Consider loose parsing modes about extwee HOT 6 CLOSED

Ah, my mistake. Twine 2 HTML must have an IFID! felt like a non-negotiable error in the console :)

from extwee.

I'm re-opening this, as I want to revise how parsing of tw-storydata and tw-passagedata HTML elements is handled. Parsing should be looser for both elements and should produce more warnings rather than errors for incoming data following the pattern of "parsing is loose and compilation is strict."

from extwee.

That's a spec issue. IFID is a required attribute.

Only name and IFID are required, though, so I'd be okay, like with with Twine, of defaulting to "Untitled Story" for the name and generating an IFID and allowing a dev to do so on their own via an exposed method.

from extwee.

That's kind of what I mean here--what if a story is parsed with warnings, but you're required to set required attributes yourself before it's re-exported out? e.g. allow a story to be temporarily invalid so that it can be repaired in code.

from extwee.

You can parse Twine 2 HTML without the required attributes. I added the word "Warning" to the existing warnings, but it was already possible:

import { parseTwine2HTML } from 'extwee';

// Without required `name` and `ifid` attributes.
const s = `<tw-storydata></tw-storydata>`;

// Parse the HTML.
// This will generate warnings for missing `name` and `ifid` attributes.
const story = parseTwine2HTML(s);

// Convert back into HTML.
// This will generate a name, Untitled Story.
// This will generate a new IFID.
console.log(story.toTwine2HTML());

from extwee.

Parsing <tw-passagedata> elements with missing ifid or name attributes will no longer throw errors; they will generate warnings and attempt to use default values.

from extwee.