Merge pull request #825 from PyThaiNLP/add-thai_synonym

Add thai_synonym

Author: wannaphong

docs/api/corpus.rst

@@ -21,6 +21,7 @@ Modules
 .. autofunction:: thai_words
 .. autofunction:: thai_wsd_dict
 .. autofunction:: thai_orst_words
+.. autofunction:: thai_synonym
 .. autofunction:: thai_syllables
 .. autofunction:: thai_negations
 .. autofunction:: thai_family_names

pythainlp/corpus/__init__.py

@@ -37,6 +37,7 @@
     "thai_female_names",
     "thai_male_names",
     "thai_negations",
+    "thai_synonym",
     "thai_stopwords",
     "thai_syllables",
     "thai_words",
@@ -110,6 +111,7 @@ def corpus_db_path() -> str:
     thai_female_names,
     thai_male_names,
     thai_negations,
+    thai_synonym,
     thai_stopwords,
     thai_syllables,
     thai_words,

pythainlp/corpus/common.py

@@ -28,6 +28,7 @@
     "thai_words",
     "thai_dict",
     "thai_wsd_dict",
+    "thai_synonym",
 ]
 
 from typing import FrozenSet, List, Union
@@ -64,6 +65,7 @@
 
 _THAI_DICT = {}
 _THAI_WSD_DICT = {}
+_THAI_SYNONYM = None
 
 
 def countries() -> FrozenSet[str]:
@@ -270,7 +272,7 @@ def thai_dict() -> dict:
     <https://pythainlp.github.io/pythainlp-corpus/thai_dict.html>`_)
 
     :return: Thai word with part-of-speech type and definition
-    :rtype: :class:`frozenset`
+    :rtype: dict
     """
     global _THAI_DICT
     if _THAI_DICT == {}:
@@ -292,7 +294,7 @@ def thai_wsd_dict() -> dict:
     <https://pythainlp.github.io/pythainlp-corpus/thai_dict.html>`_)
 
     :return: Thai word with part-of-speech type and definition
-    :rtype: :class:`frozenset`
+    :rtype: dict
     """
     global _THAI_WSD_DICT
     if _THAI_WSD_DICT == {}:
@@ -308,4 +310,27 @@ def thai_wsd_dict() -> dict:
                 _THAI_WSD_DICT["word"].append(i)
                 _THAI_WSD_DICT["meaning"].append(_use)
 
-    return _THAI_WSD_DICT
+    return _THAI_WSD_DICT
+
+
+def thai_synonym() -> dict:
+    """
+    Return Thai synonym.
+    \n(See: `thai_synonym\
+    <https://pythainlp.github.io/pythainlp-corpus/thai_synonym.html>`_)
+
+    :return: Thai word with part-of-speech type and synonym
+    :rtype: dict
+    """
+    global _THAI_SYNONYM
+    if _THAI_SYNONYM == None:
+        import csv
+        _THAI_SYNONYM = {"word":[], "pos":[], "synonym":[]}
+        with open(get_corpus_path("thai_synonym"), newline="\n", encoding="utf-8") as csvfile:
+            reader = csv.DictReader(csvfile, delimiter=",")
+            for row in reader:
+                _THAI_SYNONYM["word"].append(row["word"])
+                _THAI_SYNONYM["pos"].append(row["pos"])
+                _THAI_SYNONYM["synonym"].append(row["synonym"].split("|"))
+
+    return _THAI_SYNONYM

tests/test_corpus.py

@@ -20,6 +20,7 @@
     thai_negations,
     thai_stopwords,
     thai_syllables,
+    thai_synonym,
     thai_words,
     tnc,
     ttc,
@@ -44,6 +45,7 @@ def test_corpus(self):
         self.assertIsInstance(countries(), frozenset)
         self.assertIsInstance(provinces(), frozenset)
         self.assertIsInstance(provinces(details=True), list)
+        self.assertIsInstance(thai_synonym(), dict)
         self.assertEqual(
             len(provinces(details=False)), len(provinces(details=True))
         )