---
layout: page
title:  "Mind map generation with Python"
description: "Step 1"
date:   2019-07-15
tags: projects frc
alises: [https://retrylife.ca/blog/2019/07/15/mindmap]
---

While working on an assignment with [Coggle](https://coggle.it) today, I noticed an interesting option in the save menu. *Download as .mm file*. Having rarely worked with mind maps before, and only doing it online, it never occured to me that someone would have a file format for it. So I took a look.

## What is a .mm file?
It turns out, a `.mm` file is just some XML describing the mind map. Here is a simple mind map:

![Simple Mind Map](/images/posts/mindmap/mindmap-simple.png)

And again as a `.mm` file:

```xml
<map version="0.9.0">
    <node TEXT="Master Node" FOLDED="false" POSITION="right" ID="5d2d02b1a315dd0879f48c1c" X_COGGLE_POSX="0" X_COGGLE_POSY="0">
        <edge COLOR="#b4b4b4"/>
        <font NAME="Helvetica" SIZE="17"/>
        <node TEXT="Child branch" FOLDED="false" POSITION="right" ID="f72704969525d2a0333dd635">
            <edge COLOR="#7aa3e5"/>
            <font NAME="Helvetica" SIZE="15"/>
            <node TEXT="Children 1" FOLDED="false" POSITION="right" ID="c83826af506cae6e55761d5c">
                <edge COLOR="#7ea7e5"/>
                <font NAME="Helvetica" SIZE="13"/>
            </node>
            <node TEXT="Children 2" FOLDED="false" POSITION="right" ID="47723a4d0fb766863f70d204">
                <edge COLOR="#82aae7"/>
                <font NAME="Helvetica" SIZE="13"/>
            </node>
        </node>
    </node>
</map>
```

Neat, right?

## What can we do with it?
I have not done much research about this because I wanted to work all of this out on my own. But I know one thing as a fact: working with XML sucks (especially in Python). I decided that this would be much better if I could load `.mm` files as JSON. This would allow easy manipulation and some cool projects.

## My script
Like everything I do, I made a script to play with these files. 

It's pretty simple. First, It loads a `.mm` file, then parses it into a `list` of `xml.etree.ElementTree.Element`.

```python
raw_mm = ""

with open(args.file, "r") as fp:
    raw_mm = fp.read()
    fp.close()

xml = ET.fromstring(raw_mm)
```

The parsed `list` is then passed into a recursive function that constructs a `dict`

```python
def xmlToDict(xml):
    output = []
    for elem in list(xml):

        if "TEXT" not in elem.attrib:
            continue
        
        name = elem.attrib['TEXT']
        json_element = {"name": name}

        try:            
            json_element["children"] = xmlToDict(elem)
        except:
            continue
        
        # Detect node type
        if json_element["children"]:
            json_element["type"] = "branch"
        else:
            json_element["type"] = "leaf"
            del json_element["children"]
        
        output.append(json_element)
    
    return output
```

Finally, the `dict` is written to a file with `json.dump`

```python
json.dump(mind_map, open(args.file + ".json", "w"))
```

The whole script (with comments) can be found on my [GitHub account](https://gist.github.com/Ewpratten/0d8f7c7371380c9ca8adcfc6502ccf84#file-parser-py).

## The output
Running the `.mm` file from above through the script gives:

```json
[
    {
        "name":"Master Node",
        "children":[
            {
                "name":"Child branch",
                "children":[
                    {
                        "name":"Children 1",
                        "type":"leaf"
                    },
                    {
                        "name":"Children 2",
                        "type":"leaf"
                    }
                ],
                "type":"branch"
            }
        ],
        "type":"branch"
    }
]
```

## The next step
This script just translates a `.mm` file to JSON. Nothing else. Next, I want to convert this to a library, and add a JSON to `.mm` function as well. This leads into my ultimate goal for this project.

I want a script that I can drop in the root of any project to build a [Gource](https://gource.io/)-style visualization of the folder structure. This will give me a way to make cool visualizations for lessons on the robotics team. On top of the folder visualization, Coggle's new flowchart feature can be used to generate graphical representations of @frc5024's codebases. This could give me an interactive overview of the work being done by our team. 

### Further learning
crm.org has done a great writeup of [Coggle, and some of it's features](https://crm.org/news/free-flowin-mind-maps-with-coggle). If you are looking to learn more about the tool, I recommend taking a few minute to read their post.