abydos.phonetic._metaphone.Metaphone.encode() - Code Metrics - chrislit/abydos - Measure and Improve Code Quality continuously with Scrutinizer

abydos.phonetic._metaphone.Metaphone.encode() F
last analyzed 2020-12-31 20:10 UTC

↳ Parent: abydos.phonetic._metaphone

Complexity

Conditions

Size

Total Lines	200
Code Lines	124

Duplication

Lines	0
Ratio	0 %

Code Coverage

Tests	89
CRAP Score	79

Importance

Changes

Metric	Value
eloc	124
dl	0
loc	200
ccs	89
cts	89
cp	1
rs	0
c	0
b	0
f	0
cc	79
nop	2
crap	79

How to fix Long Method Complexity

# Copyright 2014-2020 by Christopher C. Little.
# This file is part of Abydos.
#
# Abydos is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Abydos is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Abydos. If not, see <http://www.gnu.org/licenses/>.

"""abydos.phonetic._metaphone.

Metaphone
"""

from ._phonetic import _Phonetic

__all__ = ['Metaphone']


class Metaphone(_Phonetic):
    """Metaphone.

    Based on Lawrence Philips' Pick BASIC code from 1990 :cite:`Philips:1990`,
    as described in :cite:`Philips:1990b`.
    This incorporates some corrections to the above code, particularly
    some of those suggested by Michael Kuhn in :cite:`Kuhn:1995`.

    .. versionadded:: 0.3.6
    """

    _frontv = {'E', 'I', 'Y'}
    _varson = {'C', 'G', 'P', 'S', 'T'}

    def __init__(self, max_length: int = -1) -> None:
        """Initialize AlphaSIS instance.

        Parameters
        ----------
        max_length : int
            The maximum length of the returned Metaphone code (defaults to 64,
            but in Philips' original implementation this was 4)


        .. versionadded:: 0.4.0

        """
        # Require a max_length of at least 4
        if max_length != -1:
            self._max_length = max(4, max_length)
        else:
            self._max_length = 64

    def encode(self, word: str) -> str:
        """Return the Metaphone code for a word.

        Based on Lawrence Philips' Pick BASIC code from 1990
        :cite:`Philips:1990`, as described in :cite:`Philips:1990b`.
        This incorporates some corrections to the above code, particularly
        some of those suggested by Michael Kuhn in :cite:`Kuhn:1995`.

        Parameters
        ----------
        word : str
            The word to transform

        Returns
        -------
        str
            The Metaphone value

        Examples
        --------
        >>> pe = Metaphone()
        >>> pe.encode('Christopher')
        'KRSTFR'
        >>> pe.encode('Niall')
        'NL'
        >>> pe.encode('Smith')
        'SM0'
        >>> pe.encode('Schmidt')
        'SKMTT'


        .. versionadded:: 0.1.0
        .. versionchanged:: 0.3.6
            Encapsulated in class

        """
        # As in variable sound--those modified by adding an "h"
        ename = ''.join(c for c in word.upper() if c.isalnum())

        # Delete non-alphanumeric characters and make all caps
        if not ename:
            return ''
        if ename[0:2] in {'PN', 'AE', 'KN', 'GN', 'WR'}:
            ename = ename[1:]
        elif ename[0] == 'X':
            ename = 'S' + ename[1:]
        elif ename[0:2] == 'WH':
            ename = 'W' + ename[2:]

        # Convert to metaphone
        elen = len(ename) - 1
        metaph = ''
        for i in range(len(ename)):
            if len(metaph) >= self._max_length:
                break
            if (
                ename[i] not in {'G', 'T'}
                and i > 0
                and ename[i - 1] == ename[i]
            ):
                continue

            if ename[i] in self._uc_v_set and i == 0:
                metaph = ename[i]

            elif ename[i] == 'B':
                if i != elen or ename[i - 1] != 'M':
                    metaph += ename[i]

            elif ename[i] == 'C':
                if not (
                    i > 0
                    and ename[i - 1] == 'S'
                    and ename[i + 1 : i + 2] in self._frontv
                ):
                    if ename[i + 1 : i + 3] == 'IA':
                        metaph += 'X'
                    elif ename[i + 1 : i + 2] in self._frontv:
                        metaph += 'S'
                    elif i > 0 and ename[i - 1 : i + 2] == 'SCH':
                        metaph += 'K'
                    elif ename[i + 1 : i + 2] == 'H':
                        if (
                            i == 0
                            and i + 1 < elen
                            and ename[i + 2 : i + 3] not in self._uc_v_set
                        ):
                            metaph += 'K'
                        else:
                            metaph += 'X'
                    else:
                        metaph += 'K'

            elif ename[i] == 'D':
                if (
                    ename[i + 1 : i + 2] == 'G'
                    and ename[i + 2 : i + 3] in self._frontv
                ):
                    metaph += 'J'
                else:
                    metaph += 'T'

            elif ename[i] == 'G':
                if ename[i + 1 : i + 2] == 'H' and not (
                    i + 1 == elen or ename[i + 2 : i + 3] not in self._uc_v_set
                ):
                    continue
                elif i > 0 and (
                    (i + 1 == elen and ename[i + 1] == 'N')
                    or (i + 3 == elen and ename[i + 1 : i + 4] == 'NED')
                ):
                    continue
                elif (
                    i - 1 > 0
                    and i + 1 <= elen
                    and ename[i - 1] == 'D'
                    and ename[i + 1] in self._frontv
                ):
                    continue
                elif ename[i + 1 : i + 2] == 'G':
                    continue
                elif ename[i + 1 : i + 2] in self._frontv:
                    if i == 0 or ename[i - 1] != 'G':
                        metaph += 'J'
                    else:
                        metaph += 'K'
                else:
                    metaph += 'K'

            elif ename[i] == 'H':
                if (
                    i > 0
                    and ename[i - 1] in self._uc_v_set
                    and ename[i + 1 : i + 2] not in self._uc_v_set
                ):
                    continue
                elif i > 0 and ename[i - 1] in self._varson:
                    continue
                else:
                    metaph += 'H'

            elif ename[i] in {'F', 'J', 'L', 'M', 'N', 'R'}:
                metaph += ename[i]

            elif ename[i] == 'K':
                if i > 0 and ename[i - 1] == 'C':
                    continue
                else:
                    metaph += 'K'

            elif ename[i] == 'P':
                if ename[i + 1 : i + 2] == 'H':
                    metaph += 'F'
                else:
                    metaph += 'P'

            elif ename[i] == 'Q':
                metaph += 'K'

            elif ename[i] == 'S':
                if (
                    i > 0
                    and i + 2 <= elen
                    and ename[i + 1] == 'I'
                    and ename[i + 2] in 'OA'
                ):
                    metaph += 'X'
                elif ename[i + 1 : i + 2] == 'H':
                    metaph += 'X'
                else:
                    metaph += 'S'

            elif ename[i] == 'T':
                if (
                    i > 0
                    and i + 2 <= elen
                    and ename[i + 1] == 'I'
                    and ename[i + 2] in {'A', 'O'}
                ):
                    metaph += 'X'
                elif ename[i + 1 : i + 2] == 'H':
                    metaph += '0'
                elif ename[i + 1 : i + 3] != 'CH':
                    if ename[i - 1 : i] != 'T':
                        metaph += 'T'

            elif ename[i] == 'V':
                metaph += 'F'

            elif ename[i] in 'WY':
                if ename[i + 1 : i + 2] in self._uc_v_set:
                    metaph += ename[i]

            elif ename[i] == 'X':
                metaph += 'KS'

            elif ename[i] == 'Z':
                metaph += 'S'

        return metaph


if __name__ == '__main__':
    import doctest

    doctest.testmod()


1		# Copyright 2014-2020 by Christopher C. Little.
2		# This file is part of Abydos.
3		#
4		# Abydos is free software: you can redistribute it and/or modify
5		# it under the terms of the GNU General Public License as published by
6		# the Free Software Foundation, either version 3 of the License, or
7		# (at your option) any later version.
8		#
9		# Abydos is distributed in the hope that it will be useful,
10		# but WITHOUT ANY WARRANTY; without even the implied warranty of
11		# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
12		# GNU General Public License for more details.
13		#
14		# You should have received a copy of the GNU General Public License
15		# along with Abydos. If not, see <http://www.gnu.org/licenses/>.
16
17		"""abydos.phonetic._metaphone.
18
19	1	Metaphone
20		"""
21
22		from ._phonetic import _Phonetic
23
24	1	__all__ = ['Metaphone']
25
26
27		class Metaphone(_Phonetic):
28		"""Metaphone.
29
30		Based on Lawrence Philips' Pick BASIC code from 1990 :cite:`Philips:1990`,
31	1	as described in :cite:`Philips:1990b`.
32		This incorporates some corrections to the above code, particularly
33	1	some of those suggested by Michael Kuhn in :cite:`Kuhn:1995`.
34
35	1	.. versionadded:: 0.3.6
36	1	"""
37
38	1	_frontv = {'E', 'I', 'Y'}
39		_varson = {'C', 'G', 'P', 'S', 'T'}
40
41	1	def __init__(self, max_length: int = -1) -> None:
42		"""Initialize AlphaSIS instance.
43
44		Parameters
45		----------
46		max_length : int
47		The maximum length of the returned Metaphone code (defaults to 64,
48		but in Philips' original implementation this was 4)
49
50
51		.. versionadded:: 0.4.0
52	1
53	1	"""
54		# Require a max_length of at least 4
55	1	if max_length != -1:
56		self._max_length = max(4, max_length)
57		else:
58		self._max_length = 64
59
60		def encode(self, word: str) -> str:
61		"""Return the Metaphone code for a word.
62
63		Based on Lawrence Philips' Pick BASIC code from 1990
64		:cite:`Philips:1990`, as described in :cite:`Philips:1990b`.
65		This incorporates some corrections to the above code, particularly
66		some of those suggested by Michael Kuhn in :cite:`Kuhn:1995`.
67
68		Parameters
69	1	----------
70	1	word : str
71		The word to transform
72	1
73		Returns
74	1	-------
75		str
76		The Metaphone value
77
78		Examples
79		--------
80		>>> pe = Metaphone()
81		>>> pe.encode('Christopher')
82		'KRSTFR'
83		>>> pe.encode('Niall')
84		'NL'
85		>>> pe.encode('Smith')
86		'SM0'
87		>>> pe.encode('Schmidt')
88		'SKMTT'
89
90
91		.. versionadded:: 0.1.0
92		.. versionchanged:: 0.3.6
93		Encapsulated in class
94
95		"""
96		# As in variable sound--those modified by adding an "h"
97		ename = ''.join(c for c in word.upper() if c.isalnum())
98
99		# Delete non-alphanumeric characters and make all caps
100		if not ename:
101		return ''
102		if ename[0:2] in {'PN', 'AE', 'KN', 'GN', 'WR'}:
103		ename = ename[1:]
104		elif ename[0] == 'X':
105		ename = 'S' + ename[1:]
106		elif ename[0:2] == 'WH':
107		ename = 'W' + ename[2:]
108
109		# Convert to metaphone
110		elen = len(ename) - 1
111	1	metaph = ''
112	1	for i in range(len(ename)):
113		if len(metaph) >= self._max_length:
114		break
115	1	if (
116	1	ename[i] not in {'G', 'T'}
117	1	and i > 0
118	1	and ename[i - 1] == ename[i]
119	1	):
120	1	continue
121	1
122	1	if ename[i] in self._uc_v_set and i == 0:
123		metaph = ename[i]
124
125	1	elif ename[i] == 'B':
126	1	if i != elen or ename[i - 1] != 'M':
127	1	metaph += ename[i]
128	1
129	1	elif ename[i] == 'C':
130	1	if not (
131		i > 0
132		and ename[i - 1] == 'S'
133		and ename[i + 1 : i + 2] in self._frontv
134		):
135	1	if ename[i + 1 : i + 3] == 'IA':
136		metaph += 'X'
137	1	elif ename[i + 1 : i + 2] in self._frontv:
138	1	metaph += 'S'
139		elif i > 0 and ename[i - 1 : i + 2] == 'SCH':
140	1	metaph += 'K'
141	1	elif ename[i + 1 : i + 2] == 'H':
142	1	if (
143		i == 0
144	1	and i + 1 < elen
145	1	and ename[i + 2 : i + 3] not in self._uc_v_set
146		):
147		metaph += 'K'
148		else:
149		metaph += 'X'
150	1	else:
151	1	metaph += 'K'
152	1
153	1	elif ename[i] == 'D':
154	1	if (
155	1	ename[i + 1 : i + 2] == 'G'
156	1	and ename[i + 2 : i + 3] in self._frontv
157	1	):
158		metaph += 'J'
159		else:
160		metaph += 'T'
161
162	1	elif ename[i] == 'G':
163		if ename[i + 1 : i + 2] == 'H' and not (
164	1	i + 1 == elen or ename[i + 2 : i + 3] not in self._uc_v_set
165		):
166	1	continue
167		elif i > 0 and (
168	1	(i + 1 == elen and ename[i + 1] == 'N')
169	1	or (i + 3 == elen and ename[i + 1 : i + 4] == 'NED')
170		):
171		continue
172		elif (
173	1	i - 1 > 0
174		and i + 1 <= elen
175	1	and ename[i - 1] == 'D'
176		and ename[i + 1] in self._frontv
177	1	):
178	1	continue
179		elif ename[i + 1 : i + 2] == 'G':
180		continue
181	1	elif ename[i + 1 : i + 2] in self._frontv:
182	1	if i == 0 or ename[i - 1] != 'G':
183		metaph += 'J'
184		else:
185		metaph += 'K'
186	1	else:
187	1	metaph += 'K'
188
189		elif ename[i] == 'H':
190		if (
191		i > 0
192		and ename[i - 1] in self._uc_v_set
193	1	and ename[i + 1 : i + 2] not in self._uc_v_set
194	1	):
195	1	continue
196	1	elif i > 0 and ename[i - 1] in self._varson:
197	1	continue
198	1	else:
199		metaph += 'H'
200	1
201		elif ename[i] in {'F', 'J', 'L', 'M', 'N', 'R'}:
202	1	metaph += ename[i]
203
204	1	elif ename[i] == 'K':
205	1	if i > 0 and ename[i - 1] == 'C':
206		continue
207		else:
208		metaph += 'K'
209
210	1	elif ename[i] == 'P':
211	1	if ename[i + 1 : i + 2] == 'H':
212	1	metaph += 'F'
213		else:
214	1	metaph += 'P'
215
216	1	elif ename[i] == 'Q':
217	1	metaph += 'K'
218
219	1	elif ename[i] == 'S':
220	1	if (
221	1	i > 0
222		and i + 2 <= elen
223	1	and ename[i + 1] == 'I'
224		and ename[i + 2] in 'OA'
225	1	):
226	1	metaph += 'X'
227	1	elif ename[i + 1 : i + 2] == 'H':
228		metaph += 'X'
229	1	else:
230		metaph += 'S'
231	1
232	1	elif ename[i] == 'T':
233		if (
234	1	i > 0
235	1	and i + 2 <= elen
236		and ename[i + 1] == 'I'
237		and ename[i + 2] in {'A', 'O'}
238		):
239		metaph += 'X'
240		elif ename[i + 1 : i + 2] == 'H':
241	1	metaph += '0'
242	1	elif ename[i + 1 : i + 3] != 'CH':
243	1	if ename[i - 1 : i] != 'T':
244		metaph += 'T'
245	1
246		elif ename[i] == 'V':
247	1	metaph += 'F'
248	1
249		elif ename[i] in 'WY':
250		if ename[i + 1 : i + 2] in self._uc_v_set:
251		metaph += ename[i]
252
253		elif ename[i] == 'X':
254	1	metaph += 'KS'
255	1
256	1	elif ename[i] == 'Z':
257	1	metaph += 'S'
258	1
259	1	return metaph
260
261	1
262	1	if __name__ == '__main__':
263		import doctest
264	1
265		doctest.testmod()
266

chrislit / abydos

abydos.phonetic._metaphone.Metaphone.encode() F last analyzed 2020-12-31 20:10 UTC

Complexity

Size

Duplication

Code Coverage

Importance

How to fix Long Method Complexity

Long Method

Complexity

Duplication Side-by-Side

Filter issues like

abydos.phonetic._metaphone.Metaphone.encode() F
last analyzed 2020-12-31 20:10 UTC