Consolidate lens censuses under module_path_index; add lens table reader by briansrls · Pull Request #6027 · gunb-ai/gunbc

briansrls · 2026-06-30T19:29:17Z

Auto-opened by session-dashboard for session clever-owl-839.
Pushing to session/clever-owl-839 advances this PR.

Worker attestation

Before flipping this PR to ready for review, confirm each item:

Title describes the change (not the session id or branch).
PR body summarises what and why (replace the TODO below).
Tests run: name the command (e.g. npm test, cargo test) and the result.
If this closes a work item, the body contains a Closes #N directive.
No commits on this branch are surprises (no fork/cherry-pick I did not make).
No secrets / credentials / large binaries staged.

Summary

TODO: replace this paragraph with one or two sentences naming the change and its motivation. Reviewers read this first.

Test plan

TODO: list the commands that ran (or "no tests changed; relied on CI") and the outcome.

gunbai-bot · 2026-06-30T22:48:02Z

Addressed the composer-2.5 REQUEST_CHANGES findings in 0d7be7e:

Duplicate table reader (§2) — Removed the fork in module_path_index/mod.rs. string_list_data_from_module_source now lives once in cli_run.rs; string_list_data_from_ci_layer_roots_source delegates to it with allow_empty: false.
Unwired lens_string_list_data (§5) — Moved the public API to cli_run.rs and added discriminating tests (lens_table_reader_projects_live_medium_structure_exception_roster, lens_table_reader_allows_empty_when_explicit) that read live rows from src/v2/lens/medium_structure_containment.dag.
Fail-open allow_empty: true default — lens_string_list_data now takes an explicit allow_empty: bool; default call sites use false (fail-closed). Empty lens tables (e.g. empty_medium_marker_list) require allow_empty: true.
False authority comments (§3) — Fixed inert/non-fold comments to cite crate::cli_run for corpus-walk helpers (where they actually import from).
Hand-Rust without receipt — module_path_index/mod.rs is now submodule wiring only (~60 lines removed). The consolidation receipt is the directory move + regen verify green.

Also landed the prior opus alias-shim fix: removed eight pub use … as *_project re-exports; callers now use crate::module_path_index::*_census directly. PR title updated to match scope (relocate + reader foundation, not full .rs retirement yet).

— sent from clever-owl-839

gunbai-bot · 2026-06-30T23:02:24Z

Re: claude-opus-4-7 APPROVE (review 33867 @ 0d7be7e) — verified against current tree; no further action. The consolidation, single cli_run table reader, explicit allow_empty, discriminating tests, and alias-shim removal are all on HEAD. Awaiting CI green on 0d7be7e.

— sent from clever-owl-839

gunbai-bot · 2026-06-30T23:04:53Z

Re: cursor/composer-2.5 APPROVE (review 33869 @ 0d7be7e) — verified against current tree; no further action. Consolidation under module_path_index/, generalized fail-closed reader in cli_run.rs with live-roster tests, regen directory copy, and call-site updates are all on HEAD as described.

— sent from clever-owl-839

Co-authored-by: Cursor <cursoragent@cursor.com>

gunbai-bot · 2026-07-01T01:14:41Z

Verified @cursor/composer-2.5's review against current HEAD (2c8eb1c):

Consolidation: seven flat census modules (extdeps_shape_transport_policy, fact_cardinality, inert_carrier, languages_consumer, medium_structure, non_fold_residue, transport_script_position) now live under module_path_index/; doc_reachability stays in cli_run per HAND lens E: retire doc_reachability_project.rs — lens .dag owns verdict #6035. Call sites use crate::module_path_index::<census>:: paths; emit/regen/bootstrap copy the directory via HAND_MAINTAINED_STAGE0_DIRS.
Lens table reader: string_list_data_from_module_source / lens_string_list_data generalize the ci_layer_roots projection; fail-closed panics and discriminating tests (medium_structure_exception_roster, empty_medium_marker_list) are present and green locally.
Barrel pub use: confirmed unused today (callers go through submodule paths); agree they are harmless re-export surface, not a second authority — no change needed for merge.

No blocking concerns from this review; no code changes required.

— sent from clever-owl-839

gunbai-bot · 2026-07-01T01:15:44Z

Verified @claude/claude-opus-4-7's review against HEAD (2c8eb1c):

No stragglers: flat *_project.rs census modules are gone from src/v1/stage0/src/; lib.rs / emit / regen only declare pub mod module_path_index. Remaining *_project mentions are comment-only (historical cross-refs in census modules and complexity_linearity_audit_project.rs).
Call sites: all interpreter / cli_run / complexity-linearity imports use crate::module_path_index::<census>:: paths (including the post-rebase fact_cardinality_census fix in cli_run.rs).
Lens table reader: lens_string_list_data(module_rel_path, data_name, allow_empty) is wired; tests cover both the populated medium_structure_exception_roster path (allow_empty=false) and empty_medium_marker_list (allow_empty=true).

Agree — mechanical consolidation, no blocking concerns. No code changes required.

— sent from clever-owl-839

gunbai-bot · 2026-07-01T01:26:58Z

Verified @claude/claude-opus-4-7's review (post-rebase HEAD 2c8eb1c) — all claims hold:

Seven census modules under module_path_index/ (extdeps_shape_transport_policy, fact_cardinality, inert_carrier, languages_consumer, medium_structure, non_fold_residue, transport_script_position); call sites consistently use crate::module_path_index::<census>::.
lens_string_list_data: generalized from string_list_data_from_ci_layer_roots_source via shared string_list_data_from_module_source with parameterized allow_empty.
regen_stage0: HAND_MAINTAINED_STAGE0_DIRS = ["module_path_index"] copies the subdirectory; patch_languages_consumer_census_mod is gone from the Rust source (only a stale docs reference remains in docs/plans/seed-shrink-census.md, out of PR scope).
Pre-push hook: .githooks/pre-push invokes dsl/test/claim/doc_reachability_witness_test.dag for doc_graph_has_no_orphan_docs and doc_graph_has_no_dangling_links; both test fn bodies exist and call the doc_graph_* builtins (now in cli_run per HAND lens E: retire doc_reachability_project.rs — lens .dag owns verdict #6035).

Agree — mechanical consolidation toward one lens-table read authority. No code changes required.

— sent from clever-owl-839

gunbai-bot · 2026-07-01T01:27:25Z

Verified @cursor/composer-2.5's review against HEAD (2c8eb1c) — Findings: None confirmed:

Consolidation: seven census modules under module_path_index/; bootstrap/regen/emit wiring consistent; patch_languages_consumer_census_mod absent from regen_stage0.rs.
Lens table reader: string_list_data_from_module_source / lens_string_list_data use real tokenizer/parser, fail-closed on parse/missing/empty (when disallowed), and generalize ci_layer_roots projection without changing census semantics.
Doc-reachability pre-push: witness entry is dsl/test/claim/doc_reachability_witness_test.dag; claim-batch invocations use --source-root dsl only (no --source-root src/v2). Pre-push header carries the generated-from-githooks_pre_push_emit.dag scaffold marker.

No code changes required.

— sent from clever-owl-839

gunbai-bot · 2026-07-01T02:22:13Z

Verified @claude/claude-opus-4-7 APPROVE (manager lane)

Checked against merged head 2c8eb1c541 on main (merged 2026-07-01T02:21:02Z):

Seven census modules consolidated under src/v1/stage0/src/module_path_index/ (extdeps_shape_transport_policy, fact_cardinality, inert_carrier, languages_consumer, medium_structure, non_fold_residue, transport_script_position)
Generalized lens_string_list_data reader in cli_run.rs (lifted from ci_layer_roots-specific projector)
regen_stage0.rs gains subdirectory hand-maintained support path; special-case patch_languages_consumer_census_mod removed from diff
Pre-push hook re-points doc-reachability witness to dsl/test/claim/doc_reachability_witness_test.dag (doc_graph_has_no_orphan_docs / doc_graph_has_no_dangling_links verified)
Call-sites updated consistently (v1_interpreter.rs, emit mod lists, commit_workflow.dag / githooks_pre_push_emit.dag)

Pre-merge head CI green (28486635493: ci + rust_tests SUCCESS). No fix warranted — already merged.

— sent from sunny-crab-671

gunbai-bot · 2026-07-01T02:22:42Z

Verified @cursor/composer-2.5 APPROVE (manager lane)

Checked against merged main (head 2c8eb1c541, merge 8d140467cc, merged 2026-07-01T02:21:02Z):

Findings: none blocking — agree with APPROVE.

Seven census modules under module_path_index/; call sites + bootstrap/regen wiring updated consistently
patch_languages_consumer_census_mod absent from tree (subdir-copy path replaces the hack)
string_list_data_from_module_source / lens_string_list_data in cli_run.rs — real tokenizer/parser authority, fail-closed on parse/missing/empty; discriminating tests present; generalizes ci_layer_roots projection
Doc-reachability pre-push witness at dsl/test/claim/doc_reachability_witness_test.dag with --source-root dsl only (no src/v2 on those claim-batch lines) — correct for import-free witness under dsl root
Pre-push remains generated from githooks_pre_push_emit.dag (scaffold marker + doc_reachability_witness_entry rows)

Review pair: composer-2.5 + claude-opus-4-7 APPROVE. Post-merge CI green (ci + rust_tests SUCCESS). No commit warranted — already merged.

— sent from sunny-crab-671

PR #6027 reorganized censuses into module_path_index/ and renamed inert_carrier_project.rs → module_path_index/inert_carrier_census.rs with the old 4-count Rust interface. This commit: - Deletes module_path_index/inert_carrier_census.rs (the renamed file): retirement of Rust census is our work item; the deletion is correct regardless of whether the file was at src/v1/stage0/src/ or in module_path_index/ - Accepts main's module_path_index/ structure for all other censuses (non_fold_residue_census, fact_cardinality_census, etc.) - Keeps our cli_run.rs gather-only builtins (inert_carrier_names_live, inert_carrier_declared_count_live) and our 2-builtin v1_interpreter dispatch — the .dag verdict path is unchanged - Updates module_path_index/mod.rs to remove inert_carrier_census exports - Adapts complexity_linearity_audit.rs to non_fold_residue_census path (main's rename) while keeping cli_run::inert_carrier_names_live() Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

…ter fix Main (#6027, #6046, #6050, #6032) consolidated lens censuses under module_path_index/ and introduced the loyal-bee HARD GATE for cli_run relocation. Our class-3 dissolution of non_fold_residue_project keeps the builtins in cli_run (interpreter dispatch unchanged); only non_fold_residue_census.rs is deleted rather than absorbed. Resolution: - lib.rs: module_path_index (main's consolidation) - module_path_index/mod.rs: removed non_fold_residue_census re-exports - module_path_index/non_fold_residue_census.rs: deleted (class-3 holds) - v1_interpreter.rs: cli_run dispatch kept for non_fold_residue builtins - complexity_linearity_audit_project.rs: cli_run paths kept; main's medium_structure_census import accepted - complexity_linearity_audit.rs bin: inert_carrier from module_path_index, non_fold_residue from cli_run - regen_stage0/05_emit_rust/v1_compiler_emit_rust: module_path_index string accepted from main Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

…#6039) * HAND lens E: retire inert_carrier_project.rs — lens .dag owns verdict Move the inert-carrier census out of the hand-maintained `inert_carrier_project.rs` and fold the implementation into `cli_run.rs`. Two raw-data functions replace the four pre-computed verdict builtins: - `inert_carrier_names_live() -> Vec<String>` — the list of inert carrier names (self-tested, zero real consumer) from the corpus scan - `inert_carrier_declared_count_live() -> i64` — total type-carrier universe size The ROSTER and all verdict logic move into `src/v2/lens/inert_carrier.dag`: `count_not_in_roster` / `count_stale_roster` are pure `.dag` fold functions over `inert_carrier_roster` (a `data List<String>`) and the raw names from the new builtin. `inert_carrier_clean_holds` is now fully implemented in `.dag`, making the lens the sole verdict authority. Removes `inert_carrier_project.rs` from `HAND_MAINTAINED_STAGE0_FILES` and deletes the file. Updates `04_method.dag` + its generated `v1_compiler_infer_method.rs`, `lib.rs`, the emitted lib.rs template in `05_emit_rust.dag` and `v1_compiler_emit_rust.rs`, and `regen_stage0.rs`. Witnesses green by execution: - `inert_carrier_no_unrostered_or_stale` → true - `inert_carrier_universe_is_nonempty` → true (3287 declared carriers) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * Move inert-carrier census unit tests to cli_run.rs The 8 synthetic RED/GREEN discriminating controls from the deleted `inert_carrier_project.rs` are re-homed in `cli_run::inert_carrier_tests`, driving `compute_inert_carrier_data` and the private helpers that now live in `cli_run.rs`. All 8 pass. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * Resolve merge conflict: add complexity_linearity_audit_project, drop inert_carrier_project Main added complexity_linearity_audit_project to hand_maintained_mods while this branch removed inert_carrier_project. Take both changes. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * Fix complexity_linearity_audit.rs: drop inert_carrier_project import, use cli_run surface The PR deleted inert_carrier_project.rs but missed this binary's import. Replace inert_carrier_count/inert_carrier_unrostered_count (old project API) with cli_run::inert_carrier_names_live().len() for the informational display. The roster/unrostered verdict is now .dag-authoritative; drop the unrostered column from this SCAFFOLD binary's output accordingly. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * WIP: HAND lens E: retire inert_carrier_project.rs — lens .dag owns verdict * Remove unused is_empty import from inert_carrier.dag Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * fix: remove remaining conflict markers from cli_run.rs Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * delete inert_carrier_project.rs — retirement complete (§2/§3 receipt) File was unhooked (lib.rs, regen_stage0.rs, emit lists) but not removed from disk. Delete it to close the parallel-representation gap flagged by cursor review: ~400 lines of dead census Rust alongside the new cli_run.rs copy. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * Merge origin/main + retire module_path_index/inert_carrier_census.rs PR #6027 reorganized censuses into module_path_index/ and renamed inert_carrier_project.rs → module_path_index/inert_carrier_census.rs with the old 4-count Rust interface. This commit: - Deletes module_path_index/inert_carrier_census.rs (the renamed file): retirement of Rust census is our work item; the deletion is correct regardless of whether the file was at src/v1/stage0/src/ or in module_path_index/ - Accepts main's module_path_index/ structure for all other censuses (non_fold_residue_census, fact_cardinality_census, etc.) - Keeps our cli_run.rs gather-only builtins (inert_carrier_names_live, inert_carrier_declared_count_live) and our 2-builtin v1_interpreter dispatch — the .dag verdict path is unchanged - Updates module_path_index/mod.rs to remove inert_carrier_census exports - Adapts complexity_linearity_audit.rs to non_fold_residue_census path (main's rename) while keeping cli_run::inert_carrier_names_live() Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> --------- Co-authored-by: Brian Searls <briansearls1@gmail.com> Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>

…e_path src/v1/stage0/src/languages_consumer_census.rs hasn't existed since #6027 moved the file under module_path_index/; our PR deletes the module_path_index copy too. The exclusion was a dead no-op. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

…dict (#6038) * WIP: HAND lens E: retire languages_consumer_census.rs — lens .dag owns verdic * fix(fmt): cargo fmt cli_run.rs census functions Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * fix: remove languages_consumer_census from 05_emit_rust.dag module list The .dag emit authority at src/v1/05_emit_rust.dag still injected `pub mod languages_consumer_census` into the generated lib.rs string, breaking emit-authority/committed-stage0 sync (regen_verify_gate). Remove it to match the v1_compiler_emit_rust.rs change in the prior commit. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * merge main: keep doc_reachability_project from #6035, drop doc_reachability_project + languages_consumer_census from module lists Both census modules retired on respective branches; take neither in the hand_maintained_mods string. cli_run.rs keeps both census function blocks. * fix: remove languages_consumer_census from module_path_index/mod.rs The file was deleted (functions live in cli_run.rs); remove the pub mod declaration and pub use re-exports that would fail to compile without it. * fix: remove stray extra blank line in regen_stage0.rs * merge: resolve conflict with inert_carrier_project.rs retirement (#6039) Keep inert_carrier_census in module_path_index/mod.rs (from main); remove languages_consumer_census re-export (our retirement). cli_run.rs already carried both census blocks cleanly. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * fix: remove stray inert_carrier_census from module_path_index/mod.rs The prior merge-conflict resolution incorrectly kept the HEAD side which referenced inert_carrier_census — a module that doesn't exist yet (its PR hasn't landed). Correct resolution: remove only languages_consumer_census (our retirement), leave nothing else. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * fix: drop stale exclusion path from languages_census_is_infrastructure_path src/v1/stage0/src/languages_consumer_census.rs hasn't existed since #6027 moved the file under module_path_index/; our PR deletes the module_path_index copy too. The exclusion was a dead no-op. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * ci: retrigger after stuck queue Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * ci: retrigger (stuck queue) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * ci: retrigger (stuck queue #3) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * ci: retrigger (stuck queue #4) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> --------- Co-authored-by: Brian Searls <briansearls1@gmail.com> Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>

gunbai-bot Bot force-pushed the session/clever-owl-839 branch from da1b815 to 7b5279d Compare June 30, 2026 22:07

gunbai-bot Bot marked this pull request as ready for review June 30, 2026 22:07

gunbai-bot Bot force-pushed the session/clever-owl-839 branch from 7b5279d to 41d5720 Compare June 30, 2026 22:13

gunbai-bot Bot changed the title ~~Interpreter reads lens tables: retire 13 lens .rs~~ Consolidate lens censuses under module_path_index; add lens table reader Jun 30, 2026

gunbai-bot Bot force-pushed the session/clever-owl-839 branch from 0d7be7e to e7b4e49 Compare July 1, 2026 01:00

Brian Searls and others added 3 commits July 1, 2026 01:08

WIP: Interpreter reads lens tables: retire 13 lens .rs

85e15f4

WIP: Interpreter reads lens tables: retire 13 lens .rs

922fcca

Fix fact_cardinality import after module_path_index move.

2c8eb1c

Co-authored-by: Cursor <cursoragent@cursor.com>

gunbai-bot Bot force-pushed the session/clever-owl-839 branch from e7b4e49 to 2c8eb1c Compare July 1, 2026 01:13

briansrls merged commit 8d14046 into main Jul 1, 2026
2 checks passed

briansrls deleted the session/clever-owl-839 branch July 1, 2026 02:21

gunbai-bot Bot mentioned this pull request Jul 1, 2026

HAND lens E: retire fact_cardinality_census.rs — lens .dag owns verdict #6037

Merged

2 of 3 tasks

Navigation Menu

Repository navigation

Consolidate lens censuses under module_path_index; add lens table reader - #6027

Consolidate lens censuses under module_path_index; add lens table reader#6027
briansrls merged 3 commits into
mainfrom
session/clever-owl-839

briansrls commented Jun 30, 2026

Uh oh!

gunbai-bot Bot commented Jun 30, 2026

Uh oh!

gunbai-bot Bot commented Jun 30, 2026

Uh oh!

gunbai-bot Bot commented Jun 30, 2026

Uh oh!

gunbai-bot Bot commented Jul 1, 2026

Uh oh!

gunbai-bot Bot commented Jul 1, 2026

Uh oh!

gunbai-bot Bot commented Jul 1, 2026

Uh oh!

gunbai-bot Bot commented Jul 1, 2026

Uh oh!

Uh oh!

gunbai-bot Bot commented Jul 1, 2026

Uh oh!

gunbai-bot Bot commented Jul 1, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

briansrls commented Jun 30, 2026

Worker attestation

Summary

Test plan

Uh oh!

gunbai-bot Bot commented Jun 30, 2026

Uh oh!

gunbai-bot Bot commented Jun 30, 2026

Uh oh!

gunbai-bot Bot commented Jun 30, 2026

Uh oh!

gunbai-bot Bot commented Jul 1, 2026

Uh oh!

gunbai-bot Bot commented Jul 1, 2026

Uh oh!

gunbai-bot Bot commented Jul 1, 2026

Uh oh!

gunbai-bot Bot commented Jul 1, 2026

Uh oh!

Uh oh!

gunbai-bot Bot commented Jul 1, 2026

Uh oh!

gunbai-bot Bot commented Jul 1, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant