SQL DDL → ER: a CREATE TABLE front-end with PK/FK/UK badges by clintecker · Pull Request #12 · 2389-research/MermaidKit

clintecker · 2026-07-17T01:34:11Z

A third front-end (after DOT and Dippin): point a SQL schema dump at MermaidKit and get an ER diagram. SQLDDLParser.parse(_:) turns CREATE TABLE … into the ERDiagram IR, rendered through the same layered layout as a Mermaid erDiagram.

What it parses

Tables & typed columns — CREATE TABLE t (col TYPE, …), including size clauses (VARCHAR(100), DECIMAL(10,2)).
Keys — PRIMARY KEY / UNIQUE / FOREIGN KEY, both inline and table-level (incl. CONSTRAINT name …), shown as compact PK / FK / UK badges.
Relationships — each REFERENCES / FOREIGN KEY maps to a one-to-many crow's-foot relation (parent ||--o{ child).
Dialect quoting — "x", `x`, [x]; -- and /* */ comments; unknown clauses (CHECK, INDEX, DEFAULT, engine options) ignored, never fatal.
Degrades gracefully — malformed/huge/hostile input returns nil under the shared maxTextSize/maxEdges caps; forward-progress guards throughout.

Example

CREATE TABLE customer (id INT PRIMARY KEY, email VARCHAR(255) UNIQUE, name VARCHAR(100));
CREATE TABLE orders (id INT PRIMARY KEY, customer_id INT REFERENCES customer(id), total DECIMAL(10,2));

→ two entity boxes with PK/UK/FK badges and a customer ||--o{ orders relationship (rendered image in the thread).

IR change

ERDiagram.Attribute gains keys: [Key] (PK/FK/UK), drawn as a right-aligned badge. The badge/width code is a strict no-op when keys is empty, and the Mermaid erDiagram parser still leaves it empty — so existing ER output is byte-identical (no gallery/perf regen). keys has a defaulted initializer, so no existing Attribute(type:name:) call site changes.

311 tests, 0 failures (12 new). Platform-free parser in MermaidLayout, beside DOTParser/DippinParser.

🤖 Generated with Claude Code

^{Need help on this PR? Tag /codesmith with what you need. Autofix is disabled.}

Adds SQLDDLParser.parse(_:), a Swift front-end that turns a SQL schema dump into the ERDiagram IR, so it renders through the same layered layout as a Mermaid erDiagram. Parses the structural core — CREATE TABLE, typed columns, PRIMARY / FOREIGN / UNIQUE keys (inline and table-level), and REFERENCES — mapping each foreign key to a one-to-many crow's-foot relationship. Dialect quoting ("x", `x`, [x]), comments, and unknown clauses (CHECK/INDEX/DEFAULT/engine options) are handled or ignored, never fatal; malformed/huge/hostile input returns nil under the shared maxTextSize/maxEdges caps. IR + render: ERDiagram.Attribute gains a `keys: [Key]` field (PK/FK/UK), rendered as a compact right-aligned badge in the entity box. The badge/width code is a strict no-op when keys is empty, so Mermaid erDiagram output is unchanged (its parser still leaves keys empty) — no gallery/perf regen triggered. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01JoLDcosyaHg3tAKhU5SQaw

coderabbitai · 2026-07-17T01:34:31Z

Warning

Review limit reached

You’ve reached a temporary PR review limit under our Fair Usage Limits Policy.

Your recent review volume is higher than typical usage, so adaptive limits are currently applied.

Next review available in: 22 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: 8f42940b-30f0-464c-aae3-be317be2597b

📥 Commits

Reviewing files that changed from the base of the PR and between c450e90 and 5285112.

📒 Files selected for processing (6)

Sources/MermaidLayout/DiagramLayoutBoxDiagrams.swift
Sources/MermaidLayout/MermaidModels.swift
Sources/MermaidLayout/SQLDDLParser.swift
Sources/MermaidRender/DiagramRenderer+ER.swift
Tests/MermaidLayoutTests/SQLDDLParserTests.swift
Tests/MermaidRenderTests/SQLRenderTests.swift

✨ Finishing Touches

🧪 Generate unit tests (beta)

Create PR with unit tests
Commit unit tests in branch feature/sql-ddl-frontend

_{Comment @coderabbitai help to get the list of available commands.}

chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 528511290a

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

chatgpt-codex-connector · 2026-07-17T01:37:33Z

+        let table = t[i].text; i += 1
+        guard isPunct("(", at: i) else { skipStatement(); return }   // e.g. CREATE TABLE x AS SELECT …


Parse qualified table names before requiring the body

For schema-qualified DDL such as CREATE TABLE public.customer (...), the lexer drops the dot and produces two word tokens; this line records public as the table name, and the next line then sees customer instead of ( and skips the whole statement. That means common pg_dump/MySQL dump inputs with qualified names parse as nil or miss most tables instead of producing an ER diagram.

Useful? React with 👍 / 👎.

chatgpt-codex-connector · 2026-07-17T01:37:33Z

+        case "unique":
+            p += 1
+            for col in parenCols(item, &p) { addKey(.unique, to: col, in: &attrs, indexOf: indexOf) }


Handle named UNIQUE table constraints

When a dump uses MySQL-style table constraints like UNIQUE KEY email_idx (email) or UNIQUE INDEX ..., p points at KEY/INDEX rather than (, so parenCols returns an empty list and the column never gets a UK badge. This misses a common form of table-level unique constraint even though the parser advertises table-level UNIQUE support.

Useful? React with 👍 / 👎.

chatgpt-codex-connector · 2026-07-17T01:37:33Z

+        if p < item.count, item[p].word {
+            type = item[p].text; p += 1


Preserve multi-word SQL column types

This only captures the first word of a type and only attaches a size clause when it immediately follows that word, so valid dump output like character varying(255), double precision, or timestamp without time zone is rendered as just character, double, or timestamp. For schemas using these common PostgreSQL types, the generated ER attributes show incorrect type information.

Useful? React with 👍 / 👎.

…n bad fixture - SQLDDLParserTests imported CoreGraphics unconditionally (landed in #12, macOS- only gating missed it) → Linux CI red since. Guard it like DOTParserTests (#if canImport(CoreGraphics) … #else import Foundation). Restores Linux green. - DeterminismSignatureTests: XCTUnwrap parse + rasterize instead of skip/nil, so a fixture that silently fails can't make the determinism diff pass vacuously (CodeRabbit). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01JoLDcosyaHg3tAKhU5SQaw

chatgpt-codex-connector Bot reviewed Jul 17, 2026

View reviewed changes

clintecker merged commit 30ef18e into main Jul 17, 2026
2 of 3 checks passed

clintecker deleted the feature/sql-ddl-frontend branch July 17, 2026 01:54

clintecker mentioned this pull request Jul 17, 2026

Cross-process determinism guard (closes #1) #18

Merged

Navigation Menu

Uh oh!

SQL DDL → ER: a CREATE TABLE front-end with PK/FK/UK badges - #12

SQL DDL → ER: a CREATE TABLE front-end with PK/FK/UK badges#12
clintecker merged 1 commit into
mainfrom
feature/sql-ddl-frontend

clintecker commented Jul 17, 2026 •

edited by blacksmith-sh Bot

Loading

Uh oh!

coderabbitai Bot commented Jul 17, 2026

Review limit reached

Uh oh!

chatgpt-codex-connector Bot left a comment

Uh oh!

chatgpt-codex-connector Bot Jul 17, 2026

Uh oh!

chatgpt-codex-connector Bot Jul 17, 2026

Uh oh!

chatgpt-codex-connector Bot Jul 17, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

		let table = t[i].text; i += 1
		guard isPunct("(", at: i) else { skipStatement(); return } // e.g. CREATE TABLE x AS SELECT …

		if p < item.count, item[p].word {
		type = item[p].text; p += 1

Uh oh!

Conversation

clintecker commented Jul 17, 2026 • edited by blacksmith-sh Bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What it parses

Example

IR change

Uh oh!

coderabbitai Bot commented Jul 17, 2026

Review limit reached

Uh oh!

chatgpt-codex-connector Bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!

chatgpt-codex-connector Bot Jul 17, 2026

Choose a reason for hiding this comment

Uh oh!

chatgpt-codex-connector Bot Jul 17, 2026

Choose a reason for hiding this comment

Uh oh!

chatgpt-codex-connector Bot Jul 17, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

clintecker commented Jul 17, 2026 •

edited by blacksmith-sh Bot

Loading