Merge pull request #809 from Sedictious/OE

Add Old English syllabifier
cltk · Jul 31, 2018 · 24c2740 · 24c2740
2 parents ea37e7a + 0b7b2ff
commit 24c2740
Show file tree

Hide file tree

Showing 2 changed files with 50 additions and 0 deletions.
diff --git a/cltk/corpus/old_english/syllabifier.py b/cltk/corpus/old_english/syllabifier.py
@@ -0,0 +1,35 @@
+"""
+Sonority phoneme hierarchy for Old English
+"""
+
+Syllabifier = {
+    'a': 1,
+    'e': 1,
+    'i': 1,
+    'o': 1,
+    'u': 1,
+    'ø': 1,
+    'æ': 1,
+    'm': 2,
+    'n': 2,
+    'p': 2,
+    'b': 2,
+    'c': 2,
+    't': 2,
+    'd': 2,
+    'ð': 2,
+    'k': 2,
+    'g': 2,
+    'f': 3,
+    'þ': 3,
+    's': 3,
+    'h': 3,
+    'v': 3,
+    'x': 3,
+    'z': 3,
+    'l': 4,
+    'j': 4,
+    'w': 4,
+    'ƿ': 4,
+    'r': 5
+}
diff --git a/cltk/phonology/syllabify.py b/cltk/phonology/syllabify.py
@@ -8,6 +8,7 @@
 from cltk.exceptions import InputError
 from cltk.corpus.middle_english.syllabifier import Syllabifier as ME_Syllabifier
 from cltk.corpus.middle_high_german.syllabifier import Syllabifier as MHG_Syllabifier
+from cltk.corpus.old_english.syllabifier import Syllabifier as OE_Syllabifier
 from cltk.corpus.old_norse.syllabifier import hierarchy as OLD_NORSE_HIERARCHY
 
 LOG = logging.getLogger(__name__)
@@ -80,6 +81,15 @@ def __init__(self, low_vowels=None, mid_vowels=None, high_vowels=None, flaps=Non
 
             self.set_hierarchy(hierarchy)
             self.set_vowels(hierarchy[0])
+
+        elif language == 'old english':
+            hierarchy = [[] for _ in range(len(set(OE_Syllabifier.values())))]
+
+            for k in OE_Syllabifier:
+                hierarchy[OE_Syllabifier[k] - 1].append(k)
+
+            self.set_hierarchy(hierarchy)
+            self.set_vowels(hierarchy[0])
 
         elif language == 'middle high german':
             hierarchy = [[] for _ in range(len(set(MHG_Syllabifier.values())))]
@@ -201,6 +211,11 @@ def syllabify_SSP(self, word):
             >>> s.syllabify("huntyng")
             ['hun', 'tyng']
             
+            >>> s = Syllabifier(language='old english')
+            
+            >>> s.syllabify("arcebiscop")
+            ['ar', 'ce', 'bis', 'cop']
+            
             The break_geminants parameter ensures a breakpoint is placed between geminants:
             
             >>> geminant_s = Syllabifier(break_geminants=True)