#!/usr/bin/env python

import sys
import codecs
import re
from xml.etree.ElementTree import ElementTree
from xml.etree.ElementTree import fromstring


total = []
xml_tree = fromstring(sys.stdin.read())
for entry in xml_tree.find("Lexicon"):
    lemma = entry.find("Lemma")
    baseforms = []
    latins = []
    #syns = []
    lemgram = None
    if lemma != None:
        freps = lemma.findall("FormRepresentation")
        if freps != None:
            for fr in freps:
                allfeats = fr.findall("feat")
                if allfeats != None:
                    for feat in allfeats:
                        if feat.attrib["att"] == "writtenForm":
                            baseforms.append(feat.attrib["val"])
                        elif feat.attrib["att"] == "lemgram":
                            lemgram = feat.attrib["val"]
                        elif feat.attrib["att"] == "latin":
                            latins.append(feat.attrib["val"])
                        elif feat.attrib["att"] == "syn":
                            baseforms.append(feat.attrib["val"])
    total.append((lemgram, baseforms, latins))
    #if len(baseforms) > 1:
    #    print str(len(baseforms))

for (lemgram, baseforms, latins) in total:
    if len(baseforms) == 0:
        pass#print "STRANGE: " #+ lemgram
    else:
        if len(latins) > 0:
            lats = "|".join(latins);
        else:
            lats = "-"
        for b in baseforms:
            print b.encode("utf-8") + "\t" + lats.encode("utf-8") + "\t" + lemgram.encode("utf-8")


# This is unfinished. It's supposed to be used to build a raw material from the Swedberg resource.