Task: Sync the ores.trading tree to its models

Table of Contents

This page documents a task in the Clean ores.trading to the component clean standard story. It captures the goal, current status, acceptance, and any notes or results.

1. Goal

Make the checked-in ores.trading tree equal to what its models generate, and make the component build and pass its suites at that tree. This is item group G plus the protocol items P01, P02, P03 and P05, and the wiring item W01. It changes no model and no data shape: it lands work the generator already produces and deletes what the generator supersedes.

The measured baseline is 1636 files that would change and 115 that would be created. Of the changes, 1434 are the generated-file banner alone, 78 are the eventing rename, 123 are substantive template changes and 1 is the root CMakeLists.txt. A component usually cannot be cleaned by landing that much at once, but here it is safe, because 88 per cent of the changes are attributable to one cause and the generator already reproduces the tree's semantics.

2. Status

Field Value
State BACKLOG
Parent story Clean ores.trading to the component clean standard
Now Not yet started.
Waiting on PR 1 merged, so the decisions are on the main branch.
Next Re-measure the baseline at the new head, then regenerate.
Last touched 2026-09-26

3. Acceptance

  • check_component_drift.py --component trading-cpp --dry-run reports nothing to change and nothing to create, and a second run reports the same.
  • The whole tree builds, and ores.trading.api.tests, ores.trading.core.tests and ores.trading.service.tests pass.
  • The 21 *_publishes_nats_changed_event failures are gone, because the 114 trading notify triggers emit the canonical payload.
  • regenerate_cmake_component_files.py --check passes for ores.trading.api, ores.trading.core and ores.trading.service.
  • Decision D1 is applied: the nine rates families are served by the generated per-family handlers and registrars, and the hand consolidation is deleted. No subject is registered twice.
  • Every generated file carries the marker, and every hand-written file the generation supersedes is deleted.
  • The consumers in ores.shell, ores.ore and ores.reporting are migrated in the same PR, and nothing is carried for backwards compatibility.
  • W01 is closed by reading the composition points, not by trusting the drift gate.

4. Plan

4.1. 1. Re-measure and confirm the cause of every block

Run the dry run and confirm the classification at the new head before writing anything. 

projects/ores.codegen/venv/bin/python projects/ores.codegen/scripts/check_component_drift.py --component trading-cpp --dry-run

The classification to expect: banner only, eventing rename, the trigger template, and the creates. If a block appears under a cause the PR 1 baseline does not have, stop and explain it before regenerating.

4.2. 2. Apply D1 before regenerating, not after

The generator emits per-family protocols, handlers and registrars for nine rates families: fra, vanilla swap, cap floor, swaption, balance guaranteed swap, callable swap, knock out swap, inflation swap and rpa. The tree consolidated those nine by hand into instrument_protocol.hpp plus rates_instrument_handler.hpp, registered from registrar_rates.cpp. Regenerating while the consolidation remains would subscribe the same subjects twice.

The work: let the nine per-family files be created, then delete the per-family sections of instrument_protocol.hpp, delete rates_instrument_handler.hpp and registrar_rates.cpp, and wire the generated registrars at the same composition point. The subject names do not change, so no consumer changes for this step.

4.3. 3. Regenerate

Run the drift check without --dry-run. It writes the tree. Then read the diff in the order of its causes, not in file order: the banner first, then the eventing rename, then the 114 SQL scripts, then the creates. 

projects/ores.codegen/venv/bin/python projects/ores.codegen/scripts/check_component_drift.py --component trading-cpp

If the regeneration reveals a generator defect, fix it at its source with a regression test in the codegen suite, then regenerate. Do not edit generated output.

4.4. 4. Migrate the consumers

The canonical list verb renames get_trades_request to list_trades_request and h->list to h->list_trades, and the eventing headers rename. Consumers outside the component: 38 files in ores.shell, 17 in ores.ore, 1 in ores.reporting. Migrate all of them in this PR. Nothing is kept for compatibility.

4.5. 5. Fix the CMake source lists

projects/ores.codegen/venv/bin/python projects/ores.codegen/scripts/regenerate_cmake_component_files.py --component ores.trading.api
projects/ores.codegen/venv/bin/python projects/ores.codegen/scripts/regenerate_cmake_component_files.py --component ores.trading.core
projects/ores.codegen/venv/bin/python projects/ores.codegen/scripts/regenerate_cmake_component_files.py --component ores.trading.service

4.6. 6. Build and test

Build the whole tree, not only the component, because the canonical protocol change reaches ores.shell, ores.ore and ores.reporting. Then run the four suites. 

cmake --build build/output/linux-clang-debug-make -j 3
ctest --test-dir build/output/linux-clang-debug-make -R "ores.trading|ores.shell.trading|ores.ore" --output-on-failure

4.7. 7. Prove idempotence

Run the dry run twice more and require both runs to report nothing. A second run that still changes something is a generator defect.

4.8. 8. Record the items

Update the story's checklist rows for G01, G02, G03, G04, G06, P01, P02, P03, P05, W01, V01 and V02. Record which gates were run before the component joins the registry and which are still vacuous.

5. Notes

  • The trap this PR exists for is the double registration described in step 2. It is decision D1 on the story.
  • The eventing work here closes ores.trading's share of Finish the eventing payload migration, or read the legacy shape. Read that capture before touching the trigger template.
  • The in-place drift check diffs the whole repository. Any untracked file makes it fail for reasons that are not this component's. The eleven untracked shell library directories from the scheduler pass are the known case. Keep the tree still and never run a writing gate while a build reads the tree.
  • The PR is large and mostly mechanical. The review evidence is the cause classification, the build, and the four suites, not a line by line read.
  • Two hand-written generator trees exist in the component, api/.../generator/ with 25 files and api/.../generators/ with 38. Do not collapse them here; record them for PR 6.
  • 24 hand-written domain types have no model. Do not model them here; PR 3 owns that.

6. Test Scenarios

Scenario State Notes
     

7. PRs

PR Title
  Sync ores.trading to its models

8. Review

Comment summary File Decision Notes
       

9. Result

Emacs 29.3 (Org mode 9.6.15)