Corrected ordering of rules

epitran/data/map/find_duplicates.py

@@ -0,0 +1,20 @@
+#!/usr/bin/env python3
+
+import csv
+import sys
+from collections import defaultdict
+
+def main(fn):
+    mappings = defaultdict(list)
+    with open(fn, encoding='utf-8') as f:
+        reader = csv.reader(f)
+        next(reader)
+        for orth, phon in reader:
+            mappings[orth].append(phon)
+    print(mappings)
+    for orth, phons in mappings.items():
+        if len(phons) > 1:
+            print(orth)
+
+if __name__ == '__main__':
+    main(sys.argv[1])

epitran/data/rules/pinyin-to-ipa.txt

@@ -1,14 +1,14 @@
 ng -> ŋ / _
-b -> p / _
-d -> t / _
-g -> k / _
 p -> pʰ / _
 t -> tʰ / _
 k -> kʰ / _
-zh -> t͡ʂ / _
+b -> p / _
+d -> t / _
+g -> k / _
+zh -> ʈ͡ʂ / _
 z -> t͡s / _
 j -> t͡ɕ / _
-ch -> t͡ʂʰ / _
+ch -> ʈ͡ʂʰ / _
 c -> t͡sʰ / _
 q -> t͡ɕʰ / _
 m	-> m / _
@@ -20,6 +20,7 @@ x -> ɕ / _
 h -> x / _
 l	-> l / _
 r -> ɻ / _
+eɻ -> ɻ̩ / _
 w -> 0 / _ u
 w -> w / _
 y -> 0 / _ i
@@ -30,7 +31,6 @@ i -> ɻ̩ / (t͡ʂ|t͡ʂʰ|ʂ) _
 a -> a / _
 o -> uo / (p|pʰ|f|m) _
 o -> o / _
-eɻ -> ɻ̩ / _
 e -> e / _
 i -> i / _
 u[:] -> y / _

epitran/xsampa.py

@@ -3,6 +3,7 @@
                         unicode_literals)
 
 import os.path
+import unicodedata
 
 import pkg_resources
 
@@ -54,6 +55,7 @@ def ipa2xs(self, ipa):
             Non-IPA segments are skipped.
         """
         xsampa = []
+        ipa = unicodedata.normalize('NFD', ipa)
         while ipa:
             token = self.longest_prefix(ipa)
             if token:

setup.py

@@ -1,17 +1,17 @@
 from setuptools import setup
 
 setup(name='epitran',
-      version='0.60',
+      version='0.61',
       description='Tools for transcribing languages into IPA.',
       url='http://github.com/dmort27/epitran',
-      download_url='https://github.com/dmort27/epitran/archive/0.60.tar.gz',
+      download_url='https://github.com/dmort27/epitran/archive/0.61.tar.gz',
       author='David R. Mortensen',
       author_email='[email protected]',
       license='MIT',
       install_requires=['setuptools',
                         'unicodecsv',
                         'regex',
-                        'panphon>=0.12',
+                        'panphon>=0.15',
                         'marisa_trie'],
       extras_require={':python_version<"3.0"': ['subprocess32']},
       scripts=['epitran/bin/epitranscribe.py',