修改反向标红标注bug

flask_check_bert_test.py

@@ -172,6 +172,9 @@ def original_text_contrast_func(data_sentence_dan, paper_dict):
                                                                        paper_dict[i[0]][
                                                                            3])  # text_original, bert_text, bert_text_pre
 
+        if sentence_0_bool == False or sentence_1_bool == False:
+            continue
+
         start_dan = sentence_0_dan_red.index("<red>")
         end_dan = sentence_0_dan_red.index("</red>") - len("<red>")
 
@@ -180,9 +183,6 @@ def original_text_contrast_func(data_sentence_dan, paper_dict):
         if end_dan > end:
             end = end_dan
 
-        if sentence_0_bool == False or sentence_1_bool == False:
-            continue
-
         similar_content_dan["content"] = sentence_1_dan_red
         similar_content_dan["title"] = i[3]["title"]
         similar_content_dan["author"] = i[3]["author"]
@@ -752,81 +752,85 @@ def original_text_marked_red(text_original, bert_text, bert_text_pre):
     :param bert_text_pre:
     :return:
     '''
+    try:
+        fuhao = ["\n"]
+        up_pointer = 0
+        down_pointer = 0
 
-    fuhao = ["\n"]
-    up_pointer = 0
-    down_pointer = 0
-
-    pointer_list = []
+        pointer_list = []
 
-    if len(bert_text_pre) > len(bert_text):
-        return False, ""
+        if len(bert_text_pre) > len(bert_text):
+            return False, ""
 
-    while True:
-        if down_pointer >= len(bert_text_pre):
-            break
-        elif down_pointer == len(bert_text_pre) - 1:
-            if bert_text[up_pointer] == bert_text_pre[down_pointer]:
-                pointer_list.append(up_pointer)
+        while True:
+            if down_pointer >= len(bert_text_pre):
                 break
-            else:
-                up_pointer += 1
-                down_pointer = 0
-                pointer_list = []
-
-        elif bert_text[up_pointer] in fuhao:
-            up_pointer += 1
-
-        else:
-            if bert_text[up_pointer] == bert_text_pre[down_pointer]:
-                pointer_list.append(up_pointer)
-                up_pointer += 1
-                down_pointer += 1
-            else:
-                if bert_text_pre[down_pointer:down_pointer + 5] == "[UNK]":
-                    up_pointer += 1
-                    down_pointer += 5
-                    pointer_list.append(up_pointer)
-                elif is_english_char(bert_text_pre[down_pointer]) == True:
-                    up_pointer += 1
-                    down_pointer += 1
+            elif down_pointer == len(bert_text_pre) - 1:
+                if bert_text[up_pointer] == bert_text_pre[down_pointer]:
                     pointer_list.append(up_pointer)
+                    break
                 else:
                     up_pointer += 1
                     down_pointer = 0
                     pointer_list = []
 
-    start = pointer_list[0]
-    end = pointer_list[-1]
-    bert_text_list = list(bert_text)
-    bert_text_list.insert(start, "<red>")
-    bert_text_list.insert(end + 2, "</red>")
-
-    text_original_list = list(text_original)
-
-    up = 0
-    down = 0
-
-    while True:
-        if up == len(text_original_list):
-            break
-
-        if text_original_list[up] == bert_text_list[down]:
-            up += 1
-            down += 1
+            elif bert_text[up_pointer] in fuhao:
+                up_pointer += 1
 
-        else:
-            if bert_text_list[down] == "<red>":
-                down += 1
-            elif bert_text_list[down] == "</red>":
-                down += 1
             else:
-                bert_text_list.insert(down, text_original_list[up])
+                if bert_text[up_pointer] == bert_text_pre[down_pointer]:
+                    pointer_list.append(up_pointer)
+                    up_pointer += 1
+                    down_pointer += 1
+                else:
+                    if bert_text_pre[down_pointer:down_pointer + 5] == "[UNK]":
+                        up_pointer += 1
+                        down_pointer += 5
+                        pointer_list.append(up_pointer)
+                    elif is_english_char(bert_text_pre[down_pointer]) == True:
+                        up_pointer += 1
+                        down_pointer += 1
+                        pointer_list.append(up_pointer)
+                    else:
+                        up_pointer += 1
+                        down_pointer = 0
+                        pointer_list = []
+
+        start = pointer_list[0]
+        end = pointer_list[-1]
+        bert_text_list = list(bert_text)
+        bert_text_list.insert(start, "<red>")
+        bert_text_list.insert(end + 2, "</red>")
+
+        text_original_list = list(text_original)
+
+        up = 0
+        down = 0
+
+        while True:
+            if up == len(text_original_list):
+                break
+
+            if text_original_list[up] == bert_text_list[down]:
                 up += 1
                 down += 1
 
-    bert_text = "".join(bert_text_list)
-    return True, bert_text
+            else:
+                if bert_text_list[down] == "<red>":
+                    down += 1
+                elif bert_text_list[down] == "</red>":
+                    down += 1
+                else:
+                    bert_text_list.insert(down, text_original_list[up])
+                    up += 1
+                    down += 1
+
+        bert_text = "".join(bert_text_list)
+        return True, bert_text
+    except:
+        print("句子标红报错")
+        print(text_original, bert_text, bert_text_pre)
+        return False, ""
 
 
 def biaohong_bert_predict(sentence_0_list, sentence_1_list):
@@ -927,7 +931,7 @@ def ulit_recall_paper(recall_data_list_dict):
     # return data
 
     data = []
-    for i in list(recall_data_list_dict.items()):
+    for i in list(recall_data_list_dict.items())[:5]:
         data_one = processing_one_text(i[0])
         data.extend(data_one)
 
@@ -948,7 +952,7 @@ def recall_10(queue_uuid, title, abst_zh, content):
         "content": content
     }
     print(request_json)
-    da = dialog_line_parse("http://192.168.31.145:50004/check", request_json)
+    dialog_line_parse("http://192.168.31.145:50004/check", request_json)
 
 
 
@@ -1126,6 +1130,7 @@ def classify_accurate_check():
         recall_data_list_dict = json.loads(query_recall_dict["data"])
         recall_data_list = ulit_recall_paper(recall_data_list_dict)
         data_dict_path = redis_.get(query_recall_uuid + "_request_check")
+        print(data_dict_path)
         with open(data_dict_path, encoding='utf8') as f:
             data_dict = json.loads(f.read())