Added support for slicing multiple times in Search class (#1771) (#1825)

* Added support for slicing multiple times in Search class

Fixes #799

* simplify slicing logic

* A few more slicing unit tests

* Removed unnecessary comment

* simplified slicing logic even more

* added suggested changes

* add more slicing examples to documentation

(cherry picked from commit fb57759)

Co-authored-by: Miguel Grinberg <miguel.grinberg@gmail.com>

docs/search_dsl.rst

@@ -410,16 +410,25 @@ To specify the from/size parameters, use the Python slicing API:
 
 .. code:: python
 
-  s = s[10:20]
-  # {"from": 10, "size": 10}
+   s = s[10:20]
+   # {"from": 10, "size": 10}
+
+   s = s[:20]
+   # {"size": 20}
+
+   s = s[10:]
+   # {"from": 10}
+
+   s = s[10:20][2:]
+   # {"from": 12, "size": 8}
 
 If you want to access all the documents matched by your query you can use the
 ``scan`` method which uses the scan/scroll elasticsearch API:
 
 .. code:: python
 
-  for hit in s.scan():
-      print(hit.title)
+   for hit in s.scan():
+       print(hit.title)
 
 Note that in this case the results won't be sorted.
 

elasticsearch_dsl/search_base.py

@@ -352,20 +352,34 @@ def __getitem__(self, n):
             # If negative slicing, abort.
             if n.start and n.start < 0 or n.stop and n.stop < 0:
                 raise ValueError("Search does not support negative slicing.")
-            # Elasticsearch won't get all results so we default to size: 10 if
-            # stop not given.
-            s._extra["from"] = n.start or 0
-            s._extra["size"] = max(
-                0, n.stop - (n.start or 0) if n.stop is not None else 10
-            )
-            return s
+            slice_start = n.start
+            slice_stop = n.stop
         else:  # This is an index lookup, equivalent to slicing by [n:n+1].
             # If negative index, abort.
             if n < 0:
                 raise ValueError("Search does not support negative indexing.")
-            s._extra["from"] = n
-            s._extra["size"] = 1
-            return s
+            slice_start = n
+            slice_stop = n + 1
+
+        old_from = s._extra.get("from")
+        old_to = None
+        if "size" in s._extra:
+            old_to = (old_from or 0) + s._extra["size"]
+
+        new_from = old_from
+        if slice_start is not None:
+            new_from = (old_from or 0) + slice_start
+        new_to = old_to
+        if slice_stop is not None:
+            new_to = (old_from or 0) + slice_stop
+            if old_to is not None and old_to < new_to:
+                new_to = old_to
+
+        if new_from is not None:
+            s._extra["from"] = new_from
+        if new_to is not None:
+            s._extra["size"] = max(0, new_to - (new_from or 0))
+        return s
 
     @classmethod
     def from_dict(cls, d):

tests/_async/test_search.py

@@ -363,15 +363,34 @@ def test_collapse():
 def test_slice():
     s = AsyncSearch()
     assert {"from": 3, "size": 7} == s[3:10].to_dict()
-    assert {"from": 0, "size": 5} == s[:5].to_dict()
-    assert {"from": 3, "size": 10} == s[3:].to_dict()
+    assert {"size": 5} == s[:5].to_dict()
+    assert {"from": 3} == s[3:].to_dict()
     assert {"from": 0, "size": 0} == s[0:0].to_dict()
     assert {"from": 20, "size": 0} == s[20:0].to_dict()
+    assert {"from": 10, "size": 5} == s[10:][:5].to_dict()
+    assert {"from": 10, "size": 0} == s[:5][10:].to_dict()
+    assert {"size": 10} == s[:10][:40].to_dict()
+    assert {"size": 10} == s[:40][:10].to_dict()
+    assert {"size": 40} == s[:40][:80].to_dict()
+    assert {"from": 12, "size": 0} == s[:5][10:][2:].to_dict()
+    assert {"from": 15, "size": 0} == s[10:][:5][5:].to_dict()
+    assert {} == s[:].to_dict()
+    with raises(ValueError):
+        s[-1:]
+    with raises(ValueError):
+        s[4:-1]
+    with raises(ValueError):
+        s[-3:-2]
 
 
 def test_index():
     s = AsyncSearch()
     assert {"from": 3, "size": 1} == s[3].to_dict()
+    assert {"from": 3, "size": 1} == s[3][0].to_dict()
+    assert {"from": 8, "size": 0} == s[3][5].to_dict()
+    assert {"from": 4, "size": 1} == s[3:10][1].to_dict()
+    with raises(ValueError):
+        s[-3]
 
 
 def test_search_to_dict():

tests/_sync/test_search.py

@@ -363,15 +363,34 @@ def test_collapse():
 def test_slice():
     s = Search()
     assert {"from": 3, "size": 7} == s[3:10].to_dict()
-    assert {"from": 0, "size": 5} == s[:5].to_dict()
-    assert {"from": 3, "size": 10} == s[3:].to_dict()
+    assert {"size": 5} == s[:5].to_dict()
+    assert {"from": 3} == s[3:].to_dict()
     assert {"from": 0, "size": 0} == s[0:0].to_dict()
     assert {"from": 20, "size": 0} == s[20:0].to_dict()
+    assert {"from": 10, "size": 5} == s[10:][:5].to_dict()
+    assert {"from": 10, "size": 0} == s[:5][10:].to_dict()
+    assert {"size": 10} == s[:10][:40].to_dict()
+    assert {"size": 10} == s[:40][:10].to_dict()
+    assert {"size": 40} == s[:40][:80].to_dict()
+    assert {"from": 12, "size": 0} == s[:5][10:][2:].to_dict()
+    assert {"from": 15, "size": 0} == s[10:][:5][5:].to_dict()
+    assert {} == s[:].to_dict()
+    with raises(ValueError):
+        s[-1:]
+    with raises(ValueError):
+        s[4:-1]
+    with raises(ValueError):
+        s[-3:-2]
 
 
 def test_index():
     s = Search()
     assert {"from": 3, "size": 1} == s[3].to_dict()
+    assert {"from": 3, "size": 1} == s[3][0].to_dict()
+    assert {"from": 8, "size": 0} == s[3][5].to_dict()
+    assert {"from": 4, "size": 1} == s[3:10][1].to_dict()
+    with raises(ValueError):
+        s[-3]
 
 
 def test_search_to_dict():