clean-md-fences

A desktop utility for cleaning ChatGPT-style Markdown code fences.

It removes unwanted id="..." attributes from fenced code blocks and normalizes text / markdown fences while preserving all other language identifiers.

Available as:

• Python source
• Standalone Linux AppImage
• PyInstaller executable

---

Features

• Recursive folder scanning

• Processes entire Markdown documentation repositories

• Preview files before modification

• Shows the first matching fence that triggered detection

• Optional backup creation (.bak files)

• Backup discovery and cleanup tools

• Supports:

  • ```text
  • ```markdown
  • ```python id="..."
  • ```json id="..."
  • Any other fenced language

• Preserves original line endings

• Standalone GUI application (Tkinter)

• Linux AppImage distribution

---

What It Does

Transforms:

```text id="a1b2c3"
hello
```

Into:

```
hello
```

And transforms:

```json id="cdctex"
{
  "hello": "world"
}
```

Into:

```json
{
  "hello": "world"
}
```

---

Rules

text and markdown

Language identifiers are removed entirely.

Example:

```text id="abc"
example
```

becomes:

```
example
```

And:

```markdown id="xyz"
example
```

becomes:

```
example
```

---

Other Languages

The language identifier is preserved.

Only the id="..." attribute is removed.

Example:

```python id="abc123"
print("hello")
```

becomes:

```python
print("hello")
```

---

GUI Usage

Launch the application:

python clean_md_v2.py

Workflow

1. Select a file or folder.
2. Click Scan.
3. Review files that will be modified.
4. Click Clean.
5. Optionally create backups.
6. Optionally remove backups after verification.

Backup Management

When enabled, backups are created as:

README.md.bak
guide.md.bak
api.md.bak

The application can:

• Find backup files
• Display backup files
• Delete backup files

---

Linux AppImage

A standalone AppImage is available from the GitHub Releases page:

https://github.com/TSltd/clean_md_fences/releases/tag/v1.0.0

Download:

MarkdownFenceCleaner-x86_64.AppImage

Make executable:

chmod +x MarkdownFenceCleaner-x86_64.AppImage

Run:

./MarkdownFenceCleaner-x86_64.AppImage

No Python installation is required.

---

Building from Source

Create a virtual environment:

python3 -m venv venv
source venv/bin/activate

Install build tools:

pip install pyinstaller

Build:

pyinstaller \
  --onefile \
  --windowed \
  --name MarkdownFenceCleaner \
  clean_md_v2.py

Result:

dist/MarkdownFenceCleaner

---

Example

Input

```text id="a1b2c3"
Define runtime semantic contracts
```

```json id="cdctex"
{
  "capability": "reassure"
}
```

Output

```
Define runtime semantic contracts
```

```json
{
  "capability": "reassure"
}
```

---

Requirements

Running from Source

• Python 3.8+

Standalone AppImage

No dependencies required.

---

License

MIT