worker - ectworker.py

from lxml import etree as ET
import xlrd

class dataVars():
    allowOneTwo = {'id', 'specific data 1', 'specific data 2'}
    allowThree = {'id', 'specific data 3', 'specific data 4', 'specific data 5', 'specific data 6'}
    dataOne = []
    dataTwo = []
    dataThree = []
    dataOneSet = []
    dataTwoSet = []
    dataThreeSet = []
    dataOneOutput = []
    dataTwoOutput = []

def parseOne(file1):
    rows = ET.parse(file1).findall('row')
    for row in rows:
        fields = row.findall('field')
        for field in fields:
            name = field.get('name')
            if name in dataVars.allowOneTwo:
                dataVars.dataOne.append(field.text)

def parseTwo(file2):
    rows = ET.parse(file2).findall('row')
    for row in rows:
        fields = row.findall('field')
        for field in fields:
            name = field.get('name')
            if name in dataVars.allowOneTwo:
                dataVars.dataTwo.append(field.text)

def parseThree(file3):
    allThree = []
    rows = ET.parse(file3).findall('row')
    for row in rows:
        tempThree = {}
        fields = row.findall('field')
        for field in fields:
            name = field.get('name')
            if name in dataVars.allowThree:
                tempThree[name] = field.text
        allThree.append(tempThree)

    for item in allThree:
        if item.get('specific data 5') == 'desired value' and item.get('specific data 6') == 'desired value':
            dataVars.dataThreeSet.append(item.get('id))
            dataVars.dataThree[item.get('id')] = (item.get('specific data 3'), item.get('specific data 4')

def prepareOut():
    for id in dataVars.dataThreeSet:
        if id not in set(dataVars.dataOne):
            dataVars.dataOneSet.append(id)
        if id not in set(dataVars.dataTwo):
            dataVars.dataTwoSet.append(id)

def finalOut():
    for id in dataVars.dataOneSet:
        out = {
            'ID' : '%s' % id,
            'Field_1' : '%s' % dataVars.dataThree[id][0],
            'Field_2' : '%s' % dataVars.dataThree[id][1]
        }
        dataVars.dataOneOutput.append(out)

    for id in dataVars.dataTwoSet:
        out = {
            'ID' : '%s' % id,
            'Field_1' : '%s' % dataVars.dataThree[id][0],
            'Field_2' : '%s' % dataVars.dataThree[id][1]
        }
        dataVars.dataTwoOutput.append(out)