Prepare scrape image storage schema

2026-08-05 23:47:57 +02:00
parent aa138b62ad
commit 05338a062a
3 changed files with 108 additions and 0 deletions
@@ -0,0 +1,30 @@
+// @ts-nocheck
+import test from "node:test";
+import assert from "node:assert/strict";
+import fs from "node:fs";
+
+const schema = fs.readFileSync("sql/local-scrape-db.sql", "utf8");
+
+test("local scrape schema stores image binaries with exact and visual hashes", () => {
+  assert.match(schema, /CREATE TABLE IF NOT EXISTS `scrape_image_blobs`/);
+  assert.match(schema, /`sha256` CHAR\(64\) NOT NULL/);
+  assert.match(schema, /`perceptual_hash` CHAR\(16\) NULL/);
+  assert.match(schema, /`image_data` LONGBLOB NOT NULL/);
+  assert.match(schema, /UNIQUE KEY `uq_scrape_image_blobs_sha256` \(`sha256`\)/);
+});
+
+test("local scrape schema tracks picture candidates and their target combination logic", () => {
+  assert.match(schema, /CREATE TABLE IF NOT EXISTS `scrape_picture_candidates`/);
+  assert.match(schema, /`target_strategy` ENUM\('unknown', 'color', 'size', 'combination', 'manual'\)/);
+  assert.match(schema, /`target_color` VARCHAR\(255\) NULL/);
+  assert.match(schema, /`target_size` VARCHAR\(128\) NULL/);
+  assert.match(schema, /`target_combi` VARCHAR\(255\) NULL/);
+  assert.match(schema, /`duplicate_status` ENUM\('unique', 'exact_duplicate', 'visual_duplicate', 'review'\)/);
+});
+
+test("local scrape schema can learn picture target rules from manual corrections", () => {
+  assert.match(schema, /CREATE TABLE IF NOT EXISTS `scrape_picture_target_rules`/);
+  assert.match(schema, /`product_name_pattern` VARCHAR\(255\) NULL/);
+  assert.match(schema, /`target_strategy` ENUM\('color', 'size', 'combination', 'manual'\) NOT NULL/);
+  assert.match(schema, /`examples_count` INT UNSIGNED NOT NULL DEFAULT 0/);
+});