sql name search many terms

1 files changed, 21 insertions, 2 deletions

diff --git a/megapixels/app/server/api.py b/megapixels/app/server/api.py
index 3683d5fd..cc791bb2 100644
--- a/megapixels/app/server/api.py
+++ b/megapixels/app/server/api.py
@@ -3,6 +3,7 @@ import re
 import time
 import dlib
 import numpy as np
+import operator
 from flask import Blueprint, request, jsonify
 from PIL import Image  # todo: try to remove PIL dependency
 
@@ -139,14 +140,32 @@ def name_lookup(dataset_name):
   dataset = get_dataset(dataset_name)
 
   q = request.args.get('q')
+  q = re.sub('[^a-zA-Z ]+', '*', q)
+  terms = q.split(' ')
   # print(q)
 
   query = {
     'q': q,
     'timing': time.time() - start,
   }
-  results = dataset.search_name(q + '%') if q else None
-  
+  if len(terms) == 1:
+    names = dataset.search_name('%' + term + '%') if term else []
+    results = dataset.get_uuids_for_identities(names)
+  else:
+    lookup = {}
+    results_lookup = {}
+    for i, term in enumerate(terms):
+      search_term = '%' + term + '%'
+      names = dataset.search_name() if term else []
+      for name in names:
+        if name.id in lookup:
+          lookup[name.id] += 1
+        else:
+          lookup[name.id] = 1
+          results_lookup[name.id] = name
+    top_names = [results_lookup[item[0]] for item in sorted(lookup.items(), key=operator.itemgetter(1))][0:30]
+    results = dataset.get_uuids_for_identities(top_names)
+
   # print(results)
   return jsonify({
     'query': query,