env - is this just per-render options? about markdown-it HOT 5 CLOSED

env is per render, but for different things - transfer data between stages. In theory, it's possible to add flags there, but not sure that it's a good idea to do it now.

If you have limited combination of settings - create as many instances as can be, and select appropriate, based on you state (admin permissions in your case). Cache can be done with lodash _.memioze for example.

from markdown-it.

https://markdown-it.github.io/markdown-it/#MarkdownIt.set set can help with options, but can affect speed (see note). If you don't need highest performance - use it.

from markdown-it.

There are three objects that affect parsing behavior:

1. state (inline, block) - those properties are bound with a string which parser is currently processing
2. env - those properties are bound with a current execution of the parser
3. options (or that parser class instance) - those properties are for the entire parser, and they stays the same for multiple executions

So each time you call .render(), one env object gets created, and multiple state objects, but parser class (and options) stay the same across executions.

Now, if you want to enable html in one cases, but not in others, I would probably create two separate parser instances, and use one or the other. But yes, you can use just one parser and make an env flag for this which you would set each time .render() is called.

from markdown-it.

ok having multiple parsers is okay. i just don't know wtf env is. docs are so dense man! haha

from markdown-it.

When text is parsed, the main data are tokens. Those are passed between stages in arrays. But there are some kinds of metadata, not fitting in this concept. For example - descriptions of references, abbreviations, footnotes. Those are passed via env.

In real world you will never use env directly, if you are not plugin writer. Consider it as strange shit, reserved for the future :)

from markdown-it.