Parcourir la source

[enh] stick results from the same category and template

Dalf il y a 10 ans
Parent
révision
5d1610d87a
1 fichiers modifiés avec 38 ajouts et 2 suppressions
  1. 38
    2
      searx/search.py

+ 38
- 2
searx/search.py Voir le fichier

@@ -49,7 +49,8 @@ def score_results(results):
49 49
     flat_len = len(flat_res)
50 50
     engines_len = len(results)
51 51
     results = []
52
-    # deduplication + scoring
52
+
53
+    # pass 1: deduplication + scoring
53 54
     for i, res in enumerate(flat_res):
54 55
 
55 56
         res['parsed_url'] = urlparse(res['url'])
@@ -90,7 +91,42 @@ def score_results(results):
90 91
         else:
91 92
             res['score'] = score
92 93
             results.append(res)
93
-    return sorted(results, key=itemgetter('score'), reverse=True)
94
+    results = sorted(results, key=itemgetter('score'), reverse=True)
95
+
96
+    # pass 2 : group results by category and template
97
+    gresults = []
98
+    categoryPositions = {}
99
+
100
+    for i, res in enumerate(results):
101
+        # FIXME : handle more than one category per engine
102
+        category = engines[res['engine']].categories[0] + ':' + '' if 'template' not in res else res['template'] 
103
+
104
+        current = None if category not in categoryPositions else categoryPositions[category]
105
+
106
+        # group with previous results using the same category if the group can accept more result and is not too far from the current position
107
+        if current != None and (current['count'] > 0) and (len(gresults) - current['index'] < 20):
108
+            # group with the previous results using the same category with this one
109
+            index = current['index']
110
+            gresults.insert(index, res)
111
+
112
+            # update every index after the current one (including the current one)
113
+            for k in categoryPositions:
114
+                v = categoryPositions[k]['index']
115
+                if v >= index:
116
+                    categoryPositions[k]['index'] = v+1
117
+
118
+            # update this category
119
+            current['count'] -= 1
120
+
121
+        else:
122
+            # same category
123
+            gresults.append(res)
124
+
125
+            # update categoryIndex
126
+            categoryPositions[category] = { 'index' : len(gresults), 'count' : 8 }
127
+
128
+    # return gresults
129
+    return gresults
94 130
 
95 131
 
96 132
 class Search(object):