Merge branch 'main' into main

.dockerignore

@@ -0,0 +1 @@
+*

.gitattributes

@@ -0,0 +1 @@
+tests/test_files/** linguist-vendored

Dockerfile

@@ -0,0 +1,16 @@
+FROM python:3.13-alpine
+
+USER root
+
+# Runtime dependency
+RUN apk add --no-cache ffmpeg
+
+RUN pip install markitdown
+
+# Default USERID and GROUPID
+ARG USERID=10000
+ARG GROUPID=10000
+
+USER $USERID:$GROUPID
+
+ENTRYPOINT [ "markitdown" ]

README.md

@@ -1,5 +1,7 @@
 # MarkItDown
 
+[![PyPI](https://img.shields.io/pypi/v/markitdown.svg)](https://pypi.org/project/markitdown/)
+
 The MarkItDown library is a utility tool for converting various files to Markdown (e.g., for indexing, text analysis, etc.)
 
 It presently supports:
@@ -12,6 +14,7 @@ It presently supports:
 - Audio (EXIF metadata, and speech transcription)
 - HTML (special handling of Wikipedia, etc.)
 - Various other text-based formats (csv, json, xml, etc.)
+- ZIP (Iterates over contents and converts each file)
 
 # Installation
 
@@ -27,7 +30,6 @@ or from the source
 pip install -e .
 ```
 
-
 # Usage
 The API is simple:
 
@@ -39,18 +41,44 @@ result = markitdown.convert("test.xlsx")
 print(result.text_content)
 ```
 
-You can also configure markitdown to use Large Language Models to describe images. To do so you must provide mlm_client and mlm_model parameters to MarkItDown object, according to your specific client.
+To use this as a command-line utility, install it and then run it like this:
+
+```bash
+markitdown path-to-file.pdf
+```
+
+This will output Markdown to standard output. You can save it like this:
+
+```bash
+markitdown path-to-file.pdf > document.md
+```
+
+You can pipe content to standard input by omitting the argument:
+
+```bash
+cat path-to-file.pdf | markitdown
+```
+
+You can also configure markitdown to use Large Language Models to describe images. To do so you must provide `llm_client` and `llm_model` parameters to MarkItDown object, according to your specific client.
+
 
 ```python
 from markitdown import MarkItDown
 from openai import OpenAI
 
 client = OpenAI()
-md = MarkItDown(mlm_client=client, mlm_model="gpt-4o")
+md = MarkItDown(llm_client=client, llm_model="gpt-4o")
 result = md.convert("example.jpg")
 print(result.text_content)
 ```
 
+You can also use the project as Docker Image:
+
+```sh
+docker build -t markitdown:latest .
+docker run --rm -i markitdown:latest < ~/your-file.pdf > output.md
+```
+
 ## Contributing
 
 This project welcomes contributions and suggestions.  Most contributions require you to agree to a

pyproject.toml

@@ -38,6 +38,7 @@ dependencies = [
   "youtube-transcript-api",
   "SpeechRecognition",
   "pathvalidate",
+  "charset-normalizer",
 ]
 
 [project.urls]

src/markitdown/_markitdown.py

@@ -12,8 +12,10 @@ import subprocess
 import sys
 import tempfile
 import traceback
+import zipfile
 from typing import Any, Dict, List, Optional, Union
 from urllib.parse import parse_qs, quote, unquote, urlparse, urlunparse
+from warnings import catch_warnings
 
 import mammoth
 import markdownify
@@ -26,10 +28,17 @@ import pptx
 import puremagic
 import requests
 from bs4 import BeautifulSoup
+from charset_normalizer import from_path
 
 # Optional Transcription support
 try:
-    import pydub
+    # Using warnings' catch_warnings to catch
+    # pydub's warning of ffmpeg or avconv missing
+    with catch_warnings(record=True) as w:
+        import pydub
+
+        if w:
+            raise ModuleNotFoundError
     import speech_recognition as sr
 
     IS_AUDIO_TRANSCRIPTION_CAPABLE = True
@@ -161,9 +170,7 @@ class PlainTextConverter(DocumentConverter):
         elif "text/" not in content_type.lower():
             return None
 
-        text_content = ""
-        with open(local_path, "rt", encoding="utf-8") as fh:
-            text_content = fh.read()
+        text_content = str(from_path(local_path).best())
         return DocumentConverterResult(
             title=None,
             text_content=text_content,
@@ -344,8 +351,11 @@ class YouTubeConverter(DocumentConverter):
                 assert isinstance(params["v"][0], str)
                 video_id = str(params["v"][0])
                 try:
+                    youtube_transcript_languages = kwargs.get(
+                        "youtube_transcript_languages", ("en",)
+                    )
                     # Must be a single transcript.
-                    transcript = YouTubeTranscriptApi.get_transcript(video_id)  # type: ignore
+                    transcript = YouTubeTranscriptApi.get_transcript(video_id, languages=youtube_transcript_languages)  # type: ignore
                     transcript_text = " ".join([part["text"] for part in transcript])  # type: ignore
                     # Alternative formatting:
                     # formatter = TextFormatter()
@@ -492,7 +502,9 @@ class DocxConverter(HtmlConverter):
 
         result = None
         with open(local_path, "rb") as docx_file:
-            result = mammoth.convert_to_html(docx_file)
+            style_map = kwargs.get("style_map", None)
+
+            result = mammoth.convert_to_html(docx_file, style_map=style_map)
             html_content = result.value
             result = self._convert(html_content)
 
@@ -582,6 +594,10 @@ class PptxConverter(HtmlConverter):
                         "\n" + self._convert(html_table).text_content.strip() + "\n"
                     )
 
+                # Charts
+                if shape.has_chart:
+                    md_content += self._convert_chart_to_markdown(shape.chart)
+
                 # Text areas
                 elif shape.has_text_frame:
                     if shape == title:
@@ -616,6 +632,29 @@ class PptxConverter(HtmlConverter):
             return True
         return False
 
+    def _convert_chart_to_markdown(self, chart):
+        md = "\n\n### Chart"
+        if chart.has_title:
+            md += f": {chart.chart_title.text_frame.text}"
+        md += "\n\n"
+        data = []
+        category_names = [c.label for c in chart.plots[0].categories]
+        series_names = [s.name for s in chart.series]
+        data.append(["Category"] + series_names)
+
+        for idx, category in enumerate(category_names):
+            row = [category]
+            for series in chart.series:
+                row.append(series.values[idx])
+            data.append(row)
+
+        markdown_table = []
+        for row in data:
+            markdown_table.append("| " + " | ".join(map(str, row)) + " |")
+        header = markdown_table[0]
+        separator = "|" + "|".join(["---"] * len(data[0])) + "|"
+        return md + "\n".join([header, separator] + markdown_table[1:])
+
 
 class MediaConverter(DocumentConverter):
     """
@@ -754,7 +793,7 @@ class Mp3Converter(WavConverter):
 
 class ImageConverter(MediaConverter):
     """
-    Converts images to markdown via extraction of metadata (if `exiftool` is installed), OCR (if `easyocr` is installed), and description via a multimodal LLM (if an mlm_client is configured).
+    Converts images to markdown via extraction of metadata (if `exiftool` is installed), OCR (if `easyocr` is installed), and description via a multimodal LLM (if an llm_client is configured).
     """
 
     def convert(self, local_path, **kwargs) -> Union[None, DocumentConverterResult]:
@@ -784,17 +823,17 @@ class ImageConverter(MediaConverter):
                     md_content += f"{f}: {metadata[f]}\n"
 
         # Try describing the image with GPTV
-        mlm_client = kwargs.get("mlm_client")
-        mlm_model = kwargs.get("mlm_model")
-        if mlm_client is not None and mlm_model is not None:
+        llm_client = kwargs.get("llm_client")
+        llm_model = kwargs.get("llm_model")
+        if llm_client is not None and llm_model is not None:
             md_content += (
                 "\n# Description:\n"
-                + self._get_mlm_description(
+                + self._get_llm_description(
                     local_path,
                     extension,
-                    mlm_client,
-                    mlm_model,
-                    prompt=kwargs.get("mlm_prompt"),
+                    llm_client,
+                    llm_model,
+                    prompt=kwargs.get("llm_prompt"),
                 ).strip()
                 + "\n"
             )
@@ -804,11 +843,11 @@ class ImageConverter(MediaConverter):
             text_content=md_content,
         )
 
-    def _get_mlm_description(self, local_path, extension, client, model, prompt=None):
+    def _get_llm_description(self, local_path, extension, client, model, prompt=None):
         if prompt is None or prompt.strip() == "":
             prompt = "Write a detailed caption for this image."
 
-        sys.stderr.write(f"MLM Prompt:\n{prompt}\n")
+        sys.stderr.write(f"llm Prompt:\n{prompt}\n")
 
         data_uri = ""
         with open(local_path, "rb") as image_file:
@@ -837,6 +876,124 @@ class ImageConverter(MediaConverter):
         return response.choices[0].message.content
 
 
+class ZipConverter(DocumentConverter):
+    """Converts ZIP files to markdown by extracting and converting all contained files.
+
+    The converter extracts the ZIP contents to a temporary directory, processes each file
+    using appropriate converters based on file extensions, and then combines the results
+    into a single markdown document. The temporary directory is cleaned up after processing.
+
+    Example output format:
+    ```markdown
+    Content from the zip file `example.zip`:
+
+    ## File: docs/readme.txt
+
+    This is the content of readme.txt
+    Multiple lines are preserved
+
+    ## File: images/example.jpg
+
+    ImageSize: 1920x1080
+    DateTimeOriginal: 2024-02-15 14:30:00
+    Description: A beautiful landscape photo
+
+    ## File: data/report.xlsx
+
+    ## Sheet1
+    | Column1 | Column2 | Column3 |
+    |---------|---------|---------|
+    | data1   | data2   | data3   |
+    | data4   | data5   | data6   |
+    ```
+
+    Key features:
+    - Maintains original file structure in headings
+    - Processes nested files recursively
+    - Uses appropriate converters for each file type
+    - Preserves formatting of converted content
+    - Cleans up temporary files after processing
+    """
+
+    def convert(
+        self, local_path: str, **kwargs: Any
+    ) -> Union[None, DocumentConverterResult]:
+        # Bail if not a ZIP
+        extension = kwargs.get("file_extension", "")
+        if extension.lower() != ".zip":
+            return None
+
+        # Get parent converters list if available
+        parent_converters = kwargs.get("_parent_converters", [])
+        if not parent_converters:
+            return DocumentConverterResult(
+                title=None,
+                text_content=f"[ERROR] No converters available to process zip contents from: {local_path}",
+            )
+
+        extracted_zip_folder_name = (
+            f"extracted_{os.path.basename(local_path).replace('.zip', '_zip')}"
+        )
+        new_folder = os.path.normpath(
+            os.path.join(os.path.dirname(local_path), extracted_zip_folder_name)
+        )
+        md_content = f"Content from the zip file `{os.path.basename(local_path)}`:\n\n"
+
+        # Safety check for path traversal
+        if not new_folder.startswith(os.path.dirname(local_path)):
+            return DocumentConverterResult(
+                title=None, text_content=f"[ERROR] Invalid zip file path: {local_path}"
+            )
+
+        try:
+            # Extract the zip file
+            with zipfile.ZipFile(local_path, "r") as zipObj:
+                zipObj.extractall(path=new_folder)
+
+            # Process each extracted file
+            for root, dirs, files in os.walk(new_folder):
+                for name in files:
+                    file_path = os.path.join(root, name)
+                    relative_path = os.path.relpath(file_path, new_folder)
+
+                    # Get file extension
+                    _, file_extension = os.path.splitext(name)
+
+                    # Update kwargs for the file
+                    file_kwargs = kwargs.copy()
+                    file_kwargs["file_extension"] = file_extension
+                    file_kwargs["_parent_converters"] = parent_converters
+
+                    # Try converting the file using available converters
+                    for converter in parent_converters:
+                        # Skip the zip converter to avoid infinite recursion
+                        if isinstance(converter, ZipConverter):
+                            continue
+
+                        result = converter.convert(file_path, **file_kwargs)
+                        if result is not None:
+                            md_content += f"\n## File: {relative_path}\n\n"
+                            md_content += result.text_content + "\n\n"
+                            break
+
+            # Clean up extracted files if specified
+            if kwargs.get("cleanup_extracted", True):
+                shutil.rmtree(new_folder)
+
+            return DocumentConverterResult(title=None, text_content=md_content.strip())
+
+        except zipfile.BadZipFile:
+            return DocumentConverterResult(
+                title=None,
+                text_content=f"[ERROR] Invalid or corrupted zip file: {local_path}",
+            )
+        except Exception as e:
+            return DocumentConverterResult(
+                title=None,
+                text_content=f"[ERROR] Failed to process zip file {local_path}: {str(e)}",
+            )
+
+
 class FileConversionException(BaseException):
     pass
 
@@ -852,16 +1009,18 @@ class MarkItDown:
     def __init__(
         self,
         requests_session: Optional[requests.Session] = None,
-        mlm_client: Optional[Any] = None,
-        mlm_model: Optional[Any] = None,
+        llm_client: Optional[Any] = None,
+        llm_model: Optional[Any] = None,
+        style_map: Optional[str] = None,
     ):
         if requests_session is None:
             self._requests_session = requests.Session()
         else:
             self._requests_session = requests_session
 
-        self._mlm_client = mlm_client
-        self._mlm_model = mlm_model
+        self._llm_client = llm_client
+        self._llm_model = llm_model
+        self._style_map = style_map
 
         self._page_converters: List[DocumentConverter] = []
 
@@ -880,6 +1039,7 @@ class MarkItDown:
         self.register_page_converter(Mp3Converter())
         self.register_page_converter(ImageConverter())
         self.register_page_converter(PdfConverter())
+        self.register_page_converter(ZipConverter())
 
     def convert(
         self, source: Union[str, requests.Response], **kwargs: Any
@@ -1003,7 +1163,7 @@ class MarkItDown:
                 self._append_ext(extensions, g)
 
             # Convert
-            result = self._convert(temp_path, extensions, url=response.url)
+            result = self._convert(temp_path, extensions, url=response.url, **kwargs)
         # Clean up
         finally:
             try:
@@ -1030,11 +1190,17 @@ class MarkItDown:
                     _kwargs.update({"file_extension": ext})
 
                 # Copy any additional global options
-                if "mlm_client" not in _kwargs and self._mlm_client is not None:
-                    _kwargs["mlm_client"] = self._mlm_client
+                if "llm_client" not in _kwargs and self._llm_client is not None:
+                    _kwargs["llm_client"] = self._llm_client
 
-                if "mlm_model" not in _kwargs and self._mlm_model is not None:
-                    _kwargs["mlm_model"] = self._mlm_model
+                if "llm_model" not in _kwargs and self._llm_model is not None:
+                    _kwargs["llm_model"] = self._llm_model
+
+                # Add the list of converters for nested processing
+                _kwargs["_parent_converters"] = self._page_converters
+
+                if "style_map" not in _kwargs and self._style_map is not None:
+                    _kwargs["style_map"] = self._style_map
 
                 # If we hit an error log it and keep trying
                 try:
@@ -1071,8 +1237,7 @@ class MarkItDown:
         if ext == "":
             return
         # if ext not in extensions:
-        if True:
-            extensions.append(ext)
+        extensions.append(ext)
 
     def _guess_ext_magic(self, path):
         """Use puremagic (a Python implementation of libmagic) to guess a file's extension based on the first few bytes."""

tests/test_files/test_mskanji.csv

@@ -0,0 +1,4 @@
+<EFBFBD><EFBFBD><EFBFBD>O,<EFBFBD>N<EFBFBD><EFBFBD>,<EFBFBD>Z<EFBFBD><EFBFBD>
+<EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>Y,30,<EFBFBD><EFBFBD><EFBFBD><EFBFBD>
+<EFBFBD>O<EFBFBD>؉p<EFBFBD>q,25,<EFBFBD><EFBFBD><EFBFBD><EFBFBD>
+<EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>~,35,<EFBFBD><EFBFBD><EFBFBD>É<EFBFBD>

tests/test_markitdown.py

@@ -51,12 +51,25 @@ DOCX_TEST_STRINGS = [
     "AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation",
 ]
 
+DOCX_COMMENT_TEST_STRINGS = [
+    "314b0a30-5b04-470b-b9f7-eed2c2bec74a",
+    "49e168b7-d2ae-407f-a055-2167576f39a1",
+    "## d666f1f7-46cb-42bd-9a39-9a39cf2a509f",
+    "# Abstract",
+    "# Introduction",
+    "AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation",
+    "This is a test comment. 12df-321a",
+    "Yet another comment in the doc. 55yiyi-asd09",
+]
+
 PPTX_TEST_STRINGS = [
     "2cdda5c8-e50e-4db4-b5f0-9722a649f455",
     "04191ea8-5c73-4215-a1d3-1cfb43aaaf12",
     "44bf7d06-5e7a-4a40-a2e1-a2e42ef28c8a",
     "1b92870d-e3b5-4e65-8153-919f4ff45592",
     "AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation",
+    "a3f6004b-6f4f-4ea8-bee3-3741f4dc385f",  # chart title
+    "2003",  # chart value
 ]
 
 BLOG_TEST_URL = "https://microsoft.github.io/autogen/blog/2023/04/21/LLM-tuning-math"
@@ -87,6 +100,13 @@ SERP_TEST_EXCLUDES = [
     "data:image/svg+xml,%3Csvg%20width%3D",
 ]
 
+CSV_CP932_TEST_STRINGS = [
+    "名前,年齢,住所",
+    "佐藤太郎,30,東京",
+    "三木英子,25,大阪",
+    "髙橋淳,35,名古屋",
+]
+
 
 @pytest.mark.skipif(
     skip_remote,
@@ -130,6 +150,24 @@ def test_markitdown_local() -> None:
         text_content = result.text_content.replace("\\", "")
         assert test_string in text_content
 
+    # Test DOCX processing, with comments
+    result = markitdown.convert(
+        os.path.join(TEST_FILES_DIR, "test_with_comment.docx"),
+        style_map="comment-reference => ",
+    )
+    for test_string in DOCX_COMMENT_TEST_STRINGS:
+        text_content = result.text_content.replace("\\", "")
+        assert test_string in text_content
+
+    # Test DOCX processing, with comments and setting style_map on init
+    markitdown_with_style_map = MarkItDown(style_map="comment-reference => ")
+    result = markitdown_with_style_map.convert(
+        os.path.join(TEST_FILES_DIR, "test_with_comment.docx")
+    )
+    for test_string in DOCX_COMMENT_TEST_STRINGS:
+        text_content = result.text_content.replace("\\", "")
+        assert test_string in text_content
+
     # Test PPTX processing
     result = markitdown.convert(os.path.join(TEST_FILES_DIR, "test.pptx"))
     for test_string in PPTX_TEST_STRINGS:
@@ -144,6 +182,12 @@ def test_markitdown_local() -> None:
         text_content = result.text_content.replace("\\", "")
         assert test_string in text_content
 
+    # Test ZIP file processing
+    result = markitdown.convert(os.path.join(TEST_FILES_DIR, "test_files.zip"))
+    for test_string in DOCX_TEST_STRINGS:
+        text_content = result.text_content.replace("\\", "")
+        assert test_string in text_content
+
     # Test Wikipedia processing
     result = markitdown.convert(
         os.path.join(TEST_FILES_DIR, "test_wikipedia.html"), url=WIKIPEDIA_TEST_URL
@@ -164,6 +208,12 @@ def test_markitdown_local() -> None:
     for test_string in SERP_TEST_STRINGS:
         assert test_string in text_content
 
+    ## Test non-UTF-8 encoding
+    result = markitdown.convert(os.path.join(TEST_FILES_DIR, "test_mskanji.csv"))
+    text_content = result.text_content.replace("\\", "")
+    for test_string in CSV_CP932_TEST_STRINGS:
+        assert test_string in text_content
+
 
 @pytest.mark.skipif(
     skip_exiftool,