### Complete CHM to Markdown Conversion Pipeline

Source: https://context7.com/chenchuiqing/chm-to-markdown/llms.txt

A bash script demonstrating the end-to-end workflow for converting CHM files to Markdown. It includes steps for decompiling CHM to HTML, converting HTML to Markdown, and optionally converting specific subdirectories. This pipeline preserves directory structure.

```bash
# Step 1: Navigate to project directory
cd /path/to/device-sdk-chm-to-md

# Step 2: Decompile CHM to HTML (creates out_html directory)
uv run python chm_to_html.py "DeviceNetworkSDK_Manual.chm" -o out_html

# Step 3: Convert all HTML to Markdown (creates out_md directory)
uv run python html_to_md.py --root out_html --out-root out_md

# Step 4: Convert only specific sections if needed
uv run python html_to_md.py --root out_html --out-root out_md --subdir "API"
uv run python html_to_md.py --root out_html --out-root out_md --subdir "Structures"

# Result: Directory structure preserved
# out_html/API/GetDeviceInfo.html  →  out_md/API/GetDeviceInfo.md
# out_html/Structures/DeviceConfig.html  →  out_md/Structures/DeviceConfig.md
```

--------------------------------

### Post-process Markdown Cleanup

Source: https://context7.com/chenchuiqing/chm-to-markdown/llms.txt

Cleans raw Markdown by replacing non-breaking spaces, removing control characters, converting .html links to .md, and normalizing excessive blank lines. It takes a string of Markdown as input and returns a cleaned string.

```python
from html_to_md import postprocess_markdown

raw_markdown = """
# Title


Some text with\xa0non-breaking\xa0spaces.

Check [this link](api.html) and [anchor](page.html#section).



Too many blank lines above.
"""

cleaned = postprocess_markdown(raw_markdown)
print(cleaned)
```

--------------------------------

### Decompile CHM to HTML using Python

Source: https://context7.com/chenchuiqing/chm-to-markdown/llms.txt

This script decompiles Windows CHM files into individual HTML files. It utilizes the `hh.exe` utility and can be used programmatically or via the command line. The output is organized into a specified directory, preserving the original structure. Requires Windows and `hh.exe` to be in the system PATH.

```python
# Command-line usage - Basic decompilation to default 'html' directory
# uv run python chm_to_html.py "DeviceSDK_Manual.chm"

# Command-line usage - Custom output directory
# uv run python chm_to_html.py "DeviceSDK_Manual.chm" -o out_html

# Programmatic usage
from pathlib import Path
from chm_to_html import decompile_chm, find_hh_exe

# Check if hh.exe is available
hh_path = find_hh_exe()
if hh_path:
    print(f"Found hh.exe at: {hh_path}")

    # Decompile CHM file
    chm_file = Path("DeviceSDK_Manual.chm")
    output_dir = Path("extracted_html")

    exit_code = decompile_chm(chm_file, output_dir)
    if exit_code == 0:
        print("CHM decompilation completed successfully")
    else:
        print(f"Decompilation failed with exit code: {exit_code}")
else:
    print("hh.exe not found - this script requires Windows")

# Expected output:
# Using: C:\Windows\hh.exe
# CHM file: C:\path\to\DeviceSDK_Manual.chm
# Output directory: C:\path\to\extracted_html
# Executing command: C:\Windows\hh.exe -decompile C:\path\to\extracted_html C:\path\to\DeviceSDK_Manual.chm
# CHM decompilation completed.

```

--------------------------------

### Convert HTML to Markdown with Encoding Detection using Python

Source: https://context7.com/chenchuiqing/chm-to-markdown/llms.txt

This script performs batch conversion of HTML files to Markdown. It intelligently detects and handles various Chinese character encodings (UTF-8, GBK, GB2312, GB18030), cleans up scripts and styles, and transforms links. It can process all HTML files in a directory, a specific subdirectory, or convert a single HTML string. Output can be directed to a separate directory.

```python
# Command-line usage - Convert all HTML files from default 'html' directory
# uv run python html_to_md.py

# Command-line usage - Specify HTML root directory
# uv run python html_to_md.py --root out_html

# Command-line usage - Convert only a specific subdirectory
# uv run python html_to_md.py --root out_html --subdir "Structures"

# Command-line usage - Output Markdown to separate directory
# uv run python html_to_md.py --root out_html --out-root out_md

# Programmatic usage
from pathlib import Path
from html_to_md import (
    html_to_markdown,
    detect_and_read,
    collect_html_files,
    process_file
)

# Convert a single HTML string to Markdown
html_content = """
<html>
<head><script>alert('removed')</script></head>
<body>
    <h1>API Reference</h1>
    <p>This is the <strong>main</strong> documentation.</p>
    <a href="details.html">See details</a>
</body>
</html>
"""
markdown_output = html_to_markdown(html_content)
print(markdown_output)
# Output:
# # API Reference
#
# This is the **main** documentation.
#
# [See details](details.md)

# Read file with automatic encoding detection
html_root = Path("out_html")
md_root = Path("out_md")

# Collect all HTML files in a directory
html_files = collect_html_files(html_root)
print(f"Found {len(html_files)} HTML files to convert")

# Process each file
for html_file in html_files:
    process_file(html_file, html_root, md_root)
    # Converts: out_html/Structures/DeviceInfo.html
    # To:       out_md/Structures/DeviceInfo.md

```

--------------------------------

### Detect Encoding and Read File Content using Python

Source: https://context7.com/chenchuiqing/chm-to-markdown/llms.txt

This function reads HTML files by automatically detecting their character encoding. It prioritizes common Chinese encodings like UTF-8, GBK, GB2312, and GB18030, falling back to UTF-8 with error ignoring if necessary. This ensures accurate reading of content with diverse character sets.

```python
from pathlib import Path
from html_to_md import detect_and_read

# Read a file with unknown encoding
file_path = Path("out_html/guide/introduction.html")

try:
    content = detect_and_read(file_path)
    print(f"Successfully read {len(content)} characters")
    print(content[:200])  # Preview first 200 chars
except RuntimeError as e:
    print(f"Failed to decode file: {e}")

# The function tries encodings in order:
# 1. UTF-8 (strict)
# 2. GBK (strict)
# 3. GB2312 (strict)
# 4. GB18030 (strict)
# 5. UTF-8 (ignore errors) - fallback

```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.