Reconciling Content Sources, Build Routes, and Public Sitemaps

Counting files in one directory can miss legacy arrays, generated parameters, aliases, and remote data. This paper defines a reconciliation method that treats source, build, and public inventories as related but distinct evidence.

Methodology

Enumerate every code path capable of producing an article route, including MD/MDX collections, JavaScript or TypeScript arrays, JSON data, generated files, and CMS queries. Normalize the resulting canonical paths, identify intentional aliases, and compare them with the built route manifest and every applicable public sitemap partition.

Scope

The method addresses canonical Blog and Research collections on statically generated or hybrid web applications. Static pages, taxonomy pages, and redirects are classified separately so they cannot inflate article counts.

Findings

Count agreement is necessary but not sufficient: equal totals can conceal one removal and one addition. Set correspondence exposes that substitution. An immutable deployed revision strengthens the inference that source-to-public agreement represents the active production artifact.

Inference boundaries

The reconciliation establishes route membership at the checked revision and time. It does not prove page quality, search indexing, or that a dynamic data source remained unchanged after collection.

Limitations

Runtime personalization, unpublished CMS states, and inaccessible build manifests can weaken correspondence. In those cases a complete census of the authoritative dynamic source is required. URL normalization rules must be documented to avoid merging distinct routes.

Sources

• Sitemaps XML protocol
• Next.js generateStaticParams