First cut/attempt at removing syllabipy dependency

abydos/distance/_aline.py

@@ -1459,7 +1459,10 @@ def _record(score: float, out: List[Tuple[str, str]]) -> None:
                 ):
                     loc_out = deepcopy(out)
                     loc_out.append(
-                        (src_tok[i - 1], tar_tok[j - 2] + tar_tok[j - 1],)
+                        (
+                            src_tok[i - 1],
+                            tar_tok[j - 2] + tar_tok[j - 1],
+                        )
                     )
                     _retrieve(
                         i - 1,
@@ -1500,7 +1503,10 @@ def _record(score: float, out: List[Tuple[str, str]]) -> None:
                 ):
                     loc_out = deepcopy(out)
                     loc_out.append(
-                        (src_tok[i - 2] + src_tok[i - 1], tar_tok[j - 1],)
+                        (
+                            src_tok[i - 2] + src_tok[i - 1],
+                            tar_tok[j - 1],
+                        )
                     )
                     _retrieve(
                         i - 2,

abydos/distance/_synoname.py

@@ -740,13 +740,15 @@ def _approx_c() -> Tuple[bool, float]:
                         (
                             initial_diff
                             == self._lev.dist_abs(
-                                src_initials_str, tar_initials_str,
+                                src_initials_str,
+                                tar_initials_str,
                             )
                         )
                         or (
                             initial_diff
                             == self._lev.dist_abs(
-                                src_initials_str, tar_initials_str,
+                                src_initials_str,
+                                tar_initials_str,
                             )
                         )
                     ):

abydos/phonetic/_henry_early.py

@@ -197,10 +197,15 @@ def encode(self, word: str) -> str:
             # IId
             elif char == 'H' and prev in self._uc_c_set:
                 continue
-            elif char in self._uc_c_set - {
-                'L',
-                'R',
-            } and nxch in self._uc_c_set - {'L', 'R'}:
+            elif (
+                char
+                in self._uc_c_set
+                - {
+                    'L',
+                    'R',
+                }
+                and nxch in self._uc_c_set - {'L', 'R'}
+            ):
                 continue
             elif char == 'L' and nxch in {'M', 'N'}:
                 continue

abydos/stemmer/_paice_husk.py

@@ -240,9 +240,12 @@ def stem(self, word: str) -> str:
                 if word[-n:] in self._rule_table[n]:
                     accept = False
                     for rule in self._rule_table[n][word[-n:]]:
-                        (word, accept, intact, terminate,) = self._apply_rule(
-                            word, rule, intact, terminate
-                        )
+                        (
+                            word,
+                            accept,
+                            intact,
+                            terminate,
+                        ) = self._apply_rule(word, rule, intact, terminate)
                         if accept:
                             break
 

abydos/tokenizer/_character.py

@@ -33,7 +33,8 @@ class CharacterTokenizer(_Tokenizer):
     """
 
     def __init__(
-        self, scaler: Optional[Union[str, Callable[[float], float]]] = None,
+        self,
+        scaler: Optional[Union[str, Callable[[float], float]]] = None,
     ) -> None:
         """Initialize tokenizer.
 

abydos/tokenizer/_legalipy.py

@@ -24,13 +24,13 @@
 from ._tokenizer import _Tokenizer
 
 try:
-    from syllabipy.legalipy import LegaliPy
-    from syllabipy.legalipy import getOnsets as gen_onsets  # noqa: N813
+    from nltk.tokenize import LegalitySyllableTokenizer
+    from nltk.tokenize.legality_principle import find_legal_onsets
 except ImportError:  # pragma: no cover
     # If the system lacks the SyllabiPy library, that's fine, but SyllabiPy
     # tokenization won't be supported.
-    LegaliPy = None  # type: ignore
-    gen_onsets = None  # type: ignore
+    LegalitySyllableTokenizer = None  # type: ignore
+    find_legal_onsets = None  # type: ignore
 
 
 class LegaliPyTokenizer(_Tokenizer):
@@ -40,7 +40,8 @@ class LegaliPyTokenizer(_Tokenizer):
     """
 
     def __init__(
-        self, scaler: Optional[Union[str, Callable[[float], float]]] = None,
+        self,
+        scaler: Optional[Union[str, Callable[[float], float]]] = None,
     ) -> None:
         """Initialize Tokenizer.
 
@@ -65,12 +66,11 @@ def __init__(
         .. versionadded:: 0.4.0
 
         """
-        if LegaliPy is None:
+        if LegalitySyllableTokenizer is None:
             raise TypeError(  # pragma: no cover
                 'LegaliPy tokenizer requires installation of SyllabiPy'
                 + ' package.'
             )
-
         super(LegaliPyTokenizer, self).__init__(scaler)
 
         self._onsets = ['']
@@ -99,7 +99,7 @@ def train_onsets(
         .. versionadded:: 0.4.0
 
         """
-        new_onsets = gen_onsets(text, threshold, clean)
+        new_onsets = self.tokenizer.legal_onsets
         if append:
             self._onsets = list(set(self._onsets + new_onsets))
         else:
@@ -132,8 +132,10 @@ def tokenize(self, string: str, ipa: bool = False) -> 'LegaliPyTokenizer':
         self._string = string
 
         self._ordered_tokens = []
-        for word in string.split():
-            self._ordered_tokens += LegaliPy(word, self._onsets)
+        words = string.split()
+        self.tokenizer = LegalitySyllableTokenizer(words)
+        for word in words:
+            self._ordered_tokens += self.tokenizer.tokenize(word, self._onsets)
         if not self._ordered_tokens:
             self._ordered_tokens = [self._string]
 

abydos/tokenizer/_saps.py

@@ -33,7 +33,8 @@ class SAPSTokenizer(_Tokenizer):
     """
 
     def __init__(
-        self, scaler: Optional[Union[str, Callable[[float], float]]] = None,
+        self,
+        scaler: Optional[Union[str, Callable[[float], float]]] = None,
     ) -> None:
         """Initialize Tokenizer.
 

abydos/tokenizer/_sonoripy.py

@@ -38,7 +38,8 @@ class SonoriPyTokenizer(_Tokenizer):
     """
 
     def __init__(
-        self, scaler: Optional[Union[str, Callable[[float], float]]] = None,
+        self,
+        scaler: Optional[Union[str, Callable[[float], float]]] = None,
     ) -> None:
         """Initialize Tokenizer.
 

tests/distance/test_distance_jaccard.py

@@ -164,7 +164,8 @@ def test_jaccard_tanimoto_coeff(self):
             self.cmp_q2.tanimoto_coeff('', 'neilsen'), float('-inf')
         )
         self.assertAlmostEqual(
-            self.cmp_q2.tanimoto_coeff('nelson', 'neilsen'), log2(4 / 11),
+            self.cmp_q2.tanimoto_coeff('nelson', 'neilsen'),
+            log2(4 / 11),
         )
 
         # supplied q-gram tests

tests/phonetic/test_phonetic_daitch_mokotoff.py

@@ -103,7 +103,8 @@ def test_daitch_mokotoff(self):
             '68',
         )
         self.assertEqual(
-            DaitchMokotoff(max_length=0, zero_pad=False).encode('Niall'), '68',
+            DaitchMokotoff(max_length=0, zero_pad=False).encode('Niall'),
+            '68',
         )
         self.assertEqual(
             DaitchMokotoff(max_length=0, zero_pad=True).encode('Niall'),

tox.ini

@@ -1,7 +1,7 @@
 [tox]
 envlist =
     black
-    py37
+    py38
     doctest
     regression
     fuzz
@@ -28,7 +28,7 @@ deps =
 commands = nosetests []
 
 [testenv:doctest]
-basepython = python3.7
+basepython = python3.8
 setenv =
     NOSE_WITH_COVERAGE=0
     NOSE_WITH_DOCTEST=1
@@ -44,7 +44,7 @@ passenv = {[testenv:doctest]passenv}
 commands = {[testenv:doctest]commands}
 
 [testenv:regression]
-basepython = python3.7
+basepython = python3.8
 commands = nosetests {toxinidir}/tests/regression --processes=-1 \
                --process-timeout=1200 --process-restartworker -c=0 -v []
 
@@ -55,7 +55,7 @@ deps =
 commands = {[testenv:regression]commands}
 
 [testenv:fuzz]
-basepython = python3.7
+basepython = python3.8
 commands = nosetests {toxinidir}/tests/fuzz --processes=-1 \
                --process-timeout=1200 --process-restartworker -c=0 -v []
 
@@ -68,32 +68,32 @@ commands = {[testenv:fuzz]commands}
 [testenv:black]
 depends =
 changedir = {toxinidir}
-basepython = python3.7
+basepython = python3.8
 skip_install = true
 deps = black
 commands = black .
 
 [testenv:pylint]
-basepython = python3.7
+basepython = python3.8
 skip_install = true
 deps = pylint
 commands = {toxinidir}/helpers/call_and_write_log.py \
                "pylint --rcfile=setup.cfg abydos" 0
 
 [testenv:pydocstyle]
-basepython = python3.7
+basepython = python3.8
 skip_install = true
 deps = pydocstyle
 commands = {toxinidir}/helpers/call_and_write_log.py "pydocstyle --count ." 0
 
 [testenv:pycodestyle]
-basepython = python3.7
+basepython = python3.8
 skip_install = true
 deps = pycodestyle
 commands = {toxinidir}/helpers/call_and_write_log.py "pycodestyle ." 0
 
 [testenv:flake8]
-basepython = python3.7
+basepython = python3.8
 skip_install = true
 deps =
     flake8
@@ -134,7 +134,7 @@ commands = {toxinidir}/helpers/call_and_write_log.py \
                "flake8 {toxinidir} --htmldir={toxinidir}/flake8" 0
 
 [testenv:mypy]
-basepython = python3.7
+basepython = python3.8
 deps =
     mypy
     lxml
@@ -145,14 +145,14 @@ mypy_paths =
 commands = mypy {posargs:{[testenv:mypy]mypy_paths}} --txt-report mypy --html-report mypy
 
 [testenv:doc8]
-basepython = python3.7
+basepython = python3.8
 skip_install = true
 deps = doc8
 commands = {toxinidir}/helpers/call_and_write_log.py "doc8 {toxinidir}" 0
 
 [testenv:docs]
 changedir = docs
-basepython = python3.7
+basepython = python3.8
 whitelist_externals=make
 deps =
     sphinx
@@ -176,7 +176,7 @@ commands = {toxinidir}/helpers/call_and_write_log.py \
                 "sloccount --datadir .tox/sloccount/tmp ./abydos" 0
 
 [testenv:badges]
-basepython = python3.7
+basepython = python3.8
 depends =
     pylint
     pydocstyle
@@ -188,7 +188,7 @@ deps =
 commands = python {toxinidir}/badge_update.py
 
 [testenv:build]
-basepython = python3.7
+basepython = python3.8
 skip_install = true
 deps =
     wheel
@@ -199,7 +199,7 @@ commands =
     twine check dist/*
 
 [testenv:release]
-basepython = python3.7
+basepython = python3.8
 skip_install = true
 deps =
     {[testenv:build]deps}
@@ -208,7 +208,7 @@ commands =
     twine upload --skip-existing --repository pypi dist/*
 
 [testenv:release-test]
-basepython = python3.7
+basepython = python3.8
 skip_install = true
 deps =
     {[testenv:build]deps}