perf: use only one WASM alloc for xxhash

lib/api/route/status.ts

@@ -4,6 +4,8 @@ import xxhash from 'xxhash-wasm';
 
 import cacheModule from '@/utils/cache/index';
 
+const { h64ToString } = await xxhash();
+
 const QuerySchema = z.object({
     requestPath: z.string().openapi({
         param: {
@@ -62,7 +64,6 @@ const handler: RouteHandler<typeof route> = async (ctx) => {
     }
 
     const { requestPath } = ctx.req.valid('query');
-    const { h64ToString } = await xxhash();
     const key = 'rsshub:koa-redis-cache:' + h64ToString(requestPath + ':rss');
     const cached = await cacheModule.globalCache.has(key);
 

lib/middleware/cache.ts

@@ -7,6 +7,8 @@ import type { Data } from '@/types';
 import cacheModule from '@/utils/cache/index';
 
 const bypassList = new Set(['/', '/robots.txt', '/logo.png', '/favicon.ico']);
+
+const { h64ToString } = await xxhash();
 // only give cache string, as the `!` condition tricky
 // XXH64 is used to shrink key size
 // plz, write these tips in comments!
@@ -19,13 +21,18 @@ const middleware: MiddlewareHandler = async (ctx, next) => {
     const requestPath = ctx.req.path;
     const format = `:${ctx.req.query('format') || 'rss'}`;
     const limit = ctx.req.query('limit') ? `:${ctx.req.query('limit')}` : '';
-    const { h64ToString } = await xxhash();
     const key = 'rsshub:koa-redis-cache:' + h64ToString(requestPath + format + limit);
     const controlKey = 'rsshub:path-requested:' + h64ToString(requestPath + format + limit);
 
-    const isRequesting = await cacheModule.globalCache.get(controlKey);
+    let value = await cacheModule.globalCache.get(key);
 
-    if (isRequesting === '1') {
+    // Doesn't hit the cache? Try to become the fetcher and let others know!
+    let claimed = false;
+    if (!value) {
+        claimed = await cacheModule.globalCache.claim(controlKey, config.cache.requestTimeout);
+    }
+
+    if (!value && !claimed) {
         let retryTimes = process.env.NODE_ENV === 'test' ? 1 : 10;
         let bypass = false;
         while (retryTimes > 0) {
@@ -41,10 +48,9 @@ const middleware: MiddlewareHandler = async (ctx, next) => {
         if (!bypass) {
             throw new RequestInProgressError('This path is currently fetching, please come back later!');
         }
+        value = await cacheModule.globalCache.get(key);
     }
 
-    const value = await cacheModule.globalCache.get(key);
-
     if (value) {
         ctx.status(200);
         ctx.header('RSSHub-Cache-Status', 'HIT');
@@ -53,8 +59,10 @@ const middleware: MiddlewareHandler = async (ctx, next) => {
         return;
     }
 
-    // Doesn't hit the cache? We need to let others know!
-    await cacheModule.globalCache.set(controlKey, '1', config.cache.requestTimeout);
+    if (!claimed) {
+        // waited out a stale claim without finding a cache entry, take over the fetch
+        await cacheModule.globalCache.set(controlKey, '1', config.cache.requestTimeout);
+    }
 
     // let routers control cache
     ctx.set('cacheKey', key);

lib/routes/economist/espresso.ts

@@ -11,6 +11,8 @@ import { parseDate } from '@/utils/parse-date';
 
 const link = 'https://www.economist.com/the-world-in-brief';
 
+const { h64ToString } = await xxhash();
+
 export const route: Route = {
     path: '/espresso',
     categories: ['traditional-media'],
@@ -37,7 +39,6 @@ export const route: Route = {
 };
 
 async function handler() {
-    const { h64ToString } = await xxhash();
     const nextData = await cache.tryGet(
         link,
         async () => {

lib/utils/cache/index.ts

@@ -11,10 +11,16 @@ const globalCache: {
     get: (key: string) => Promise<string | null | undefined> | string | null | undefined;
     has: (key: string) => Promise<boolean> | boolean;
     set: (key: string, value?: string | Record<string, any>, maxAge?: number) => any;
+    /**
+     * Atomically set `key` to '1' and return true, unless it is already '1' (return false).
+     * A get-then-set in the caller races: two same-tick requests would both read "not '1'".
+     */
+    claim: (key: string, maxAge: number) => Promise<boolean> | boolean;
 } = {
     get: () => null,
     has: () => false,
     set: () => null,
+    claim: () => true,
 };
 
 const noopCacheModule: CacheModule = {
@@ -54,6 +60,13 @@ if (isWorker) {
                 return false;
             };
             globalCache.set = cacheModule.set;
+            globalCache.claim = async (key, maxAge) => {
+                if (!key || !cacheModule.status.available || !redisClient) {
+                    return true;
+                }
+                const result = await redisClient.eval("if redis.call('GET', KEYS[1]) == '1' then return 0 end redis.call('SET', KEYS[1], '1', 'EX', ARGV[1]) return 1", 1, key, maxAge);
+                return result === 1;
+            };
             break;
         }
         case 'http':
@@ -75,6 +88,17 @@ if (isWorker) {
                     return cacheModule.set(key, value, maxAge);
                 }
             };
+            globalCache.claim = async (key, maxAge) => {
+                if (!key || !cacheModule.status.available) {
+                    return true;
+                }
+                // best effort: the HTTP cache protocol has no atomic operation
+                if ((await cacheModule.get(key, false)) === '1') {
+                    return false;
+                }
+                await cacheModule.set(key, '1', maxAge);
+                return true;
+            };
             break;
         case 'memory': {
             cacheModule = memory;
@@ -102,6 +126,17 @@ if (isWorker) {
                     return memoryCache.set(key, value, { ttl: maxAge * 1000 });
                 }
             };
+            // fully synchronous, so nothing can interleave between the read and the write
+            globalCache.claim = (key, maxAge) => {
+                if (!key || !cacheModule.status.available || !memoryCache) {
+                    return true;
+                }
+                if (memoryCache.get(key, { updateAgeOnGet: false }) === '1') {
+                    return false;
+                }
+                memoryCache.set(key, '1', { ttl: maxAge * 1000 });
+                return true;
+            };
             break;
         }
         default: