fix(subtitle): preserve literal backslashes in correction values

tests/tools/test_subtitle_corrections.py

@@ -0,0 +1,29 @@
+"""Correction values are literal text, not regular-expression replacements."""
+
+import pytest
+
+from tools.subtitle.subtitle_gen import SubtitleGen
+
+
+@pytest.mark.parametrize("replacement", [r"C:\Users\speaker", r"\1", r"line\ntext"])
+def test_segment_corrections_preserve_backslashes(tmp_path, replacement):
+    segments = [{"text": "Use folder.", "start": 0, "end": 1}]
+    output = tmp_path / "captions.srt"
+    result = SubtitleGen().execute({
+        "segments": segments,
+        "corrections": {"folder": replacement},
+        "output_path": str(output),
+    })
+    assert result.success, result.error
+    assert f"Use {replacement}." in output.read_text()
+    assert segments[0]["text"] == "Use folder."
+
+
+def test_segment_and_word_corrections_agree():
+    replacement = r"C:\Users\speaker"
+    segments = [
+        {"text": "FOLDER", "start": 0, "end": 1},
+        {"words": [{"word": "FOLDER", "start": 1, "end": 2}]},
+    ]
+    corrected = SubtitleGen._apply_corrections(segments, {"folder": replacement})
+    assert corrected[0]["text"] == corrected[1]["words"][0]["word"] == replacement

tools/subtitle/subtitle_gen.py

@@ -158,7 +158,7 @@ def _apply_corrections(
                     import re as _re
                     seg["text"] = _re.sub(
                         r"\b" + _re.escape(wrong) + r"\b",
-                        right,
+                        lambda match, replacement=right: replacement,
                         seg["text"],
                         flags=_re.IGNORECASE,
                     )