Allow BQ tables to be suffixed with _latest

Rakefile

@@ -82,13 +82,8 @@ task :update_bigquery_tables do
     config.version = `bundle exec ruby -e 'puts DfE::ReferenceData::VERSION'`.chomp
     config.commit = `git rev-parse HEAD`.chomp
 
-    # Suffix table names with the major version number, so multiple release
-    # branches can coexist peacefully
-    version_parts = config.version.split('.')
-    major_version = version_parts[0]
-    minor_version = version_parts[1] || '0'
-
-    config.table_name_suffix = "_v#{major_version}_#{minor_version}"
+    # Suffix table names with '_latest' to push to the same table
+    config.table_name_suffix = '_latest'
 
     puts "Updating #{config.project}.#{config.dataset} with version #{config.version}:"
   end

docs/bigquery.md

@@ -13,9 +13,7 @@ The following lists are, on each release, automatically written into BigQuery un
 Edit `BIGQUERY_TABLES` in the Rakefile to change this list.
 
 The table name is made by taking the base table name from the table above, and
-appending a version based on the major version of the release, for instance
-`_v1`. This is to allow existing users of older releases to continue operating,
-rather than changing the data structure out underneath them.
+appending `_latest`. This ensures that the data is always pushed to the same table, providing a consistent and up-to-date reference.
 
 ### Internals
 
@@ -38,5 +36,5 @@ If you have a `dfe-reference-data_bigquery_api_key.json` file or the
 test of the BigQuery importer will be run. This will create a randomly-named
 table in the `cross-teacher-services.dfe_reference_data_dev` dataset and,
 hopefully, delete it afterwards. The project and dataset used for the test can
-be overriden with the `BIGQUERY_QA_PROJECT` and `BIGQUERY_QA_DATASET`
+be overridden with the `BIGQUERY_QA_PROJECT` and `BIGQUERY_QA_DATASET`
 environment variables, respectively.

docs/lists_qualifications.md

@@ -28,4 +28,3 @@ This list is [autocomplete compatible](autocomplete_compatability.md).
 | `match_synonyms` | string array | A list of common alternative names that are equivalent to this type. An answer matching a match synonym can be safely matched to this type.|
 | `level` | enumerated string | The level of the qualification. `entry`, `1`, `2`, `3`, `4`, `5`, `6`, or `7`. |
 | `degree` | optional enumerated symbol | If the qualification type is a degree, the level of degree - `foundation`, `bachelor`, `master` or `doctor` |
-

lib/dfe/reference_data/bigquery/importer.rb

@@ -21,7 +21,7 @@ class Config
         self.max_retry_sleep = 60 # seconds
         self.bigquery_retries = 10
         self.bigquery_timeout = 10
-        self.table_name_suffix = ''
+        self.table_name_suffix = '_latest' # Default suffix
 
         def self.configure
           yield(self)
@@ -236,7 +236,7 @@ def convert_list_to_bigquery_format(list)
         end
 
         def update_reference_list_into_bigquery_table(dataset, table_name, list)
-          table = dataset.table table_name
+          table = dataset.table(table_name)
 
           # NOTE: TO FUTURE SELF: If we start to deal with very large reference
           # lists we might start to hit BigQuery API limits documented here: