tutorial/001-convert-md-to-substance [master] » node convert.js lorem_ipsum.md
/home/joel/Repositories/tutorial/001-convert-md-to-substance/node_modules/substance-converter/src/pandoc_importer.js:196:15 (topLevelNode )
/home/joel/Repositories/tutorial/001-convert-md-to-substance/node_modules/substance-converter/src/pandoc_importer.js:146:23 (document )
/home/joel/Repositories/tutorial/001-convert-md-to-substance/node_modules/substance-converter/src/pandoc_importer.js:116:17 (import )
/home/joel/Repositories/tutorial/001-convert-md-to-substance/node_modules/substance-converter/src/converter.js:81:34 (convert )
/home/joel/Repositories/tutorial/001-convert-md-to-substance/node_modules/substance-converter/src/converter.js:60:9 (ChildProcess. )
events.js:98:17 (ChildProcess.EventEmitter.emit )
child_process.js:789:12 (Process.ChildProcess._handle.onexit )
{ message: 'Node not supported: t',
  name: 'ImporterError',
  code: -1,
  __stack: 
   [ { func: 'topLevelNode ',
       file: '/home/joel/Repositories/tutorial/001-convert-md-to-substance/node_modules/substance-converter/src/pandoc_importer.js',
       line: '196',
       col: '15' },
     { func: 'document ',
       file: '/home/joel/Repositories/tutorial/001-convert-md-to-substance/node_modules/substance-converter/src/pandoc_importer.js',
       line: '146',
       col: '23' },
     { func: 'import ',
       file: '/home/joel/Repositories/tutorial/001-convert-md-to-substance/node_modules/substance-converter/src/pandoc_importer.js',
       line: '116',
       col: '17' },
     { func: 'convert ',
       file: '/home/joel/Repositories/tutorial/001-convert-md-to-substance/node_modules/substance-converter/src/converter.js',
       line: '81',
       col: '34' },
     { func: 'ChildProcess. ',
       file: '/home/joel/Repositories/tutorial/001-convert-md-to-substance/node_modules/substance-converter/src/converter.js',
       line: '60',
       col: '9' },
     { func: 'ChildProcess.EventEmitter.emit ',
       file: 'events.js',
       line: '98',
       col: '17' },
     { func: 'Process.ChildProcess._handle.onexit ',
       file: 'child_process.js',
       line: '789',
       col: '12' } ] }

When doing conversion within node, sometimes i get this little weird error.

/Users/michael/io/node_modules/substance-converter/src/converter.js:81:40 (convert )
/Users/michael/io/node_modules/substance-converter/src/converter.js:60:9 (ChildProcess.<anonymous> )
events.js:98:17 (ChildProcess.EventEmitter.emit )
child_process.js:789:12 (Process.ChildProcess._handle.onexit )

I wonder if we should rather create an image node type when we import from markdown. Then we can treat a figure as the full-fledged resource-ish thing.

Markdown image captions aren't really used like captions. They primarily serve the purpose of specifying the alt text (e.g. on mouseover). At least that's what they mean to me. ;)

If you specified in your markdown:

# Intro {#intro}

# Summary {#summary}

Node id's would still be artificial: heading_1, heading_2. That becomes a problem when we do deep links like:

http://substance.io/#substance/manual/heading_2

And in an update to the document those are likely to be invalidated (e.g. when someone inserts a heading intro and summary. We should use the explicitly given id's when possible. We need to make sure though that we normalize the ids so they can only have alphanumerics and underscores (we need html id compatibility).

From the Pandoc documentation here

This Markdown input brought up an error:

- Foo
- Bar

- Bla
    - Blupp

E.g., these are inline HTML comments.

How can I convert a table?

i get a error downing that from markdown to json.

.md :

+-------------------------+-------------------------+-------------------------+
| O | W | B |
+-------------------------+-------------------------+-------------------------+
| 1 | 1.1 Make | Md |
| | | LL |
| | | |
| | | •        the |
| | | czxc cxzc, |
| | | czxc of |
| | | czxc czxc |
| | | schoon voor czxn |
| | | czxc. |
| | | |
| | | •        czxc de |
| | | czxc op peil. |
| | | |
| | | •        czxc het |
| | | czxc op czxc. |
+-------------------------+-------------------------+-------------------------+

An interesting document base is arxiv.
We need to evaluate how Pandoc can deal with it.

Here are the available formats for most of the articles: http://arxiv.org/format/1401.0013v2

Already supported by the Substance.Article but not yet considered by the converter (I think).

Since you don't respond to e-mails, I will try to write here.

As I wrote the basic functionality of substance to pandoc converter is implemented. This means that text blocks decomposes to JSON representation of abstract syntax tree. You can try to use it in playground.js and convert substance document to HTML or markdown. There is still no inline converter, but I hope to implement it soon.

Now we have to decide few things:

• Where we should (or we shouldn't) make substance document from JSON. Currently there is Document.fromSnapshot function running in playground.js. I thought that if we should implement export from some substance app, there'll be already substance documents, not plain JSON. Correct me if I was wrong.
• We should think of what to do with other formats which need some directory to output, e.g. with formats which is not working with stdout (docx, pdf etc).

...waiting for your response... :)

I've introduced code and codeblock types to the schema. Construction looks like this:

For convering Inline code annotations (backticks in the markdown)

doc.apply(["insert", {
  "type": "code",
  "data": {
    "source": "text:14",
    "pos": [30, 5]
   }
});

And for code blocks:

doc.apply(["insert", {
  "type": "codeblock",
  "target": ["figures", "back"],
  "data": {
    "content": "function() bla\ndaefdasdfdfasdf"
   }
});

Then we also have images (![name](url)):

doc.apply(["insert", {
  "type": "image",
  "target": ["figures", "back"],
  "data": {
        "large_url": "http://bla.me/bla.png",
        "medium_url": "http://bla.me/large_bla.png"
   }
});