Module xml 

uor_addr::xml — the XML realization of UOR-ADDR (ARCHITECTURE.md “Format-specific realizations” § uor-addr-xml).

XML typed-input content-addressing under a subset of W3C Canonical XML 1.1 (XML-C14N 1.1), with the σ-projection bound to prism::crypto::Sha256Hasher.

Authoritative sources

• W3C Canonical XML Version 1.1 — Recommendation 2 May 2008 (https://www.w3.org/TR/xml-c14n11/).
• W3C XML 1.0 (Fifth Edition) — base syntax (https://www.w3.org/TR/xml/).
• SHA-256 σ-projection — NIST FIPS 180-4 (https://nvlpubs.nist.gov/nistpubs/FIPS/NIST.FIPS.180-4.pdf).

Supported canonical-XML subset

Per ARCHITECTURE.md the typed-input shape XmlValue is a partition_coproduct! over five XML grammar cases (Element, Attribute, Text, CDATA, ProcessingInstruction). The shipped conformance subset implements the rules in XML-C14N 1.1 §1.1 that apply to structurally-typed XML — without external entities, DTDs, or namespace-prefix manipulation, which are out of scope for typed content-addressing:

1. Elements are emitted as <name>...</name> with no namespace prefixes (the typed input shape does not carry namespace qualification).
2. Attributes are emitted in lexicographic byte order by attribute name per XML-C14N 1.1 §1.1 rule 3.
3. Attribute values are double-quoted; the value bytes are character-escaped per XML-C14N 1.1 §1.1 rule 4 (< → <, > → >, & → &, " → ", \t → 	, \n → 
, \r → 
).
4. Text content is character-escaped per the C14N 1.1 §1.1 rule 5 (< → <, > → >, & → &, \r → 
).
5. CDATA sections are expanded to text content before canonicalization (XML-C14N 1.1 §1.1 — CDATA is informationally equivalent to escaped text).
6. Processing instructions are emitted as <?target data?>.

Out-of-scope rules (documented deviations)

The shipped canonicalizer does not handle: namespace prefix rewriting (we don’t admit namespaced input), DTD-internal entity resolution (we don’t admit DTDs), xml: reserved attributes beyond the structural typing the parser enforces, and document whitespace outside element content (we admit only element content). These rules apply to deserialization from arbitrary XML 1.0 documents — out of scope for the typed-input pipeline. Conformance tests pin the supported subset; the deviation is documented in STANDARDS.md.

Re-exports

pub use model::AddressModel;

pub use model::AddressModelBlake3;

pub use model::AddressModelKeccak256;

pub use model::AddressModelSha3_256;

pub use model::AddressModelSha512;

pub use model::AddressRoute;

pub use pipeline::address;

pub use pipeline::address_blake3;

pub use pipeline::address_keccak256;

pub use pipeline::address_sha3_256;

pub use pipeline::address_sha512;

pub use pipeline::AddressFailure;

pub use pipeline::AddressOutcome;

pub use pipeline::AddressWitness;

pub use pipeline::VerifyError;

pub use shapes::MAX_XML_DEPTH;

pub use value::canonicalize;

pub use value::XmlValue;

pub use verbs::address_inference;

pub use verbs::VERB_TERMS_ADDRESS_INFERENCE;

pub use crate::resolvers::AddressResolverTuple;

Modules

model

xml::AddressModel* — the xml realization’s PrismModel declarations, one per admissible σ-axis (crate::hash). Each binds the shared AddressResolverTuple ψ-tower and the axis’s capacity profile (AddrBounds for the 32-byte axes, AddrBounds64 for sha512). AddressModel (sha256) is the default.

pipeline

xml::address — the XML realization’s public entry point.

shapes

Grammar constant for the XML realization. The capacity profile is the shared crate::bounds::AddrBounds; only the recursion stack-safety bound lives here.

value

XmlValue — the typed XML input handle (ADR-023 amended by ADR-060) with W3C Canonical XML 1.1 (subset) byte-output discipline.

verbs

XML realization’s ψ-chain content-address derivation verb (wiki ADR-024 + ADR-035 + ADR-036 + ARCHITECTURE.md “Common verb arena”).