[biobank] bugfix: Update DB schema, and fix CandidateID references (#10037)

SQL/0000-00-06-BiobankTables.sql

@@ -182,7 +182,9 @@ CREATE TABLE `biobank_specimen_protocol_attribute_rel` (
   `SpecimenAttributeID` int(10) unsigned NOT NULL,
   `Required` tinyint(1) DEFAULT NULL,
   `showInDataTable` tinyint(1) DEFAULT NULL,
+  `OrderIndex` INT UNSIGNED NOT NULL,
   PRIMARY KEY (`SpecimenProtocolID`,`SpecimenAttributeID`),
+  UNIQUE KEY `UK_SpecimenProtocolId_OrderIndex` (`SpecimenProtocolID`, `OrderIndex`),
   KEY `FK_biobank_specimen_protocol_attribute_rel_SpecimenAttributeID` (`SpecimenAttributeID`),
   CONSTRAINT `FK_biobank_specimen_protocol_attribute__rel_SpecimenProtocolID` FOREIGN KEY (`SpecimenProtocolID`) REFERENCES `biobank_specimen_protocol` (`SpecimenProtocolID`),
   CONSTRAINT `FK_biobank_specimen_protocol_attribute_rel_SpecimenAttributeID` FOREIGN KEY (`SpecimenAttributeID`) REFERENCES `biobank_specimen_attribute` (`SpecimenAttributeID`)

SQL/New_patches/2025-10-29_add-order-index-to-specimen-protocols.sql

@@ -0,0 +1,41 @@
+-- Step 1: Add the column allowing NULL (temporarily)
+-- This allows the operation to succeed on a table that already has existing rows.
+ALTER TABLE biobank_specimen_protocol_attribute_rel
+ADD COLUMN OrderIndex INT UNSIGNED NULL;
+
+-- Step 2: Populate the existing rows with unique, non-NULL OrderIndex values.
+-- This uses a variable-based method to assign a sequential number (0, 1, 2, ...)
+-- to each row within the same SpecimenProtocolID group.
+
+SET @r := -1;
+SET @g := 0;
+
+-- A temporary table/derived table is used to calculate the unique index for each group
+UPDATE biobank_specimen_protocol_attribute_rel AS t1
+INNER JOIN (
+    SELECT
+        t2.SpecimenProtocolID,
+        t2.SpecimenAttributeID,
+        @r := CASE WHEN @g = t2.SpecimenProtocolID THEN @r + 1 ELSE 0 END AS new_OrderIndex,
+        @g := t2.SpecimenProtocolID AS current_group
+    FROM
+        biobank_specimen_protocol_attribute_rel AS t2
+    ORDER BY
+        t2.SpecimenProtocolID, t2.SpecimenAttributeID
+
+) AS ranked_data
+ON t1.SpecimenProtocolID = ranked_data.SpecimenProtocolID
+AND t1.SpecimenAttributeID = ranked_data.SpecimenAttributeID
+SET t1.OrderIndex = ranked_data.new_OrderIndex;
+
+-- Step 3: Enforce the constraints (NOT NULL and UNIQUE).
+-- Now that every row has a valid, unique value, these operations will succeed.
+
+-- 3a. Add the UNIQUE Constraint
+ALTER TABLE biobank_specimen_protocol_attribute_rel
+ADD CONSTRAINT UK_SpecimenProtocolId_OrderIndex
+UNIQUE (SpecimenProtocolID, OrderIndex);
+
+-- 3b. Change the Column to NOT NULL
+ALTER TABLE biobank_specimen_protocol_attribute_rel
+MODIFY COLUMN OrderIndex INT UNSIGNED NOT NULL;