#!/usr/bin/env python

# Converts a CSV dataset to XML using a mappings spreadsheet

import csv

import re

import sys

from copy import deepcopy

from xml.dom.minidom import getDOMImplementation

from XpathParser import XpathParser, XpathElem

from xml_dom import by_path

def main():

    prog_name = sys.argv.pop(0)

    try:

        dest = sys.argv.pop(0)

        mappings_path = sys.argv.pop(0)

    except IndexError: raise Exception('Usage: '+prog_name

        +' dest_mappings_column mappings_path <dataset')

    # Get mappings

    mappings = {}

    stream = open(mappings_path, 'rb')

    reader = csv.reader(stream, delimiter=',', quotechar='"')

    fieldnames = reader.next()

    src = fieldnames[0]

    dest_idx = fieldnames.index(dest)

    for row in reader:

        name = row[0]

        path = row[dest_idx]

        if path != '':

            if path.startswith('?'): path = path[1:]

            path = path.replace('<name>', name)

            path = re.sub(r'(?<=/)\*(?=s/(\w+))', r'\1', path)

            mappings[name] = XpathParser(path).parse()

    stream.close()

    # Process dataset

    doc = getDOMImplementation().createDocument(None, dest, None)

    stream = sys.stdin

    reader = csv.reader(stream, delimiter=',', quotechar='"')

    fieldnames = reader.next()

    row_idx = 0

    for row in reader:

        row_id = src+'_'+str(row_idx)

        for idx, name in enumerate(fieldnames):

            value = row[idx]

            if value and name in mappings:

                path = deepcopy(mappings[name]) # don't modify main value!

                path[1].attrs.append([XpathElem('id', row_id, is_attr=True)])

                path[-1].value = value

                by_path(doc, path, True)

        row_idx += 1

    doc.writexml(sys.stdout, addindent='    ', newl='\n')

main()