abbr not supported in markdown

[oupala]

markdown has been recently added as a new feature of this theme, it works well and is a pleasure to use.

Unfortunately, the markdown parser does not support the abbr, which identified as a bug in marked: markedjs/marked#27

The issue suggest to use another markdown parser known as markdown-it: markdown-it.

I let you decide what to do.

[mudassir0909]

oh, I didn't know about this, my reply would be same as here :)

[oupala]

@mudassir0909 it seems that you closed the issue a bit too fast!

I tried to use abbr, but I did not succeeded due to some limitations. Let me explain them.

First, I think that each string is parsed from markdown to html separately. This means that you can't define the abbrs somewhere in the footer and use them anywhere else in the resume.

Second, my tests show that the html generated by markdown-it is escaped by the elegant jsonresume theme. It should produce the following html:

The <abbr title="Hyper Text Markup Language">HTML</abbr> specification.

but the produced html is the following:

The <abbr title="Hyper Text Markup Language">HTMLspecification.

We can see that < is converted to <, > to >, </abbr> is stripped (including the trailing space).

abbr is not usable at this time, we should reopen this issue.

[mudassir0909]

Well, I tested it with the following string & it worked perfectly.

 "*[HTML]: Hyper Text Markup Language.\nThe HTML specification is maintained by the W3C."

Added the above string inside the highlights array & it worked as expected

[mudassir0909]

And yes, each string is parsed separately because there is no other way to globally compile everything & put them back to where they came from. For example, work[0].highlights is an array, you can't just compile the entire array, you have to individually compile each item in the array unless you have any other idea for the same.

Also can you share what is your resume.json structure so that I can get the idea of what you're trying to do

[oupala]

I tried with the exact same string as you:

 "*[HTML]: Hyper Text Markup Language.\nThe HTML specification is maintained by the W3C."

and I get the following generated html:

<li class="mop-wrapper">
    <p>@@@@~*[HTML]: Hyper Text Markup Language.
        <br>The HTML specification is maintained by the W3C.~@@@@</p>
</li>

Of course, the abbr is not displayed.

When I use this string:

"this is HTML !\n*[HTML]: Hyper Text Markup Language\n"

and I get the following generated html:

<li class="mop-wrapper">
    <p>@@@@~this is <abbr title="Hyper Text Markup Language">HTML</abbr> !</p>
    <p>~@@@@</p>
</li>

This time, the abbr is not displayed, but there is still some strange strings: @@@@~ and ~@@@@.

It seems that the final \n is important.

[mudassir0909]

strange! let me check this weekend

[oupala]

Hi @mudassir0909, did you had time to test it?

[mudassir0909]

Yes I did test it & this is what I'm seeing

for the json

Maybe you could share your jsonresume profile? that'd be more helpful

[oupala]

I tried with the markdown that you copied:

      "summary": "conception and QA",
      "highlights": [
        "this is HTML !\n*[HTML]: Hyper Text Markup Language\n",
        "conception of a distributed system"

And the result is showing some annoying @@@@~ and ~@@@@ strings:

I don't have any idea on where it could come from. I'm using debian jessie with a standard nvm install of npm.

Searching @@@@~ string is hard with Google...

Any help would be gladly welcome!

[oupala]

Further testing shows me that the strange strings are due to the final \n.

Without the final \n, the abbr is not rendered, but the @@@@~ disappear.

[mudassir0909]

Ah! Maybe compatibility issue with the OS, if you have a VM, can you try publishing the same from ubuntu & see if the issue still persists?

Also, if possible open up node console & include mardown-it, markdown-it-abbr. Try to convert the markdown in the console & if the issue is at package level, maybe we can raise the issue in their github repository

[oupala]

The problem does not come from debian: I've just tested with Ubuntu Xenial (a real Ubuntu, not a VM) and I get the same result. What OS are you using?

Where does that come from?

I thought it could come from the encoding:

$ file resume.json 
resume.json: UTF-8 Unicode text, with very long lines

What is your encoding?

[oupala]

Could you help me test markdown-it and markdown-it-abbr from the node console? I would like to test if only I knew how to do it.

[mudassir0909]

Assume you have node installed on your machine, you can open node console using the following command

node

Once you enter the console you could do

var markdown = require('markdown-it')({
    breaks: true
}).use(require('markdown-it-abbr'));

markdown.render("this is HTML !\n*[HTML]: Hyper Text Markup Language\n")

See if the output exports special characters

P.S. Ensure you have markdown-it & markdown-it-abbr installed globally

[mudassir0909]

By the way here is my encoding

$ file resume.json
resume.json: ASCII English text

I'm on OS X El Capitan

[mudassir0909]

Also, did you publish your resume somewhere? can I have a look at it?

[oupala]

Here is the result of my test:

>node
> var markdown = require('markdown-it')({
...     breaks: true
... }).use(require('markdown-it-abbr'));
undefined
> markdown.render("this is HTML !\n*[HTML]: Hyper Text Markup Language\n")
'<p>this is <abbr title="Hyper Text Markup Language">HTML</abbr> !</p>\n'
> .exit

So, it seems to work!

As my file is in UTF-8 and yours is in ASCII, this could explain the different behaviour. Could please add an é character somewhere in your resume, save it in UTF-8, and try to render it again?

My resume is not published anywhere at this time (it is not finished yet as there is still some bugs), but I can send it to by private mail.

[mudassir0909]

shoot me an email at mudassir.dev @ gmail.com, attach your resume.json file, it'll make things easier for me

[oupala]

It might be due to differences in system's locales.

Here are mines:

>locale
LANG=fr_FR.utf8
LANGUAGE=
LC_CTYPE="fr_FR.utf8"
LC_NUMERIC="fr_FR.utf8"
LC_TIME="fr_FR.utf8"
LC_COLLATE="fr_FR.utf8"
LC_MONETARY="fr_FR.utf8"
LC_MESSAGES="fr_FR.utf8"
LC_PAPER="fr_FR.utf8"
LC_NAME="fr_FR.utf8"
LC_ADDRESS="fr_FR.utf8"
LC_TELEPHONE="fr_FR.utf8"
LC_MEASUREMENT="fr_FR.utf8"
LC_IDENTIFICATION="fr_FR.utf8"
LC_ALL=

What are yours?

[mudassir0909]

LANG="en_US.UTF-8"
LC_COLLATE="en_US.UTF-8"
LC_CTYPE="en_US.UTF-8"
LC_MESSAGES="en_US.UTF-8"
LC_MONETARY="en_US.UTF-8"
LC_NUMERIC="en_US.UTF-8"
LC_TIME="en_US.UTF-8"
LC_ALL="en_US.UTF-8"

[mudassir0909]

And even then this issue is probably related to markdown-it-abbr npm package & has nothing to with this them, I feel

[oupala]

As explained in my previous comment, everything seems to work very well with markdown-it and markdown-it-abbr.

Do you really think I should try to have a deeper look at markdown-it-abbr?

[mudassir0909]

As mentioned in the email, it could be due to hackmyresume. If you are not having any issues when you host your resume(jsonresume) locally then it's probably hackmyresume

See https://github.com/mudassir0909/jsonresume-theme-elegant#contributing on how to host your resume locally

[oupala]

It works very well with jsonresume-theme-elegant self-hosted. But how do you manage to do that without resume-cli?

I'll have to dive into hackmyresume to find out where it does come from... (see hacksalot/HackMyResume#168)

Thanks a lot for your patience!

EDIT: replace json-server by resume-cli

[mudassir0909]

no problem :) closing the issue then

[mudassir0909]

how do you manage to do that without json-server?

I didn't get this part

[oupala]

I edited the previous comment.

How do you manage to render the resume without resume-cli?

[mudassir0909]

Aha, this even I've no clue, although this might help

https://github.com/mudassir0909/jsonresume-theme-elegant/blob/master/serve.js#L47

We use this script to render the theme locally, you could do the similar thing

[oupala]

So... What is resume-cli for if jsonresume-theme-elegant is a standalone tool?

I mean, it's more than a theme, it is a fully usable standalone tool!