add find_job_location_position

VerdictCut/tool/__init__.py

@@ -1,3 +1,4 @@
 from .normalize_position import normalize_position
 from .find_name_law import find_name_and_law
-from .get_units_list import get_units_list
+from .get_units_list import get_units_list
+from .find_job_location import find_job_location

VerdictCut/tool/find_job_location.py

@@ -0,0 +1,82 @@
+import re
+
+def find_job_location_position(judgement,job_list,location_list, break_line='\r\n'):
+    all_data_dict={}
+    clean_judgement = re.sub(break_line, "", re.sub(r"\s+", "", judgement))
+    judgement_list=re.split("，|。",clean_judgement)
+    # 先找有matched的
+    matched_location_list=[]
+    matched_job_list=[]
+    unmatched_location_list=location_list.copy()
+    unmatched_job_list=job_list.copy()
+    tuple_list=[]
+    for paragraph in judgement_list:
+        for job in job_list:
+            for location in location_list:
+                if location in paragraph and job in paragraph:
+                    if (location,job) not in tuple_list:
+                        tuple_list.append((location,job))
+                        matched_location_list.append(location)
+                        matched_job_list.append(job)       
+                    if location in unmatched_location_list:
+                        unmatched_location_list.remove(location)
+                    if job in unmatched_job_list:
+                        unmatched_job_list.remove(job)
+    # 長度要一樣
+    assert len(matched_location_list)==len(matched_job_list)
+
+    all_matched_location_list=find_position_dict("location",matched_location_list,judgement)
+    all_matched_job_list=find_position_dict("job",matched_job_list,judgement)
+    all_matched_pair_list=[]
+    for i in range(len(all_matched_location_list)):
+        matched_pair_list=[]
+        matched_pair_list.append(all_matched_location_list[i])
+        matched_pair_list.append(all_matched_job_list[i])
+        all_matched_pair_list.append(matched_pair_list.copy())
+    all_data_dict["matched"]=all_matched_pair_list
+
+    all_unmatched_location_list=find_position_dict("location",unmatched_location_list,judgement)
+    all_unmatched_job_list=find_position_dict("job",unmatched_job_list,judgement)
+
+    all_unmatched_dict={}
+    all_unmatched_dict["location"]=all_unmatched_location_list
+    all_unmatched_dict["job"]=all_unmatched_job_list
+    all_data_dict["unmatched"]=all_unmatched_dict
+
+    return all_data_dict
+
+def regularize(text):
+    regex_list="\s*"
+    regex_text=""
+    for i,token in enumerate(text):
+        if i==0:
+            regex_text=regex_list+token+regex_list
+        else:
+            regex_text=regex_text+token+regex_list
+
+    return regex_text
+
+def find_position_dict(kind,text_list,judgement):
+    if len(text_list)==0:
+        return []
+    position_dict_list=[]
+    for text in text_list:
+        text_position_dict={}
+        text_position=re.search(regularize(text),judgement)
+        if text_position==None:
+            continue
+        text_position_dict["type"]=kind
+        text_position_dict["name"]=text
+        text_position_dict["start"]=text_position.start()
+        text_position_dict["end"]=text_position.end()
+        position_dict_list.append(text_position_dict.copy())
+    return position_dict_list
+
+if __name__ == "__main__":
+    f = open(
+        "/root/project/law_data/5d30de97cbd1c48dc98a329e.txt", 'r', newline="", encoding='utf-8')
+    judgement = f.read()
+    job_list=["檢察官","公務員","車手","車手頭","監管科人員","青壯年","包括"]
+    location_list=["詐欺集團","臺灣臺北地方檢察署","臺灣新竹地方檢察署","政府機關","本院","新竹縣政府警察局","竹北"]
+    # 先去空白 再去\r\n
+    find_job_location_position(judgement,job_list,location_list)