markitdown-pro 1.1.2__tar.gz → 1.2.3__tar.gz

{markitdown_pro-1.1.2 → markitdown_pro-1.2.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: markitdown-pro
-Version: 1.1.2
+Version: 1.2.3
 Summary: A package that converts almost any file format to Markdown.
 Author: Developer
 Classifier: Programming Language :: Python :: 3

{markitdown_pro-1.1.2 → markitdown_pro-1.2.3}/markitdown_pro/converters/azure_doc_intel_wrapper.py

@@ -5,9 +5,14 @@ from .base import ConverterWrapper
 
 
 class DocIntelligenceWrapper(ConverterWrapper):
+
+    SUPPORTED_EXTENSIONS = DocumentIntelligenceHandler.SUPPORTED_EXTENSIONS
+
     def __init__(self):
         super().__init__("Azure Document Intelligence")
         self.converter = DocumentIntelligenceHandler()
 
     async def convert(self, file_path: str) -> Optional[str]:
-        return await self.converter.convert_to_md(file_path)
+        markdown = await self.converter.convert_to_md(file_path)
+        await self.converter.aclose()
+        return markdown

markitdown_pro-1.2.3/markitdown_pro/converters/tabular_wrapper.py

@@ -0,0 +1,31 @@
+from typing import Optional
+
+from ..common.logger import logger
+from ..common.utils import clean_markdown
+from ..handlers.tabular_handler import TabularHandler
+from .base import ConverterWrapper
+
+
+class TabularWrapper(ConverterWrapper):
+    SUPPORTED_FORMATS = (
+        "csv",
+        "tsv",
+        "xls",
+        "xlsx",
+    )
+
+    def __init__(self):
+        super().__init__("TabularHandler")
+        self.tabular_handler = TabularHandler()
+
+    async def convert(self, file_path: str) -> Optional[str]:
+        logger.debug(f"TabularWrapper: Converting {file_path} to markdown")
+        file_extension = file_path.split(".")[-1].lower()
+
+        if file_extension not in self.SUPPORTED_FORMATS:
+            logger.warning(f"TabularWrapper: Unsupported file format: {file_extension}")
+            return None
+
+        markdown = await self.tabular_handler.handle(file_path)
+        markdown = clean_markdown(markdown) if markdown else None
+        return markdown

{markitdown_pro-1.1.2 → markitdown_pro-1.2.3}/markitdown_pro/handlers/markup_handler.py

@@ -18,7 +18,7 @@ from .base_handler import BaseHandler
 class MarkupHandler(BaseHandler):
     """Handler for .html, .xml, .json, .ndjson, .yaml, .yml files."""
 
-    extensions = frozenset([".html", ".htm", ".xml", ".json", ".ndjson", ".yaml", ".yml"])
+    SUPPORTED_EXTENSIONS = frozenset([".html", ".htm", ".xml", ".json", ".ndjson", ".yaml", ".yml"])
 
     async def handle(self, file_path: str, *args, **kwargs) -> Optional[str]:
         logger.info(f"Processing markup file: {file_path}")

{markitdown_pro-1.1.2 → markitdown_pro-1.2.3}/markitdown_pro/handlers/office_handler.py

@@ -5,6 +5,7 @@ from ..common.utils import ensure_minimum_content
 from ..converters.azure_doc_intel_wrapper import DocIntelligenceWrapper
 from ..converters.base import ConverterWrapper
 from ..converters.markitdown_wrapper import MarkItDownWrapper
+from ..converters.tabular_wrapper import TabularWrapper
 from ..converters.unstructured_wrapper import UnstructuredWrapper
 from .base_handler import BaseHandler
 
@@ -13,9 +14,10 @@ class OfficeHandler(BaseHandler):
     """
     Convert Office documents to Markdown by trying a series of converters in order:
 
-      1) MarkItDown (fast, local parsing for Office formats)
-      2) Azure Document Intelligence (cloud extraction + OCR for embedded images)
-      3) Unstructured (robust fallback parser)
+      1) TabularHandler (specialized handler for tabular data)
+      2) MarkItDown (fast, local parsing for Office formats)
+      3) Azure Document Intelligence (cloud extraction + OCR for embedded images)
+      4) Unstructured (robust fallback parser)
 
     The first converter that returns non-trivial Markdown (as judged by
     `ensure_minimum_content`) wins. If all fail, `None` is returned.
@@ -34,12 +36,14 @@ class OfficeHandler(BaseHandler):
         super().__init__(*args, **kwargs)
 
         # Ordered, fastest → most comprehensive
+        self.tabular = TabularWrapper()
         self.markitdown = MarkItDownWrapper()
         self.doc_intelligence = DocIntelligenceWrapper()
         self.unstructured = UnstructuredWrapper()
 
         # (converter, human_readable_name) in the exact order to try
         self._pipeline: list[tuple[ConverterWrapper, str]] = [
+            (self.tabular, "TabularHandler"),
             (self.markitdown, "MarkItDown"),
             (self.doc_intelligence, "Azure Document Intelligence"),
             (self.unstructured, "Unstructured"),

{markitdown_pro-1.1.2 → markitdown_pro-1.2.3}/markitdown_pro/handlers/pdf_handler.py

@@ -89,10 +89,10 @@ class PDFHandler(BaseHandler):
             # Choose pipeline based on detected content type
             if pdf_type == PDFType.TEXT_ONLY:
                 pipeline = self.text_pipeline
-            elif pdf_type == PDFType.ALL_IMAGES:
+            elif pdf_type == PDFType.ALL_IMAGES or pdf_type == PDFType.TEXT_PLUS_IMAGES:
                 pipeline = self.image_pipeline
-            elif pdf_type == PDFType.TEXT_PLUS_IMAGES:
-                pipeline = self.text_pipeline + self.image_pipeline
+            # elif pdf_type == PDFType.TEXT_PLUS_IMAGES:
+            #     pipeline = self.text_pipeline + self.image_pipeline
             else:
                 # Fallback to text pipeline if detection returned an unexpected value
                 pipeline = self.text_pipeline

markitdown_pro-1.2.3/markitdown_pro/handlers/tabular_handler.py

@@ -0,0 +1,96 @@
+import asyncio
+import csv
+import os
+from io import StringIO
+from typing import Optional
+
+from openpyxl import load_workbook
+
+from ..common.logger import logger
+from .base_handler import BaseHandler
+
+
+def _escape_md_cell(s: str) -> str:
+    return s.replace("|", r"\|").replace("\n", " ").replace("\r", " ").strip()
+
+
+def _csv_to_markdown(path: str, delimiter: str) -> str:
+    out = StringIO()
+    with open(path, "r", newline="", encoding="utf-8") as f:
+        reader = csv.reader(f, delimiter=delimiter)
+        try:
+            header = next(reader)
+        except StopIteration:
+            return ""  # empty file
+        header = ["" if h is None else str(h) for h in header]
+        out.write("| " + " | ".join(_escape_md_cell(h) for h in header) + " |\n")
+        out.write("| " + " | ".join(["---"] * len(header)) + " |\n")
+        for row in reader:
+            row = ["" if c is None else str(c) for c in row]
+            out.write("| " + " | ".join(_escape_md_cell(c) for c in row) + " |\n")
+    return out.getvalue()
+
+
+def _xlsx_to_markdown(path: str) -> str:
+    wb = load_workbook(filename=path, read_only=True, data_only=True)
+    out = StringIO()
+    for sheet_name in wb.sheetnames:
+        ws = wb[sheet_name]
+        out.write(f"\n\n### Sheet: {sheet_name}\n\n")
+        rows_iter = ws.iter_rows(values_only=True)
+        try:
+            header_row = next(rows_iter)
+        except StopIteration:
+            out.write("_(empty sheet)_\n")
+            continue
+        if header_row is None:
+            out.write("_(empty sheet)_\n")
+            continue
+        header_vals = ["" if v is None else str(v) for v in header_row]
+        out.write("| " + " | ".join(_escape_md_cell(v) for v in header_vals) + " |\n")
+        out.write("| " + " | ".join(["---"] * len(header_vals)) + " |\n")
+        for row in rows_iter:
+            if row is None:
+                out.write("| " + " | ".join("" for _ in header_vals) + " |\n")
+                continue
+            cells = ["" if v is None else str(v) for v in row]
+            # pad/truncate to header length so column count stays consistent
+            if len(cells) < len(header_vals):
+                cells += [""] * (len(header_vals) - len(cells))
+            elif len(cells) > len(header_vals):
+                cells = cells[: len(header_vals)]
+            out.write("| " + " | ".join(_escape_md_cell(v) for v in cells) + " |\n")
+    return out.getvalue()
+
+
+class TabularHandler(BaseHandler):
+    """Handler for .csv, .tsv, .xls, .xlsx files."""
+
+    extensions = frozenset([".csv", ".tsv", ".xls", ".xlsx"])
+
+    async def handle(self, file_path: str, *args, **kwargs) -> Optional[str]:
+        logger.info(f"Processing tabular file: {file_path}")
+        try:
+            ext = os.path.splitext(file_path)[1].lower()
+            if ext == ".csv":
+                md = await asyncio.to_thread(_csv_to_markdown, file_path, ",")
+            elif ext == ".tsv":
+                md = await asyncio.to_thread(_csv_to_markdown, file_path, "\t")
+            elif ext in [".xlsx"]:
+                md = await asyncio.to_thread(_xlsx_to_markdown, file_path)
+            elif ext in [".xls"]:
+                # Fallback: many environments lack a safe streaming .xls reader.
+                # If .xls appears, let pandas handle it but still offload to a thread.
+                import pandas as pd
+
+                def _xls_to_md(p: str) -> str:
+                    df = pd.read_excel(p)
+                    return df.to_markdown(index=False)
+
+                md = await asyncio.to_thread(_xls_to_md, file_path)
+            else:
+                raise RuntimeError("Unsupported tabular format")
+            return md or ""
+        except Exception as e:
+            logger.error(f"Error processing tabular file {file_path}: {e}")
+            return None

{markitdown_pro-1.1.2 → markitdown_pro-1.2.3}/markitdown_pro/handlers/text_handler.py

@@ -11,7 +11,7 @@ from .base_handler import BaseHandler
 class TextHandler(BaseHandler):
     """Handler for .txt, .md, .py, .go, and other text/code files."""
 
-    extensions = frozenset([".txt", ".md", ".py", ".go"])
+    SUPPORTED_EXTENSIONS = frozenset([".txt", ".md", ".py", ".go"])
 
     async def handle(self, file_path: str, *args, **kwargs) -> Optional[str]:
         logger.info(f"Processing text file: {file_path}")

{markitdown_pro-1.1.2 → markitdown_pro-1.2.3}/markitdown_pro/services/azure_doc_intelligence.py

@@ -59,7 +59,7 @@ class DocumentIntelligenceHandler:
             ".heic",
         }
     )
-    extensions: frozenset[str] = frozenset(
+    SUPPORTED_EXTENSIONS: frozenset[str] = frozenset(
         list(pdf_extensions) + list(office_extensions) + list(image_extensions)
     )
 
@@ -112,7 +112,7 @@ class DocumentIntelligenceHandler:
             return None
 
         ext = file_path.suffix.lower()
-        if ext not in self.extensions:
+        if ext not in self.SUPPORTED_EXTENSIONS:
             logger.warning(f"DocumentIntelligenceHandler: unsupported extension '{ext}'.")
             return None
 
@@ -178,4 +178,4 @@ class DocumentIntelligenceHandler:
     @classmethod
     def supports(cls, path: str | Path) -> bool:
         """Return True if the file extension is supported by this handler."""
-        return Path(path).suffix.lower() in cls.extensions
+        return Path(path).suffix.lower() in cls.SUPPORTED_EXTENSIONS

{markitdown_pro-1.1.2 → markitdown_pro-1.2.3}/markitdown_pro.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: markitdown-pro
-Version: 1.1.2
+Version: 1.2.3
 Summary: A package that converts almost any file format to Markdown.
 Author: Developer
 Classifier: Programming Language :: Python :: 3

{markitdown_pro-1.1.2 → markitdown_pro-1.2.3}/markitdown_pro.egg-info/SOURCES.txt

@@ -21,6 +21,7 @@ markitdown_pro/converters/base.py
 markitdown_pro/converters/gpt_vision_wrapper.py
 markitdown_pro/converters/markitdown_wrapper.py
 markitdown_pro/converters/pymupdf_wrapper.py
+markitdown_pro/converters/tabular_wrapper.py
 markitdown_pro/converters/unstructured_wrapper.py
 markitdown_pro/converters/youtube_wrapper.py
 markitdown_pro/handlers/__init__.py
@@ -45,7 +46,5 @@ tests/test_conversion_pipeline.py
 tests/handlers/__init__.py
 tests/handlers/test_email_handler.py
 tests/handlers/test_markup_handler.py
-tests/handlers/test_office_handler.py
-tests/handlers/test_pdf_handler.py
 tests/handlers/test_pst_handler.py
 tests/handlers/test_text_handler.py

{markitdown_pro-1.1.2 → markitdown_pro-1.2.3}/setup.py

@@ -2,7 +2,7 @@ from setuptools import find_packages, setup
 
 setup(
     name="markitdown-pro",
-    version="1.1.2",
+    version="1.2.3",
     author="Developer",
     description="A package that converts almost any file format to Markdown.",
     long_description=open("README.md").read(),

{markitdown_pro-1.1.2 → markitdown_pro-1.2.3}/tests/handlers/test_markup_handler.py

@@ -9,25 +9,15 @@ from tests.utils import list_files
 
 log = logging.getLogger(__name__)
 
-INCLUDE_EXTENSIONS = [
-    ".html",
-    ".htm",
-    ".xml",
-    ".json",
-    ".ndjson",
-    ".yaml",
-    ".yml",
-]
-
 TEST_FILES_PATH = data_path()
-ALL_FILES = list_files(TEST_FILES_PATH, include_ext=INCLUDE_EXTENSIONS, recursive=True)
+ALL_FILES = list_files(TEST_FILES_PATH, include_ext=MarkupHandler.extensions, recursive=True)
 log.info(f"Found {len(ALL_FILES)} test files in {TEST_FILES_PATH}.")
 
 
 @pytest.mark.asyncio
 @pytest.mark.parametrize("file_path", ALL_FILES, ids=[pretty_id(p) for p in ALL_FILES])
 async def test_files(file_path: Path, markup_handler: MarkupHandler):
-    log.info(f"Testing file: {file_path}")
+    log.info(f"{MarkupHandler.__name__}: Testing file {file_path}")
 
     markdown_text = await markup_handler.handle(str(file_path))
     assert markdown_text is not None, f"Conversion returned None for {file_path.name}"

{markitdown_pro-1.1.2 → markitdown_pro-1.2.3}/tests/handlers/test_text_handler.py

@@ -9,17 +9,17 @@ from tests.utils import list_files
 
 log = logging.getLogger(__name__)
 
-INCLUDE_EXTENSIONS = [".txt", ".md", ".py", ".go"]
-
 TEST_FILES_PATH = data_path()
-ALL_FILES = list_files(TEST_FILES_PATH, include_ext=INCLUDE_EXTENSIONS, recursive=True)
+ALL_FILES = list_files(
+    TEST_FILES_PATH, include_ext=TextHandler.SUPPORTED_EXTENSIONS, recursive=True
+)
 log.info(f"Found {len(ALL_FILES)} test files in {TEST_FILES_PATH}.")
 
 
 @pytest.mark.asyncio
 @pytest.mark.parametrize("file_path", ALL_FILES, ids=[pretty_id(p) for p in ALL_FILES])
 async def test_files(file_path: Path, text_handler: TextHandler):
-    log.info(f"Testing file: {file_path}")
+    log.info(f"{TextHandler.__name__}: Testing file {file_path}")
 
     markdown_text = await text_handler.handle(str(file_path))
     assert markdown_text is not None, f"Conversion returned None for {file_path.name}"

markitdown_pro-1.1.2/markitdown_pro/handlers/tabular_handler.py

@@ -1,31 +0,0 @@
-import os
-from typing import Optional
-
-import pandas as pd
-
-from ..common.logger import logger
-from .base_handler import BaseHandler
-
-
-class TabularHandler(BaseHandler):
-    """Handler for .csv, .tsv, .xls, .xlsx files."""
-
-    extensions = frozenset([".csv", ".tsv", ".xls", ".xlsx"])
-
-    async def handle(self, file_path: str, *args, **kwargs) -> Optional[str]:
-        logger.info(f"Processing tabular file: {file_path}")
-        try:
-            ext = os.path.splitext(file_path)[1].lower()
-            if ext in [".csv", ".tsv"]:
-                delimiter = "\t" if ext == ".tsv" else ","
-                df = pd.read_csv(file_path, delimiter=delimiter)
-            elif ext in [".xls", ".xlsx"]:
-                df = pd.read_excel(file_path)
-            else:
-                raise RuntimeError("Unsupported tabular format")
-
-            md = df.to_markdown(index=False)
-            return md
-        except Exception as e:
-            logger.error(f"Error processing tabular file {file_path}: {e}")
-            return None

markitdown_pro-1.1.2/tests/handlers/test_office_handler.py

@@ -1,26 +0,0 @@
-import logging
-from pathlib import Path
-
-import pytest
-
-from markitdown_pro.handlers.office_handler import OfficeHandler
-from tests.fixtures import data_path, pretty_id
-from tests.utils import list_files
-
-log = logging.getLogger(__name__)
-
-INCLUDE_EXTENSIONS = [".docx", ".xlsx", ".pptx"]
-
-TEST_FILES_PATH = data_path()
-ALL_FILES = list_files(TEST_FILES_PATH, include_ext=INCLUDE_EXTENSIONS, recursive=True)
-log.info(f"Found {len(ALL_FILES)} test files in {TEST_FILES_PATH}.")
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize("file_path", ALL_FILES, ids=[pretty_id(p) for p in ALL_FILES])
-async def test_files(file_path: Path, office_handler: OfficeHandler):
-    log.info(f"Testing file: {file_path}")
-
-    markdown_text = await office_handler.handle(str(file_path))
-    assert markdown_text is not None, f"Conversion returned None for {file_path.name}"
-    assert isinstance(markdown_text, str), f"Output type is not str for {file_path.name}"

markitdown_pro-1.1.2/tests/handlers/test_pdf_handler.py

@@ -1,26 +0,0 @@
-import logging
-from pathlib import Path
-
-import pytest
-
-from markitdown_pro.handlers.pdf_handler import PDFHandler
-from tests.fixtures import data_path, pretty_id
-from tests.utils import list_files
-
-log = logging.getLogger(__name__)
-
-INCLUDE_EXTENSIONS = [".pdf"]
-
-TEST_FILES_PATH = data_path()
-ALL_FILES = list_files(TEST_FILES_PATH, include_ext=INCLUDE_EXTENSIONS, recursive=True)
-log.info(f"Found {len(ALL_FILES)} test files in {TEST_FILES_PATH}.")
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize("file_path", ALL_FILES, ids=[pretty_id(p) for p in ALL_FILES])
-async def test_files(file_path: Path, pdf_handler: PDFHandler):
-    log.info(f"Testing file: {file_path}")
-
-    markdown_text = await pdf_handler.handle(str(file_path))
-    assert markdown_text is not None, f"Conversion returned None for {file_path.name}"
-    assert isinstance(markdown_text, str), f"Output type is not str for {file_path.name}"