Add some examples in Chinese

Makefile

@@ -26,8 +26,9 @@ setup-virtualenv:
 # Setup prerequisites
 setup: setup-virtualenv
 	$(pip) install --requirement=requirements.txt
-	$(python) -m spacy download en
-	$(python) -m spacy download de
+	$(python) -m spacy download en_core_web_md
+	$(python) -m spacy download de_core_news_md
+	$(python) -m spacy download zh_core_web_md
 
 # Setup Flair
 setup-flair: setup-virtualenv

README.rst

@@ -74,6 +74,10 @@ Usage
     python wq.py Berliner Tagestemperatur um 11:00 Uhr
     python wq.py Berliner Tagestemperaturen um 11:00 Uhr
 
+    # Chinese
+    python wq.py 成都的雨
+    python wq.py 南昌的温度
+
 
 .. _Flair: https://pypi.org/project/flair/
 .. _spaCy: https://pypi.org/project/spacy/

requirements.txt

@@ -1,2 +1,7 @@
+numpy==1.21.2
 spacy<3
 spacy-langdetect
+
+# https://github.com/lancopku/pkuseg-python/issues/148
+# https://github.com/explosion/spaCy/discussions/7370#discussioncomment-455375
+https://github.com/lancopku/pkuseg-python/archive/d581c95.zip

test_wq.py

@@ -88,3 +88,31 @@ def test_german_ozon():
 def test_german_particulates():
     result = analyze_spacy("Feinstaub in Stuttgart am 17.09.2020")
     assert result == Result(where="Stuttgart", when="am 17.09.2020", what="Feinstaub")
+
+
+def test_chinese_rain():
+    # Rain in Chengdu
+    # Translated using DeepL
+    result = analyze_spacy("成都的雨")
+    assert result == Result(where='成都', when='now', what='雨')
+
+
+def test_chinese_temperature_now():
+    # Temperature in Nanchang
+    # Translated using DeepL
+    result = analyze_spacy("南昌的温度")
+    assert result == Result(where='南昌', when='now', what='温度')
+
+
+def test_chinese_temperature_tomorrow():
+    # Temperature in Nanchang tomorrow
+    # Translated using DeepL
+    result = analyze_spacy("南昌明天的温度")
+    assert result == Result(where='南昌', when='明天', what='温度')
+
+
+def test_chinese_temperature_on_date():
+    # Temperature in Nanchang on 2020-09-17
+    # Translated using DeepL
+    result = analyze_spacy("2020年9月17日南昌市的温度")
+    assert result == Result(where='南昌市', when='2020年9月17日', what='温度')

wq.py

@@ -61,6 +61,14 @@ def detect_language(text: str):
     doc = nlp(text)
     language = doc._.language["language"]
 
+    # Correct language misdetections.
+    if language == "ko":
+        language = "zh-cn"
+
+    # Map language to model.
+    if language == "zh-cn":
+        language = "zh_core_web_md"
+
     return language
 
 
@@ -199,6 +207,13 @@ def improve_with_heuristics(nlp, expression, sentence):
         except IndexError:
             pass
 
+    # "Temperature in Nanchang on 2020-09-17" in Chinese: "2020年9月17日南昌市的温度"
+    if what in when:
+        for noun in dh.find_tokens("NOUN"):
+            if noun.lemma_ in when:
+                continue
+            what = noun.lemma_
+
     result = Result(where=where, when=when, what=what)
     return result