hcrosse commented on code in PR #2445:
URL: https://github.com/apache/iceberg-go/pull/2445#discussion_r4605992023


##########
table/equality_delete_bytes_fastpath_test.go:
##########
@@ -265,6 +265,182 @@ func TestProcessEqualityDeletesSingleBytesDictionary(t 
*testing.T) {
        }
 }
 
+func TestProcessEqualityDeletesSingleBytesDictionaryAcrossBatches(t 
*testing.T) {
+       const apple, null = "\x01\x00\x00\x00\x05apple", "\x00"
+       type batch struct {
+               dictionary string
+               indices    string
+               offset     int
+               wantIDs    []int64
+       }
+       tests := []struct {
+               name         string
+               dictionaries map[string]string
+               // dictionaryOffset slices this many leading values off every 
dictionary.
+               dictionaryOffset int
+               sets             [][]string
+               batches          []batch
+       }{
+               {
+                       name:         "different_dictionary_same_length",
+                       dictionaries: map[string]string{"a": `["apple", "bob", 
"cherry"]`, "b": `["bob", "apple", "cherry"]`},
+                       sets:         [][]string{{apple}},
+                       batches: []batch{
+                               {dictionary: "a", indices: "[0, 1]", wantIDs: 
[]int64{2}},
+                               {dictionary: "b", indices: "[0, 1]", wantIDs: 
[]int64{1}},
+                       },
+               },
+               {
+                       name:         "shorter_dictionary",
+                       dictionaries: map[string]string{"a": `["apple", "bob", 
"cherry", "date"]`, "b": `["bob", "cherry", "apple"]`},
+                       sets:         [][]string{{apple}},
+                       batches: []batch{
+                               {dictionary: "a", indices: "[0, 1]", wantIDs: 
[]int64{2}},
+                               {dictionary: "b", indices: "[0, 2]", wantIDs: 
[]int64{1}},
+                       },
+               },
+               {
+                       name:         "null_index_and_null_entry_with_null_key",
+                       dictionaries: map[string]string{"a": `["apple", null, 
"cherry", "date", "egg"]`},
+                       sets:         [][]string{{null}},
+                       batches:      []batch{{dictionary: "a", indices: "[0, 
null, 1, 2]", wantIDs: []int64{1, 4}}},
+               },
+               {
+                       name:         
"null_index_and_null_entry_without_null_key",
+                       dictionaries: map[string]string{"a": `["apple", null, 
"cherry", "date", "egg"]`},
+                       sets:         [][]string{{apple}},
+                       batches:      []batch{{dictionary: "a", indices: "[0, 
null, 1, 2]", wantIDs: []int64{2, 3, 4}}},
+               },
+               {
+                       name:         "sliced_over_shared_dictionary",
+                       dictionaries: map[string]string{"a": `["apple", "bob", 
"cherry", "date", "egg"]`},
+                       sets:         [][]string{{apple}},
+                       batches: []batch{
+                               {dictionary: "a", indices: "[1, 2, 0, 0, 3, 
1]", offset: 3, wantIDs: []int64{5, 6}},
+                               {dictionary: "a", indices: "[4, 0, 2]", offset: 
1, wantIDs: []int64{3}},
+                       },
+               },
+               {
+                       name:             "sliced_dictionary_values",
+                       dictionaries:     map[string]string{"a": `["zzz", 
"apple", "bob", "cherry", "date"]`},
+                       dictionaryOffset: 1,
+                       sets:             [][]string{{apple}},
+                       batches:          []batch{{dictionary: "a", indices: 
"[0, 1]", wantIDs: []int64{2}}},
+               },
+               {
+                       // The second set's row 1 is already deleted by the 
first set, so
+                       // the second set never looks up apple.
+                       name:         "delete_sets_share_dictionary",
+                       dictionaries: map[string]string{"a": `["apple", "bob", 
"cherry", "date"]`},
+                       sets:         [][]string{{apple}, 
{"\x01\x00\x00\x00\x03bob"}},
+                       batches: []batch{
+                               {dictionary: "a", indices: "[0, 1]", wantIDs: 
nil},
+                               {dictionary: "a", indices: "[1, 0, 2]", 
wantIDs: []int64{3}},
+                       },
+               },
+               {
+                       // The second batch is long enough for the per-batch 
entry bitmap,
+                       // so the third batch must not reuse results from the 
first.
+                       name:         "entry_bitmap_between_batches",
+                       dictionaries: map[string]string{"a": `["apple", "bob", 
"cherry"]`, "b": `["cherry", "bob", "apple"]`},
+                       sets:         [][]string{{apple}},
+                       batches: []batch{
+                               {dictionary: "a", indices: "[0, 1]", wantIDs: 
[]int64{2}},
+                               {dictionary: "b", indices: "[0, 1, 2, 0]", 
wantIDs: []int64{1, 2, 4}},
+                               {dictionary: "b", indices: "[0, 2]", wantIDs: 
[]int64{1}},
+                       },
+               },
+       }
+       for _, tt := range tests {
+               t.Run(tt.name, func(t *testing.T) {
+                       mem := 
memory.NewCheckedAllocator(memory.DefaultAllocator)
+                       defer mem.AssertSize(t, 0)
+                       ctx := compute.WithAllocator(t.Context(), mem)
+
+                       dictionaries := make(map[string]arrow.Array, 
len(tt.dictionaries))
+                       for name, values := range tt.dictionaries {
+                               full, _, err := array.FromJSON(mem, 
arrow.BinaryTypes.String, strings.NewReader(values))

Review Comment:
   Both tests now run every case for `String`, `LargeString`, `Binary` and 
`LargeBinary`, with the sliced dictionary and null cases kept.



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to