feat(iid, iid-spec): typed Wikidata identity int:wd:<slug>:Q… — spec amendment and reference implementation (P16)

packages/iid-spec/README.md

@@ -72,6 +72,25 @@ import vectors from '@0xintuition/iid-spec/conformance/canonicalization.json'
 **Class C — derived** (a hash of the few facts we actually have)
 [`gen1`](./schemes/gen1.md)
 
+## Typed Wikidata identifiers
+
+[`wd`](./schemes/wd.md) is a value-typed scheme: bare `int:wd:Q42` remains polymorphic and floors at P1; active `<slug>:Q…` bindings carry the type and qualify for P0. The active slugs are `film`, `television-series`, and `human`. Dormant bindings remain valid for parsing and validation but are not mintable. URLs canonicalize to bare QIDs and never infer a slug.
+
+The reference implementation reports typing per value:
+
+```ts
+import { inspectIntuitionId } from '@0xintuition/iid'
+
+inspectIntuitionId('int:wd:Q42')
+// { valid: true, typing: 'polymorphic', anchorEligible: false,
+//   anchorIneligibilityReason: 'polymorphic-scheme', … }
+
+inspectIntuitionId('int:wd:film:Q188035')
+// { valid: true, typing: 'unambiguous', anchorEligible: true, wdSlug: 'film', … }
+```
+
+The binding table and canonicalization rules are in [`schemes/wd.md`](./schemes/wd.md); per-value typing is defined in [§7.3](./spec/07-representation-profiles.md). This additive extension follows [§9.3.1](./spec/09-registry-governance.md), preserving existing bare identifiers and their eligibility.
+
 ## The conformance corpus
 
 The [`conformance/`](./conformance/) directory is the machine-readable half of this specification. Every vector is normative, generated from the reference implementation, and cross-checked against the golden values stated in the specification prose.
@@ -84,7 +103,7 @@ The [`conformance/`](./conformance/) directory is the machine-readable half of t
 | `gen1.json` | Full `gen1` derivations: recipe fields, exact preimage, and resulting IID |
 | `parse.json` | Grammar/parsing vectors, including colon-bearing values and rejections |
 | `validation.json` | Full-IID validity and P0 anchor-eligibility vectors |
-| `schemes.json` | The registry snapshot: identity class and scheme typing per scheme |
+| `schemes.json` | The registry snapshot: identity class, scheme typing, and optional `valueTyped` marker per scheme |
 
 Fixture files are versioned (`specVersion`, `fixtureVersion`) and append-only: changing an existing vector's expected output is an identity-impacting change and follows the governance process in [§9](./spec/09-registry-governance.md).
 
@@ -101,7 +120,7 @@ Every test vector in this specification is executable against the reference impl
 
 ## Status
 
-**Version 0.1.0 — Draft.** The grammar, the class system, NORM-1, and the `gen1` algorithm are stable and in production use. Scheme canonicalization rules are frozen under the [freeze rule](./spec/09-registry-governance.md): once ratified, a scheme's rules never change in place — a change ships as a new scheme name or version.
+**Version 0.1.0 — Draft.** The grammar, the class system, NORM-1, and the `gen1` algorithm are stable and in production use. Scheme canonicalization rules are frozen under the [freeze rule](./spec/09-registry-governance.md): changes to existing canonical values require a new scheme name or version; additive value-grammar extensions follow [§9.3.1](./spec/09-registry-governance.md).
 
 This specification is published for review and adoption. It is not yet submitted to any standards body.
 

packages/iid-spec/conformance/canonicalization.json

@@ -709,6 +709,104 @@
 			"raw": "movie:r4:FB681AFE7D438CAD73AE90A70F1CC55A",
 			"canonical": null,
 			"note": "hash must be lowercase"
+		},
+		{
+			"id": "canon-wd-102",
+			"scheme": "wd",
+			"raw": "film:q42",
+			"canonical": "film:Q42",
+			"note": "typed QID uppercases"
+		},
+		{
+			"id": "canon-wd-103",
+			"scheme": "wd",
+			"raw": " \tfilm:q42\r\n",
+			"canonical": "film:Q42",
+			"note": "edge whitespace trims"
+		},
+		{
+			"id": "canon-wd-104",
+			"scheme": "wd",
+			"raw": "https://www.wikidata.org/wiki/Q188035",
+			"canonical": "Q188035",
+			"note": "URL remains bare; never infers a slug"
+		},
+		{
+			"id": "canon-wd-105",
+			"scheme": "wd",
+			"raw": "bogus:Q1",
+			"canonical": null,
+			"note": "unknown binding rejects"
+		},
+		{
+			"id": "canon-wd-106",
+			"scheme": "wd",
+			"raw": "written-worK:Q42",
+			"canonical": null,
+			"note": "non-ASCII slug rejects before case folding"
+		},
+		{
+			"id": "canon-wd-107",
+			"scheme": "wd",
+			"raw": "film:Q42",
+			"canonical": "film:Q42",
+			"note": "U+FEFF trims like other edge whitespace (base behavior preserved)"
+		},
+		{
+			"id": "canon-wd-108",
+			"scheme": "wd",
+			"raw": "film:Q042",
+			"canonical": null,
+			"note": "typed QID leading zero rejects"
+		},
+		{
+			"id": "canon-wd-109",
+			"scheme": "wd",
+			"raw": "FILM:q42",
+			"canonical": "film:Q42",
+			"note": "ASCII slug lowercases"
+		},
+		{
+			"id": "canon-wd-110",
+			"scheme": "wd",
+			"raw": "written-work:q42",
+			"canonical": "written-work:Q42",
+			"note": "dormant binding remains canonical and parse-only"
+		},
+		{
+			"id": "canon-wd-111",
+			"scheme": "wd",
+			"raw": "Q42",
+			"canonical": "Q42",
+			"note": "BOM-wrapped bare QID keeps its base canonicalization"
+		},
+		{
+			"id": "canon-isni-url-112",
+			"scheme": "isni",
+			"raw": "https://isni.org/isni/0000000121032683",
+			"canonical": "0000000121032683",
+			"note": "ISNI URL prefix strips to the existing bare canonical value (section 9.3.1)"
+		},
+		{
+			"id": "canon-isni-url-113",
+			"scheme": "isni",
+			"raw": "https://www.isni.org/isni/0000000121032683",
+			"canonical": "0000000121032683",
+			"note": "ISNI URL accepts the www host alias"
+		},
+		{
+			"id": "canon-isni-url-114",
+			"scheme": "isni",
+			"raw": "HTTPS://ISNI.ORG/isni/0000000121032683",
+			"canonical": "0000000121032683",
+			"note": "ISNI URL scheme and host are case-insensitive"
+		},
+		{
+			"id": "canon-isni-url-115",
+			"scheme": "isni",
+			"raw": "https://isni.org/ISNI/0000000121032683",
+			"canonical": "0000000121032683",
+			"note": "ISNI URL path segment is matched case-insensitively as well (whole-prefix fold, as the reference implementation does)"
 		}
 	]
 }

packages/iid-spec/conformance/parse.json

@@ -79,6 +79,20 @@
 			"scheme": null,
 			"value": null,
 			"note": "value exceeds 220-char limit"
+		},
+		{
+			"id": "parse-012",
+			"input": "int:wd:film:Q188035",
+			"scheme": "wd",
+			"value": "film:Q188035",
+			"note": "typed Wikidata value survives first-two-colon split"
+		},
+		{
+			"id": "parse-013",
+			"input": "int:wd:Q42",
+			"scheme": "wd",
+			"value": "Q42",
+			"note": "legacy bare Wikidata parse remains unchanged"
 		}
 	]
 }

packages/iid-spec/conformance/schema.json

@@ -140,6 +140,10 @@
 							"id": { "$ref": "#/$defs/vectorId" },
 							"iid": { "type": "string" },
 							"valid": { "type": "boolean" },
+							"typing": {
+								"enum": ["unambiguous", "polymorphic"],
+								"description": "Effective typing of THIS identifier; permitted only for a scheme marked `valueTyped`; defaults to the scheme's typing."
+							},
 							"anchorEligible": {
 								"type": "boolean",
 								"description": "Whether the IID may be minted as a bare P0 anchor (spec §7.2)."
@@ -163,7 +167,11 @@
 						"additionalProperties": false,
 						"properties": {
 							"class": { "enum": ["A", "B", "C"] },
-							"typing": { "enum": ["unambiguous", "polymorphic"] }
+							"typing": { "enum": ["unambiguous", "polymorphic"] },
+							"valueTyped": {
+								"type": "boolean",
+								"description": "Bare values are polymorphic, but a value MAY carry the entity type inside the value (spec §7.3); validation vectors for such a scheme MAY state per-value typing."
+							}
 						}
 					}
 				}

packages/iid-spec/conformance/schemes.json

@@ -105,7 +105,8 @@
 		},
 		"wd": {
 			"class": "A",
-			"typing": "polymorphic"
+			"typing": "polymorphic",
+			"valueTyped": true
 		}
 	}
 }

packages/iid-spec/conformance/validation.json

@@ -43,7 +43,7 @@
 			"iid": "int:wd:Q42",
 			"valid": true,
 			"anchorEligible": false,
-			"note": "valid but polymorphic — floors at P1"
+			"note": "valid but bare — polymorphic, floors at P1"
 		},
 		{
 			"id": "valid-007",
@@ -100,6 +100,59 @@
 			"valid": true,
 			"anchorEligible": true,
 			"note": "strong acct form is anchor eligible"
+		},
+		{
+			"id": "valid-015",
+			"iid": "int:wd:film:Q188035",
+			"valid": true,
+			"typing": "unambiguous",
+			"anchorEligible": true,
+			"note": "active film binding is P0 eligible"
+		},
+		{
+			"id": "valid-016",
+			"iid": "int:wd:human:Q42",
+			"valid": true,
+			"typing": "unambiguous",
+			"anchorEligible": true,
+			"note": "active human binding is P0 eligible"
+		},
+		{
+			"id": "valid-017",
+			"iid": "int:wd:written-work:Q47461344",
+			"valid": true,
+			"typing": "unambiguous",
+			"anchorEligible": false,
+			"note": "dormant written-work binding is valid but not mintable"
+		},
+		{
+			"id": "valid-018",
+			"iid": "int:wd:film:q42",
+			"valid": false,
+			"anchorEligible": false,
+			"note": "typed QID must be canonical byte-exactly"
+		},
+		{
+			"id": "valid-019",
+			"iid": "int:wd:bogus:Q1",
+			"valid": false,
+			"anchorEligible": false,
+			"note": "unknown binding rejects"
+		},
+		{
+			"id": "valid-020",
+			"iid": "int:wd:television-series:Q137400033",
+			"valid": true,
+			"typing": "unambiguous",
+			"anchorEligible": true,
+			"note": "active television series binding is P0 eligible"
+		},
+		{
+			"id": "valid-isni-url",
+			"iid": "int:isni:https://isni.org/isni/0000000121032683",
+			"valid": false,
+			"anchorEligible": false,
+			"note": "ISNI URL is accepted raw input but is not a canonical IID value"
 		}
 	]
 }

packages/iid-spec/schemes/README.md

@@ -15,7 +15,7 @@ This directory is the normative scheme registry referenced by [§9.1](../spec/09
 | [`gtin`](./gtin.md) | A | T1 | unambiguous | A trade item (product) |
 | [`doi`](./doi.md) | A | T1 | polymorphic | A registered digital object |
 | [`eidr`](./eidr.md) | A | T1 | unambiguous | An audiovisual work |
-| [`wd`](./wd.md) | A | T2 | polymorphic | Any Wikidata entity |
+| [`wd`](./wd.md) | A | T2 | polymorphic (bare) · unambiguous (typed, active binding) | Any Wikidata entity (value-typed) |
 | [`mbid`](./mbid.md) | A | T2 | unambiguous | A MusicBrainz entity (type in-value) |
 | [`olid`](./olid.md) | A | T2 | unambiguous | An OpenLibrary work, edition, or author |
 | [`imdb`](./imdb.md) | A | T3 | polymorphic | An IMDb title or name |

packages/iid-spec/schemes/isni.md

@@ -22,6 +22,12 @@ That is what "polymorphic" means here, and it has a hard consequence: an `isni`
 
 with the additional constraint that the final character is the correct ISO 7064 mod 11-2 check character over the first 15 digits. The grammar alone is not sufficient — a conforming validator recomputes the check character.
 
+## Accepted input forms
+
+The existing bare 16-character form and separator-containing display forms remain accepted. The `https://isni.org/isni/<id>` URL form, with or without `www.`, is also accepted; the whole prefix (scheme, host and the `/isni/` path segment) is matched case-insensitively, so `https://isni.org/ISNI/<id>` is accepted as well; the identifier itself is then canonicalized as below. Strip that URL prefix before applying the existing canonicalization below. Canonical output and existing trim behavior remain unchanged; a URL is an input form, never a canonical IID value.
+
+This follows [§9.3.1](../spec/09-registry-governance.md): an additive value-grammar extension accepts a previously-invalid raw form without changing any existing canonical bytes. The URL form mirrors v2 commit `83a003bc1` (R24, default-in-effect).
+
 ## Canonicalization
 
 1. **Strip separators.** Remove every hyphen, dot, and whitespace character (`[-\s.]`). ISNIs are conventionally displayed in four space-separated groups of four; the grouping carries no identity.