Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

GODRIVER-1831 Avoid accidental extra getMore when using limit on Find #562

Merged
merged 4 commits into from
Jan 27, 2021
+98 −1
Merged

GODRIVER-1831 Avoid accidental extra getMore when using limit on Find #562

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
d5740e6
GODRIVER-1831 Add find limit test with unset batch size
Jan 19, 2021
1bae72e
Add more test coverage
Jan 21, 2021
d1d7aee
Test limit < expected docs
Jan 21, 2021
42fb8ee
Use names for testCases
Jan 27, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 14 additions & 0 deletions mongo/integration/collection_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -883,6 +883,20 @@ func TestCollection(t *testing.T) {
_, ok := err.(mongo.CommandError)
assert.True(mt, ok, "expected error type %v, got %v", mongo.CommandError{}, err)
})
mt.Run("unset batch size does not surpass limit", func(mt *mtest.T) {
initCollection(mt, mt.Coll)
for _, limit := range []int64{2, 3, 4} {
opts := options.Find().SetSkip(0).SetLimit(limit)
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Similarly, I suggest adding a name to the test cases. Otherwise, it'd be harder to track down which case failed if L949 errors.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Sounds good! Used the limit number as the name for each case.

cursor, err := mt.Coll.Find(mtest.Background, bson.D{}, opts)
assert.Nil(mt, err, "Find error with limit: %v", err)

var docs []interface{}
err = cursor.All(mtest.Background, &docs)
assert.Nil(mt, err, "All error with limit: %v", err)

assert.Equal(mt, int(limit), len(docs), "expected number of docs to be %v, got %v", limit, len(docs))
}
})
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Given that this is a bit tricky to reason about, I suggest:

  1. Insert more than 101 documents in the collection. That is the default batch size for find, so we can test what happens on a getMore when using the default batch size.
  2. Adding a table test parameterized on: skip, limit, and batchSize which can check for an expected result count. That will make it easy to add future tests as well.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Added both!

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks!

})
mt.RunOpts("find one", noClientOpts, func(mt *mtest.T) {
mt.Run("limit", func(mt *mtest.T) {
Expand Down
2 changes: 1 addition & 1 deletion x/mongo/driver/batch_cursor.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -241,7 +241,7 @@ func (bc *BatchCursor) getMore(ctx context.Context) {

// Required for legacy operations which don't support limit.
numToReturn := bc.batchSize
if bc.limit != 0 && bc.numReturned+bc.batchSize > bc.limit {
if bc.limit != 0 && bc.numReturned+bc.batchSize >= bc.limit {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Based on my reading of https://github.com/mongodb/specifications/blob/master/source/crud/crud.rst#combining-limit-and-batch-size-for-the-wire-protocol I believe it is ok to use batchSize=0 when a limit is set. So I think this might be ok? I'm still not entirely sure where the limit is enforced after, when batchSize=0 and more than the limit documents are returned. I think tests which query a collection containing more than 101 documents may verify that behavior.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm not sure either, but the tests now query a collection containing more than 101 documents and pass...

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Great, the new tests cover the case of a getMore being run.

numToReturn = bc.limit - bc.numReturned
if numToReturn <= 0 {
err := bc.Close(ctx)
Expand Down