fix: avoid mutating Document.from_dict input (#11330)

Co-authored-by: Julian Risch <julian.risch@deepset.ai>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>

Author: 3 people

haystack/dataclasses/document.py

@@ -149,6 +149,7 @@ def from_dict(cls, data: dict[str, Any]) -> "Document":
 
         The `blob` field is converted to its original type.
         """
+        data = data.copy()
         if blob := data.get("blob"):
             data["blob"] = ByteStream.from_dict(blob)
         if sparse_embedding := data.get("sparse_embedding"):

releasenotes/notes/document-from-dict-copy-input-9b7d2a8e6f4c1d03.yaml

@@ -0,0 +1,4 @@
+---
+fixes:
+  - |
+    Prevent ``Document.from_dict()`` from mutating the input dictionary during deserialization.

test/dataclasses/test_document.py

@@ -3,6 +3,7 @@
 # SPDX-License-Identifier: Apache-2.0
 
 import warnings
+from copy import deepcopy
 from dataclasses import replace
 
 import pytest
@@ -213,7 +214,7 @@ def test_from_dict():
     assert Document.from_dict({}) == Document()
 
 
-def from_from_dict_with_parameters():
+def test_from_dict_with_parameters():
     blob_data = b"some bytes"
     assert Document.from_dict(
         {
@@ -234,6 +235,40 @@ def from_from_dict_with_parameters():
     )
 
 
+def test_from_dict_does_not_mutate_input():
+    blob_data = b"some bytes"
+    data = {
+        "content": "test text",
+        "blob": {"data": list(blob_data), "mime_type": "text/markdown"},
+        "score": 0.812,
+        "embedding": [0.1, 0.2, 0.3],
+        "sparse_embedding": {"indices": [0, 2, 4], "values": [0.1, 0.2, 0.3]},
+        "date": "10-10-2023",
+        "type": "article",
+    }
+    original_data = deepcopy(data)
+
+    assert Document.from_dict(data) == Document(
+        content="test text",
+        blob=ByteStream(blob_data, mime_type="text/markdown"),
+        score=0.812,
+        embedding=[0.1, 0.2, 0.3],
+        sparse_embedding=SparseEmbedding(indices=[0, 2, 4], values=[0.1, 0.2, 0.3]),
+        meta={"date": "10-10-2023", "type": "article"},
+    )
+    assert data == original_data
+
+
+def test_from_dict_does_not_mutate_input_with_explicit_meta():
+    data = {"content": "test text", "meta": {"date": "10-10-2023", "type": "article"}, "score": 0.812}
+    original_data = deepcopy(data)
+
+    assert Document.from_dict(data) == Document(
+        content="test text", meta={"date": "10-10-2023", "type": "article"}, score=0.812
+    )
+    assert data == original_data
+
+
 def test_from_dict_with_legacy_fields():
     assert Document.from_dict(
         {