Database Reference
In-Depth Information
{
"datasetReference": {
"datasetId": "scratch",
"projectId": "bigquery-e2e"
},
"friendlyName": "my dataset",
"access": [
{
"role": "READER",
"specialGroup": "projectReaders"
},
{
"role": "WRITER",
"specialGroup": "projectWriters"
},
{
"role": "OWNER",
"specialGroup": "projectOwners"
}
],
"creationTime": "1376367421192",
"lastModifiedTime": "1376369547951"
}
Datasets.patch()
Patch operates just like update, except you don't need to provide all the
fields; instead you just need the ones you care about updating. This is more
than just convenience; it makes it possible for multiple updaters to operate
on the same object without worrying that they clobber each other.
For example, suppose Alice was updating the ACL for the
application_logs
dataset. She executes a
Datasets.read()
operation to get the current state, applies her ACL changes, and then calls
Datasets.update()
to commit her changes. At the same time, however,
Bob wants to update the dataset's friendly name to Bob's Dataset. If he reads
the dataset before Alice's changes have been made but writes his update
after her changes have finished, the ACL updates will be wiped out. The
patch
command gives a way to avoid that problem, by only specifying the
