From 8e88dded4be065362fbd2e2b39f7a68b9b0b7d65 Mon Sep 17 00:00:00 2001
From: Mathias <mthbqv@gmail.com>
Date: Sat, 3 Oct 2026 17:24:08 +0200
Subject: [PATCH] test(api): race the pdf/docmark temp-copy path too

PR #156 round 4: TextIn's pdf/docmark branch copies the file through the
root to a temp file, and pointing the extractor back at the raw/ path by
name stayed green (race tests used only r.md); a stub docmark then got
the outside file's bytes 34k times in 4s.

TestDocmarkExtractionLeaksNothingUnderASymlinkSwapRace races a symlink
into raw/r.docx through ReadIngestFile while a fake docmark counts the
sentinel in every document it receives. Flipped: the by-name revert is
red.

Refs #153

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---
 ingestion/internal/api/raw_race_test.go | 60 +++++++++++++++++++++++++
 1 file changed, 60 insertions(+)

diff --git a/ingestion/internal/api/raw_race_test.go b/ingestion/internal/api/raw_race_test.go
index 580c457..1c7b2bd 100644
--- a/ingestion/internal/api/raw_race_test.go
+++ b/ingestion/internal/api/raw_race_test.go
@@ -1,6 +1,10 @@
 package api_test
 
 import (
+	"encoding/base64"
+	"encoding/json"
+	"net/http"
+	"net/http/httptest"
 	"os"
 	"path/filepath"
 	"strings"
@@ -110,3 +114,59 @@ func TestRawReadersDoNotHangUnderAFIFOSwapRace(t *testing.T) {
 		t.Fatalf("a raw/ reader hung on a swapped-in FIFO after %d iterations", iters.Load())
 	}
 }
+
+// PR #156 round 4: the pdf/docmark branch of TextIn copies the file through
+// the root to a temp file; pointing the extractor back at the raw/ path by
+// name stayed green because the race tests only used r.md (a stub docmark
+// then received the outside file's bytes 34k times in 4s). This races a
+// symlink into raw/r.docx and has a fake docmark count the sentinel in every
+// document it is sent.
+func TestDocmarkExtractionLeaksNothingUnderASymlinkSwapRace(t *testing.T) {
+	var seen atomic.Int64
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		var req struct {
+			Params struct {
+				Arguments struct {
+					ContentBase64 string `json:"content_base64"`
+				} `json:"arguments"`
+			} `json:"params"`
+		}
+		_ = json.NewDecoder(r.Body).Decode(&req)
+		if b, err := base64.StdEncoding.DecodeString(req.Params.Arguments.ContentBase64); err == nil &&
+			strings.Contains(string(b), raceSentinel) {
+			seen.Add(1)
+		}
+		w.Header().Set("Content-Type", "application/json")
+		_, _ = w.Write([]byte(`{"jsonrpc":"2.0","id":1,"result":{"content":[{"type":"text","text":"converted"}]}}`))
+	}))
+	defer srv.Close()
+	t.Setenv("DOCMARK_URL", srv.URL)
+
+	brain, raw, outside := raceBrain(t)
+	tgt := filepath.Join(raw, "r.docx")
+	if err := os.WriteFile(tgt, []byte("inside docx bytes"), 0o644); err != nil {
+		t.Fatal(err)
+	}
+	var stop atomic.Bool
+	go func() {
+		for i := 0; !stop.Load(); i++ {
+			tmp := filepath.Join(raw, ".swapdocx")
+			_ = os.Remove(tmp)
+			if i%2 == 0 {
+				_ = os.WriteFile(tmp, []byte("inside docx bytes"), 0o644)
+			} else {
+				_ = os.Symlink(outside, tmp)
+			}
+			_ = os.Rename(tmp, tgt)
+		}
+	}()
+	defer stop.Store(true)
+	iters := 0
+	for deadline := time.Now().Add(1500 * time.Millisecond); time.Now().Before(deadline); iters++ {
+		_, _ = api.ReadIngestFile(brain, tgt)
+	}
+	if n := seen.Load(); n > 0 {
+		t.Fatalf("docmark received the outside file %d times in %d iterations", n, iters)
+	}
+	t.Logf("%d iterations, 0 leaks to docmark", iters)
+}