upgrade laws parser of docx (#1332)

### What problem does this PR solve?


### Type of change

- [x] Refactoring

api/apps/chunk_app.py

@@ -20,7 +20,7 @@
 from elasticsearch_dsl import Q
 
 from rag.app.qa import rmPrefix, beAdoc
-from rag.nlp import search, rag_tokenizer
+from rag.nlp import search, rag_tokenizer, keyword_extraction
 from rag.utils.es_conn import ELASTICSEARCH
 from rag.utils import rmSpace
 from api.db import LLMType, ParserType
@@ -268,6 +268,10 @@ def retrieval_test():
             rerank_mdl = TenantLLMService.model_instance(
                 kb.tenant_id, LLMType.RERANK.value, llm_name=req["rerank_id"])
 
+        if req.get("keyword", False):
+            chat_mdl = TenantLLMService.model_instance(kb.tenant_id, LLMType.CHAT)
+            question += keyword_extraction(chat_mdl, question)
+
         ranks = retrievaler.retrieval(question, embd_mdl, kb.tenant_id, [kb_id], page, size,
                                       similarity_threshold, vector_similarity_weight, top,
                                       doc_ids, rerank_mdl=rerank_mdl)

api/db/services/dialog_service.py

@@ -23,7 +23,7 @@
 from api.db.services.llm_service import LLMService, TenantLLMService, LLMBundle
 from api.settings import chat_logger, retrievaler
 from rag.app.resume import forbidden_select_fields4resume
-from rag.nlp.rag_tokenizer import is_chinese
+from rag.nlp import keyword_extraction
 from rag.nlp.search import index_name
 from rag.utils import rmSpace, num_tokens_from_string, encoder
 
@@ -121,6 +121,8 @@ def chat(dialog, messages, stream=True, **kwargs):
     if "knowledge" not in [p["key"] for p in prompt_config["parameters"]]:
         kbinfos = {"total": 0, "chunks": [], "doc_aggs": []}
     else:
+        if prompt_config.get("keyword", False):
+            questions[-1] += keyword_extraction(chat_mdl, questions[-1])
         kbinfos = retrievaler.retrieval(" ".join(questions), embd_mdl, dialog.tenant_id, dialog.kb_ids, 1, dialog.top_n,
                                         dialog.similarity_threshold,
                                         dialog.vector_similarity_weight,

rag/app/laws.py

@@ -54,62 +54,44 @@ def __call__(self, filename, binary=None, from_page=0, to_page=100000):
         self.doc = Document(
             filename) if not binary else Document(BytesIO(binary))
         pn = 0
-        last_question, last_answer, last_level = "", "", -1
         lines = []
-        root = DocxNode()
-        point = root
         bull = bullets_category([p.text for p in self.doc.paragraphs])
         for p in self.doc.paragraphs:
             if pn > to_page:
                 break
-            question_level, p_text = 0, ''
-            if from_page <= pn < to_page and p.text.strip():
-                question_level, p_text = docx_question_level(p, bull)
-            if not question_level or question_level > 6: # not a question
-                last_answer = f'{last_answer}\n{p_text}'
-            else:   # is a question
-                if last_question:
-                    while last_level <= point.level:
-                        point = point.parent
-                    new_node = DocxNode(last_question, last_answer, last_level, [], point)
-                    point.childs.append(new_node)
-                    point = new_node
-                    last_question, last_answer, last_level = '', '', -1
-                last_level = question_level
-                last_answer = ''
-                last_question = p_text
-
+            question_level, p_text = docx_question_level(p, bull)
+            if not p_text.strip("\n"):continue
+            lines.append((question_level, p_text))
+
             for run in p.runs:
                 if 'lastRenderedPageBreak' in run._element.xml:
                     pn += 1
                     continue
                 if 'w:br' in run._element.xml and 'type="page"' in run._element.xml:
                     pn += 1
-        if last_question:
-            while last_level <= point.level:
-                point = point.parent
-            new_node = DocxNode(last_question, last_answer, last_level, [], point)
-            point.childs.append(new_node)
-            point = new_node
-            last_question, last_answer, last_level = '', '', -1
-        traversal_queue = [root]
-        while traversal_queue:
-            current_node: DocxNode = traversal_queue.pop()
-            sum_text = f'{self.__clean(current_node.question)}\n{self.__clean(current_node.answer)}'
-            if not current_node.childs and not current_node.answer.strip():
-                continue
-            for child in current_node.childs:
-                sum_text = f'{sum_text}\n{self.__clean(child.question)}'
-                traversal_queue.insert(0, child)
-            lines.append(self.__clean(sum_text))
-        return [l for l in lines if l]
-class DocxNode:
-    def __init__(self, question: str = '', answer: str = '', level: int = 0, childs: list = [], parent = None) -> None:
-        self.question = question
-        self.answer = answer
-        self.level = level
-        self.childs = childs
-        self.parent = parent
+
+        visit = [False for _ in range(len(lines))]
+        sections = []
+        for s in range(len(lines)):
+            e = s + 1
+            while e < len(lines):
+                if lines[e][0] <= lines[s][0]:
+                    break
+                e += 1
+            if e - s == 1 and visit[s]: continue
+            sec = []
+            next_level = lines[s][0] + 1
+            while not sec and next_level < 22:
+                for i in range(s+1, e):
+                    if lines[i][0] != next_level: continue
+                    sec.append(lines[i][1])
+                    visit[i] = True
+                next_level += 1
+            sec.insert(0, lines[s][1])
+
+            sections.append("\n".join(sec))
+        return [l for l in sections if l]
+
     def __str__(self) -> str:
         return f'''
             question:{self.question},

rag/nlp/__init__.py

@@ -514,16 +514,19 @@ def add_chunk(t, pos):
 
     return cks
 
+
 def docx_question_level(p, bull = -1):
+    txt = re.sub(r"\u3000", " ", p.text).strip()
     if p.style.name.startswith('Heading'):
-        return int(p.style.name.split(' ')[-1]), re.sub(r"\u3000", " ", p.text).strip()
+        return int(p.style.name.split(' ')[-1]), txt
     else:
         if bull < 0:
-            return 0, re.sub(r"\u3000", " ", p.text).strip()
+            return 0, txt
         for j, title in enumerate(BULLET_PATTERN[bull]):
-            if re.match(title, re.sub(r"\u3000", " ", p.text).strip()):
-                return j+1, re.sub(r"\u3000", " ", p.text).strip()
-    return 0, re.sub(r"\u3000", " ", p.text).strip()
+            if re.match(title, txt):
+                return j+1, txt
+    return len(BULLET_PATTERN[bull]), txt
+
 
 def concat_img(img1, img2):
     if img1 and not img2:
@@ -544,6 +547,7 @@ def concat_img(img1, img2):
 
     return new_image
 
+
 def naive_merge_docx(sections, chunk_token_num=128, delimiter="\n。；！？"):
     if not sections:
         return []
@@ -573,4 +577,15 @@ def add_chunk(t, image, pos=""):
     for sec, image in sections:
         add_chunk(sec, image, '')
 
-    return cks, images
+    return cks, images
+
+
+def keyword_extraction(chat_mdl, content):
+    prompt = """
+You're a question analyzer. 
+1. Please give me the most important keyword/phrase of this question.
+Answer format: (in language of user's question)
+ - keyword: 
+"""
+    kwd, _ = chat_mdl.chat(prompt, [{"role": "user",  "content": content}], {"temperature": 0.2})
+    return kwd