DataValidationCompiler

Create a data validation compiler.

Compile validation for one filled data graph.

Compile validation for one filled root data collection.

Get the paths restored after validation.

A constructor input is restored only when no compiled rule applies to it or to anything beneath it; otherwise the validated result owns it.

Compile one data node into the root rule graph.

Compile nested rules for one data iterable.

Compile one data iterable from its current structure path.

Compile dynamic items at one wildcard path when their complete output matches.

Build the rules for one property: inferred presence and type rules, then declared attributes.

Get the fresh validation attributes declared on a property.

A Rule attribute's rules become typed attributes where possible, so they replace inferred rules of their type.

Infer fixed presence and type rules for one property.

Run the configured rule inferrers over one property's inferred and attribute rules.

Declared attributes follow the inferred rules and replace inferred rules of the same type.

Resolve the configured rule inferrers for one compilation.

They are resolved per compilation, so scoped inferrers follow the current coroutine.

Apply class-owned rule replacements in PHP property-name space.

Collapse concrete translations onto an existing authoritative wildcard rule.

Apply class-owned messages and attribute labels in PHP property-name space.

Invoke one array-returning validation lifecycle method.

Resolve a field reference's property names to input names.

Unlike rule keys, references are not filtered, since a condition may read input that has no rules.

Translate a class rule key to its observed wire paths.

Recursively translate class-rule segments through Data metadata.

Translate the rest of a key that reaches a finished value.

A finished value is not validated internally, so only a key ending at the value applies, or a message key ending with one rule for it.

Append rule segments that no longer describe Data properties.

Append wildcard identity markers after every rule replacement is complete.

Determine if a rule list contains a distinct rule.

Merge rule lists without duplicating identical string rules.

Determine if a rule list explicitly controls field presence.

Get the keyword of a string or structured validation rule.

Determine if item rules can differ across one collection.

Get one unambiguous type validation rule.

Get the type rule for one declared named type.

Get the inferred type rule for one property value.

A union value uses the rule of the declared type that holds it, so numeric size rules keep their numeric meaning. Other values use the declaration's rule.

Get the wire key selected during Fill or its canonical fallback.

Get the data object or data collection type that receives a property's value on the current node.

A union value uses the type Fill selected from the raw value, because Fill replaced a data value with its normalized input. Without a value, the declaration's one data type applies.

Record an exact field or opaque subtree excluded from rule compilation.

Determine if a skipped property can contain nested input.

Determine if a validated property has no recursive schema.

Determine if a named object type has no Data-owned child schema.