Add live-dead syllable classification

.github/workflows/deploy_docs.yml

@@ -7,13 +7,13 @@ on:
 jobs:
   release:
     name: Build
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     steps:
     - uses: actions/checkout@v1
-    - name: Set up Python 3.7
+    - name: Set up Python 3.8
       uses: actions/setup-python@v1
       with:
-        python-version: 3.7
+        python-version: 3.8
     - name: Install dependencies
       run: |
         python -m pip install --upgrade pip

.github/workflows/lint.yml

@@ -20,7 +20,7 @@ jobs:
     runs-on: ubuntu-20.04
     strategy:
       matrix:
-        python-version: [3.7]
+        python-version: [3.8]
 
     steps:
     - uses: actions/checkout@v2

.github/workflows/macos-test.yml

@@ -20,7 +20,7 @@ jobs:
       fail-fast: false
       matrix:
         os: [macos-latest]
-        python-version: [3.7]
+        python-version: [3.8]
 
     steps:
     - uses: actions/checkout@v2

.github/workflows/pypi-publish.yml

@@ -7,10 +7,10 @@ on:
 jobs:
   deploy:
 
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     strategy:
       matrix:
-        python-version: [3.7]
+        python-version: [3.8]
 
     steps:
     - uses: actions/checkout@v2

.github/workflows/pypi-test.yml

@@ -7,10 +7,10 @@ on:
 jobs:
   build:
 
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     strategy:
       matrix:
-        python-version: [3.7]
+        python-version: [3.8]
 
     steps:
     - uses: actions/checkout@v2

.github/workflows/test.yml

@@ -15,10 +15,10 @@ on:
 jobs:
   build:
 
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     strategy:
       matrix:
-        python-version: [3.7]
+        python-version: [3.8]
 
     steps:
     - uses: actions/checkout@v2

Dockerfile

@@ -1,4 +1,4 @@
-FROM python:3.7-slim-buster
+FROM python:3.8-slim-buster
 
 COPY . .
 

docker_requirements.txt

@@ -24,8 +24,9 @@ pyicu==2.8
 deepcut==0.7.0.0
 h5py==3.1.0
 tensorflow==2.7.2
-pandas==0.24
+pandas==1.4.*
 tltk==1.3.8
 OSKut==1.3
 nlpo3==1.2.2
 thai-nner==0.3
+spacy==2.3.*

docs/api/util.rst

@@ -32,6 +32,7 @@ Modules
 .. autofunction:: remove_tonemark
 .. autofunction:: remove_zw
 .. autofunction:: reorder_vowels
+.. autofunction:: sound_syllable
 .. autofunction:: text_to_arabic_digit
 .. autofunction:: text_to_num
 .. autofunction:: text_to_thai_digit

pythainlp/util/__init__.py

@@ -43,6 +43,7 @@
     "time_to_thaiword",
     "text_to_num",
     "words_to_num",
+    "sound_syllable",
 ]
 
 from pythainlp.util.collate import collate
@@ -89,3 +90,4 @@
 from pythainlp.util.time import thai_time, thaiword_to_time, time_to_thaiword
 from pythainlp.util.trie import Trie, dict_trie
 from pythainlp.util.wordtonum import thaiword_to_num, text_to_num, words_to_num
+from pythainlp.util.syllable import sound_syllable

pythainlp/util/syllable.py

@@ -0,0 +1,113 @@
+# -*- coding: utf-8 -*-
+"""
+Syllable tools
+"""
+import re
+from pythainlp import thai_consonants
+
+spelling_class = {
+    "กง": list("ง"),
+    "กม": list("ม"),
+    "เกย": list("ย"),
+    "เกอว": list("ว"),
+    "กน": list("นญณรลฬ"),
+    "กก": list("กขคฆ"),
+    "กด": list("ดจชซฎฏฐฑฒตถทธศษส"),
+    "กบ": list("บปภพฟ")
+}
+
+thai_consonants_all = list(thai_consonants)
+thai_consonants_all.remove("อ")
+
+_temp = list(
+    "".join(["".join(spelling_class[i]) for i in spelling_class.keys()])
+)
+not_spelling_class = [j for j in thai_consonants_all if j not in _temp]
+
+# vowel's short sound
+short = "ะัิึุ"
+re_short = re.compile("เ(.*)ะ|แ(.*)ะ|เ(.*)อะ|โ(.*)ะ|เ(.*)าะ", re.U)
+pattern = re.compile("เ(.*)า", re.U)  # เ-า is live syllable
+
+_check_1 = []
+# these spelling consonant are live syllable.
+for i in ["กง", "กน", "กม", "เกย", "เกอว"]:
+    _check_1.extend(spelling_class[i])
+# these spelling consonant are dead syllable.
+_check_2 = spelling_class["กก"]+spelling_class["กบ"]+spelling_class["กด"]
+
+
+def sound_syllable(syllable: str) -> str:
+    """
+    Sound syllable classification
+
+    This function is sound syllable classification.
+    It is live syllable or dead syllable.
+
+    :param str syllable: Thai syllable
+    :return: syllable's type (live or dead)
+    :rtype: str
+
+    :Example:
+    ::
+
+        from pythainlp.util import sound_syllable
+
+        print(sound_syllable("มา"))
+        # output: live
+
+        print(sound_syllable("เลข"))
+        # output: dead
+    """
+    # get consonants
+    consonants = [i for i in syllable if i in list(thai_consonants_all)]
+    # get spelling consonants
+    spelling_consonant = consonants[-1]
+    # if len of syllable < 2
+    if len(syllable) < 2:
+        return "dead"
+    elif (
+        (
+            spelling_consonant in _check_2)
+            and
+            (
+                any((c in set("าีืแูาเโ")) for c in syllable) == False
+                and any((c in set("ำใไ")) for c in syllable) == False
+                and pattern.findall(syllable) != True
+            )
+    ):
+        return "dead"
+    elif any((c in set("าีืแูาโ")) for c in syllable):  # in syllable:
+        if spelling_consonant != syllable[-1]:
+            return "live"
+        elif spelling_consonant in _check_1:
+            return "live"
+        elif spelling_consonant in _check_2:
+            return "dead"
+        elif (
+            re_short.findall(syllable)
+            or
+            any((c in set(short)) for c in syllable)
+        ):
+            return "dead"
+        return "live"
+    elif any((c in set("ำใไ")) for c in syllable):
+        return "live"  # if these vowel's long sound are live syllable
+    elif pattern.findall(syllable):  # if it is เ-า
+        return "live"
+    elif spelling_consonant in _check_1:
+        if (
+            re_short.findall(syllable)
+            or
+            any((c in set(short)) for c in syllable)
+        ) and len(consonants) < 2:
+            return "dead"
+        return "live"
+    elif (
+        re_short.findall(syllable)  # if found vowel's short sound
+        or
+        any((c in set(short)) for c in syllable)  # consonant in short
+    ):
+        return "dead"
+    else:
+        return "dead"

tests/test_util.py

@@ -48,6 +48,7 @@
     thai_keyboard_dist,
     text_to_num,
     words_to_num,
+    sound_syllable,
 )
 
 
@@ -661,3 +662,41 @@ def test_emoji_to_thai(self):
             emoji_to_thai("🇹🇭 นี่คือธงประเทศไทย"),
             ":ธง_ไทย: นี่คือธงประเทศไทย",
         )
+
+    def test_sound_syllable(self):
+        test = [
+            ("มา", "live"),
+            ("ดู", "live"),
+            ("ปู", "live"),
+            ("เวลา", "live"),
+            ("ปี", "live"),
+            ("จำ", "live"),
+            ("น้ำ", "live"),
+            ("ใช่", "live"),
+            ("เผ่า", "live"),
+            ("เสา", "live"),
+            ("ไป", "live"),
+            ("จริง", "live"),
+            ("กิน", "live"),
+            ("กำ", "live"),
+            ("มา", "live"),
+            ("สาว", "live"),
+            ("ฉุย", "live"),
+            ("ธุ", "dead"),
+            ("ระ", "dead"),
+            ("กะ", "dead"),
+            ("ทิ", "dead"),
+            ("เกะ", "dead"),
+            ("กะ", "dead"),
+            ("บท", "dead"),
+            ("บาท", "dead"),
+            ("ลาภ", "dead"),
+            ("เมฆ", "dead"),
+            ("เลข", "dead"),
+            ("ธูป", "dead"),
+            ("บ", "dead"),
+            ("บ่", "dead"),
+            ("ก็", "dead"),
+        ]
+        for i, j in test:
+            self.assertEqual(sound_syllable(i), j)