Fixed format errors, tests, and read charset when available.

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

@@ -2,6 +2,7 @@ import sys
 import csv
 import io
 from typing import BinaryIO, Any
+from charset_normalizer import from_bytes
 from ._html_converter import HtmlConverter
 from .._base_converter import DocumentConverter, DocumentConverterResult
 from .._stream_info import StreamInfo
@@ -12,10 +13,12 @@ ACCEPTED_MIME_TYPE_PREFIXES = [
 ]
 ACCEPTED_FILE_EXTENSIONS = [".csv"]
 
+
 class CsvConverter(DocumentConverter):
     """
     Converts CSV files to Markdown tables.
     """
+
     def __init__(self):
         super().__init__()
 
@@ -41,33 +44,36 @@ class CsvConverter(DocumentConverter):
         **kwargs: Any,  # Options to pass to the converter
     ) -> DocumentConverterResult:
         # Read the file content
-        content = file_stream.read().decode('utf-8', errors='replace')
+        if stream_info.charset:
-        
+            content = file_stream.read().decode(stream_info.charset)
+        else:
+            content = str(from_bytes(file_stream.read()).best())
+
         # Parse CSV content
         reader = csv.reader(io.StringIO(content))
         rows = list(reader)
-        
+
         if not rows:
             return DocumentConverterResult(markdown="")
-        
+
         # Create markdown table
         markdown_table = []
-        
+
         # Add header row
         markdown_table.append("| " + " | ".join(rows[0]) + " |")
-        
+
         # Add separator row
         markdown_table.append("| " + " | ".join(["---"] * len(rows[0])) + " |")
-        
+
         # Add data rows
         for row in rows[1:]:
             # Make sure row has the same number of columns as header
             while len(row) < len(rows[0]):
                 row.append("")
             # Truncate if row has more columns than header
-            row = row[:len(rows[0])]
+            row = row[: len(rows[0])]
             markdown_table.append("| " + " | ".join(row) + " |")
-        
+
         result = "\n".join(markdown_table)
-        
+
         return DocumentConverterResult(markdown=result)

packages/markitdown/src/markitdown/converters/_doc_intel_converter.py

@@ -24,14 +24,28 @@ try:
 except ImportError:
     # Preserve the error and stack trace for later
     _dependency_exc_info = sys.exc_info()
+
     # Define these types for type hinting when the package is not available
-    class AzureKeyCredential: pass
+    class AzureKeyCredential:
-    class TokenCredential: pass
+        pass
-    class DocumentIntelligenceClient: pass
+
-    class AnalyzeDocumentRequest: pass
+    class TokenCredential:
-    class AnalyzeResult: pass
+        pass
-    class DocumentAnalysisFeature: pass
+
-    class DefaultAzureCredential: pass
+    class DocumentIntelligenceClient:
+        pass
+
+    class AnalyzeDocumentRequest:
+        pass
+
+    class AnalyzeResult:
+        pass
+
+    class DocumentAnalysisFeature:
+        pass
+
+    class DefaultAzureCredential:
+        pass
 
 
 # TODO: currently, there is a bug in the document intelligence SDK with importing the "ContentFormat" enum.