Merge pull request #14463 from ankush/fuzzy_website_search

feat: allow fuzzy search in website search

frappe/search/full_text_search.py

@@ -7,7 +7,7 @@ from frappe.utils import update_progress_bar
 from whoosh.index import create_in, open_dir, EmptyIndexError
 from whoosh.fields import TEXT, ID, Schema
 from whoosh.qparser import MultifieldParser, FieldsPlugin, WildcardPlugin
-from whoosh.query import Prefix
+from whoosh.query import Prefix, FuzzyTerm
 from whoosh.writing import AsyncWriter
 
 
@@ -23,6 +23,9 @@ class FullTextSearch:
 	def get_schema(self):
 		return Schema(name=ID(stored=True), content=TEXT(stored=True))
 
+	def get_fields_to_search(self):
+		return ["name", "content"]
+
 	def get_id(self):
 		return "name"
 
@@ -120,8 +123,15 @@ class FullTextSearch:
 		results = None
 		out = []
 
+		search_fields = self.get_fields_to_search()
+		fieldboosts = {}
+
+		# apply reducing boost on fields based on order. 1.0, 0.5, 0.33 and so on
+		for idx, field in enumerate(search_fields, start=1):
+			fieldboosts[field] = 1.0 / idx
+
 		with ix.searcher() as searcher:
-			parser = MultifieldParser(["title", "content"], ix.schema)
+			parser = MultifieldParser(search_fields, ix.schema, termclass=FuzzyTermExtended, fieldboosts=fieldboosts)
 			parser.remove_plugin_class(FieldsPlugin)
 			parser.remove_plugin_class(WildcardPlugin)
 			query = parser.parse(text)
@@ -136,5 +146,13 @@ class FullTextSearch:
 
 		return out
 
+
+class FuzzyTermExtended(FuzzyTerm):
+	def __init__(self, fieldname, text, boost=1.0, maxdist=2, prefixlength=1,
+			constantscore=True):
+		super().__init__(fieldname, text, boost=boost, maxdist=maxdist,
+				prefixlength=prefixlength, constantscore=constantscore)
+
+
 def get_index_path(index_name):
 	return frappe.get_site_path("indexes", index_name)

frappe/search/test_full_text_search.py

@@ -125,4 +125,4 @@ def get_documents():
 			deploy business applications with Rich Admin Interface. CommonSearchTerm"""
 	})
 
-	return docs
+	return docs

frappe/search/website_search.py

@@ -21,6 +21,9 @@ class WebsiteSearch(FullTextSearch):
 			title=TEXT(stored=True), path=ID(stored=True), content=TEXT(stored=True)
 		)
 
+	def get_fields_to_search(self):
+		return ["title", "content"]
+
 	def get_id(self):
 		return "path"
 

frappe/tests/test_global_search.py

@@ -88,13 +88,13 @@ class TestGlobalSearch(unittest.TestCase):
 		event = frappe.get_doc('Event', event_name)
 		test_subject = event.subject
 		results = global_search.search(test_subject)
-		self.assertEqual(len(results), 1)
+		self.assertTrue(any(r["name"] == event_name for r in results), msg="Failed to search document by exact name")
 
 		frappe.delete_doc('Event', event_name)
 		global_search.sync_global_search()
 
 		results = global_search.search(test_subject)
-		self.assertEqual(len(results), 0)
+		self.assertTrue(all(r["name"] != event_name for r in results), msg="Deleted documents appearing in global search.")
 
 	def test_insert_child_table(self):
 		frappe.db.delete("Event")