rebolek / lest Goto Github PK

Essentially this means that certain & in the source text should not be escaped as & in the HTML.

It's a little tricky and I don't know how CommonMark is going about it. What if you write I enjoy Q&A; it's fun? That works, but if you write I enjoy Q& it's fun StackOverflow won't escape (when writing questions). It's as if it has a list of what the valid entities are, and will only keep from escaping if it's one of those.

(I know there's lots of Markdown stuff, just mentioning things that I use. e.g. this breaks δ and → on the nocycle page)

Current implementation holds templates in different complex than other variables. This is problematic for commands, debugging, etc.

downloaded R3 from http://www.rebolsource.net/

after do %lest.reb

I got that error: compile-rules word is not bound to a context

I just discovered Lest and I think it's awesome. However, I'm thinking about using Cheyenne on the backend and it runs on Rebol 2. Is there any technical issue that prevents Lest from being ported to Rebol 2? Thanks!

• FROM should support TIMES
• full OFFSET support
• WITH without TIMES (return NONE to stop the loop)

Following code should work:

x: either [condition]
    [div "true"]
    [div "false"]

[[1]](http://example.com) for instance is not recognized as a link, and is shown as plain text.

(Both GitHub and Stackoverflow allow brackets in links, and I cannot tell you how they decide what is o.k. or not, but CommonMark spec may have something to say on it.)

Some thing cannot be parametrized yet, for example value in form elements. See hidden for example kludge implementation.

x: "some value"
hidden name x

Generally speaking in markdown, you can use markdown inside of a link (as long as it isn't a link, I guess?)

[**hostilefork**](http://hostilefork.com) => hostilefork

Currently under md.reb, that produces a link with the text **hostilefork**

Following code:

b ("hello") span "world"

produces wrong result:

<b>hello<span>world</span></b>

Should be:

<b>hello</b><span>world</span>

>> markdown "***FOO***"
== "<strong>*FOO/**"

(What it's supposed to do is be both bold and italic.)

Generally it seems there should be some sort of protection or assert against making unpaired tags in a single call to markdown, as that allows errors to propagate pretty far in the document (perhaps some way of designing it so that can't happen?)

submit with action "info" "Informujte se"
p "Ano, mám zájem o zasílání informací o vašich kurzech."

This requires enclosing SUBMIT into DIV

Some values are not initialized when calling LEST, so creating second page throws an error and %lest.reb must be reloaded to work again. This is not problem with simple code snippets.

SET css-path http://myweb.cz/css/
STYLESHEET css-path/my-style.css

If you write: "X -- *(It marks the spot!)*"

The markdown rules will turn that into: "X <hr />(It marks the spot!)*"

Expected it just to italicize "(It marks the spot!)"

This result comes from the horizontal-mark rule.

I haven't gotten around to installing bootstrap yet, but there probably is a better way to set these plugin paths

lest.reb

compile-rules.reb

lest.reb

To assign one user word to value of another, following code needs to be used:

x: [b "first user word"]
y: [x]

This should be simplified as:

x: [b "first user word"]
y: x

If I call markdown on "*[hostilefork](http://hostilefork.com)*" I get"<em>[foo](http://example.com)</em>"

Expected the inner link to be processed.

Similar problems happen if the outer style is done with double asterisks for bold style, etc. Backticks should not recognize markdown inside.

(Note: I assume it's probably obvious...but what happens here is that text is captured without then recursing. So what I'm doing as a workaround is taking context that capture into text, which then put text into the HTML, and replacing it with either xml? [markdown/xml text] [markdown text]...with a recursion-safe modification to markdown. But I'm probably not doing it how you'd want it done.)

See following code:

x: [b "test] 
x: [i "test"] 
x

This returns <b>test</b> instead of <i>test</i>.

RADIO uses custom rule instead of INPUT-PARAMETERS which can be source of problems.