Raz Monsonego
Raz Monsonego
Adds support for indexing empty JSON arrays and objects (i.e., `JSON.SET j $ '{"t":[]}'`, `JSON.SET j $ '{"t":{}}'`) and separated empty tag values (e.g, `HSET h1 t 'a,,c'`). **Mark if...
Adds a test to the short-read tests as part of the EMPTY TAG indexing. Update: We also refactor the path of the test-rdbs to a non-production s3 bucket, so that...
Adds a validation that when we get an error for a command with a non-zero offset and a zero limit - which today is interpreted as `LIMIT 0 0` regardless...
Currently both the serialized and the expanded format of a JSON value are being computed and saved in the loading process. This is both costly and inefficient, since only one...
Hi, There is somewhat of an inconsistency between the interpretation of literals as lists by `UNWIND` and `FOREACH`. `UNWIND` will interpret them as lists, but `FOREACH` will not. Example: `UNWIND...
Fixes the construction of the Unwind operation, by building the ops corresponding to the path comprehension that may be inside it. Fixes #3032.
Validates that when an aggregation is used in a projection of a WITH\RETURN clause followed by an ORDER-BY clause, then the variables used in the ORDER BY clause are one...
**Describe the changes in the pull request** Currently we discard empty strings for TEXT fields in our index, i.e., we don't index them and don't allow querying for them as...
**Describe the changes in the pull request** Adds support for `EMPTY TEXT` indexing - please refer to #4622's comment for more information. Also: * Removes support of `EMPTY TAG` indexing...