From 900516f2251be9d3d98fcd095c60fbdbb900f500 Mon Sep 17 00:00:00 2001
From: tdurieux <durieuxthomas@hotmail.com>
Date: Sun, 6 Sep 2026 09:16:23 +0200
Subject: [PATCH] fix: omit source content length for rewritten file responses

---
 src/core/AnonymizedFile.ts          |  4 ---
 test/production-regressions.test.js | 44 +++++++++++++++++++++++++++++
 2 files changed, 44 insertions(+), 4 deletions(-)
 create mode 100644 test/production-regressions.test.js

diff --git a/src/core/AnonymizedFile.ts b/src/core/AnonymizedFile.ts
index 45d7b46..877bf48 100644
--- a/src/core/AnonymizedFile.ts
+++ b/src/core/AnonymizedFile.ts
@@ -477,10 +477,6 @@ export default class AnonymizedFile {
           if (!mime && data.isText) {
             res.contentType("text/plain");
           }
-          if (!data.wasAnonimized && this._file?.size) {
-            // the text files may be anonymized and therefore the size may be different
-            res.header("Content-Length", this._file?.size.toString());
-          }
         });
         const content = await this.content();
         function handleStreamError(error: Error) {
diff --git a/test/production-regressions.test.js b/test/production-regressions.test.js
new file mode 100644
index 0000000..05f81dd
--- /dev/null
+++ b/test/production-regressions.test.js
@@ -0,0 +1,44 @@
+const { setTimeout, setImmediate } = require("timers");
+const { expect } = require("chai");
+require("ts-node/register/transpile-only");
+const { Readable } = require("stream");
+const { once } = require("events");
+const config = require("../src/config").default;
+const db = require("../src/server/database");
+const gh = require("../src/core/GitHubUtils");
+const RepoModel = require("../src/core/model/anonymizedRepositories/anonymizedRepositories.model").default;
+const Repository = require("../src/core/Repository").default;
+const GitHubStream = require("../src/core/source/GitHubStream").default;
+const { ContentAnonimizer, AnonymizeTransformer, anonymizePath } = require("../src/core/anonymize-utils");
+
+async function collect(stream) {
+  const chunks = [];
+  for await (const chunk of stream) chunks.push(Buffer.from(chunk));
+  return Buffer.concat(chunks).toString();
+}
+
+describe("production regressions", function () {
+  const restores = [];
+  function stub(object, key, value) {
+    const old = object[key]; restores.push(() => { object[key] = old; }); object[key] = value;
+  }
+  afterEach(() => { while (restores.length) restores.pop()(); });
+  it("omits an upstream length when later text is rewritten", async function () {
+    const File = require("../src/core/AnonymizedFile").default;
+    stub(config, "STREAMER_ENTRYPOINT", "");
+    const repo = new Repository(new RepoModel({ repoId: "repo", options: { terms: ["Alice"], image: true, link: true }, source: {} }));
+    const file = new File({ repository: repo, anonymizedPath: "file.txt" });
+    const input = "z".repeat(9000) + " Alice";
+    file._file = { name: "file.txt", path: "", size: input.length };
+    file.content = async () => Readable.from([Buffer.from(input.slice(0, 5000)), Buffer.from(input.slice(5000))]);
+    const res = new (require("stream").PassThrough)();
+    const headers = {};
+    res.header = (key, value) => { headers[key] = value; return res; };
+    res.contentType = () => res;
+    const output = collect(res);
+    await file.send(res);
+    expect(await output).to.equal("z".repeat(9000) + " XXXX-1");
+    expect(headers).not.to.have.property("Content-Length");
+  });
+
+});