feat: Update Dockerfile and server configuration for improved health checks and error handling

server/source.js

@@ -1,22 +1,19 @@
 import { XMLParser } from "fast-xml-parser";
-import { execFile } from "node:child_process";
-import { promisify } from "node:util";
 import { readFile, writeFile, mkdir } from "node:fs/promises";
-import { fileURLToPath } from "node:url";
 import { createHash } from "node:crypto";
 import { parseWeeksDocx, parseBellsDocx } from "./calendar.js";
 import { apiError, sourceError } from "./errors.js";
+import { DATA } from "./paths.js";
 import {
   academicStartYear,
   assignSemesters,
   todayIso,
   weekIndexOf,
 } from "../shared/weeks.js";
-const run = promisify(execFile);
-const ORIGIN = "https://nextcloud.psu.by";
+// The university's Nextcloud; tests point PSU_SOURCE_ORIGIN at a local stand-in.
+const ORIGIN = process.env.PSU_SOURCE_ORIGIN || "https://nextcloud.psu.by";
 const SHARE = "/public.php/dav/files/no5zKbMrns6j5jQ/";
 const YEAR_FOLDER = /^(\d{4})-(\d{4})$/;
-const DATA = fileURLToPath(new URL("../.data/", import.meta.url));
 const parser = new XMLParser({ removeNSPrefix: true, ignoreAttributes: false });
 const arr = (value) => (value ? (Array.isArray(value) ? value : [value]) : []);
 const DISCOVERY_TTL = 15 * 60 * 1000,
@@ -39,6 +36,8 @@ const status = { catalogError: null, calendarError: null };
 const failedAt = { catalog: 0, calendar: 0 };
 const backingOff = (what) => Date.now() - failedAt[what] < MIN_REFRESH;
 
+const TIMEOUT = Number(process.env.PSU_SOURCE_TIMEOUT_MS) || 25000;
+const MAX_SIZE = 10 * 1024 * 1024;
 export async function request(path, method = "GET") {
   // Only the university share is reachable; listings and catalog entries supply the paths.
   if (
@@ -48,31 +47,35 @@ export async function request(path, method = "GET") {
     !["GET", "PROPFIND"].includes(method)
   )
     throw new Error("Недопустимый путь документа.");
-  const url = ORIGIN + path;
-  // curl also works with the system trust store on macOS. Certificate checks stay enabled.
-  const args = [
-    "--fail",
-    "--silent",
-    "--show-error",
-    "--max-time",
-    "25",
-    "--max-filesize",
-    "10485760",
-    "--max-redirs",
-    "0",
-    "--request",
-    method,
-  ];
-  if (method === "PROPFIND") args.push("--header", "Depth: 1");
-  args.push(url);
-  const { stdout } = await run("curl", args, {
-    encoding: "buffer",
-    maxBuffer: 11 * 1024 * 1024,
-    timeout: 28000,
-  }).catch((error) => {
+  // Node's fetch keeps connections to the university open between requests, so a catalog
+  // crawl of a hundred listings pays for one TLS handshake, not a hundred. Certificates are
+  // checked, redirects are refused, and the answer is cut off past 10 MB or 25 seconds.
+  try {
+    const response = await fetch(ORIGIN + path, {
+      method,
+      headers: method === "PROPFIND" ? { Depth: "1" } : {},
+      redirect: "manual",
+      signal: AbortSignal.timeout(TIMEOUT),
+    });
+    if (!response.ok) {
+      await response.body?.cancel();
+      throw new Error(`HTTP ${response.status}`);
+    }
+    if (Number(response.headers.get("content-length")) > MAX_SIZE) {
+      await response.body?.cancel();
+      throw new Error("Документ больше 10 МБ.");
+    }
+    const chunks = [];
+    let size = 0;
+    for await (const chunk of response.body || []) {
+      size += chunk.length;
+      if (size > MAX_SIZE) throw new Error("Документ больше 10 МБ.");
+      chunks.push(chunk);
+    }
+    return Buffer.concat(chunks);
+  } catch (error) {
     throw sourceError(error);
-  });
-  return stdout;
+  }
 }
 export function decodeListing(xml, root = SHARE) {
   const entries = arr(parser.parse(xml).multistatus?.response);