Merge pull request #3817 from github/esbena/fetch-logs-through-dca

PoC: compare-performance of external logs

Author: asgerf

extensions/ql-vscode/src/common/jsonl-reader.ts

@@ -1,26 +1,66 @@
-import { readFile } from "fs-extra";
+import { statSync } from "fs";
+import { createReadStream } from "fs-extra";
+import { createInterface } from "readline";
 
 /**
  * Read a file consisting of multiple JSON objects. Each object is separated from the previous one
  * by a double newline sequence. This is basically a more human-readable form of JSONL.
  *
- * The current implementation reads the entire text of the document into memory, but in the future
- * it will stream the document to improve the performance with large documents.
- *
  * @param path The path to the file.
  * @param handler Callback to be invoked for each top-level JSON object in order.
  */
 export async function readJsonlFile<T>(
   path: string,
   handler: (value: T) => Promise<void>,
+  logger?: { log: (message: string) => void },
 ): Promise<void> {
-  const logSummary = await readFile(path, "utf-8");
+  function parseJsonFromCurrentLines() {
+    try {
+      return JSON.parse(currentLineSequence.join("\n")) as T;
+    } catch (e) {
+      void logger?.log(
+        // eslint-disable-next-line @typescript-eslint/no-explicit-any
+        `Error: Failed to parse at line ${lineCount} of ${path} as JSON: ${(e as any)?.message ?? "UNKNOWN REASON"}. Problematic line below:\n${JSON.stringify(currentLineSequence, null, 2)}`,
+      );
+      throw e;
+    }
+  }
 
-  // Remove newline delimiters because summary is in .jsonl format.
-  const jsonSummaryObjects: string[] = logSummary.split(/\r?\n\r?\n/g);
+  function logProgress() {
+    void logger?.log(
+      `Processed ${lineCount} lines with ${parseCounts} parses...`,
+    );
+  }
 
-  for (const obj of jsonSummaryObjects) {
-    const jsonObj = JSON.parse(obj) as T;
-    await handler(jsonObj);
+  void logger?.log(
+    `Parsing ${path} (${statSync(path).size / 1024 / 1024} MB)...`,
+  );
+  const fileStream = createReadStream(path, "utf8");
+  const rl = createInterface({
+    input: fileStream,
+    crlfDelay: Infinity,
+  });
+
+  let lineCount = 0;
+  let parseCounts = 0;
+  let currentLineSequence: string[] = [];
+  for await (const line of rl) {
+    if (line === "") {
+      // as mentioned above: a double newline sequence indicates the end of the current JSON object, so we parse it and pass it to the handler
+      await handler(parseJsonFromCurrentLines());
+      parseCounts++;
+      currentLineSequence = [];
+    } else {
+      currentLineSequence.push(line);
+    }
+    lineCount++;
+    if (lineCount % 1000000 === 0) {
+      logProgress();
+    }
+  }
+  // in case the file is not newline-terminated, we need to handle the last JSON object
+  if (currentLineSequence.length > 0) {
+    await handler(parseJsonFromCurrentLines());
   }
+  logProgress();
 }