Target all work for the project to a new bigquery-b2 branch.

P0 Tasks

• Rename job classes: LoadTableFromStorageJob -> LoadJob, ExtractTableToStorageJob -> ExtractJob (PR Rename job classes #3797)
• Rename ACL entry: dataset.AccessGrant -> dataset.AccessEntry, and likewise Dataset.access_grants -> Dataset.access_entries. (PR Rename class: 'dataset.AccessGrant' -> 'dataset.AccessEntry'. #3798)
• Client.get_job: construct concrete job instance from job ID via jobs.get API (PR Add 'Client.get_job' API wrapper. #3804)
• QueryJob: add result method returning future-ish object which polls both the job and the result until done; then returns rows via Table.fetch_data on the destination table.
• Implement high-level Client.query method: waits for completion, returns rows
• Generate job IDs only on Client-side if not supplied by the user.
• Add support for retries to streaming insert
• Add support for external tables (issue Add support for "external" tables #2372)
• Add etag support to Dataset.patch / Dataset.update
• Add etag support to Table.patch / Table.update
• Add support for Table.streamingBuffer
• Add CRUD for Dataset labels (issue BigQuery: support labels field of Datasets and Table resources #2931)
• Add CRUD for Table labels (issue BigQuery: support labels field of Datasets and Table resources #2931)

P1 Tasks

• Add support for client-side job ID generation with user-supplied prefix
• Add support for Dataset ACLs
• Expose returned query plan (Job.statistics.query.queryPlan) (PR Add 'QueryJob.query_plan' property. #3799)
• Add support for creating table on insert (TableData.insertAll to non-existent table)

Other tasks

• Migrate new job statistics work (from PR Add properties for new job statistics #3721) (PR Add 'QueryJob.query_plan' property. #3799, Add new scalar statistics properties to 'QueryJob' #3800, Add 'QueryJob.referenced_tables' property. #3801, Add 'QueryJob.undeclared_query_parameters' property. #3802, Add 'ExtractTableStorageJob.destination_uri_file_counts' property. #3803)