feat: consolidate backend and docker-compose setup

backend/pfm-web-app/test_exec.js

@@ -0,0 +1,87 @@
+const http = require("http");
+
+function dockerRequest(path, method, body = null) {
+  return new Promise((resolve, reject) => {
+    const options = {
+      socketPath: "/var/run/docker.sock",
+      path: path,
+      method: method,
+      headers: {
+        "Content-Type": "application/json",
+      },
+    };
+
+    const req = http.request(options, (res) => {
+      let data = "";
+      res.on("data", (chunk) => (data += chunk));
+      res.on("end", () => {
+        if (res.statusCode >= 200 && res.statusCode < 300) {
+          try {
+            resolve(data ? JSON.parse(data) : null);
+          } catch (e) {
+            resolve(data);
+          }
+        } else {
+          reject(new Error(`Docker API Error ${res.statusCode}: ${data}`));
+        }
+      });
+    });
+
+    req.on("error", (err) => reject(err));
+    if (body) {
+      req.write(JSON.stringify(body));
+    }
+    req.end();
+  });
+}
+
+async function runExec(containerName, cmd) {
+  try {
+    // 1. Create exec instance
+    const execConfig = {
+      AttachStdout: true,
+      AttachStderr: true,
+      Cmd: cmd,
+    };
+    const createRes = await dockerRequest(`/containers/${containerName}/exec`, "POST", execConfig);
+    const execId = createRes.Id;
+    
+    // 2. Start exec instance
+    // Note: Start API returns raw stream, so we use http.request directly to read it
+    return new Promise((resolve, reject) => {
+      const options = {
+        socketPath: "/var/run/docker.sock",
+        path: `/exec/${execId}/start`,
+        method: "POST",
+        headers: {
+          "Content-Type": "application/json",
+        },
+      };
+
+      const req = http.request(options, (res) => {
+        let output = "";
+        res.on("data", (chunk) => (output += chunk));
+        res.on("end", () => {
+          // Docker multiplexes stdout/stderr in the stream.
+          // First 8 bytes of each frame contain header info: [stream_type, 0, 0, 0, size1, size2, size3, size4]
+          // For simple outputs, we can clean up non-printable characters or parse directly
+          resolve(output);
+        });
+      });
+
+      req.on("error", (err) => reject(err));
+      req.write(JSON.stringify({ Detach: false, Tty: false }));
+      req.end();
+    });
+  } catch (err) {
+    throw err;
+  }
+}
+
+async function main() {
+  const result = await runExec("paddleocr-vllm-server", ["nvidia-smi", "--query-gpu=index,name,utilization.gpu,utilization.memory,memory.total,memory.used,memory.free,uuid", "--format=csv,noheader,nounits"]);
+  console.log("Exec output:");
+  console.log(result);
+}
+
+main().catch(console.error);