Improve .dl file layout

[emilkovacev]

Is your feature request related to a problem? Please describe.

The .dl file layout works well for small datasets, but with larger datasets with lots of text and many parameters, it begins to fall apart.

Describe the solution you'd like

A new .dl filetype should be developed. The current parser should be moved to a new file, and expanded to include the following:

• [ ] Check .dl version number in header example header: version: 1.5
• [ ] Use .dl version number to redirect to a parser for that specific version
• [ ] For the v2 parser, develop the new layout, as specified below:

version: 2.1
---
1 | Castle
node_type: location
text-interaction: You run from the dragon.

2 | Health Potion
node_type: potion
text-interaction: You take the health potion
health: +25
xp: +100
coins: -5

3 | Dungeon
node_type: Location
text-interaction: You take the health potion
health: +0
xp: +0
coins: +100
---
1 -> 2
2 -> 3
1 \-> 3

[WeirdAlex03]

I imagine we'll probably be updating the file layout rather frequently this early in development. I think it would be better to have the file version in two parts, as in 0.2. This would follow the standard major.minor pattern, with all pre-release changes as 0.x. The 3-part version number, major.minor.bugfix is also an option, but I don't think we'll need the third class for the file type. Maybe for the parser, but not for the file itself.

[emilkovacev]

Good point! I'll revise the file format to match your suggestion

[emilkovacev]

I would also like to add parser specifications:

• v2.x will have 3 sections to the .dl file:
  • Section 1: Metadata (version number, game title, etc.)
  • Section 2: Game (node_type, node_alias, parameters)
  • Section 3: Connections (list of connected node_aliases)
• Each section is separated by three dash characters, ---
• Empty newlines separating sections should be ignored
• Empty spaces, tabs, newlines separating data should be ignored
• Spaces between words should not be removed, spaces around sentences should be stripped
• Headers with values that start with either + or - should be parsed into the appropriate numbers. Positive numbers must always start with +
• Comprehensive error checking for files should be employed