docs(archive): Preserve PR #5263

docs/pr-discussions/PR-5263-fix-b-0823-suffix-pattern-match-for-nixpkgs-25-11-store-hash.md

@@ -0,0 +1,88 @@
+---
+pr_number: 5263
+title: "fix(B-0823): suffix-pattern match for nixpkgs 25.11 store-hashed kernel/initrd paths"
+author: "AceHack"
+state: "MERGED"
+created_at: "2026-05-26T17:54:20Z"
+merged_at: "2026-05-26T18:03:19Z"
+closed_at: "2026-05-26T18:03:19Z"
+head_ref: "otto-cli/b0823-iso-suffix-match-25-11-store-hashed-paths-2026-05-26"
+base_ref: "main"
+archived_at: "2026-05-26T18:07:03Z"
+archive_tool: "tools/pr-preservation/archive-pr.ts"
+---
+
+# PR #5263: fix(B-0823): suffix-pattern match for nixpkgs 25.11 store-hashed kernel/initrd paths
+
+## PR description
+
+## Summary
+
+Diagnostic dump from build run [26465084701](https://github.com/Lucent-Financial-Group/Zeta/actions/runs/26465084701) confirmed nixpkgs 25.11 places kernel + initrd at **store-hashed paths**:
+
+\`\`\`
+kernel: boot/nix/store/<hash>-linux-<version>/bzImage
+initrd: boot/nix/store/<hash>-initrd-linux-<version>/initrd
+\`\`\`
+
+The hash varies per build, so exact-path lookup IS impossible by construction.
+
+## Fix
+
+Converted \`REQUIRED_KERNEL_ANY\` + \`REQUIRED_INITRD_ANY\` from exact-path to **suffix-pattern matching**:
+
+\`\`\`typescript
+{prefix: \"boot/\", suffix: \"/bzImage\"}   // matches any boot/...path.../bzImage
+{prefix: \"boot/\", suffix: \"/initrd\"}    // matches any boot/...path.../initrd
+\`\`\`
+
+Handles both legacy (24.11 \`boot/bzImage\` at top-level) AND new (25.11 store-hashed \`boot/nix/store/<hash>-linux-<ver>/bzImage\`) layouts uniformly.
+
+Added \`matchesAny()\` helper — generic suffix-pattern matcher (\`entry.startsWith(prefix) AND entry.endsWith(suffix)\`).
+
+## Composes with
+
+- B-0823 (substrate-layer follow-up; now empirically confirmed via diagnostic dump)
+- B-0818 (sibling 25.11 isoName regression — same root cause class; same fix-pattern)
+- B-0800 (nixpkgs 25.11 EOL recovery)
+
+The diagnostic dump pattern (from #5251) worked as designed — revealed actual 25.11 paths on first failure, enabling targeted fix without log-archaeology.
+
+## Test plan
+
+- [ ] build-iso passes on this PR's commit (the actual test)
+
+🤖 Generated with [Claude Code](https://claude.com/claude-code)
+
+## Reviews
+
+### COMMENTED — @copilot-pull-request-reviewer (2026-05-26T17:56:29Z)
+
+## Pull request overview
+
+Updates the installer ISO content audit to tolerate nixpkgs 25.11’s store-hashed kernel/initrd paths by switching from exact-path checks to prefix+suffix pattern matching, so CI can still assert the ISO is bootable across channel bumps.
+
+**Changes:**
+- Converted kernel/initrd “any-of” checks from exact `path` matches to `{ prefix, suffix }` pattern candidates.
+- Added a local `matchesAny()` helper to check ISO entries via `startsWith(prefix) && endsWith(suffix)`.
+- Updated failure messages to report pattern candidates (e.g., `boot/*/bzImage`) instead of exact paths.
+
+## Review threads
+
+### Thread 1: tools/ci/audit-installer-iso-content.ts:122 (resolved)
+
+**@copilot-pull-request-reviewer** (2026-05-26T17:56:28Z):
+
+The "legacy exact-path" candidate uses {prefix: "", suffix: "boot/bzImage"}, but the matcher only checks startsWith/endsWith, so this is not actually an exact match (it would also match e.g. "some/dir/boot/bzImage"). If you need an exact legacy check, either constrain prefix to "boot/" (or remove this entry since the {prefix:"boot/", suffix:"/bzImage"} candidate already covers "boot/bzImage"), or special-case exact matching when prefix is empty.
+
+### Thread 2: tools/ci/audit-installer-iso-content.ts:128 (resolved)
+
+**@copilot-pull-request-reviewer** (2026-05-26T17:56:29Z):
+
+Same issue as the kernel legacy entry: {prefix: "", suffix: "boot/initrd"} is described as an exact legacy check, but the matcher will accept any path that merely ends with "boot/initrd". Consider using prefix "boot/" (or relying on the existing {prefix:"boot/", suffix:"/initrd"} which already matches "boot/initrd") if you want to avoid false positives.
+
+## General comments
+
+### @chatgpt-codex-connector (2026-05-26T17:54:26Z)
+
+You have reached your Codex usage limits for code reviews. You can see your limits in the [Codex usage dashboard](https://chatgpt.com/codex/cloud/settings/usage).