feat(csv): add CSV conversion support

- Introduce a new CsvConverter for converting CSV files into Markdown tables.
- Register CsvConverter in MarkItDown’s converter initialization in _markitdown.py.
- Update converters/__init__.py to export CsvConverter.
- Add _csv_converter.py to implement CSV conversion using pandas and openpyxl.
- Revise test vectors to expect the Markdown table format produced by CsvConverter.
- Adjust test_cli_vectors header for proper encoding handling.

This change enables CSV file support in MarkItDown while keeping existing features intact.

packages/markitdown/src/markitdown/_markitdown.py

@@ -32,6 +32,7 @@ from .converters import (
     BingSerpConverter,
     PdfConverter,
     DocxConverter,
+    CsvConverter,
     XlsxConverter,
     XlsConverter,
     PptxConverter,
@@ -185,6 +186,7 @@ class MarkItDown:
             self.register_converter(YouTubeConverter())
             self.register_converter(BingSerpConverter())
             self.register_converter(DocxConverter())
+            self.register_converter(CsvConverter())
             self.register_converter(XlsxConverter())
             self.register_converter(XlsConverter())
             self.register_converter(PptxConverter())

packages/markitdown/src/markitdown/converters/__init__.py

@@ -12,6 +12,7 @@ from ._bing_serp_converter import BingSerpConverter
 from ._pdf_converter import PdfConverter
 from ._docx_converter import DocxConverter
 from ._xlsx_converter import XlsxConverter, XlsConverter
+from ._csv_converter import CsvConverter
 from ._pptx_converter import PptxConverter
 from ._image_converter import ImageConverter
 from ._audio_converter import AudioConverter
@@ -33,6 +34,7 @@ __all__ = [
     "BingSerpConverter",
     "PdfConverter",
     "DocxConverter",
+    "CsvConverter",
     "XlsxConverter",
     "XlsConverter",
     "PptxConverter",

packages/markitdown/src/markitdown/converters/_csv_converter.py

@@ -0,0 +1,75 @@
+import sys
+from typing import BinaryIO, Any
+from ._html_converter import HtmlConverter
+from .._base_converter import DocumentConverter, DocumentConverterResult
+from .._stream_info import StreamInfo
+from .._exceptions import MissingDependencyException, MISSING_DEPENDENCY_MESSAGE
+
+# Try loading optional (but in this case, required) dependencies
+# Save reporting of any exceptions for later
+_dependency_exc_info = None
+try:
+    import pandas as pd
+except ImportError:
+    _dependency_exc_info = sys.exc_info()
+
+
+ACCEPTED_CSV_MIME_TYPE_PREFIXES = [
+    "text/csv",
+    "application/csv"
+]
+ACCEPTED_CSV_FILE_EXTENSIONS = [".csv"]
+
+class CsvConverter(DocumentConverter):
+    """
+    Converts CSV files to Markdown.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self._html_converter = HtmlConverter()
+
+    def accepts(
+        self,
+        file_stream: BinaryIO,
+        stream_info: StreamInfo,
+        **kwargs: Any,  # Options to pass to the converter
+    ) -> bool:
+        mimetype = (stream_info.mimetype or "").lower()
+        extension = (stream_info.extension or "").lower()
+
+        if extension in ACCEPTED_CSV_FILE_EXTENSIONS:
+            return True
+
+        for prefix in ACCEPTED_CSV_MIME_TYPE_PREFIXES:
+            if mimetype.startswith(prefix):
+                return True
+
+        return False
+
+    def convert(
+        self,
+        file_stream: BinaryIO,
+        stream_info: StreamInfo,
+        **kwargs: Any,  # Options to pass to the converter
+    ):
+        # Check: the dependencies
+        if _dependency_exc_info is not None:
+            raise MissingDependencyException(
+                MISSING_DEPENDENCY_MESSAGE.format(
+                    converter=type(self).__name__,
+                    extension=".csv",
+                    feature="csv",
+                )
+            ) from _dependency_exc_info[
+                1
+            ].with_traceback(  # type: ignore[union-attr]
+                _dependency_exc_info[2]
+            )
+
+        encoding = "utf-8" if stream_info.charset is None else stream_info.charset
+        content = pd.read_csv(file_stream, encoding=encoding)
+        md_content = self._html_converter.convert_string(
+            content.to_html(index=False), **kwargs
+        ).markdown.strip()
+        return DocumentConverterResult(markdown=md_content)

packages/markitdown/tests/_test_vectors.py

@@ -144,10 +144,11 @@ GENERAL_TEST_VECTORS = [
         charset="cp932",
         url=None,
         must_include=[
-            "名前,年齢,住所",
+            "| 名前 | 年齢 | 住所 |",
-            "佐藤太郎,30,東京",
+            "| --- | --- | --- |",
-            "三木英子,25,大阪",
+            "| 佐藤太郎 | 30 | 東京 |",
-            "髙橋淳,35,名古屋",
+            "| 三木英子 | 25 | 大阪 |",
+            "| 髙橋淳 | 35 | 名古屋 |",
         ],
         must_not_include=[],
     ),