Merge lp:~dholbach/harvest/974431 into lp:harvest

[Daniel Holbach]

The API now does slicing and sorting now properly.

Test-case:
wget -O - "http://127.0.0.1:8000/services/sourcepackages/?_sortby=id"
wget -O - "http://127.0.0.1:8000/services/sourcepackages/?_offset=500&_sortby=id"

[lp:~dholbach/harvest/974431]

332. By Daniel Holbach on 2012-04-11

handle length argument properly, thanks mhall119

[Michael Hall (mhall119)]

Looks good to me

review: Approve

Unmerged revisions

=== modified file 'harvest/services/__init__.py'
--- harvest/services/__init__.py	2010-12-08 08:15:50 +0000
+++ harvest/services/__init__.py	2012-04-11 12:33:18 +0000
@@ -80,14 +80,16 @@
     for key in search_values.keys():
         if key.startswith('_'):
             extra_arguments[key] = search_values.pop(key)
+    if '_sortby' in request and request['_sortby'] not in EMPTY_VALUES:
+        results = model.objects.filter(**search_values).order_by(request.get('_sortby'))
+    
     offset = parse_number_argument(extra_arguments, '_offset')
     length = parse_number_argument(extra_arguments, '_length')
+    start = offset if offset else 0
     if not length or length > CHUNK_SIZE:
         length = CHUNK_SIZE
-    results = model.objects.filter(**search_values)[offset:length]
-
-    if '_sortby' in request and request['_sortby'] not in EMPTY_VALUES:
-        results = results.order_by(request.get('_sortby'))
+    end = start + length
+    results = results[start:end]
     return results
 
 def get_model_id(request, url):