Package: model

Built-in attribute keys used in requirement attr blocks. They areinjected into every schema by internal/schema.Compile and accessed bygraph checks and HTML export.

Built-in attribute keys used in requirement attr blocks. They areinjected into every schema by internal/schema.Compile and accessed bygraph checks and HTML export.

Built-in attribute keys used in requirement attr blocks. They areinjected into every schema by internal/schema.Compile and accessed bygraph checks and HTML export.

Built-in attribute keys used in requirement attr blocks. They areinjected into every schema by internal/schema.Compile and accessed bygraph checks and HTML export.

Built-in attribute keys used in requirement attr blocks. They areinjected into every schema by internal/schema.Compile and accessed bygraph checks and HTML export.

Built-in attribute keys used in requirement attr blocks. They areinjected into every schema by internal/schema.Compile and accessed bygraph checks and HTML export.

Built-in attribute keys used in requirement attr blocks. They areinjected into every schema by internal/schema.Compile and accessed bygraph checks and HTML export.

Tool-owned attribute keys used on synthesized verification-resultpseudo-requirements (internal/verify). They are namespaced underx-reqmd.* — the same convention as reqmd-code's provenance attributes —so they can never collide with authored schema attributes.

Tool-owned attribute keys used on synthesized verification-resultpseudo-requirements (internal/verify). They are namespaced underx-reqmd.* — the same convention as reqmd-code's provenance attributes —so they can never collide with authored schema attributes.

Tool-owned attribute keys used on synthesized verification-resultpseudo-requirements (internal/verify). They are namespaced underx-reqmd.* — the same convention as reqmd-code's provenance attributes —so they can never collide with authored schema attributes.

Tool-owned attribute keys used on synthesized verification-resultpseudo-requirements (internal/verify). They are namespaced underx-reqmd.* — the same convention as reqmd-code's provenance attributes —so they can never collide with authored schema attributes.

StripPin removes a trailing ~N version pin from a requirement ref.Returns the bare ref, the pin value, and true if a numeric pin waspresent. A ~ followed by a non-numeric suffix is NOT treated as apin (e.g. "UP-001abc" returns ("UP-001abc", 0, false)) — thismatches the graph layer's parsing and prevents mis-stripping refsthat contain ~ for other reasons.

This is the single source of truth for ~N stripping across thecodebase: graph (Pass 2 pin extraction), repin (newRefFrom, stripPinfor display), and verify (MergeLatest key normalization).

Built-in status enum values. The default (when no status attr isdeclared) is StatusApproved — see StatusDefault.

Built-in status enum values. The default (when no status attr isdeclared) is StatusApproved — see StatusDefault.

StatusDefault is the implicit status when none is declared on arequirement. The built-in status attribute is enum [draft, approved];approved is the coverage-providing value.

FormatScalar converts a primitive scalar value to its stringrepresentation. It handles the types that YAML decoding produces forattribute values: string, bool, int, int64, float64. Non-scalar types(slices, maps) return "" — callers should handle those withencoding/json or a domain-specific renderer.

This is the shared scalar formatter used by the exporter and reporterpackages to avoid duplicating the same type switch in four places.

Attribute accessors. Attr values come from the YAML attr block and aretyped by the parser (string, bool, int, []any of those). These helpersare the single place that type-switches on them; every consumer (graphchecks, exporters, reporters) reads attrs through them instead ofreimplementing the same switch.AttrString returns the string value of attrs[key], or "" when the keyis absent or holds a non-string value.

AttrStrings reads an array attr as []string. Returns nil when the key isabsent (so callers can tell "not set" from "explicitly empty") and[]string{} when the key is present but the array is empty. Non-stringitems are dropped.

AttrList returns an array attr in its native []any form, converting a[]string. Returns nil when the key is absent or not an array.

AttrInt reads an integer attr, handling the types that appear inpractice: int and int64 (YAML decode) and float64 (JSON round-trip).Returns 0 when the key is absent or undecodable.

FormatAttr renders an attribute value for display: scalars throughFormatScalar, slices and maps as compact JSON. Unknown types render as"" so a table cell never shows Go syntax.

Kind discriminates the node types in a document's content tree.

Node is a single content node in a parsed Markdown document. Everyheading becomes a node; a node is a requirement iff it has an adjacentattr block. Prose and other blocks attach to the nearest open heading'sBody. Nodes form a tree via Children, preserving document order.

XReqmd holds reqmd-specific directory metadata from schema.yaml.All fields map directly to the x-reqmd extension block.

TraceUpstream declares the expected upstream layer and source directories.

SourceConfig describes where to find and how to parse the originating artefact.

Document groups all content sharing a single schema.yaml. Nodes holdsthe ordered content tree; every heading in every .md file is a nodeexcept the level-1 document title (see parser). Title is the firstfile's h1 heading, when present.

Synthetic marks documents that do not originate from authored Markdownon disk — currently the pseudo-document of verification resultssynthesized by internal/verify, later also synthesized test cases. Nodesof a synthetic document carry no schema, no ID prefix, and no source filethat tooling may rewrite (e.g. repin must skip them).

CollectRequirements walks a node tree depth-first and returns allrequirement-kind nodes in document order.