Chapter 5: Proving Your Work

Workbook validation, auto-recovery, quality flags, and keeping your map clean.

Building a map is one thing. Knowing it's correct is another. As a project grows — schemas evolve, columns get reassigned, configuration changes — it becomes easy for the workbook to accumulate subtle inconsistencies: mappings referencing tables that no longer exist, transformation logic that was never filled in, source columns that could truncate when moved to a shorter target. Bridger's validation system surfaces these issues before they become problems in review or downstream in development.

Validation on Open

Every time you open a workbook, Bridger runs validation automatically. If issues are found, a summary notice appears telling you how many were detected and at what severity. It isn't the full report — it's a prompt to go get one. The message directs you to Tools → Validate Workbook… for the complete breakdown.

This automatic check is intentionally lightweight. It tells you something needs attention without requiring you to read a full report just to open a file.

Auto-Recovery

Alongside validation, Bridger runs an auto-recovery pass when a workbook is opened. Some structural issues — status references that no longer match a configured status code, notation type references that point to a type that was renamed or removed, orphaned indicator values — can be repaired automatically rather than left broken.

When auto-recovery finds and fixes something, the Auto-Recovery Results dialog appears summarizing what was repaired. This dialog is informational — the fixes have already been applied — but it is worth reading carefully. Auto-recovered items are usually a signal that configuration drifted away from what the mappings expect: a status was renamed, a notation type was deleted, or an indicator definition changed. The right response is to open the Configuration area and formalize the recovered items, or use bulk operations to bring the affected mappings into alignment with the current configuration.

Note Auto-recovery repairs the workbook in memory. Save after reviewing the results to make the fixes permanent. If you close without saving, the same issues will be recovered again on next open.

Running Validation

To run a full validation at any time, choose Tools → Validate Workbook…. The dialog opens with two options for what to validate: the current open workbook, or a workbook file on disk. The file-from-disk option is useful for checking a workbook before sharing it or after receiving one from a colleague — it validates the raw file without loading it into the workspace.

Click Validate and the report appears in the dialog. A clean workbook shows a simple pass confirmation. A workbook with issues shows a summary count by severity followed by a detailed breakdown of every issue found, including the project name and mapping involved.

The report can be copied to the clipboard with the Copy button. Paste it into an email, a ticket, or a shared document to use as a punch list. This is particularly useful in team settings — a team lead can validate a contributor's workbook before a merge and share the report to flag issues or drift from standards before consolidating the work.

The Three Severity Tiers

Every issue in the validation report falls into one of three severity tiers.

Critical issues are data integrity problems that must be resolved. Mappings that are structurally incomplete — missing a default value or lacking the source or target information needed to be valid — fall here, as do duplicate mappings between the same column pair. These can result from file handling issues or edge cases in certain operations. A workbook with critical issues has mappings that cannot be trusted to produce correct output.

Warnings are functional issues that won't prevent the workbook from being used but indicate something worth investigating. Library sync drift — project tables that have diverged from their library counterparts — falls here, as do unresolved source group members and invalid auto-mapping rule configurations.

Quality issues are informational flags about mapping content rather than structure. Missing transformation logic, data type mismatches between source and target columns, nullability mismatches (a nullable source feeding a non-nullable target), length mismatches that could cause truncation, and missing or malformed default values all appear here. These don't indicate a broken workbook — they indicate mappings that may need attention before the map is ready for review.

Quality Flags

When you validate the current workbook, Bridger stamps quality flags directly onto the affected mappings. Any mapping with an issue gets an orange warning triangle visible in the mappings panel. Hover over the triangle to see a tooltip describing what's wrong.

The triangles make it practical to work through issues without keeping the validation report open. Scan the mappings panel for orange triangles, open each affected mapping, fix the issue, and move on. The mapping filter is also useful here — filtering by quality flag lets you isolate the affected rows without scrolling through a long list.

Once you've addressed the issues, re-run validation. If the fixes were effective the triangles disappear and the report comes back clean. If new issues have surfaced — schema changes sometimes introduce fresh problems while you're fixing old ones — they'll appear in the new report.

Tip Quality flag icons can be hidden via Settings if you prefer a less busy workspace during active mapping work. Hiding them doesn't affect validation — the checks still run, and the icons are visible whenever the setting is turned back on.

What's Next

A validated workbook is a workbook you can stand behind. Chapter 6: Reporting Your Progress covers getting that work in front of reviewers: generating Excel exports, the spreadsheet round-trip for SME feedback, and report templates that give you control over what gets included and how it's organized.