Second checkpoint

migrations/00001_enable_postgis.sql

@@ -0,0 +1,6 @@
+-- +goose Up
+-- Enable PostGIS so datasets can carry a spatial geometry column.
+CREATE EXTENSION IF NOT EXISTS postgis;
+
+-- +goose Down
+DROP EXTENSION IF EXISTS postgis;

migrations/00002_create_categories_table.sql

@@ -0,0 +1,14 @@
+-- +goose Up
+CREATE TABLE categories (
+    id          UUID PRIMARY KEY      DEFAULT gen_random_uuid(),
+    parent_id   UUID                  REFERENCES categories (id) ON DELETE RESTRICT,
+    name        VARCHAR(255) NOT NULL,
+    description TEXT         NOT NULL  DEFAULT '',
+    created_at  TIMESTAMPTZ  NOT NULL  DEFAULT now(),
+    updated_at  TIMESTAMPTZ  NOT NULL  DEFAULT now()
+);
+
+CREATE INDEX idx_categories_parent_id ON categories (parent_id);
+
+-- +goose Down
+DROP TABLE categories;

migrations/00003_create_datasets_table.sql

@@ -0,0 +1,39 @@
+-- +goose Up
+CREATE TYPE file_type AS ENUM ('vector_with_kato', 'vector', 'raster');
+
+CREATE TABLE datasets (
+    id           UUID PRIMARY KEY       DEFAULT gen_random_uuid(),
+    category_id  UUID         NOT NULL  REFERENCES categories (id) ON DELETE RESTRICT,
+    code         VARCHAR(255) NOT NULL,
+    name         VARCHAR(255) NOT NULL,
+    description  TEXT,
+    unit         VARCHAR(255),
+    filename     VARCHAR(255) NOT NULL,
+    storage_key  TEXT         NOT NULL,
+    cog_storage_key TEXT,                -- Cloud-Optimized GeoTIFF (raster only)
+    file_type    file_type    NOT NULL,
+    size_bytes   BIGINT       NOT NULL  DEFAULT 0,
+    content_type VARCHAR(255) NOT NULL  DEFAULT '',
+    properties   JSONB,
+    meta         JSONB,
+    automated    BOOLEAN      NOT NULL  DEFAULT false,
+    status       VARCHAR(255) NOT NULL  DEFAULT 'pending',
+    -- vector_with_kato attribute-table parsing + mapping
+    attribute_columns JSONB,            -- detected columns: [{name, samples}]
+    kato_column       VARCHAR(255),     -- user-selected KATO column
+    year_columns      JSONB,            -- mapping: [{column, date}]
+    parse_error       TEXT,             -- failure reason when status = 'failed'
+    -- Footprint/extent in EPSG:4326 (any geometry type). The SRID is constrained
+    -- so geometry and the derived bbox are always comparable.
+    geometry     geometry(Geometry, 4326),
+    created_at   TIMESTAMPTZ  NOT NULL  DEFAULT now(),
+    updated_at   TIMESTAMPTZ  NOT NULL  DEFAULT now()
+);
+
+CREATE INDEX idx_datasets_category_id ON datasets (category_id);
+CREATE INDEX idx_datasets_geometry ON datasets USING GIST (geometry);
+CREATE INDEX idx_datasets_created_at ON datasets (created_at DESC);
+
+-- +goose Down
+DROP TABLE datasets;
+DROP TYPE file_type;

migrations/00004_create_events_table.sql

@@ -0,0 +1,12 @@
+-- +goose Up
+-- Sink for the generic example RabbitMQ consumer. Safe to drop once a real
+-- async use case replaces the scaffold.
+CREATE TABLE events (
+    id         UUID PRIMARY KEY      DEFAULT gen_random_uuid(),
+    kind       VARCHAR(255) NOT NULL,
+    payload    JSONB        NOT NULL  DEFAULT '{}'::jsonb,
+    created_at TIMESTAMPTZ  NOT NULL  DEFAULT now()
+);
+
+-- +goose Down
+DROP TABLE events;

migrations/00005_create_dataset_observations_table.sql

@@ -0,0 +1,20 @@
+-- +goose Up
+-- Long-format (tidy) values unpivoted from a vector_with_kato dataset's
+-- attribute table: one row per (KATO code, date). value holds numeric cells,
+-- value_text holds non-numeric ones.
+CREATE TABLE dataset_observations (
+    id         UUID PRIMARY KEY      DEFAULT gen_random_uuid(),
+    dataset_id UUID NOT NULL         REFERENCES datasets (id) ON DELETE CASCADE,
+    kato_code  TEXT NOT NULL,
+    date       DATE NOT NULL,
+    value      DOUBLE PRECISION,
+    value_text TEXT,
+    created_at TIMESTAMPTZ NOT NULL  DEFAULT now()
+);
+
+CREATE UNIQUE INDEX uq_dataset_observations ON dataset_observations (dataset_id, kato_code, date);
+CREATE INDEX idx_dataset_observations_dataset ON dataset_observations (dataset_id);
+CREATE INDEX idx_dataset_observations_kato_date ON dataset_observations (kato_code, date);
+
+-- +goose Down
+DROP TABLE dataset_observations;

migrations/20260604134433_create_datasets_table.sql

@@ -1,11 +0,0 @@
--- +goose Up
-CREATE TABLE datasets (
-    id UUID PRIMARY KEY default gen_random_uuid(),
-    name varchar(255),
-    description text,
-    created_at timestamp default now(),
-    updated_at timestamp default now()
-);
-
--- +goose Down
-DROP TABLE datasets;

migrations/20260604141221_create_categories_table.sql

@@ -1,12 +0,0 @@
--- +goose Up
-CREATE TABLE categories
-(
-    id          UUID PRIMARY KEY default gen_random_uuid(),
-    name        varchar(255),
-    description text,
-    created_at  timestamp        default now(),
-    updated_at  timestamp        default now()
-);
-
--- +goose Down
-DROP TABLE categories;

migrations/20260604141656_create_files_table.sql

@@ -1,22 +0,0 @@
--- +goose Up
-CREATE TYPE file_type AS ENUM ('vector_with_table', 'vector', 'raster');
-CREATE TYPE file_validation_status AS ENUM ('pending', 'valid', 'failed');
-
-CREATE TABLE files (
-    id                UUID PRIMARY KEY DEFAULT gen_random_uuid(),
-    filename          VARCHAR(255) NOT NULL,
-    storage_key       TEXT NOT NULL,
-    file_type         file_type NOT NULL,
-    validation_status file_validation_status NOT NULL DEFAULT 'pending',
-    validation_error  TEXT,
-    kato_column       VARCHAR(255),
-    crs               VARCHAR(64),
-    feature_count     INTEGER,
-    uploaded_at       TIMESTAMP NOT NULL DEFAULT now(),
-    updated_at        TIMESTAMP NOT NULL DEFAULT now()
-);
-
--- +goose Down
-DROP TABLE files;
-DROP TYPE file_validation_status;
-DROP TYPE file_type;

migrations/embed.go

@@ -0,0 +1,10 @@
+// Package migrations embeds the SQL migration files so they can be applied from
+// the single application binary (via the `gis migrate` subcommand).
+package migrations
+
+import "embed"
+
+// FS holds the embedded goose migration files.
+//
+//go:embed *.sql
+var FS embed.FS