From bba84e551721c1bda779968b3c09247653af2dbd Mon Sep 17 00:00:00 2001
From: Sebastian Husch Lee <sjrl@users.noreply.github.com>
Date: Tue, 28 Jan 2025 01:29:55 -0800
Subject: [PATCH] fix: Fix JSONConverter to properly skip files that are not
 utf-8 encoded (#8775)

* Small fix

* Add reno

* Trying out license header fix here
---
 haystack/components/converters/json.py        |  1 +
 ...n-converter-non-utf8-3a755df732a8cbd5.yaml |  4 ++++
 test/components/converters/test_json.py       | 19 +++++++++++++++++++
 3 files changed, 24 insertions(+)
 create mode 100644 releasenotes/notes/fix-json-converter-non-utf8-3a755df732a8cbd5.yaml

diff --git a/haystack/components/converters/json.py b/haystack/components/converters/json.py
index 3a8c6f52f0..6d3781e4e9 100644
--- a/haystack/components/converters/json.py
+++ b/haystack/components/converters/json.py
@@ -194,6 +194,7 @@ def _get_content_and_meta(self, source: ByteStream) -> List[Tuple[str, Dict[str,
                 source=source.meta["file_path"],
                 error=exc,
             )
+            return []
 
         meta_fields = self._meta_fields or set()
 
diff --git a/releasenotes/notes/fix-json-converter-non-utf8-3a755df732a8cbd5.yaml b/releasenotes/notes/fix-json-converter-non-utf8-3a755df732a8cbd5.yaml
new file mode 100644
index 0000000000..2c475d201e
--- /dev/null
+++ b/releasenotes/notes/fix-json-converter-non-utf8-3a755df732a8cbd5.yaml
@@ -0,0 +1,4 @@
+---
+fixes:
+  - |
+    Fixed JSONConverter to properly skip converting JSON files that are not utf-8 encoded.
diff --git a/test/components/converters/test_json.py b/test/components/converters/test_json.py
index f9dcf2fa0c..5419fa812a 100644
--- a/test/components/converters/test_json.py
+++ b/test/components/converters/test_json.py
@@ -236,6 +236,25 @@ def test_run_with_bad_filter(tmpdir, caplog):
     assert result == {"documents": []}
 
 
+def test_run_with_bad_encoding(tmpdir, caplog):
+    test_file = Path(tmpdir / "test_file.json")
+    test_file.write_text(json.dumps(test_data[0]), "utf-16")
+
+    sources = [test_file]
+    converter = JSONConverter(".laureates")
+
+    caplog.clear()
+    with caplog.at_level(logging.WARNING):
+        result = converter.run(sources=sources)
+
+    records = caplog.records
+    assert len(records) == 1
+    assert records[0].msg.startswith(
+        f"Failed to extract text from {test_file}. Skipping it. Error: 'utf-8' codec can't decode byte"
+    )
+    assert result == {"documents": []}
+
+
 def test_run_with_single_meta(tmpdir):
     first_test_file = Path(tmpdir / "first_test_file.json")
     second_test_file = Path(tmpdir / "second_test_file.json")