diff --git a/umbra/form_checker.py b/umbra/form_checker.py
index cf3583431b8fa1506f1efe90f8fe406509f89482..e5e823d7e75f8081e54c9fd78607f37df7385028 100644
--- a/umbra/form_checker.py
+++ b/umbra/form_checker.py
@@ -8,32 +8,38 @@ class FormChecker:
"""
Constructor
"""
- self.initialize_frames()
- self._prefixes = ['','ge','be','ver','on','ont']
- self._affixes = ['','en','t','te','ten','de','den','s',"'s"]
- self._verb_prefixes = ['ge', 'ver', 'be', 'ont']
+ self._initialize_frames()
+ self._prefixes = ['', 'ge', 'be', 'ver', 'on', 'ont']
+ self._affixes = ['', 'en', 't', 'te', 'ten', 'de', 'den', 's', "'s"]
+ self._verb_prefixes = ['ge', 'ver', 'be', 'ont', 'her', 'mis']
self._verb_affixes = ['den', 'ten', 'de', 'te', 'en', 't', 'd', 'n',
'']
self._consonants = ['b', 'd', 'f', 'g', 'k', 'l', 'm', 'n', 'p', 'r',
's', 't', 'z', 'c', 'h', 'j', 'q', 'v', 'w', 'x']
self._vowels = ['a', 'e', 'o', 'u']
- def initialize_frames(self):
+ def _initialize_frames(self):
"""
Help the constructor by initializing the frames for irregular verbs.
"""
self._irr_verbs_frame = pd.read_csv(ut.get_path(
'resources/irregular_verbs.csv'))
- self._irr_verbs_vte = self._irr_verbs_frame.copy()
- self._irr_verbs_vtm = self._irr_verbs_frame.copy()
- self._irr_verbs_vdw = self._irr_verbs_frame.copy()
+ self._irr_verbs_vte = self._create_frame('verleden tijd enkelvoud')
+ self._irr_verbs_vtm = self._create_frame('verleden tijd meervoud')
+ self._irr_verbs_vdw = self._create_frame('voltooid deelwoord')
self._irr_verbs_frame = self._irr_verbs_frame.values
- self._irr_verbs_vte = self._irr_verbs_vte.sort_values(
- 'verleden tijd enkelvoud').values
- self._irr_verbs_vtm = self._irr_verbs_vtm.sort_values(
- 'verleden tijd meervoud').values
- self._irr_verbs_vdw = self._irr_verbs_vdw.sort_values(
- 'voltooid deelwoord').values
+
+ def _create_frame(self, key):
+ """
+ Create a new dataframe based on the key value given as parameter
+
+ Args:
+ key: Key value for which column to sort the dataframe on.
+
+ Returns:
+ A new dataframe sorted on the column corresponding to the key
+ """
+ return self._irr_verbs_frame.copy().sort_values(key).values
def form_related(self, source_word, shadow_word):
"""
@@ -46,11 +52,11 @@ class FormChecker:
Returns:
A boolean value indicating whether the arguments are form-related
"""
- return (self.fix_related(source_word, shadow_word)
- or self.irr_verb_related(source_word, shadow_word)
- or self.reg_verb_related(source_word, shadow_word))
+ return self._fix_related(source_word, shadow_word) \
+ or self._irr_verb_related(source_word, shadow_word) \
+ or self._reg_verb_related(source_word, shadow_word)
- def irr_verb_related(self, source_word, shadow_word):
+ def _irr_verb_related(self, source_word, shadow_word):
"""
Check whether the arguments are different versions of the same
irregular verb.
@@ -63,18 +69,13 @@ class FormChecker:
A boolean value indicating whether the arguments are different
versions of the same irregular verb.
"""
- return (self.irr_verb_helper(self._irr_verbs_frame, 0,
- source_word, shadow_word)
- or self.irr_verb_helper(self._irr_verbs_frame, 0,
- source_word, shadow_word)
- or self.irr_verb_helper(self._irr_verbs_vte, 1,
- source_word, shadow_word)
- or self.irr_verb_helper(self._irr_verbs_vtm, 2,
- source_word, shadow_word)
- or self.irr_verb_helper(self._irr_verbs_vdw, 3,
- source_word, shadow_word))
+ for key in range(4):
+ if self._irr_verb_helper(self._irr_verbs_frame, key, source_word,
+ shadow_word):
+ return True
+ return False
- def irr_verb_helper(self, frame, key, source_word, shadow_word):
+ def _irr_verb_helper(self, frame, key, source_word, shadow_word):
"""
Search in one frame to see whether the shadow word and source word are
in the same row.
@@ -93,13 +94,35 @@ class FormChecker:
i = bisect_left(frame[key], source_word)
return i != len(frame[key]) and shadow_word in frame[i]
- def reg_verb_related(self, source_word, shadow_word):
- return self.reg_verb_helper(source_word, shadow_word) or \
- self.reg_verb_helper(shadow_word, source_word)
+ def _reg_verb_related(self, source_word, shadow_word):
+ """
+ Check whether two words are different conjugations of the same verb
- def reg_verb_helper(self, word1, word2):
- stem = self.get_stem(word1)
- vdw1, vdw2 = self.create_vdws(stem)
+ Args:
+ source_word: lemma of type string, corresponding to a source word
+ shadow_word: lemma of type string, corresponding to a shadow word
+
+ Returns:
+ A boolean value indicating whether the two input words are
+ different conjugations of the same verb.
+ """
+ return self._reg_verb_helper(source_word, shadow_word) or \
+ self._reg_verb_helper(shadow_word, source_word)
+
+ def _reg_verb_helper(self, word1, word2):
+ """
+ Help the _reg_verb_related function, this function is one-directional
+
+ Args:
+ word1: lemma of type string
+ word2: lemma of type string
+
+ Returns:
+ A boolean value indicating whether word1 and word2 are different
+ conjugations of the same verb.
+ """
+ stem = self._get_stem(word1)
+ vdw1, vdw2 = self._create_vdws(stem)
if vdw1 == word2 or vdw2 == word2:
return True
for affix in self._verb_affixes:
@@ -107,51 +130,91 @@ class FormChecker:
return True
return False
- def create_vdws(self, stem):
+ def _create_vdws(self, stem):
+ """
+ Creates the perfect time form of the verb based on the stem
+
+ Args:
+ stem: the stem of a word as determined by other functions
+
+ Returns:
+ The two possible perfect time forms based on the stem
+
+ """
word = 'ge' + stem
for v_prefix in self._verb_prefixes:
- if stem[:len(v_prefix)] == v_prefix:
- word = stem[2:]
+ if stem[:len(v_prefix)] == v_prefix and len(stem) > \
+ len(v_prefix) + 1:
+ word = word[2:]
word_1 = word+'d'
word_2 = word+'t'
return word_1, word_2
- def get_stem(self, word):
+ def _get_stem(self, word):
+ """
+ Find the stem of the word given as input
+
+ Args:
+ word: Lemma of type String
+
+ Returns:
+ The stem of the word given as input
+ """
stem = ''
for affix in self._verb_affixes:
length = len(affix)
if (word[-length:] == affix or affix == '') \
- and not self.is_vdw(word):
- if affix == '':
- stem = word
- elif len(word) == len(affix):
+ and not self._is_vdw(word):
+ if affix == '' or len(word) == length:
stem = word
else:
stem = word[:-length]
- stem = self.create_single_stem(stem, affix)
+ stem = self._create_single_stem(stem, affix)
return stem
elif word[-length:] == affix:
stem = word[2:-length]
return stem
return stem
- def is_vdw(self, word):
- return word[:2] == 'ge' and (word[-1]=='t' or word[-1]=='d')
+ def _is_vdw(self, word):
+ """
+ Check whether the input word is in perfect time form
- def create_single_stem(self, stem, affix):
+ Args:
+ word: Lemma of type string
+
+ Returns:
+ A boolean value indicating whether the input word is in perfect
+ time form.
+ """
+ return word[:2] == 'ge' and (word[-1] == 't' or word[-1] == 'd')
+
+ def _create_single_stem(self, stem, affix):
+ """
+ Find the single-person stem based on the stem and affix
+
+ Args:
+ stem: The stem of a word, of type String
+ affix: The affix which was found, which was used in creating the
+ stem
+
+ Returns:
+ The single-person stem of the stem given as input, based on the
+ stem itself and the affix
+ """
if affix == 'en' and len(stem)>2 and stem != affix:
if stem[-1] in self._consonants and stem[-1] == stem[-2]:
- stem= stem[:-1]
- elif stem[-1] in self._consonants and stem[-2] in self._vowels and \
- stem[-3] in self._consonants:
- stem= stem[:-2] + stem[-2] + stem[-2:]
+ stem = stem[:-1]
+ elif stem[-1] in self._consonants and stem[-2] in self._vowels \
+ and stem[-3] in self._consonants:
+ stem = stem[:-1] + stem[-2:]
if stem[-1] == 'z':
return stem[:-1] + 's'
elif stem[-1] == 'v':
return stem[:-1] + 'f'
return stem
- def fix_related(self, source_word, shadow_word):
+ def _fix_related(self, source_word, shadow_word):
"""
Check whether the arguments are related in terms of their prefixes and
affixes.
@@ -164,10 +227,10 @@ class FormChecker:
A boolean value indicating whether the arguments are related in
terms of their prefixes and affixes.
"""
- return self.prefix_related(source_word, shadow_word) or \
- self.affix_related(source_word, shadow_word)
+ return self._prefix_related(source_word, shadow_word) or \
+ self._affix_related(source_word, shadow_word)
- def prefix_related(self, source_word, shadow_word):
+ def _prefix_related(self, source_word, shadow_word):
"""
Check whether the arguments are related in terms of their prefixes
@@ -188,7 +251,7 @@ class FormChecker:
return True
return False
- def affix_related(self, source_word, shadow_word):
+ def _affix_related(self, source_word, shadow_word):
"""
Check whether the arguments are related in terms of their prefixes and
postfixes.
@@ -208,4 +271,4 @@ class FormChecker:
for aff2 in self._affixes:
if shadow_rest + aff2 == source_word:
return True
- return False
\ No newline at end of file
+ return False