8 releases

0.3.0 May 30, 2020
0.2.5 May 8, 2020
0.2.1 Apr 30, 2020
0.1.0 Apr 28, 2020

#1110 in Parser implementations

MIT license

170KB
3.5K SLoC

tweep

Tweep is a parser for the Twee 3 interactive fiction format

What is Twee?

Twee is "a text format for marking up the source code of Twine stories." It is an alternate way to produce interactive fiction stories for the Twine 2 platform, using plain text files instead of a graphical environment. The specification for Twee 3, supported by tweep, can be found here

Goals

The goal of tweep is to provide a fully standards-compliant Twee 3 parser that provides helpful warning and error messages for bad practices and common mistakes, which can be used as a backend for a compiler as well as other novel applications of the Twee 3 format.

What it's not

  • A compiler - while a corresponding compiler front end is in the works, this is not it. tweep only produces rust objects, not html
  • A Twee v1 or v2 parser - currently, there are no plans for supporting any version of the Twee specification other than Twee 3

Getting started

To use tweep in your Rust project, simply add the following to your Cargo.toml:

[dependencies]
tweep = "0.2"

For basic parsing, the main entry point into tweep is through the Story struct, which provides utility methods for parsing out a complete story from a String or a Path representing a file or directory. When given a directory, tweep will parse all files ending in .tw or .twee and merge them into a single output story. For more advanced parsing, such as if the tags or metadata attached to a special passage is needed, StoryPassages provides the same interface, but provides Passage objects in places where usually unnecessary information is stripped out.

Examples

use tweep::Story;
let input = r#":: StoryTitle
RustDoc Sample Story

:: StoryData
{
 "ifid": "D674C58C-DEFA-4F70-B7A2-27742230C0FC",
 "format": "SugarCube",
 "format-version": "2.28.2",
 "start": "My Starting Passage",
 "tag-colors": {
   "tag1": "green",
   "tag2": "red",
   "tag3": "blue"
 },
 "zoom": 0.25
}

:: My Starting Passage [ tag1 tag2 ]
This is the starting passage, specified by the start attribute of StoryData.
Alternately, we could remove that attribute and rename the passage to Start.

It has tags and links to:
 [[Another passage]]
 [[Here too!|Another passage]]
 [[A third passage<-And a different passage]]

:: Another passage {"position":"600,400","size":"100,200"}
This passage has some metadata attached to it

:: A third passage [tag3] { "position": "400,600" }
This passage has both tags and metadata. The size attribute of the metadata
isn't overridden, so it will be set to the default value.
"#.to_string();

// Parse the input into an Output<Result<Story, ErrorList>>
let out = Story::from_string(input);
assert!(!out.has_warnings());

// Move the Result out of the Output
let (res, _) = out.take();
assert!(res.is_ok());

// Get the Story object
let story = res.ok().unwrap();

// StoryTitle and StoryData contents are parsed into special fields
assert_eq!(story.title.unwrap(), "RustDoc Sample Story");
assert_eq!(story.data.unwrap().ifid, "D674C58C-DEFA-4F70-B7A2-27742230C0FC");

// Other passages are parsed into a map, keyed by the passage name
assert_eq!(story.passages["My Starting Passage"].tags(), &vec!["tag1", "tag2"]);
let metadata = story.passages["A third passage"].metadata();
assert_eq!(metadata["size"], "100,100");
assert_eq!(metadata["position"], "400,600");

License: MIT

Dependencies

~0.7–1.4MB
~33K SLoC