def get_completions(args):
tries = corpusTries[args["corpus"]]
mode = args["trie_mode"]
if mode in ("words", "pattern"):
trie = tries[0]
prefix = args["prefix"]
else:
trie = tries[1]
prefix = tuple(args["prefix"].split())
max_results = args["max_results"]
if max_results == 0:
max_results = None
if mode == "pattern":
r = [
word for word, freq in lab.word_filter(trie, prefix)[:max_results]
]
elif args["autocorrect"] and mode == "words":
r = lab.autocorrect(trie, prefix, max_results)
else:
if mode == "sentences":
r = [
' '.join(result)
for result in lab.autocomplete(trie, prefix, max_results)
]
else:
r = lab.autocomplete(trie, prefix, max_results)
return r
def run_test(input_data):
trie = lab.generate_trie(input_data["words"])
if input_data["f"] == "autocorrect": # Tests for correct autocorrect
return lab.autocorrect(trie, input_data["prefix"], input_data["N"])
if input_data["f"] == "autocomplete": # Tests for correct autocomplete
return lab.autocomplete(trie, input_data["prefix"], input_data["N"])
else: # Tests just for producing the trie
return trie
def run_test( input_data ):
trie = lab.generate_trie(input_data["words"])
if input_data["f"] == "autocorrect": # Tests for correct autocorrect
return lab.autocorrect(trie, input_data["prefix"], input_data["N"])
if input_data["f"] == "autocomplete": # Tests for correct autocomplete
return lab.autocomplete(trie, input_data["prefix"], input_data["N"])
else: # Tests just for producing the trie
return trie
def autocorrect(input_data):
global trie
if trie is None:
words = []
print "LOADING CORPUS"
with open("resources/words.json", "r") as f:
words = json.load(f)
trie = lab.generate_trie(words)
return lab.autocorrect(trie, input_data["prefix"], input_data["N"])
def autocorrect( input_data ):
global trie
if trie is None:
words = []
print "LOADING CORPUS"
with open("resources/words.json", "r") as f:
words = json.load(f)
trie = lab.generate_trie(words)
return lab.autocorrect(trie, input_data["prefix"], input_data["N"])
def test_01_autocorrect(self):
# Autocorrect on cat in small corpus
trie = lab.make_word_trie("cats cattle hat car act at chat crate act car act")
result = lab.autocorrect(trie, 'cat',4)
self.assertIsInstance(result,list,"result not a list.")
for w in result:
self.assertIsInstance(w,str,"expecting list of strings.")
result.sort()
expect = ["act", "car", "cats", "cattle"]
self.assertEqual(expect,result,msg="incorrect result from autocorrect.")
def test_02_big_autocorrect(self):
nums = {'thin': [0, 8, 10, None],
'tom': [0, 2, 4, None],
'mon': [0, 2, 15, 17, 20, None]}
with open(os.path.join(TEST_DIRECTORY, 'testing_data', 'frankenstein.txt'), encoding='utf-8') as f:
text = f.read()
w = lab.make_word_trie(text)
for i in sorted(nums):
for n in nums[i]:
result = lab.autocorrect(w, i, n)
expected = read_expected('frank_autocorrect_%s_%s.pickle' % (i, n))
self.assertEqual(len(result), len(expected), msg='wrong autocorrect of '+repr(i)+' with maxcount = '+str(n))
self.assertEqual(set(result), set(expected), msg='wrong autocorrect of '+repr(i)+' with maxcount = '+str(n))
def test_02_big_autocorrect(self):
nums = {
'thin': [0, 8, 10, None],
'tom': [0, 2, 4, None],
'mon': [0, 2, 15, 17, 20, None]
}
with open(os.path.join(TEST_DIRECTORY, 'resources', 'testing_data',
'frankenstein.txt'),
encoding='utf-8') as f:
text = f.read()
w = lab.make_word_trie(text)
for i in sorted(nums):
for n in nums[i]:
result = lab.autocorrect(w, i, n)
expected = read_expected('frank_autocorrect_%s_%s.pickle' %
(i, n))
self.assertEqual(len(result), len(expected), msg=('missing' if len(result) < len(expected)\
else 'too many') + ' autocorrect results for ' + repr(i) + ' with macount = ' + str(n))
self.assertEqual(set(result), set(expected), msg='autocorrect included ' + repr(set(result) - set(expected))\
+ ' instead of ' + repr(set(expected) - set(result)) + ' for ' + repr(i) + ' with maxcount = '+str(n))
def test_autocorrect_small():
# Autocorrect on cat in small corpus
trie = lab.make_word_trie(
"cats cattle hat car act at chat crate act car act")
result = lab.autocorrect(trie, 'cat', 4)
assert set(result) == {"act", "car", "cats", "cattle"}