support images

packages/markitup/src/markitup/_markitup.py

@@ -9,6 +9,7 @@ from ._schemas import StreamInfo, Config
 from .converters import (
     PlainTextConverter,
     HtmlConverter,
+    ImageConverter,
     PdfConverter,
     DocxConverter,
     XlsxConverter,
@@ -58,11 +59,16 @@ class MarkItUp:
                     return CsvConverter().convert(stream, stream_info), stream_info
                 case "docx":
                     return DocxConverter(config=self.config).convert(stream, stream_info), stream_info
+                case "image":
+                    return ImageConverter(config=self.config).convert(stream, stream_info), stream_info
                 case _:
                     match stream_info.category:
                         case "ppt":
                             raise UnsupportedFormatException(
                                 ".ppt files are not supported, try .pptx instead")
+                        case "doc":
+                            raise UnsupportedFormatException(
+                                ".doc files are not supported, try .docx instead")
                         case "other":
                             raise UnsupportedFormatException(
                                 f"{stream_info.magic_type} files are not supported")
@@ -84,7 +90,10 @@ class MarkItUp:
 
         # Determine file category based on magic_type
         if magic_type.startswith("image/"):
-            category = "image"
+            if magic_type in ["image/webp", "image/jpeg", "image/png", "image/jpg"]:
+                category = "image"
+            else:
+                category = "other"
         elif magic_type.startswith("audio/"):
             category = "audio"
         elif magic_type.startswith("video/"):

packages/markitup/src/markitup/_schemas.py

@@ -13,3 +13,4 @@ class Config:
     modalities: List[Literal["image", "audio"]] = field(
         default_factory=lambda: ["image", "audio"]
     )
+    image_use_webp: bool = True  # TODO: support files contains images

packages/markitup/src/markitup/converters/__init__.py

@@ -10,6 +10,7 @@ from ._xlsx_converter import XlsxConverter, XlsConverter
 from ._pptx_converter import PptxConverter
 from ._audio_converter import AudioConverter
 from ._csv_converter import CsvConverter
+from ._image_converter import ImageConverter
 from ._markdownify import _CustomMarkdownify
 
 __all__ = [
@@ -19,6 +20,7 @@ __all__ = [
     "_CustomMarkdownify",
     "WikipediaConverter",
     "YouTubeConverter",
+    "ImageConverter"
     "IpynbConverter",
     "BingSerpConverter",
     "PdfConverter",

packages/markitup/src/markitup/converters/_image_converter.py

@@ -0,0 +1,46 @@
+from typing import BinaryIO, Any
+from .._base_converter import DocumentConverter, DocumentConverterResult
+from .._schemas import StreamInfo, Config
+import base64
+
+
+class ImageConverter(DocumentConverter):
+    """
+    Converts image files to markdown with embedded base64 image.
+    """
+
+    def __init__(self, config: Config):
+        self.config = config
+
+    def convert(
+        self,
+        file_stream: BinaryIO,
+        stream_info: StreamInfo,
+        **kwargs: Any,  # Options to pass to the converter
+    ) -> DocumentConverterResult:
+        # Read the image data
+        image_bytes = file_stream.read()
+
+        # Determine image extension from magic_type
+        image_ext = "png"  # Default extension
+        match stream_info.magic_type:
+            case "image/jpeg" | "image/jpg":
+                image_ext = "jpeg"
+            case "image/png":
+                image_ext = "png"
+            case "image/webp":
+                image_ext = "webp"
+
+        if 'image' in self.config.modalities:
+            img_base64 = base64.b64encode(image_bytes).decode('utf-8')
+
+            # Create markdown with embedded image
+            markdown_content = f"![Image](data:image/{image_ext};base64,{img_base64})\n\n"
+
+            return DocumentConverterResult(
+                markdown=markdown_content,
+            )
+        else:
+            return DocumentConverterResult(
+                markdown="No Image read as the supported modalities do not include 'image'",
+            )