highlights and weights

searching/__init__.py

@@ -24,17 +24,25 @@ def es_query(query:str):
         query ={
         "multi_match":{ 
                 "query": query,
-                "fields": ["title^20","title.ngrams^10","text","text.ngrams"],
-                "type": "most_fields"
+                "fields": ["title^20","title.ngrams^10","text^5","text.ngrams"],
+                "tie_breaker": 0.3
+                #"type": "most_fields"
                 }
         }
+        return query
+
+
 def es_highlight():
         highlight = {
               "fields": {
                 "title": {},
-                "text": {}
+                "text": {"fragment_size" : 150},
+                "title.ngrams": {},
+                "text.ngrams": {"fragment_size" : 150},
+                
               }
             }
+        return highlight
 
 def es_search(query:str):
     with es_client() as client:    
@@ -57,4 +65,5 @@ if __name__ =="__main__":
                 hit.get("_source",{}).get("title",""),
                  " ".join(hit.get("highlight",{}).get("title",[""])),
                 " ".join(hit.get("highlight",{}).get("text",[""]))
-        ))
+        ))
+        

searching/index.py

@@ -4,7 +4,7 @@ import elasticsearch
 
 settings = {
   "index":
-  {"max_ngram_diff": 3
+  {"max_ngram_diff": 7
   },
   "analysis": {
       "analyzer": {
@@ -19,10 +19,10 @@ settings = {
         "my_tokenizer": {
           "type": "ngram",
           "min_gram": 3,
-          "max_gram": 6,
+          "max_gram": 10,
           "token_chars": [
             "letter",
-            "digit"
+            "digit", "symbol"
           ]
         }
       }
@@ -55,7 +55,7 @@ mapping = {
        "published": {"type": "date", "format": "date_optional_time"},
        "updated_at": {"type": "date", "format": "date_optional_time"},
       "raw": {
-         "type": "text",
+         "type": "text",  
          "index": False       
        },
     }