import argparse

from codecs import open

from json import dumps,JSONEncoder

from os import chdir,walk

from os.path import isdir,join

from sys import exit,stderr

from xml.etree.ElementTree import parse

# Define namespaces.

ns = {

'adsml-ma' : 'http://www.adsml.org/adsmlmaterials/2.5',

'adsml-sd' : 'http://www.adsml.org/adsmlstructureddescriptions/1.0',

'adsml-cv' : 'http://www.adsml.org/controlledvocabularies/3.0',

'adsml' : 'http://www.adsml.org/typelibrary/2.0',

'booking' : 'http://www.adsml.org/adsmlbookings/2.5'

}

# Known labeled properties.

labels = ['Package', "Title", "Zone"]

# Custom JSON encoder.

class OrderLineEncoder(JSONEncoder):

def default(self, o):

return o.__dict__

# Things i'd like to know.

class OrderLine:

def __init__(self, orderSerial, orderId, adId, adTitle, insertionDate, insertionStatus, titleCode, editionCode, packageCode):

self.orderSerial = orderSerial

self.orderId = orderId

self.adId = adId

self.adTitle = adTitle

self.insertionDate = insertionDate

self.insertionStatus = insertionStatus

self.titleCode = titleCode

self.editionCode = editionCode

self.packageCode = packageCode

def getTsvHeader(self):

return '{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}'.format('ORDERSERIAL', 'ORDERID', 'ADID', 'ADTITLE', 'INSERTIONDATE', 'INSERTIONSTATUS', 'TITLECODE', 'EDITIONCODE', 'PACKAGECODE')

def toTsv(self):

return '{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}'.format(self.orderSerial, self.orderId, self.adId, self.adTitle, self.insertionDate, self.insertionStatus, self.titleCode, self.editionCode, self.packageCode)

def toJson(self):

return dumps({

'orderSerial': self.orderSerial,

'orderId': self.orderId,

'adId': self.adId,

'adTitle': self.adTitle,

'insertionDate': self.insertionDate,

'insertionStatus': self.insertionStatus,

'titleCode': self.titleCode,

'editionCode': self.editionCode,

'packageCode': self.packageCode

})

def eprint(*args, **kwargs):

print(*args, file=stderr, **kwargs)

def getOrderLines(files, encoding):

# Create set for order lines.

orderLines = []

eprint('Using {} encoding to read files...'.format(encoding.upper()))

eprint('Processing {} XML document(s). Please be patient...'.format(len(files)))

# Reset values, python does not do 'block' scope.

orderId = 'NONE'

orderSerial = 'NONE'

# The entire xml element tree.

for file in files:

target = open(file, mode='r', encoding=encoding)

root = parse(target).getroot()

# The order part of the tree. Should always be singular.

order = root.find('./booking:AdOrder', ns)

# The order serial number.

orderSerial = order.attrib['{http://www.adsml.org/typelibrary/2.0}messageID']

# The order identifier (our correlation id).

orderId = order.find('booking:BookingIdentifier', ns).text

# An order can contain multiple placements.

placements = order.findall('booking:Placement.NewspaperMagazine', ns)

# Reset values, python does not do 'block' scope.

adId = 'NONE'

adTitle = 'NONE'

for placement in placements:

adId = placement.find('booking:PlacementIdentifier', ns).text

adTitle = placement.find('adsml:DescriptionLine', ns).text

# A placement can contain multiple insertions.

insertions = placement.findall('booking:InsertionPeriod', ns)

for insertion in insertions:

# The insertion date.

elements = insertion.find('booking:PreDefinedPeriod', ns)

# Reset values, python does not do 'block' scope.

insertionDate = 'NONE'

for element in elements:

if element.tag == '{http://www.adsml.org/typelibrary/2.0}CodeList':

label = element.text

if element.tag == '{http://www.adsml.org/typelibrary/2.0}CodeValue':

value = element.text

if label == 'Insertiondate':

insertionDate = value

# The insertion status.

elements = insertion.find('adsml:Status', ns)

# Reset values, python does not do 'block' scope.

insertionStatus = 'NONE'

for element in elements:

if element.tag == '{http://www.adsml.org/typelibrary/2.0}CodeValue':

insertionStatus = element.text

# The insertion properties (title, zone and package codes).

labeledProperties = insertion.findall('adsml:Properties/adsml:LabeledProperty', ns)

# Reset values, python does not do 'block' scope.

titleCode = 'NONE'

editionCode = 'NONE'

packageCode = 'NONE'

for labeledProperty in labeledProperties:

for element in labeledProperty:

if element.tag == '{http://www.adsml.org/typelibrary/2.0}Label':

label = element.text

if element.tag == '{http://www.adsml.org/typelibrary/2.0}Value':

value = element.text

if label in labels:

if label == 'Title':

titleCode = value

if label == 'Zone':

editionCode = value

if label == 'Package':

packageCode = value

# Now we know everything, create an Orderline Object:

orderLine = OrderLine(orderSerial, orderId, adId, adTitle, insertionDate, insertionStatus, titleCode, editionCode, packageCode)

orderLines.append(orderLine)

return orderLines

def toJson(fileList, encoding):

return dumps(getOrderLines(fileList, encoding), cls=OrderLineEncoder)

def toTsv(fileList, encoding):

orderLines = getOrderLines(fileList, encoding)

tsvLines = []

tsvLines.append(orderLines[0].getTsvHeader())

for orderLine in orderLines:

tsvLines.append(orderLine.toTsv())

return tsvLines

def output(fileList, encoding, format):

if format == 'json':

print(toJson(fileList, encoding))

if format == 'tsv':

for line in toTsv(fileList, encoding):

print(line)

def main():

parser = argparse.ArgumentParser()

parser.add_argument("adsml", help="An AdsML document or directory of AdsML documents to parse")

parser.add_argument("-e", "--encoding", type=str, choices=['cp1252', 'utf8'],

help="Specify the encoding of the input files.", default='utf8')

parser.add_argument("-f", "--format", type=str, choices=['tsv', 'json'],

help="Specify the output format. Default is tab-separated.", default='tsv')

args = parser.parse_args()

if not args.adsml:

exit(1)

if isdir(args.adsml):

for root, dirs, files in walk(args.adsml):

chdir(root)

output(files, args.encoding, args.format)

else:

output([args.adsml], args.encoding, args.format)

if __name__ == "__main__":

main()