Print branch

[MareoRaft]

Print

This branch which I haven't even started yet allows a user to print a node or nodes or a curriculum. The idea is simple. Browsers are already capable of printing what you see on an HTML page very effectively. Therefore, an easy and straightforward solution is to produce a basic HTML page from a node (or nodes, or curriculum) that is almost the same as we already have, but with a few things simplified.

Instead of using blinds, we can simply have headers/subheaders using <h2>. There is no need for node-template icons such as "learn" and "edit" and "save". Other than that, everything is pretty much identical to what we already have, including the running of mathjax to get pretty math.

The HTML that we produce (after a user clicks "printable page") is opened in a separate page, just like I did in Andrew's groceries repository.

[MareoRaft]

This is now working AFAIK. A little cleanup, then merge with master. Merged.

[MareoRaft]

I FORGOT. We still need to

1. run MARKED on the server-side in order for this to be completely finished.
2. say "thm" or "defn", etc instead of "description" for the description attribute.

[MareoRaft]

2. finished. Still need 1.

[MareoRaft]

Let's also add

3. the number, but smaller font (use span and CSS) immediately below the title(name).

[MareoRaft]

3. finished. Still need 1.

[MareoRaft]

We shall now run ./Markdown.pl (in server/lib) on an input

https://github.com/JanSchulz/knitpy

i'm so lost. I was just trying to use Markdown.pl directly, but tempfile module is misbehaving. the Markdown reads the file, it reads it as blank :( Opened a SO question about it:

https://stackoverflow.com/questions/45517816/python-write-to-temp-file-and-read-from-command-line

[MareoRaft]

This markdown to html thingy is apparently more widely used, and I should now try it out: https://github.com/trentm/python-markdown2

[MareoRaft]

It seems to be the same behavior as the JS marked library, so let's move forward with it! Now remember that we customized markdown, so I have forked the library and I must make the following customizations:

4. SINGLE underscores do NOTHING instead of ITALICS.
5. DOUBLE underscores do UNDERLINING instead of BOLD

This probably needs to be done in python-markdown2/lib/markdown2.py at def _do_italics_and_bold(self, text).

[MareoRaft]

This is WORKING. 4 and 5 resolved. But markdown2 has this one side effect where displaystyle math delimited by

\[ and \]

no longer gets processed by mathjax. Somehow, markdown causes the \[, for example, to be relaced by [ in the ultimate HTML.

I started a GitHub issue for this: trentm/python-markdown2#266

[MareoRaft]

Just for kicks, I was able to get the original Markdown.pl to work. It turns out to have the same behavior as marked2. So this weird behavior above is actually standard. It must actually be marked (in JS) that behaves differently than the "standard". Anyway, the marked behavior is what we want... :(

[MareoRaft]

If I want to change the behavior of marked2 manually, I think the relevant code is

# Table of hash values for escaped characters:
g_escape_table = dict([(ch, _hash_text(ch))
    for ch in '\\`*_{}[]()>#+-.!'])

and

def _encode_backslash_escapes(self, text):
    for ch, escape in list(self._escape_table.items()):
        text = text.replace("\\"+ch, escape)
    return text

[MareoRaft]

Okay so I just replaced \[ and \] with an escape sequence before running markdown, and then changed it back after markdown. Hope there are no issues.