[DRAFT] chore(bigtable): add samples for sync client

[daniel-sanche]

This PR covers sample gaps between the sync data client/generated admin client

TODO: review

[gemini-code-assist]

Code Review

This pull request adds new Python samples and tests for the Google Cloud Bigtable admin and data clients, while updating legacy client samples with _legacy region tags. The review feedback highlights several critical API usage issues in the new data client snippets, including incorrect attribute access on Row and Cell objects, the use of the non-existent DeleteRangeFromColumn class (which should be DeleteCells), and type mismatches where regex filters expect bytes instead of strings (or vice-versa). Additionally, the feedback suggests minor cleanups for unused imports, unused variables, and redundant fully qualified imports.

[gemini-code-assist]

There are multiple issues in this snippet:

1. Row objects in the new data client are not subscriptable; you should access cells via .cells.
2. Cell objects do not have a row_key attribute; the row key belongs to the Row object.
3. Cell objects cannot be directly cast to int. Since incremented values are stored as 8-byte big-endian integers, you should decode the bytes using int.from_bytes.

Suggested change

	result_row = table.read_modify_write_row(row_key, increment_rule)
	cell = result_row[0]
	print(f"{cell.row_key} value: {int(cell)}")
	result_row = table.read_modify_write_row(row_key, increment_rule)
	cell = result_row.cells[0]
	val = int.from_bytes(cell.value, byteorder="big", signed=True)
	print(f"{result_row.row_key!r} value: {val}")

[gemini-code-assist]

The class DeleteRangeFromColumn does not exist in google.cloud.bigtable.data. To delete cells from a column, use DeleteCells instead.

Suggested change

	from google.cloud.bigtable.data import (
	BigtableDataClient,
	DeleteRangeFromColumn,
	)
	client = BigtableDataClient(project=project_id)
	table = client.get_table(instance_id, table_id)
	table.mutate_row(
	"phone#4c410523#20190501",
	DeleteRangeFromColumn(family="cell_plan", qualifier=b"data_plan_01gb"),
	)
	from google.cloud.bigtable.data import (
	BigtableDataClient,
	DeleteCells,
	)
	client = BigtableDataClient(project=project_id)
	table = client.get_table(instance_id, table_id)
	table.mutate_row(
	"phone#4c410523#20190501",
	DeleteCells(family="cell_plan", qualifier=b"data_plan_01gb"),
	)

[gemini-code-assist]

The class DeleteRangeFromColumn does not exist in google.cloud.bigtable.data. To delete cells from a column, use DeleteCells instead.

Suggested change

	from google.cloud.bigtable.data import (
	BigtableDataClient,
	DeleteRangeFromColumn,
	ReadRowsQuery,
	RowMutationEntry,
	)
	client = BigtableDataClient(project=project_id)
	table = client.get_table(instance_id, table_id)
	with table.mutations_batcher() as batcher:
	for row in table.read_rows(ReadRowsQuery(limit=10)):
	batcher.append(
	RowMutationEntry(
	row.row_key,
	DeleteRangeFromColumn(
	family="cell_plan", qualifier=b"data_plan_01gb"
	),
	)
	from google.cloud.bigtable.data import (
	BigtableDataClient,
	DeleteCells,
	ReadRowsQuery,
	RowMutationEntry,
	)
	client = BigtableDataClient(project=project_id)
	table = client.get_table(instance_id, table_id)
	with table.mutations_batcher() as batcher:
	for row in table.read_rows(ReadRowsQuery(limit=10)):
	batcher.append(
	RowMutationEntry(
	row.row_key,
	DeleteCells(
	family="cell_plan", qualifier=b"data_plan_01gb"
	),
	)
	)

[gemini-code-assist]

The class DeleteRangeFromColumn does not exist in google.cloud.bigtable.data. To delete cells from a column, use DeleteCells instead.

Suggested change

	from google.cloud.bigtable.data import (
	BigtableDataClient,
	DeleteRangeFromColumn,
	)
	from google.cloud.bigtable.data.row_filters import LiteralValueFilter
	client = BigtableDataClient(project=project_id)
	table = client.get_table(instance_id, table_id)
	table.check_and_mutate_row(
	"phone#4c410523#20190501",
	predicate=LiteralValueFilter("PQ2A.190405.003"),
	true_case_mutations=DeleteRangeFromColumn(
	family="cell_plan", qualifier=b"data_plan_01gb"
	),
	)
	from google.cloud.bigtable.data import (
	BigtableDataClient,
	DeleteCells,
	)
	from google.cloud.bigtable.data.row_filters import LiteralValueFilter
	client = BigtableDataClient(project=project_id)
	table = client.get_table(instance_id, table_id)
	table.check_and_mutate_row(
	"phone#4c410523#20190501",
	predicate=LiteralValueFilter("PQ2A.190405.003"),
	true_case_mutations=DeleteCells(
	family="cell_plan", qualifier=b"data_plan_01gb"
	),
	)

[gemini-code-assist]

In the Python Bigtable data client, both ValueRegexFilter and ColumnQualifierRegexFilter expect bytes as their patterns because cell values and column qualifiers are stored as bytes. Passing strings will cause runtime errors or matching failures.

Suggested change

	query = ReadRowsQuery(
	row_filter=row_filters.RowFilterUnion(
	filters=[
	row_filters.ValueRegexFilter("true"),
	row_filters.ColumnQualifierRegexFilter("os_build"),
	]
	)
	)
	query = ReadRowsQuery(
	row_filter=row_filters.RowFilterUnion(
	filters=[
	row_filters.ValueRegexFilter(b"true"),
	row_filters.ColumnQualifierRegexFilter(b"os_build"),
	]
	)
	)

[gemini-code-assist]

In the Python Bigtable data client, both ColumnQualifierRegexFilter and ValueRegexFilter expect bytes as their patterns because column qualifiers and cell values are stored as bytes. Passing strings will cause runtime errors or matching failures.

Suggested change

	query = ReadRowsQuery(
	row_filter=row_filters.ConditionalRowFilter(
	predicate_filter=row_filters.RowFilterChain(
	filters=[
	row_filters.ColumnQualifierRegexFilter("data_plan_10gb"),
	row_filters.ValueRegexFilter("true"),
	]
	),
	true_filter=row_filters.ApplyLabelFilter(label="passed-filter"),
	false_filter=row_filters.ApplyLabelFilter(label="filtered-out"),
	)
	)
	query = ReadRowsQuery(
	row_filter=row_filters.ConditionalRowFilter(
	predicate_filter=row_filters.RowFilterChain(
	filters=[
	row_filters.ColumnQualifierRegexFilter(b"data_plan_10gb"),
	row_filters.ValueRegexFilter(b"true"),
	]
	),
	true_filter=row_filters.ApplyLabelFilter(label="passed-filter"),
	false_filter=row_filters.ApplyLabelFilter(label="filtered-out"),
	)
	)

[gemini-code-assist]

The variable cluster_path is defined here but is never used anywhere in the run_instance_operations function. Consider removing it to clean up the code.

[gemini-code-assist]

The datetime module is imported here but is never used in this file. Consider removing this unused import.

[gemini-code-assist]

Since row_filters is imported directly on line 31, you can use it here instead of the fully qualified path. This also allows you to remove the dummy statement and comment on lines 37-38.

Suggested change

	row_filter = bigtable.data.row_filters.CellsColumnLimitFilter(1)
	row_filter = row_filters.CellsColumnLimitFilter(1)

[gemini-code-assist]

In the Python Bigtable data client, column family names are strings, so FamilyNameRegexFilter expects a string pattern rather than bytes. Passing bytes may cause type mismatches or matching failures.

Suggested change

	row_filter=row_filters.FamilyNameRegexFilter("stats_.*$".encode("utf-8"))
	row_filter=row_filters.FamilyNameRegexFilter("stats_.*$")