Updated the project structure, fixed code

.gitignore

@@ -161,3 +161,4 @@ cython_debug/
 
 path/*
 *.docx
+.pypirc

pyproject.toml

@@ -0,0 +1,28 @@
+[build-system]
+requires = ["setuptools>=45", "wheel", "setuptools_scm>=6.2"]
+build-backend = "setuptools.build_meta"
+
+[tool.black]
+line-length = 88
+include = '\.pyi?$'
+extend-exclude = '''
+# Add files to exclude from formatting
+'''
+
+[tool.isort]
+profile = "black"
+multi_line_output = 3
+
+[tool.mypy]
+python_version = "3.8"
+warn_return_any = true
+warn_unused_configs = true
+disallow_untyped_defs = true
+check_untyped_defs = true
+
+[tool.pytest.ini_options]
+minversion = "6.0"
+addopts = "-ra -q"
+testpaths = [
+    "tests",
+]

requirements-dev.txt

@@ -1,10 +1,9 @@
 -r requirements.txt
-pytest>=7.0.0
+black>=22.3.0
+isort>=5.10.1
+mypy>=0.981
+pytest>=7.1.3
 pytest-cov>=3.0.0
-black>=22.0.0
-isort>=5.10.0
-flake8>=4.0.0
-mypy>=0.950
-tox>=3.24.0
-build>=0.7.0
-twine>=3.8.0
+flake8>=4.0.1
+tox>=3.24.5
+pre-commit>=2.20.0

requirements.txt

@@ -1,4 +1,4 @@
-beautifulsoup4>=4.9.3
+beautifulsoup4>=4.9.3,<5.0.0
 python-docx>=0.8.11
 requests>=2.25.1
 click>=8.0.0

setup.py

@@ -1,27 +1,24 @@
+"""Package configuration for boxtodocx."""
 from setuptools import setup, find_packages
 
 with open("README.md", "r", encoding="utf-8") as fh:
     long_description = fh.read()
 
-with open("requirements.txt", "r", encoding="utf-8") as fh:
-    requirements = [line.strip() for line in fh if line.strip() and not line.startswith("#")]
-
 setup(
     name="boxtodocx",
-    version="1.0.0",
+    version="0.1.0",
     author="Ujjwal Kumar",
     author_email="[email protected]",
-    description="Convert Box Notes to Microsoft Word (docx) documents",
+    description="Convert Box documents to HTML and DOCX formats",
     long_description=long_description,
     long_description_content_type="text/markdown",
-    url="https://github.com/ujjwal-ibm/boxtodocx",
+    url="https://github.ibm.com/Ujjwal-Kumar1/boxtodocx",
     project_urls={
-        "Bug Tracker": "https://github.com/ujjwal-ibm/boxtodocx/issues",
-        "Documentation": "https://github.com/ujjwal-ibm/boxtodocx",
+        "Bug Tracker": "https://github.ibm.com/Ujjwal-Kumar1/boxtodocx/issues",
     },
     classifiers=[
-        "Development Status :: 5 - Production/Stable",
-        "Intended Audience :: End Users",
+        "Development Status :: 4 - Beta",
+        "Intended Audience :: Developers",
         "License :: OSI Approved :: MIT License",
         "Operating System :: OS Independent",
         "Programming Language :: Python :: 3",
@@ -30,17 +27,25 @@
         "Programming Language :: Python :: 3.10",
         "Programming Language :: Python :: 3.11",
         "Topic :: Office/Business",
-        "Topic :: Text Processing",
+        "Topic :: Text Processing :: Markup :: HTML",
     ],
     package_dir={"": "src"},
     packages=find_packages(where="src"),
     python_requires=">=3.8",
-    install_requires=requirements,
+    install_requires=[
+        "beautifulsoup4>=4.9.3",
+        "python-docx>=0.8.11",
+        "requests>=2.25.1",
+        "click>=8.0.0",
+        "typing-extensions>=4.0.0",
+        "colorlog>=6.7.0",
+        "Pillow>=10.0.0",
+        "yattag>=1.16.0",
+        "selenium>=4.0.0",
+    ],
     entry_points={
         "console_scripts": [
             "boxtodocx=boxtodocx.cli:main",
         ],
     },
-    include_package_data=True,
-)
-
+)

src/boxtodocx/cli.py

@@ -1,37 +1,142 @@
+"""Command-line interface for Box document conversion."""
+from typing import Optional
 import click
-import os
 from pathlib import Path
-from .converter import BoxNoteConverter
-from .utils.logger import get_logger
 
-logger = get_logger(__name__)
+from .convertor import BoxNoteConverter
+from .utils.logger import setup_logger
+from .utils.constants import DEFAULT_OUTPUT_DIR
 
-@click.command()
-@click.version_option()
-@click.argument('input_path', type=click.Path(exists=True))
-@click.option('-d', '--dir', help='Work directory for temporary files')
-@click.option('-t', '--token', help='Box access token')
-@click.option('-o', '--output', help='Output file name')
-@click.option('-u', '--user', help='Box user id')
-@click.option('-v', '--verbose', is_flag=True, help='Enable verbose logging')
-def main(input_path, dir, token, output, user, verbose):
+logger = setup_logger()
+
+@click.argument(
+    "input_path",
+    type=click.Path(exists=True, path_type=Path)
+)
+@click.option(
+    "--dest-dir", "-d",
+    type=click.Path(path_type=Path),
+    default=DEFAULT_OUTPUT_DIR,
+    help="Destination directory for output files"
+)
+@click.option(
+    "--export-images",
+    is_flag=True,
+    default=False,
+    help="Export images from Box documents"
+)
+@click.option(
+    "--box-id",
+    help="Box login email"
+)
+@click.option(
+    "--box-pwd",
+    help="Box login password"
+)
+@click.option(
+    "--link",
+    help="Div ID for additional login step"
+)
+@click.option(
+    "--id",
+    help="User ID field for Box login"
+)
+@click.option(
+    "--pwd",
+    help="Password field for Box login"
+)
+@click.option(
+    "--mfa-link",
+    help="Div ID to click for MFA"
+)
+@click.option(
+    "--mfa-otp-id",
+    help="ID of input field for MFA code"
+)
+@click.option(
+    "--mfa-btn-id",
+    help="ID of button to submit MFA code"
+)
+@click.option(
+    "--mfa-otp",
+    help="MFA code (one-time password)"
+)
+@click.option(
+    "--directory",
+    is_flag=True,
+    default=False,
+    help="Process all Box documents in a directory"
+)
+@click.command(help="Convert Box documents to HTML and DOCX formats with image support.")
+def main(
+    input_path: Path,
+    dest_dir: Path,
+    export_images: bool,
+    box_id: Optional[str],
+    box_pwd: Optional[str],
+    link: Optional[str],
+    id: Optional[str],
+    pwd: Optional[str],
+    mfa_link: Optional[str],
+    mfa_otp_id: Optional[str],
+    mfa_btn_id: Optional[str],
+    mfa_otp: Optional[str],
+    directory: bool
+) -> None:
+    """
+    Convert Box documents to HTML and DOCX formats.
+    
+    Args:
+        input_path: Path to Box document or directory
+    """
     try:
-        converter = BoxNoteConverter()
-        input_path = Path(input_path)
+        credentials = None
+        if export_images:
+            if not all([box_id, box_pwd, link, id, pwd]):
+                raise click.UsageError(
+                    "All Box login parameters must be provided when exporting images"
+                )
+            credentials = {
+                "user_id": box_id,
+                "password": box_pwd,
+                "link_id": link,
+                "user_field_id": id,
+                "password_field_id": pwd,
+                "mfa_link": mfa_link,
+                "mfa_otp_id": mfa_otp_id,
+                "mfa_btn_id": mfa_btn_id,
+                "mfa_otp": mfa_otp
+            }
 
-        if input_path.is_file():
-            _convert_file(converter, input_path, output or dir)
-        else:
-            _convert_directory(converter, input_path)
+        converter = BoxNoteConverter(dest_dir)
+
+        if directory:
+            if not input_path.is_dir():
+                raise click.BadParameter(f"Not a directory: {input_path}")
+
+            results = converter.convert_directory(input_path, credentials)
+            logger.info(f"Processed {len(results)} files in {input_path}")
 
+        else:
+            if not input_path.is_file():
+                raise click.BadParameter(f"Not a file: {input_path}")
+
+            if not BoxNoteConverter.validate_boxnote(input_path):
+                raise click.BadParameter(f"Invalid Box document: {input_path}")
+
+            html_path, docx_path, image_paths = converter.convert(
+                input_path,
+                credentials
+            )
+            logger.info(f"Converted {input_path} to:")
+            logger.info(f"  HTML: {html_path}")
+            logger.info(f"  DOCX: {docx_path}")
+            if image_paths:
+                logger.info(f"  Images: {len(image_paths)} files")
+
     except Exception as e:
-        logger.error(f"Error: {str(e)}")
+        logger.error(str(e))
         raise click.Abort()
 
-def _convert_file(converter, file_path, output_dir=None):
-    output_dir = output_dir or file_path.parent
-    converter.convert(str(file_path), str(output_dir))
-
-def _convert_directory(converter, dir_path):
-    for file in dir_path.glob('**/*.boxnote'):
-        _convert_file(converter, file)
+if __name__ == "__main__":
+    main()