Fit a lot of encodings for text file. (#458)

### What problem does this PR solve?

#384

### Type of change

- [x] Performance Improvement

rag/nlp/__init__.py

@@ -6,6 +6,35 @@ from . import huqie
 import re
 import copy
 
+all_codecs = [
+    'utf-8', 'gb2312', 'gbk', 'utf_16', 'ascii', 'big5', 'big5hkscs',
+    'cp037', 'cp273', 'cp424', 'cp437',
+    'cp500', 'cp720', 'cp737', 'cp775', 'cp850', 'cp852', 'cp855', 'cp856', 'cp857',
+    'cp858', 'cp860', 'cp861', 'cp862', 'cp863', 'cp864', 'cp865', 'cp866', 'cp869',
+    'cp874', 'cp875', 'cp932', 'cp949', 'cp950', 'cp1006', 'cp1026', 'cp1125',
+    'cp1140', 'cp1250', 'cp1251', 'cp1252', 'cp1253', 'cp1254', 'cp1255', 'cp1256',
+    'cp1257', 'cp1258', 'euc_jp', 'euc_jis_2004', 'euc_jisx0213', 'euc_kr',
+    'gb2312', 'gb18030', 'hz', 'iso2022_jp', 'iso2022_jp_1', 'iso2022_jp_2',
+    'iso2022_jp_2004', 'iso2022_jp_3', 'iso2022_jp_ext', 'iso2022_kr', 'latin_1',
+    'iso8859_2', 'iso8859_3', 'iso8859_4', 'iso8859_5', 'iso8859_6', 'iso8859_7',
+    'iso8859_8', 'iso8859_9', 'iso8859_10', 'iso8859_11', 'iso8859_13',
+    'iso8859_14', 'iso8859_15', 'iso8859_16', 'johab', 'koi8_r', 'koi8_t', 'koi8_u',
+    'kz1048', 'mac_cyrillic', 'mac_greek', 'mac_iceland', 'mac_latin2', 'mac_roman',
+    'mac_turkish', 'ptcp154', 'shift_jis', 'shift_jis_2004', 'shift_jisx0213',
+    'utf_32', 'utf_32_be', 'utf_32_le''utf_16_be', 'utf_16_le', 'utf_7'
+]
+
+
+def find_codec(blob):
+    global all_codecs
+    for c in all_codecs:
+        try:
+            blob.decode(c)
+            return c
+        except Exception as e:
+            pass
+    return "utf-8"
+
 
 BULLET_PATTERN = [[
     r"第[零一二三四五六七八九十百0-9]+(分?编|部分)",

rag/nlp/huqie.py

@@ -8,6 +8,7 @@ import re
 import string
 import sys
 from hanziconv import HanziConv
+from huggingface_hub import snapshot_download
 from nltk import word_tokenize
 from nltk.stem import PorterStemmer, WordNetLemmatizer
 from api.utils.file_utils import get_project_base_directory

rag/nlp/search.py

@@ -68,7 +68,7 @@ class Dealer:
         pg = int(req.get("page", 1)) - 1
         ps = int(req.get("size", 1000))
         topk = int(req.get("topk", 1024))
-        src = req.get("fields", ["docnm_kwd", "content_ltks", "kb_id", "img_id",
+        src = req.get("fields", ["docnm_kwd", "content_ltks", "kb_id", "img_id", "title_tks", "important_kwd",
                                  "image_id", "doc_id", "q_512_vec", "q_768_vec", "position_int",
                                  "q_1024_vec", "q_1536_vec", "available_int", "content_with_weight"])
 
@@ -289,8 +289,18 @@ class Dealer:
                 sres.field[i].get("q_%d_vec" % len(sres.query_vector), "\t".join(["0"] * len(sres.query_vector)))) for i in sres.ids]
         if not ins_embd:
             return [], [], []
-        ins_tw = [sres.field[i][cfield].split(" ")
-                  for i in sres.ids]
+
+        for i in sres.ids:
+            if isinstance(sres.field[i].get("important_kwd", []), str):
+                sres.field[i]["important_kwd"] = [sres.field[i]["important_kwd"]]
+        ins_tw = []
+        for i in sres.ids:
+            content_ltks = sres.field[i][cfield].split(" ")
+            title_tks = [t for t in sres.field[i].get("title_tks", "").split(" ") if t]
+            important_kwd = sres.field[i].get("important_kwd", [])
+            tks = content_ltks + title_tks + important_kwd
+            ins_tw.append(tks)
+
         sim, tksim, vtsim = self.qryr.hybrid_similarity(sres.query_vector,
                                                         ins_embd,
                                                         keywords,
@@ -368,7 +378,7 @@ class Dealer:
 
     def sql_retrieval(self, sql, fetch_size=128, format="json"):
         from api.settings import chat_logger
-        sql = re.sub(r"[ ]+", " ", sql)
+        sql = re.sub(r"[ `]+", " ", sql)
         sql = sql.replace("%", "")
         es_logger.info(f"Get es sql: {sql}")
         replaces = []