From 2aa4353a62fa2e42139f5083a7f12bcd6923dc7c Mon Sep 17 00:00:00 2001
From: "David R. Mortensen" <dmortens@cs.cmu.edu>
Date: Thu, 11 Jul 2019 17:14:01 -0400
Subject: [PATCH] Corrected ordering of rules

---
 epitran/data/map/find_duplicates.py  | 20 ++++++++++++++++++++
 epitran/data/rules/pinyin-to-ipa.txt | 12 ++++++------
 epitran/xsampa.py                    |  2 ++
 setup.py                             |  6 +++---
 4 files changed, 31 insertions(+), 9 deletions(-)
 create mode 100644 epitran/data/map/find_duplicates.py

diff --git a/epitran/data/map/find_duplicates.py b/epitran/data/map/find_duplicates.py
new file mode 100644
index 00000000..92e62741
--- /dev/null
+++ b/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])
diff --git a/epitran/data/rules/pinyin-to-ipa.txt b/epitran/data/rules/pinyin-to-ipa.txt
index 1435c838..3157d63e 100644
--- a/epitran/data/rules/pinyin-to-ipa.txt
+++ b/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 / _
diff --git a/epitran/xsampa.py b/epitran/xsampa.py
index 393cc0a9..1e8bd215 100644
--- a/epitran/xsampa.py
+++ b/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:
diff --git a/setup.py b/setup.py
index a4453d16..65e6fe7d 100644
--- a/setup.py
+++ b/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='dmortens@cs.cmu.edu',
       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',