mirror of
https://github.com/windmill-labs/windmill.git
synced 2026-09-06 08:01:35 +00:00
Ruben FiszelandClaude Opus 4.8
af4ca8fdf1
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
59 lines
2.9 KiB
SQL
59 lines
2.9 KiB
SQL
-- Execution DAG edges declared via `// on <asset | schedule | ...>`
|
|
-- annotations.
|
|
-- For `trigger_kind='asset'`: trigger_ref is `<kind>://<path>` (kind from
|
|
-- parse_asset_syntax, so downstream lookups match the `asset` table).
|
|
-- The other kinds mirror the keywords the annotation parser recognises in
|
|
-- `// on <kind> <ref>` lines (every non-integration trigger kind; their
|
|
-- trigger_ref is the trigger row path, or empty for marker-only forms).
|
|
--
|
|
-- Per-edge columns that are in fact script-level properties (every row for
|
|
-- a given runnable carries the same value, set once at deploy) but live on
|
|
-- the edge so the dispatcher reads everything from a single query:
|
|
-- join_all `// trigger all` AND-join barrier (else OR, the default).
|
|
-- debounce_s `// on debounce=` (else script-level `// debounce`); only
|
|
-- asset-cascade edges carry one. NULL = no debounce.
|
|
-- retry_* `// retry <count> [<delay>]` cascade retry. NULL = none.
|
|
--
|
|
-- The idempotency guards (IF NOT EXISTS / duplicate_object) are load-bearing:
|
|
-- this migration squashes several pre-release ones, so databases migrated
|
|
-- from the unsquashed history already contain the final objects and
|
|
-- re-applying must be a no-op.
|
|
DO $$ BEGIN
|
|
CREATE TYPE SCRIPT_TRIGGER_KIND AS ENUM (
|
|
'asset', 'schedule', 'webhook', 'email', 'kafka', 'mqtt', 'nats',
|
|
'postgres', 'sqs', 'gcp');
|
|
EXCEPTION WHEN duplicate_object THEN NULL;
|
|
END $$;
|
|
|
|
CREATE TABLE IF NOT EXISTS script_trigger (
|
|
id BIGSERIAL PRIMARY KEY,
|
|
workspace_id VARCHAR(50) NOT NULL REFERENCES workspace(id) ON DELETE CASCADE ON UPDATE CASCADE,
|
|
runnable_kind ASSET_USAGE_KIND NOT NULL,
|
|
runnable_path VARCHAR(255) NOT NULL,
|
|
trigger_kind SCRIPT_TRIGGER_KIND NOT NULL,
|
|
trigger_ref TEXT NOT NULL,
|
|
join_all BOOLEAN NOT NULL DEFAULT FALSE,
|
|
debounce_s INTEGER,
|
|
retry_count SMALLINT,
|
|
retry_delay_s INTEGER
|
|
);
|
|
|
|
-- Per-runnable lookup (wipe-on-deploy, list-triggers-for-script).
|
|
CREATE INDEX IF NOT EXISTS idx_script_trigger_runnable
|
|
ON script_trigger (workspace_id, runnable_kind, runnable_path);
|
|
|
|
-- Reverse lookup: "which scripts are triggered by asset X?" (the asset → script
|
|
-- edges in the graph). trigger_ref is unbounded text so can't share the
|
|
-- asset_kind btree, but this covers the common prefix-scan use case.
|
|
CREATE INDEX IF NOT EXISTS idx_script_trigger_ref
|
|
ON script_trigger (workspace_id, trigger_kind, trigger_ref);
|
|
|
|
-- Fast lookups for:
|
|
-- 1. "does folder F have a pipeline?" (exists check on prefix)
|
|
-- 2. "list all folders with a pipeline" (distinct folder from path)
|
|
-- The partial predicate keeps the index tiny on workspaces with few
|
|
-- pipeline scripts, and text_pattern_ops lets 'f/foo/%' LIKE scans use it.
|
|
CREATE INDEX IF NOT EXISTS idx_script_pipeline_path
|
|
ON script (workspace_id, path text_pattern_ops)
|
|
WHERE auto_kind = 'pipeline' AND archived = false AND deleted = false;
|