals) : diffs = {} for id in uniq(oldVals.keys()+newVals.keys()) : if id not in oldVals : raise NotImplementedError("Creation not yet supported") elif id not in newVals : diffs[id] = {'action': 'delete'} else : if (oldVals[id]['tags'] == newVals[id]['tags']) : continue diff = { 'action': 'modify', 'setTags': dict([(name, newValue) for (name, newValue) in newVals[id]['tags'].items() if (oldVals[id]['tags'].get(name) != newValue)]), 'delTags': [name for name in oldVals[id]['tags'].keys() if (name not in newVals[id]['tags'])], } if diff['setTags'] or diff['delTags'] : diffs[id] = diff return diffs def genChanges(oldVals, newVals) : diffs = dict([(t,{}) for t in elemTypes]) for type_ in elemTypes : diffs[type_] = genChangesType(oldVals[type_], newVals[type_]) return diffs def applyDiff(oldNode, diff) : newNode=oldNode.cloneNode(deep=True) for (name, value) in diff['setTags'].items() : xml.xpath.Evaluate('[EMAIL PROTECTED]"%s"]' % (name,), newNode)[0].setAttribute("v", value) for name in diff['delTags'] : oldTag = xml.xpath.Evaluate('[EMAIL PROTECTED]"%s"]' % (name,), newNode) newNode.removeChild(oldTag) oldTag.unlink() return newNode def formatChangeOldNew(type_, id, xmlNode, diff) : lines = ['<%s id="%d" action="%s">' % (type_, id, diff['action'])] if (diff['action'] == 'modify') : lines += ['<old>', xmlNode.toxml(encoding='utf-8'), '</old>'] lines += ['<new>', applyDiff(xmlNode, diff).toxml(encoding='utf-8'), '</old>'] lines.append('</%s>' % (type_,)) return lines def formatChangesOldNew(oldXml, diffsByType) : lines=["<?xml version='1.0' encoding='UTF-8'?>", '<osmOldNew version="1.0" generator="%s">' % (generatorName,)] for type_ in elemTypes : for (id, diff) in diffsByType[type_].items() : elemNode = xml.xpath.Evaluate("/osm/[EMAIL PROTECTED]'%d']" % (type_,id), oldXml)[0] lines += formatChangeOldNew(type_, id, elemNode, diff) lines.append('</osmOldNew>') return "\n".join(lines) def formatChangesNew(oldXml, diffsByType) : lines=["<?xml version='1.0' encoding='UTF-8'?>", '<osm version="0.5" generator="%s">' % (generatorName,)] for type_ in elemTypes : for (id, diff) in diffsByType[type_].items() : elemNode = xml.xpath.Evaluate("/osm/[EMAIL PROTECTED]'%d']" % (type_,id), oldXml)[0] lines.append(applyDiff(elemNode, diff).toxml(encoding='utf-8')) lines.append('</osm>') return "\n".join(lines) def printSyntax(myName) : print "Syntax: "+myName+" <outputFormat> <old.xml> <old.csv> <new.csv>" print "Available output formats: oldNew, new" return 100 def main(myName, args) : if len(args) != 4 : return printSyntax(myName) outputFormat, oldXmlFName, oldCsvFName, newCsvFName = args oldXml = xml.dom.minidom.parse(oldXmlFName) if outputFormat == "oldNew" : print formatChangesOldNew(oldXml, genChanges(*readCSVs(oldCsvFName, newCsvFName))) elif outputFormat == "new" : print formatChangesNew(oldXml, genChanges(*readCSVs(oldCsvFName, newCsvFName))) else : raise ValueError("Unknown output format: "+`outputFormat`) sys.exit(main(sys.argv[0], sys.argv[1:]))
old.xml
Description: XML document
type,id,name,is_in node,2,Wiesbaden-Naurod,"Wiesbaden;Hessen,Germany,Europe"
type,id,name,is_in node,2,Naurod,"Wiesbaden,Hessen,Bundesrepublik Deutschland,Europe"
new.xml
Description: XML document
oldNew.xml
Description: XML document
signature.asc
Description: Digital signature
_______________________________________________ Talk-de mailing list Talk-de@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-de
