What validation-first conversion improves
- Find unsupported commands and profile conflicts before publishing.
- Catch the rendering mismatches that don't show up until QA reviews the converted manual.
- Provide reviewable output for engineering and documentation stakeholders.
- Re-run the same checks each revision cycle and get the same results.
How validation works in Reforge
Validation runs against the detected or explicitly selected CGM profile. Each file produces a result with three severity bands: errors (constructs that will fail or misrender on conversion), warnings (constructs that may degrade output quality), and informational notices. Severity filtering lets QA reviewers focus on blocking issues without noise from advisory notices.
- Profile detection runs before validation, scoring the file against all supported profile signatures and reporting High, Medium, or Low confidence. The primary match and a runner-up candidate are both shown.
- Each issue includes a code, description, line number where available, and a remediation hint that explains what to fix in the source file.
- Validation results can be exported as CSV or JSON for QA handoff, audit trail, and integration with downstream issue-tracking tools.
Issue Explainer: what issues mean and how to fix them
Every validation issue code in Reforge links to the Issue Explainer — a built-in reference covering 150+ distinct issue codes across all supported profiles. For each code the explainer provides:
- What the issue is — a plain-language explanation of the construct or constraint involved.
- Why it matters — how the issue affects conversion output or standards compliance downstream.
- How to fix it — specific guidance on resolving the issue in the source CGM file.
- Search online — a shortcut to find the relevant specification clause without leaving the workflow.
This removes the manual spec lookup step from the QA cycle. Engineers reviewing exception reports no longer need to cross-reference the CGM or S1000D specification to understand what a validation flag means.
Profile detection and confidence scoring
When no profile is explicitly set, Reforge auto-detects the CGM profile using a weighted scoring engine that recognises 50+ known CGM-producing tool signatures. The confidence level — High, Medium, or Low — is shown for each file alongside a runner-up candidate when the file sits near a profile boundary.
This matters for mixed-origin archives: understanding detection confidence early helps QA teams decide when to apply a forced profile override before committing to a full batch conversion. Medium or Low confidence files are candidates for manual review before export.
Advanced validation checks (Professional and Enterprise)
The following check categories run additional profile-specific constraints. They are available on Professional and Enterprise plans, and can be selectively skipped via --skip-rules on the CLI or toggled in the Convert tab's validation panel:
- Color Spaces — validates color space declarations against the profile requirement; catches device-dependent color usage that can cause rendering inconsistencies.
- Text Encoding — validates character set declarations against the profile; catches non-compliant encoding that can cause text rendering failures in downstream viewers.
- References — verifies that all internal and external references (hotspot targets, APS links, DM references) resolve correctly.
- Coordinates — validates coordinate precision and range against profile tolerances (off by default in the CLI; available in the desktop app).
CLI validation and export formats
The reforge-cli validate command provides standalone validation without conversion — useful for pre-flight checks on incoming file sets before a batch run.
Exit codes: 0 = valid (no errors), 1 = errors present (or warnings when using --strict), 2 = warnings only (normal mode). The --strict flag promotes all warnings to errors, useful for CI/CD gates where zero warnings is required. Use --skip-rules to exclude specific check categories from the run.
Practical compliance workflow
Most teams validate at intake, fix the exceptions in the source graphics, then convert the approved batch. Doing it in that order is cheaper than fixing it in QA.
- Validate source files against expected CGM profile behavior.
- Flag high-risk files for manual review before export.
- Convert approved sets to SVG for publication and portal delivery.
- Retain validation artifacts (CSV or JSON export) as part of QA documentation.
Evaluation tip: Start a trial to validate representative legacy files from your own archive. See download.
Test validation on representative legacy files
Use files that commonly trigger downstream QA issues to confirm warning visibility, workflow fit, and review handoff quality.
Need automation or profile-specific rollout guidance?
Validation works best when paired with a clear workflow for profile handling and batch execution in your publication process.