Query parameters

This section contains information about additional query parameters. For more information about commands, see the Commands section.

Prerequisites

Here is a list of the commands for which you can specify success conditions:

• set;
• remove;
• create;
• lock;
• copy;
• move;
• link;
• create.

There are two types of conditions:

• prerequisite_revisions: A set of hints including path, transaction_id, revision. The query will be executed only if all the paths provided in the hint exist (in the context of the given transactions) and the revisions are as hinted.
• prerequisite_transactions includes a list of transactions that must exist when the query is executed.

Example:

revision = client.get("//tmp/my_table/@revision")
#... some calculations ...
revision_parameter = yt.create_revision_parameter(path="//tmp/my_table", revision=revision)
revision_client = yt.create_client_with_command_params(prerequisite_revisions=[revision_parameter])
revision_client.move("//tmp/transformed_table", "//tmp/my_table")

Data source for reading

In non-mutating queries to Cypress, you can specify a data source for reading, using the read_from parameter.

A popular value for the read_from parameter is cache. In this case, the response is read from the cache (usually hosted on dedicated instances in the cluster). This allows decreasing the load on the master server. This is important when running numerous heavy queries against Cypress: for example, if you need to traverse a large subtree.

Example:

for obj in client.search("//tmp/my_table", read_from="cache"):
    if "abc" in str(obj):
        ...