def generateDataBase(self):
"""
This function generates word database base on list of text files
Input:
Nothing
Returns:
Word database
"""
from database import WordDataBase
from wordparser import WordParser
from record import Record
words = set()
for file_path in self._file_list:
contents = self._readFileContents(file_path)
words = words.union(WordParser(contents).wordsSet())
word_db = WordDataBase()
word_db.addWord(Record("")) # Index 0 is not used.
for word in words:
word_db.addWord(word)
return word_db
def setUp(self):
from trainer import Trainer
from database import TrainingDataBase,WordDataBase,WordRecord
self.tr_empty = Trainer(WordDataBase(),TrainingDataBase())
wdb = WordDataBase()
wdb.addWord(WordRecord("aaa"))
wdb.addWord(WordRecord("bbb"))
wdb.addWord(WordRecord("ccc"))
tdb = TrainingDataBase()
tdb.add([WordRecord("aaa"),WordRecord("bbb"),WordRecord("ccc")],[WordRecord("ccc"),WordRecord("bbb")])
tdb.add([WordRecord("aaa"),WordRecord("ccc")],[WordRecord("ccc"),WordRecord("ccc")])
self.tr_notempty = Trainer(wdb,tdb)
class WordDataBaseTest(unittest.TestCase):
def setUp(self):
from database import WordDataBase
from database import WordRecord
self.db = WordDataBase()
self.db.addWord(WordRecord("ccc"))
self.db.addWord(WordRecord("bbb"))
self.db.addWord(WordRecord("aaa"))
def test_wordId_invalidword(self):
from database import DataBaseException
from database import WordRecord
with self.assertRaises(DataBaseException):
self.db.wordId(None)
with self.assertRaises(DataBaseException):
self.db.wordId(WordRecord("ddd"))
def test_wordId_validword(self):
from database import DataBaseException
from database import WordRecord
self.assertEqual(self.db.wordId(WordRecord("ccc")),0)
self.assertEqual(self.db.wordId(WordRecord("bbb")),1)
self.assertEqual(self.db.wordId(WordRecord("aaa")),2)
def test_addWord_invalidword(self):
from database import WordRecord
from database import DataBaseException
with self.assertRaises(DataBaseException):
self.db.addWord(None)
with self.assertRaises(DataBaseException):
self.db.addWord(WordRecord("ccc"))
def test_idWord_invalidinput(self):
from database import DataBaseException
with self.assertRaises(DataBaseException):
self.db.idWord(None)
with self.assertRaises(DataBaseException):
self.db.idWord(3)
with self.assertRaises(DataBaseException):
self.db.idWord(-5)
def test_multipleWordId(self):
from database import WordRecord
ids = self.db.multipleWordId([WordRecord("aaa"),WordRecord("bbb"),WordRecord("ccc")])
self.assertEqual(ids[0],2)
self.assertEqual(ids[1],1)
self.assertEqual(ids[2],0)
def test_multipleWordId_exception(self):
from database import WordRecord
from database import DataBaseException
with self.assertRaises(DataBaseException):
ids = self.db.multipleWordId([WordRecord("kkk"),WordRecord("hhh"),WordRecord("www")])
def test_multipleIdWord(self):
from database import WordRecord
words = self.db.multipleIdWord([2,1,0,0])
self.assertEqual(words[0],WordRecord("aaa"))
self.assertEqual(words[1],WordRecord("bbb"))
self.assertEqual(words[2],WordRecord("ccc"))
self.assertEqual(words[3],WordRecord("ccc"))
def test_multipleIdWord_exception(self):
from database import DataBaseException
with self.assertRaises(DataBaseException):
words = self.db.multipleIdWord([5,5,5])