Task: Split the ER diagram by package
Table of Contents
This page documents a task in the Split the ER diagram by group story. It captures the goal, current status, acceptance, and any notes or results.
1. Goal
The generator writes one diagram per package plus an index, so a reader opens the part of the schema they want and every file renders.
2. Status
| Field | Value |
|---|---|
| State | DONE |
| Parent story | Split the ER diagram by group |
| Now | Nothing. |
| Waiting on | Nothing. |
| Next | Nothing. |
| Last touched | 2026-10-09 |
3. Acceptance
- One .puml per package, named for it, plus an index that names the groups and the tables each holds.
- Every generated diagram renders on this machine, with no raster overflow.
- The policy for an edge that leaves its package is stated and implemented.
4. Plan
(Implementation strategy. Written when work starts; key decisions
are distilled into the parent story's * Decisions at close, but the
plan itself stays — it is the historical record of what we did.)
5. Measurements
Every number this task claims about a tree, with the command that produced it
and the commit it was measured at. A number without both is not a measurement,
because a clean-up record describes a tree and trunk moves under a branch: a
consumer count, a gate's component count and a suite's size all change with no
code changing, and the reader of a bare number cannot tell which tree it came
from. Where the raw output matters, put it in an evidence/ file beside the
task and name that file in the Command column.
| What | Value | Command | Commit |
|---|---|---|---|
| Diagrams generated | 22: one per package and the index | plantuml_er_generate.py --index-template |
334f9acbfc |
| Diagrams that render | 22 of 22, no exception | plantuml -tpng at PLANTUML_LIMIT_SIZE=131072 over every file |
334f9acbfc |
| Largest diagram | dq_staging, 20436 x 11090 (226.6M px) |
same | 334f9acbfc |
| Next largest | refdata 4432 x 27026, trading 2743 x 26303 |
same | 334f9acbfc |
| Index | 6709 x 2483 (16.7M px) | same | 334f9acbfc |
| The old single diagram | failed at about 2.68e9 px | same | - |
| codegen-tests | pass, 51.9 s | compass check --only codegen-tests |
da901c61e8 |
| er-diagram, tangle-templates, org-links | pass | compass check --only er-diagram --only tangle-templates --only org-links |
da901c61e8 |
6. Notes
7. Test Scenarios
Manual QA scenarios (scaffolded via compass add test_scenario)
that verify this task. Link
new ones here as they're created; the scenario doc itself links back
via its "Verifies task" field.
| Scenario | State | Notes |
|---|---|---|
8. PRs
| PR | Title |
|---|---|
| #2937 | [codegen,sql] Split the ER diagram into one file per group |
9. Review
| Comment summary | File | Decision | Notes |
|---|---|---|---|
10. Result
The diagram is drawn one file per package, from the grouping the model
already carried, and ores_schema.puml is the index over them. Every one
of the 22 renders; the largest, dq_staging, is 226.6M pixels, against
the single 463-table diagram that overflowed a 32-bit pixel count at
about 2.68e9.
The policy for a key that leaves its package. A group draws its own
tables and the keys with both ends inside it. A key that leaves is not
drawn, because its other end is in another file; the group's note names
those tables and the group each belongs to, so the reader knows where to
go next. The note is emitted from both sides, so alpha names c_tbl
(beta) and beta names b_tbl (alpha).
The index. It names every group, the file that group is drawn in and the tables it holds, and draws no columns and no keys, so it stays the starting point without being another picture to read.
The generator writes the group files beside --output and the index to
--output itself, so the shell script, the CMake target and the check
needed only the new --index-template. --check now walks every file
and reports each stale or missing one rather than the first.