Merge branch 'fixProjectCodeInSampler' into '2.8.2.3-DEVELOPMENT'

Fix the core sampler access to project code linking table.

See merge request !1589

apps/cli/utilities/core_sampler/core_sampler/core_sampler.py

@@ -170,6 +170,15 @@ class CoreSampler:
         self.visited.clear()
         self.save(ancillary_rows)
 
+        # the science products table doesn't directly connect to the projects table via a key,
+        # so the sampler can't find them by default; grab the project code entries manually
+        products_projects = self.table("science_products_projects")
+        finder = ProductsProjectsFinder(self.connection)
+        project_codes = finder.find_project_codes_for_product(science_product_locators)
+        products_projects_rows = products_projects.fetch_from_many(
+            {"project_code": project_codes, "science_product_locator": science_product_locators})
+        self.save(products_projects_rows)
+
         # finally, grab the files associated with the entries
         filegroup_ids = [row["filegroup_id"] for row in science_product_rows] + \
                         [row["filegroup_id"] for row in ancillary_rows]
@@ -207,10 +216,19 @@ class CoreSampler:
         ancillary_product_locators = finder.find_ancillary_product_locators(product_group_id, science_product_locators)
         requested_ancillary_rows = ancillary_products.fetch_from_many({"ancillary_product_locator": ancillary_product_locators})
 
+        # the science products table doesn't directly connect to the projects table via a key,
+        # so the sampler can't find them by default; grab the project code entries manually
+        products_projects = self.table("science_products_projects")
+        finder = ProductsProjectsFinder(self.connection)
+        project_codes = finder.find_project_codes_for_product(science_product_locators)
+        requested_project_product_rows = products_projects.fetch_from_many(
+            {"project_code": project_codes, "science_product_locator": science_product_locators})
+
         # Create one dictionary pairing table names to row sets
         rows_dict = self.create_row_dict(requested_product_group_rows, {})
         rows_dict = self.create_row_dict(requested_product_rows, rows_dict)
         rows_dict = self.create_row_dict(requested_ancillary_rows, rows_dict)
+        rows_dict = self.create_row_dict(requested_project_product_rows, rows_dict)
 
         # the correct table deletion order is not the same as the reverse of the topographic map, set the order
         deletion_order = ["ancillary_products", "science_products_projects", "science_products_product_groups",
@@ -326,6 +344,27 @@ class CoreSampler:
         rows.write_to(self.writer)
 
 
+class ProductsProjectsFinder:
+    """Looks up project codes for a list of spls"""
+
+    def __init__(self, connection: MDDBConnector):
+        self.connection = connection
+
+    def find_project_codes_for_product(self, spls: [str]) -> [str]:
+        cursor = self.connection.cursor()
+        sql = """
+            SELECT DISTINCT project_code FROM science_products_projects
+            WHERE science_product_locator IN %(spls)s
+        """
+        cursor.execute(sql, {"spls": tuple(spls)})
+        data = cursor.fetchall()
+
+        if len(data) != 1:
+            raise Exception("Either too many project codes or no project codes were found for product group.")
+
+        # All spls will have the same project code, but return a list with a length matching the spls list
+        return [data[0]["project_code"]] * len(spls)
+
 class ProductFinder:
     """Looks up science product locators for a product group id"""