fix(node): ANS-104 codec and recovery policy types (#26 split 2/4)

crates/gitlawb-node/src/arweave_v2.rs

@@ -0,0 +1,62 @@
+//! #26 Split PR 2 — three-outcome recovery policy types.
+//!
+//! This module owns the recovery classification policy the reviewer
+//! demanded: only a trustworthy, protocol-defined absence authorizes
+//! a paid re-upload. Everything else stays non-terminal.
+//!
+//! Deliberately narrow: the ANS-104 codec lives in `ans104`, and the
+//! gateway probe, the `verify_anchor` path, and the public verify
+//! endpoint are deferred until the uploader, retrieval proof, and
+//! recovery consumer can be proven as one vertical slice (reviewer 2,
+//! round 5). A probe against a real provider contract — same-network
+//! upload/read pair, redirect-safe client, envelope-supplying read
+//! API, provider-documented absence evidence — belongs in that slice,
+//! not here. What ships here is the policy type plus the rule that
+//! only `DefinitivelyAbsent` permits spending another upload.
+
+/// Outcome of a gateway probe for a persisted `item_id`.
+// Policy surface for the recovery slice: no construction on this
+// head (the probe arrives with the provider contract); the unit
+// test pins the rule meanwhile.
+#[allow(dead_code)]
+#[derive(Debug, Clone, Copy, PartialEq, Eq)]
+pub enum ProbeOutcome {
+    /// The persisted item was served back with a verifiable envelope
+    /// bound to the requested id. No re-upload allowed.
+    Present,
+    /// Trustworthy, protocol-defined proof the item was never
+    /// accepted. Authorizes re-upload. The exact evidence that
+    /// establishes this is defined by the provider contract in the
+    /// vertical slice that introduces the probe — never by a
+    /// mock-only body shape.
+    DefinitivelyAbsent,
+    /// Anything else: transport failure, oversized body, bad
+    /// signature, id mismatch, ambiguous gateway response. The
+    /// outbox stays non-terminal.
+    Indeterminate,
+}
+
+impl ProbeOutcome {
+    /// True iff the recovery code is allowed to spend another paid
+    /// upload request. Only `DefinitivelyAbsent` qualifies.
+    #[allow(dead_code)] // the recovery consumer lives in a later slice
+    pub fn permits_reupload(self) -> bool {
+        matches!(self, ProbeOutcome::DefinitivelyAbsent)
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    /// Only `DefinitivelyAbsent` authorizes a paid re-upload. This is
+    /// the policy the reviewer named: collapsing any ambiguous outcome
+    /// to absent spends a second immutable upload for the same
+    /// transition.
+    #[test]
+    fn only_definitively_absent_authorizes_reupload() {
+        assert!(!ProbeOutcome::Present.permits_reupload());
+        assert!(ProbeOutcome::DefinitivelyAbsent.permits_reupload());
+        assert!(!ProbeOutcome::Indeterminate.permits_reupload());
+    }
+}

crates/gitlawb-node/src/main.rs

@@ -1,5 +1,7 @@
+mod ans104;
 mod api;
 mod arweave;
+mod arweave_v2;
 mod auth;
 mod bootstrap;
 mod cert;

crates/gitlawb-node/src/server.rs

@@ -231,6 +231,10 @@ pub fn build_router(state: AppState) -> Router {
         .merge(Router::new().route("/api/v1/ipfs/pins", get(ipfs::list_pins)));
 
     // ── Arweave permanent anchors ──────────────────────────────────────────
+    // List endpoint only (public; issue #134 tracks surfacing
+    // visibility rules on list). The verify endpoint and its probe
+    // are deferred to the vertical slice with the uploader and a
+    // real provider contract (reviewer 2, #26 split 2/4 round 5).
     let arweave_routes = Router::new().route("/api/v1/arweave/anchors", get(arweave::list_anchors));
 
     // ── Bounty routes (write — require HTTP Signature) ─────────────────

scripts/ans104_golden.mjs

@@ -0,0 +1,33 @@
+import { createData, EthereumSigner } from "arbundles";
+import { createHash } from "node:crypto";
+import base64url from "base64url";
+
+const data = "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789!@#$%^&*()_+-=[]{};':\",./<>?`~";
+const tags = [
+  { name: "tag1", value: "value1" },
+  { name: "tag2", value: "value2" },
+];
+const anchor = "thisSentenceIs32BytesLongTrustMe";
+const target = "OXcT1sVRSA5eGwt2k6Yuz8-3e3g9WJi5uSE99CWqsBs";
+const signer = new EthereumSigner("8da4ef21b864d2cc526dbdb2a120bd2874c36c9d0a1fb7f8c63d7f7a8b41de8f");
+
+const item = createData(data, signer, { anchor, target, tags });
+// Sign so the captured signature is real: `createData` alone leaves a
+// zeroed placeholder, whose id is just sha256(zeros).
+await item.sign(signer);
+const raw = item.getRaw();
+const id = item.id;
+
+// `item.signature` is base64url(rawSignature). Decode and sha256.
+const sigBytes = base64url.toBuffer(item.signature);
+const idCheck = base64url.encode(createHash("sha256").update(sigBytes).digest());
+if (idCheck !== id) {
+  console.error(`MISMATCH: id=${id} sha256(sig)=${idCheck}`);
+  process.exit(1);
+}
+
+console.log("id           =", id);
+console.log("signature_b64=", item.signature);
+console.log("signature_len=", sigBytes.length);
+console.log("binary_len   =", raw.length);
+console.log("binary_hex   =", raw.toString("hex"));

scripts/ans104_golden_ed25519.mjs

@@ -0,0 +1,40 @@
+import { createData, SolanaSigner } from "arbundles";
+import bs58 from "bs58";
+import { getPublicKey } from "@noble/ed25519";
+import { createHash } from "node:crypto";
+import base64url from "base64url";
+
+// Deterministic Ed25519 keypair from seed 0x01 * 32.
+const seed = Buffer.alloc(32, 0x01);
+const pub = await getPublicKey(seed);
+const secret64 = Buffer.concat([seed, Buffer.from(pub)]);
+const signer = new SolanaSigner(bs58.encode(secret64));
+
+const data = "hello gitlawb ed25519 golden";
+const tags = [
+  { name: "App-Name", value: "gitlawb" },
+  { name: "Schema", value: "gitlawb/ref-update/v1" },
+];
+
+const item = createData(data, signer, { tags });
+await item.sign(signer);
+const raw = Buffer.from(item.getRaw());
+const id = item.id;
+const sigBytes = base64url.toBuffer(item.signature);
+const idCheck = base64url.encode(createHash("sha256").update(sigBytes).digest());
+if (idCheck !== id) {
+  console.error(`MISMATCH: id=${id} sha256(sig)=${idCheck}`);
+  process.exit(1);
+}
+const sigData = await item.getSignatureData();
+console.log("id           =", id);
+console.log("signature_b64=", item.signature);
+console.log("signature_len=", sigBytes.length);
+console.log("owner_b64    =", item.owner);
+console.log("owner_len    =", base64url.toBuffer(item.owner).length);
+console.log("sigtype      =", item.signatureType);
+console.log("binary_len   =", raw.length);
+console.log("binary_hex   =", raw.toString("hex"));
+console.log("deephash_hex =", Buffer.from(sigData).toString("hex"));
+console.log("data_b64     =", base64url.encode(Buffer.from(data)));
+console.log("isValid      =", await item.isValid());

scripts/ans104_golden_output.txt

@@ -0,0 +1,37 @@
+ANS-104 golden vectors, captured from arbundles 0.10.x.
+
+Ed25519 vector (the interop pin used by
+`ans104::tests::dataitem_matches_arbundles_golden_vector`):
+  captured by  = scripts/ans104_golden_ed25519.mjs (deterministic seed
+                 0x01 * 32 via SolanaSigner, sigtype 2, `await item.sign`)
+  data         = "hello gitlawb ed25519 golden"
+  tags         = [{name:"App-Name",value:"gitlawb"},
+                  {name:"Schema",value:"gitlawb/ref-update/v1"}]
+  target       = absent
+  anchor       = absent
+
+Outputs:
+  signature_type = 2 (Ed25519)
+  signature_len  = 64 bytes
+  owner_len      = 32 bytes
+  binary_len     = 192 bytes
+  id             = SGrcBs-ITTyzvd7eIB5kk2GdBWoxB_iTi9iJ6KOe_RE (base64url of sha256(signature))
+  deephash_hex   = f15c82431767f14ac9e66ab8e995a8cd08e094be3773245163b53c12feb50aefc55d9f8c1098fabcfbf11a462706d347
+                   (arbundles `getSignatureData` 8-element fold)
+
+binary_hex:
+0200da41825fd44ca3b2705af18fce86ed6d04d0204331965d9af5d5cb1a740fcc6587ee81501b1d7928c54c0f174fde8893560d785db4d988a3161113b10a2028038a88e3dd7409f195fd52db2d3cba5d72ca6709bf1d94121bf3748801b40f6f5c00000200000000000000300000000000000004104170702d4e616d650e6769746c6177620c536368656d612a6769746c6177622f7265662d7570646174652f76310068656c6c6f206769746c617762206564323535313920676f6c64656e
+
+The Rust side will:
+1. Parse binary via DataItem::from_binary (signature_type, signature,
+   owner, target, anchor, tags, data).
+2. Compute deep_hash via DataItem::deep_hash (the 8-element arbundles
+   fold) and compare to deephash_hex above.
+3. Verify the Ed25519 signature via DataItem::verify_data_item WITHOUT
+   re-signing.
+4. Assert sha256(signature_bytes) base64url-encodes to the captured id.
+5. Assert to_binary round-trips byte-exact (signature slot preserved).
+
+Legacy Ethereum script (scripts/ans104_golden.mjs) now signs before
+capture as well; its output is illustrative only — the node verifies
+Ed25519 (sigtype 2) on the verify path.