.dvc/.gitignore

+/config.local
+/tmp
+/cache

.dvcignore

+# Add patterns of files dvc should ignore, which could improve
+# the performance. Learn more at
+# https://dvc.org/doc/user-guide/dvcignore

.gitignore

 __pycache__/
 .ipynb_checkpoints/
 *.ipynb
+/examples

app.py

+
+# Copyright 2022, Frank Sauerburger
+
+"""Movie title suggestion api"""
+
+import json
+from flask import Flask, request
+import qgram
+
+app = Flask(__name__)
+
+with open("examples/titles.json") as fileobj:
+    titles = json.load(fileobj)
+
+index = qgram.QGramIndex()
+for title, _ in titles.values():
+    index.add_term(title)
+
+@app.route("/")
+def suggest():
+    query = request.args.get("q")
+    if not query:
+        return {}
+
+    results = index.search(query)
+    return {
+        "suggestions": results[:100]
+    }
+

examples.dvc

+outs:
+- md5: 89bb55be2fbe9a824be916d12042252b.dir
+  size: 394035
+  nfiles: 2
+  path: examples

qgram.py

@@ -40,7 +40,7 @@ class QGramIndex:
             jaccard = count / (n_term + n_query - count)
             result.append((jaccard, term))
 
-        result.sort(key=lambda x: x[0])
+        result.sort(key=lambda x: -x[0])
 
         return result
 
@@ -55,7 +55,7 @@ class QGramIndex:
         The beginning and end of the string is denoted by a dollar sign ($).
         If the term is too short, the qgrams might be shorter than q.
         """
-        term = f"${term}$"
+        term = f"${term}$".lower()
         qgrams = []
 
         if len(term) < self.q_param:

requirements.txt

+flask~=2.1.2

test_qgram.py

@@ -9,6 +9,12 @@ from qgram import QGramIndex
 class QGramChunkTest(unittest.TestCase):
     """Test the  implementation"""
 
+    def test_case(self):
+        """Test the chunk converts to lower case"""
+        index = QGramIndex()
+        qgrams = index._chunk("WeaTHer")
+        self.assertEqual(qgrams, ["$we", "wea", "eat", "ath", "the", "her", "er$"])
+
     def test_long(self):
         """Test the chunk method with a long string"""
         index = QGramIndex()