From 3e2a8020efcd6141d7b101268085d87c21855865 Mon Sep 17 00:00:00 2001
From: Kyou0203 <kyou12138@gmail.com>
Date: Thu, 16 Jul 2026 23:18:24 +0800
Subject: [PATCH] fix(tools): make RAG file extension detection
 case-insensitive

DataTypes.from_content() compared path.endswith(ext) against a
lowercase-only map, so .PDF/.CSV/.DOCX were misclassified as text
(or website for URLs). Lowercase the path before matching.

Closes #6399
---
 .../src/crewai_tools/rag/data_types.py        |  6 +-
 .../tools/rag/test_data_types_from_content.py | 69 +++++++++++++++++++
 2 files changed, 74 insertions(+), 1 deletion(-)
 create mode 100644 lib/crewai-tools/tests/tools/rag/test_data_types_from_content.py

diff --git a/lib/crewai-tools/src/crewai_tools/rag/data_types.py b/lib/crewai-tools/src/crewai_tools/rag/data_types.py
index 0fcf07e121..057f234838 100644
--- a/lib/crewai-tools/src/crewai_tools/rag/data_types.py
+++ b/lib/crewai-tools/src/crewai_tools/rag/data_types.py
@@ -113,6 +113,9 @@ def from_content(content: str | Path | None = None) -> DataType:
                 pass
 
         def get_file_type(path: str) -> DataType | None:
+            # Case-insensitive extension match: Windows scanners/exports often
+            # produce uppercase extensions (.PDF, .CSV, .DOCX) that must not fall
+            # through to the plain-text loader.
             mapping = {
                 ".pdf": DataType.PDF_FILE,
                 ".csv": DataType.CSV,
@@ -123,8 +126,9 @@ def get_file_type(path: str) -> DataType | None:
                 ".xml": DataType.XML,
                 ".txt": DataType.TEXT_FILE,
             }
+            lowered = path.lower()
             for ext, dtype in mapping.items():
-                if path.endswith(ext):
+                if lowered.endswith(ext):
                     return dtype
             return None
 
diff --git a/lib/crewai-tools/tests/tools/rag/test_data_types_from_content.py b/lib/crewai-tools/tests/tools/rag/test_data_types_from_content.py
new file mode 100644
index 0000000000..ae2fe9cd59
--- /dev/null
+++ b/lib/crewai-tools/tests/tools/rag/test_data_types_from_content.py
@@ -0,0 +1,69 @@
+"""Tests for DataTypes.from_content extension auto-detection."""
+
+from pathlib import Path
+
+import pytest
+
+from crewai_tools.rag.data_types import DataType, DataTypes
+
+
+class TestFromContentCaseInsensitiveExtensions:
+    """Regression tests for #6399: uppercase/mixed-case extensions."""
+
+    @pytest.mark.parametrize(
+        ("path", "expected"),
+        [
+            ("report.pdf", DataType.PDF_FILE),
+            ("report.PDF", DataType.PDF_FILE),
+            ("report.Pdf", DataType.PDF_FILE),
+            ("data.csv", DataType.CSV),
+            ("data.CSV", DataType.CSV),
+            ("doc.docx", DataType.DOCX),
+            ("doc.DOCX", DataType.DOCX),
+            ("notes.md", DataType.MDX),
+            ("notes.MD", DataType.MDX),
+            ("notes.mdx", DataType.MDX),
+            ("notes.MDX", DataType.MDX),
+            ("payload.json", DataType.JSON),
+            ("payload.JSON", DataType.JSON),
+            ("tree.xml", DataType.XML),
+            ("tree.XML", DataType.XML),
+            ("readme.txt", DataType.TEXT_FILE),
+            ("readme.TXT", DataType.TEXT_FILE),
+        ],
+    )
+    def test_extension_detection_is_case_insensitive(
+        self, path: str, expected: DataType, tmp_path: Path
+    ) -> None:
+        """Local file paths with mixed-case extensions map to the right DataType."""
+        file_path = tmp_path / path
+        file_path.write_bytes(b"dummy")
+        assert DataTypes.from_content(str(file_path)) == expected
+
+    @pytest.mark.parametrize(
+        ("url", "expected"),
+        [
+            ("https://example.com/file.pdf", DataType.PDF_FILE),
+            ("https://example.com/file.PDF", DataType.PDF_FILE),
+            ("https://example.com/data.CSV", DataType.CSV),
+            ("https://example.com/doc.DOCX", DataType.DOCX),
+            ("https://example.com/payload.JSON", DataType.JSON),
+        ],
+    )
+    def test_url_extension_detection_is_case_insensitive(
+        self, url: str, expected: DataType
+    ) -> None:
+        """URL path extensions must also be matched case-insensitively."""
+        assert DataTypes.from_content(url) == expected
+
+    def test_url_without_known_extension_stays_website(self) -> None:
+        """URLs without a known file extension still classify as website."""
+        assert (
+            DataTypes.from_content("https://example.com/docs/page") == DataType.WEBSITE
+        )
+
+    def test_path_object_with_uppercase_extension(self, tmp_path: Path) -> None:
+        """Path objects with uppercase suffixes are handled correctly."""
+        file_path = tmp_path / "Report.PDF"
+        file_path.write_bytes(b"%PDF")
+        assert DataTypes.from_content(file_path) == DataType.PDF_FILE