From 3691a4248d7fab7842f05164760395f1351bd05a Mon Sep 17 00:00:00 2001
From: Gerardo Bort <gerardobort@gmail.com>
Date: Sun, 18 Feb 2018 11:31:13 -0300
Subject: [PATCH] Adding properties param for tokensregex, semgrex and regex

---
 example.py           |  5 +++++
 pycorenlp/corenlp.py | 20 ++++++++++++--------
 2 files changed, 17 insertions(+), 8 deletions(-)

diff --git a/example.py b/example.py
index 4b7a358..82b3430 100644
--- a/example.py
+++ b/example.py
@@ -14,3 +14,8 @@
     print(output)
     output = nlp.semgrex(text, pattern='{tag: VBD}', filter=False)
     print(output)
+    output = nlp.semgrex(text, pattern='{ner: PERS}', filter=False, properties={
+        'annotators': 'tokenize,ssplit,ner,depparse',
+        'pipelineLanguage': 'en',
+    })
+    print(output)
diff --git a/pycorenlp/corenlp.py b/pycorenlp/corenlp.py
index 6eb2175..7176ea3 100644
--- a/pycorenlp/corenlp.py
+++ b/pycorenlp/corenlp.py
@@ -36,21 +36,25 @@ def annotate(self, text, properties=None):
                 pass
         return output
 
-    def tokensregex(self, text, pattern, filter):
-        return self.regex('/tokensregex', text, pattern, filter)
+    def tokensregex(self, text, pattern, filter, properties = None):
+        return self.regex('/tokensregex', text, pattern, filter, properties)
 
-    def semgrex(self, text, pattern, filter):
-        return self.regex('/semgrex', text, pattern, filter)
+    def semgrex(self, text, pattern, filter, properties = None):
+        return self.regex('/semgrex', text, pattern, filter, properties)
 
-    def regex(self, endpoint, text, pattern, filter):
+    def regex(self, endpoint, text, pattern, filter, properties = None):
+        assert isinstance(text, str)
+        data = text.encode()
         r = requests.get(
             self.server_url + endpoint, params={
-                'pattern':  pattern,
+                'pattern': pattern,
+                'properties': str(properties or {}),
                 'filter': filter
-            }, data=text)
+            }, data=data)
+        r.encoding = 'utf-8'
         output = r.text
         try:
-            output = json.loads(r.text)
+            output = json.loads(r.text, encoding='utf-8', strict=True)
         except:
             pass
         return output