#!/usr/bin/env python

# Inner-joins two map spreadsheets A->B and B->C to A->C

import csv

import os.path

import sys

import warnings

sys.path.append(os.path.dirname(__file__)+"/../lib")

import maps

def main():

    try: _prog_name, map_1_path = sys.argv

    except ValueError:

        raise SystemExit('Usage: '+sys.argv[0]+' <map_0 map_1 [| '+sys.argv[0]

            +' map_2]... >joined_map')

    # Get map 1

    map_1 = {}

    stream = open(map_1_path, 'rb')

    reader = csv.reader(stream)

    map_1_cols = reader.next()

    for row in reader:

        if row[0] != '': map_1[row[0]] = row

    stream.close()

    # Join map 1 to map 0

    reader = csv.reader(sys.stdin)

    writer = csv.writer(sys.stdout)

    map_0_cols = reader.next()

    if not map_0_cols[1] == map_1_cols[0]: raise SystemExit('Map error: '

        'Map 0 output column name doesn\'t match map 1 input column name')

    writer.writerow(maps.merge_mappings(map_0_cols, map_1_cols))

    for row in reader:

        if row[1] != '':

            # Look for a match

            out_row = None

            suffix = ''

            while True:

                try:

                    out_row = map_1[row[1]]

                    break

                except KeyError:

                    # Heuristically look for a match on a parent path.

                    # If this produces a syntactically invalid parent path (e.g.

                    # there is a / within []), it will be ignored since there

                    # wouldn't be a an entry in map_1.

                    row[1], sep, new_suffix = row[1].rpartition('/')

                    if sep == '': break

                    suffix = sep+new_suffix+suffix # prepend new suffix

            # Write new mapping

            if out_row != None:

                row = maps.merge_mappings(row, out_row)

                row[1] += suffix # don't modify out_row!

            else:

                msg = 'No join mapping for '+row[1]

                warnings.warn(UserWarning(msg))

                row[2] = '** '+msg+' ** '+row[2]

                row[1] = ''

        writer.writerow(row)

main()