feat(images): Add local image storage with on-demand resizing

- Store product images locally with hierarchy: /images/products/<state>/<store>/<brand>/<product>/
- Add /img/* proxy endpoint for on-demand resizing via Sharp
- Implement per-product image checking to skip existing downloads
- Fix pathToUrl() to correctly generate /images/... URLs
- Add frontend getImageUrl() helper with preset sizes (thumb, medium, large)
- Update all product pages to use optimized image URLs
- Add stealth session support for Dutchie GraphQL crawls
- Include test scripts for crawl and image verification

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
This commit is contained in:
Kelly
2025-12-09 11:04:42 -07:00
parent aa776226b0
commit 91efd1d03d
28 changed files with 2027 additions and 205 deletions

View File

@@ -1,26 +1,29 @@
/**
* Local Image Storage Utility
*
* Downloads and stores product images to local filesystem.
* Replaces MinIO-based storage with simple local file storage.
* Downloads and stores product images to local filesystem with proper hierarchy.
*
* Directory structure:
* /images/products/<dispensary_id>/<product_id>.webp
* /images/products/<dispensary_id>/<product_id>-thumb.webp
* /images/products/<dispensary_id>/<product_id>-medium.webp
* /images/brands/<brand_slug>.webp
* /images/products/<state>/<store_slug>/<brand_slug>/<product_id>/image.webp
* /images/products/<state>/<store_slug>/<brand_slug>/<product_id>/image-medium.webp
* /images/products/<state>/<store_slug>/<brand_slug>/<product_id>/image-thumb.webp
* /images/brands/<brand_slug>/logo.webp
*
* This structure allows:
* - Easy migration to MinIO/S3 (bucket per state)
* - Browsing by state/store/brand
* - Multiple images per product (future: gallery)
*/
import axios from 'axios';
import sharp from 'sharp';
// @ts-ignore - sharp module typing quirk
const sharp = require('sharp');
import * as fs from 'fs/promises';
import * as path from 'path';
import { createHash } from 'crypto';
// Base path for image storage - configurable via env
// Uses project-relative paths by default, NOT /app or other privileged paths
function getImagesBasePath(): string {
// Priority: IMAGES_PATH > STORAGE_BASE_PATH/images > ./storage/images
if (process.env.IMAGES_PATH) {
return process.env.IMAGES_PATH;
}
@@ -35,16 +38,28 @@ const IMAGES_BASE_PATH = getImagesBasePath();
const IMAGES_PUBLIC_URL = process.env.IMAGES_PUBLIC_URL || '/images';
export interface LocalImageSizes {
full: string; // URL path: /images/products/123/456.webp
medium: string; // URL path: /images/products/123/456-medium.webp
thumb: string; // URL path: /images/products/123/456-thumb.webp
original: string; // URL path to original image
// Legacy compatibility - all point to original until we add image proxy
full: string;
medium: string;
thumb: string;
}
export interface DownloadResult {
success: boolean;
urls?: LocalImageSizes;
localPaths?: LocalImageSizes;
error?: string;
bytesDownloaded?: number;
skipped?: boolean; // True if image already exists
}
export interface ProductImageContext {
stateCode: string; // e.g., "AZ", "CA"
storeSlug: string; // e.g., "deeply-rooted"
brandSlug: string; // e.g., "high-west-farms"
productId: string; // External product ID
dispensaryId?: number; // For backwards compat
}
/**
@@ -58,6 +73,17 @@ async function ensureDir(dirPath: string): Promise<void> {
}
}
/**
* Sanitize a string for use in file paths
*/
function slugify(str: string): string {
return str
.toLowerCase()
.replace(/[^a-z0-9]+/g, '-')
.replace(/^-+|-+$/g, '')
.substring(0, 50) || 'unknown';
}
/**
* Generate a short hash from a URL for deduplication
*/
@@ -81,53 +107,30 @@ async function downloadImage(imageUrl: string): Promise<Buffer> {
}
/**
* Process and save image in multiple sizes
* Returns the file paths relative to IMAGES_BASE_PATH
* Process and save original image (convert to webp for consistency)
*
* We store only the original - resizing will be done on-demand via
* an image proxy service (imgproxy, thumbor, or similar) in the future.
*/
async function processAndSaveImage(
buffer: Buffer,
outputDir: string,
baseFilename: string
): Promise<{ full: string; medium: string; thumb: string; totalBytes: number }> {
): Promise<{ original: string; totalBytes: number }> {
await ensureDir(outputDir);
const fullPath = path.join(outputDir, `${baseFilename}.webp`);
const mediumPath = path.join(outputDir, `${baseFilename}-medium.webp`);
const thumbPath = path.join(outputDir, `${baseFilename}-thumb.webp`);
const originalPath = path.join(outputDir, `${baseFilename}.webp`);
// Process images in parallel
const [fullBuffer, mediumBuffer, thumbBuffer] = await Promise.all([
// Full: max 1200x1200, high quality
sharp(buffer)
.resize(1200, 1200, { fit: 'inside', withoutEnlargement: true })
.webp({ quality: 85 })
.toBuffer(),
// Medium: 600x600
sharp(buffer)
.resize(600, 600, { fit: 'inside', withoutEnlargement: true })
.webp({ quality: 80 })
.toBuffer(),
// Thumb: 200x200
sharp(buffer)
.resize(200, 200, { fit: 'inside', withoutEnlargement: true })
.webp({ quality: 75 })
.toBuffer(),
]);
// Convert to webp, preserve original dimensions, high quality
const originalBuffer = await sharp(buffer)
.webp({ quality: 90 })
.toBuffer();
// Save all sizes
await Promise.all([
fs.writeFile(fullPath, fullBuffer),
fs.writeFile(mediumPath, mediumBuffer),
fs.writeFile(thumbPath, thumbBuffer),
]);
const totalBytes = fullBuffer.length + mediumBuffer.length + thumbBuffer.length;
await fs.writeFile(originalPath, originalBuffer);
return {
full: fullPath,
medium: mediumPath,
thumb: thumbPath,
totalBytes,
original: originalPath,
totalBytes: originalBuffer.length,
};
}
@@ -135,47 +138,107 @@ async function processAndSaveImage(
* Convert a file path to a public URL
*/
function pathToUrl(filePath: string): string {
// Find /products/ or /brands/ in the path and extract from there
const productsMatch = filePath.match(/(\/products\/.*)/);
const brandsMatch = filePath.match(/(\/brands\/.*)/);
if (productsMatch) {
return `${IMAGES_PUBLIC_URL}${productsMatch[1]}`;
}
if (brandsMatch) {
return `${IMAGES_PUBLIC_URL}${brandsMatch[1]}`;
}
// Fallback: try to replace base path (works if paths match exactly)
const relativePath = filePath.replace(IMAGES_BASE_PATH, '');
return `${IMAGES_PUBLIC_URL}${relativePath}`;
}
/**
* Download and store a product image locally
* Build the directory path for a product image
* Structure: /images/products/<state>/<store>/<brand>/<product>/
*/
function buildProductImagePath(ctx: ProductImageContext): string {
const state = slugify(ctx.stateCode || 'unknown');
const store = slugify(ctx.storeSlug || 'unknown');
const brand = slugify(ctx.brandSlug || 'unknown');
const product = slugify(ctx.productId || 'unknown');
return path.join(IMAGES_BASE_PATH, 'products', state, store, brand, product);
}
/**
* Download and store a product image with proper hierarchy
*
* @param imageUrl - The third-party image URL to download
* @param dispensaryId - The dispensary ID (for directory organization)
* @param productId - The product ID or external ID (for filename)
* @param ctx - Product context (state, store, brand, product)
* @param options - Download options
* @returns Download result with local URLs
*/
export async function downloadProductImage(
imageUrl: string,
dispensaryId: number,
productId: string | number
ctx: ProductImageContext,
options: { skipIfExists?: boolean } = {}
): Promise<DownloadResult> {
const { skipIfExists = true } = options;
try {
if (!imageUrl) {
return { success: false, error: 'No image URL provided' };
}
const outputDir = buildProductImagePath(ctx);
const urlHash = hashUrl(imageUrl);
const baseFilename = `image-${urlHash}`;
// Check if image already exists
if (skipIfExists) {
const existingPath = path.join(outputDir, `${baseFilename}.webp`);
try {
await fs.access(existingPath);
// Image exists, return existing URL
const url = pathToUrl(existingPath);
return {
success: true,
skipped: true,
urls: {
original: url,
full: url,
medium: url,
thumb: url,
},
localPaths: {
original: existingPath,
full: existingPath,
medium: existingPath,
thumb: existingPath,
},
};
} catch {
// Image doesn't exist, continue to download
}
}
// Download the image
const buffer = await downloadImage(imageUrl);
// Organize by dispensary ID
const outputDir = path.join(IMAGES_BASE_PATH, 'products', String(dispensaryId));
// Use product ID + URL hash for uniqueness
const urlHash = hashUrl(imageUrl);
const baseFilename = `${productId}-${urlHash}`;
// Process and save
// Process and save (original only)
const result = await processAndSaveImage(buffer, outputDir, baseFilename);
const url = pathToUrl(result.original);
return {
success: true,
urls: {
full: pathToUrl(result.full),
medium: pathToUrl(result.medium),
thumb: pathToUrl(result.thumb),
original: url,
full: url,
medium: url,
thumb: url,
},
localPaths: {
original: result.original,
full: result.original,
medium: result.original,
thumb: result.original,
},
bytesDownloaded: result.totalBytes,
};
@@ -188,33 +251,70 @@ export async function downloadProductImage(
}
/**
* Download and store a brand logo locally
* Legacy function - backwards compatible with old signature
* Maps to new hierarchy using dispensary_id as store identifier
*/
export async function downloadProductImageLegacy(
imageUrl: string,
dispensaryId: number,
productId: string | number
): Promise<DownloadResult> {
return downloadProductImage(imageUrl, {
stateCode: 'unknown',
storeSlug: `store-${dispensaryId}`,
brandSlug: 'unknown',
productId: String(productId),
dispensaryId,
});
}
/**
* Download and store a brand logo
*
* @param logoUrl - The brand logo URL
* @param brandId - The brand ID or slug
* @param brandSlug - The brand slug/ID
* @returns Download result with local URL
*/
export async function downloadBrandLogo(
logoUrl: string,
brandId: string
brandSlug: string,
options: { skipIfExists?: boolean } = {}
): Promise<DownloadResult> {
const { skipIfExists = true } = options;
try {
if (!logoUrl) {
return { success: false, error: 'No logo URL provided' };
}
const safeBrandSlug = slugify(brandSlug);
const outputDir = path.join(IMAGES_BASE_PATH, 'brands', safeBrandSlug);
const urlHash = hashUrl(logoUrl);
const baseFilename = `logo-${urlHash}`;
// Check if logo already exists
if (skipIfExists) {
const existingPath = path.join(outputDir, `${baseFilename}.webp`);
try {
await fs.access(existingPath);
return {
success: true,
skipped: true,
urls: {
full: pathToUrl(existingPath),
medium: pathToUrl(existingPath),
thumb: pathToUrl(existingPath),
},
};
} catch {
// Logo doesn't exist, continue
}
}
// Download the image
const buffer = await downloadImage(logoUrl);
// Brand logos go in /images/brands/
const outputDir = path.join(IMAGES_BASE_PATH, 'brands');
// Sanitize brand ID for filename
const safeBrandId = brandId.replace(/[^a-zA-Z0-9-_]/g, '_');
const urlHash = hashUrl(logoUrl);
const baseFilename = `${safeBrandId}-${urlHash}`;
// Process and save (single size for logos)
// Brand logos in their own directory
await ensureDir(outputDir);
const logoPath = path.join(outputDir, `${baseFilename}.webp`);
@@ -243,20 +343,16 @@ export async function downloadBrandLogo(
}
/**
* Check if a local image already exists
* Check if a product image already exists
*/
export async function imageExists(
dispensaryId: number,
productId: string | number,
export async function productImageExists(
ctx: ProductImageContext,
imageUrl: string
): Promise<boolean> {
const outputDir = buildProductImagePath(ctx);
const urlHash = hashUrl(imageUrl);
const imagePath = path.join(
IMAGES_BASE_PATH,
'products',
String(dispensaryId),
`${productId}-${urlHash}.webp`
);
const imagePath = path.join(outputDir, `image-${urlHash}.webp`);
try {
await fs.access(imagePath);
return true;
@@ -266,24 +362,27 @@ export async function imageExists(
}
/**
* Delete a product's local images
* Get the local image URL for a product (if exists)
*/
export async function deleteProductImages(
dispensaryId: number,
productId: string | number,
imageUrl?: string
): Promise<void> {
const productDir = path.join(IMAGES_BASE_PATH, 'products', String(dispensaryId));
const prefix = imageUrl
? `${productId}-${hashUrl(imageUrl)}`
: String(productId);
export async function getProductImageUrl(
ctx: ProductImageContext,
imageUrl: string
): Promise<LocalImageSizes | null> {
const outputDir = buildProductImagePath(ctx);
const urlHash = hashUrl(imageUrl);
const imagePath = path.join(outputDir, `image-${urlHash}.webp`);
try {
const files = await fs.readdir(productDir);
const toDelete = files.filter(f => f.startsWith(prefix));
await Promise.all(toDelete.map(f => fs.unlink(path.join(productDir, f))));
await fs.access(imagePath);
const url = pathToUrl(imagePath);
return {
original: url,
full: url,
medium: url,
thumb: url,
};
} catch {
// Directory might not exist, that's fine
return null;
}
}
@@ -296,19 +395,17 @@ export function isImageStorageReady(): boolean {
/**
* Initialize the image storage directories
* Does NOT throw on failure - logs warning and continues
*/
export async function initializeImageStorage(): Promise<void> {
try {
await ensureDir(path.join(IMAGES_BASE_PATH, 'products'));
await ensureDir(path.join(IMAGES_BASE_PATH, 'brands'));
console.log(`Image storage initialized at ${IMAGES_BASE_PATH}`);
console.log(`[ImageStorage] Initialized at ${IMAGES_BASE_PATH}`);
imageStorageReady = true;
} catch (error: any) {
console.warn(`⚠️ WARNING: Could not initialize image storage at ${IMAGES_BASE_PATH}: ${error.message}`);
console.warn(' Image upload/processing is disabled. Server will continue without image features.');
console.warn(`[ImageStorage] WARNING: Could not initialize at ${IMAGES_BASE_PATH}: ${error.message}`);
console.warn(' Image features disabled. Server will continue without image downloads.');
imageStorageReady = false;
// Do NOT throw - server should still start
}
}
@@ -316,34 +413,43 @@ export async function initializeImageStorage(): Promise<void> {
* Get storage stats
*/
export async function getStorageStats(): Promise<{
productsDir: string;
brandsDir: string;
basePath: string;
productCount: number;
brandCount: number;
totalSizeBytes: number;
}> {
const productsDir = path.join(IMAGES_BASE_PATH, 'products');
const brandsDir = path.join(IMAGES_BASE_PATH, 'brands');
let productCount = 0;
let brandCount = 0;
let totalSizeBytes = 0;
try {
const productDirs = await fs.readdir(productsDir);
for (const dir of productDirs) {
const files = await fs.readdir(path.join(productsDir, dir));
productCount += files.filter(f => f.endsWith('.webp') && !f.includes('-')).length;
}
} catch { /* ignore */ }
async function countDir(dirPath: string): Promise<{ count: number; size: number }> {
let count = 0;
let size = 0;
try {
const entries = await fs.readdir(dirPath, { withFileTypes: true });
for (const entry of entries) {
const fullPath = path.join(dirPath, entry.name);
if (entry.isDirectory()) {
const sub = await countDir(fullPath);
count += sub.count;
size += sub.size;
} else if (entry.name.endsWith('.webp') && !entry.name.includes('-')) {
count++;
const stat = await fs.stat(fullPath);
size += stat.size;
}
}
} catch { /* ignore */ }
return { count, size };
}
try {
const brandFiles = await fs.readdir(brandsDir);
brandCount = brandFiles.filter(f => f.endsWith('.webp')).length;
} catch { /* ignore */ }
const products = await countDir(path.join(IMAGES_BASE_PATH, 'products'));
const brands = await countDir(path.join(IMAGES_BASE_PATH, 'brands'));
return {
productsDir,
brandsDir,
productCount,
brandCount,
basePath: IMAGES_BASE_PATH,
productCount: products.count,
brandCount: brands.count,
totalSizeBytes: products.size + brands.size,
};
}