sqlglot incompatibilites with INFORMATION_SCHEMA identifiers

[scholtzan]

A recent update to sqlglot results in INFORMATION_SCHEMA table identifiers being parsed differently. The INFORMATION_SCHEMA part is now always quoted and parts are interpreted differently. This messes with some of our substitution logic in

bigquery-etl/bigquery_etl/util/common.py

Lines 245 to 256 in 9814402

	# but information_schema region may also be parsed as catalog
	if table_expr.catalog.startswith("region-"):
	project_name = f"{target_project}.{table_expr.catalog}"
	elif table_expr.catalog == "": # no project id
	project_name = target_project
	else: # project id exists
	continue
	# fully qualified table ref
	replacement_string = f"`{project_name}.{table_expr.db or default_dataset}.{table_expr.name}`"
	table_replacements.add((reference_string, replacement_string))

The sqlglot change that caused this: tobymao/sqlglot@84f78aa#diff-3cdd876ac935bf47c0213f4f49aac3319401a3f4d0012d58d716d9c814a6e2faR609

We should try updating this logic so we can update to the most recent sqlglot version

┆Issue is synchronized with this Jira Task