Importing Markdown

In preparation for Scrivener for iOS, I’m trying to get my WIP imported cleanly into Scrivener. There are some things I’m pretty sure I’m messing up at the moment.

I’m using the latest Scrivener for Windows, running under Wine on Linux Mint 18, but this behavior is the same using Scrivener in Windows (via a Windows 10 virtual machines). I’m using the standard novel project file.

My document is a simple Markdown file. It has standard headers: a single hashmark for chapters, dual hashmarks for scenes, triple hashmarks for a third header level. Each scene is terminated with a “#”, which I use as the marker to use to split during formatting. This file runs through Pandoc to create a Word document with properly formatted headers.

When I import the Markdown files to Scrivener, it splits as expected, but each split is simply a plain text file. In the Binder, I have # - Chapter 1, ## - Scene 1, ## - Scene 2, etc. No formatting of headers, no hierarchical division, certainly nothing that looks like the user guide project, where I see no raw text header divisions at all.

When I compile this project to Docx (admittedly without doing anything special to the compiler), I get strange results: the first Chapter starts out with a properly formatted and centered “Chapter 1” with “# Chapter 1” below it, and each following scene having “## Scene 1,” “## Scene 2”, etc.

Perhaps I don’t understand how MultiMarkdown support in Scrivener is supposed to work, but I thought it would know how to import my Markdown formatted file. Do I need to reformat the headings manually? It’s not a problem if I do, but want to make sure there’s not a way of saving myself some work.

Others may have a more helpful answer, but I’ll add my bit. I’ve used markdown for several years in Scrivener, maintaining all of my drafts in markdown code. I’ve never seen or heard of Scrivener able to “import” a markdown file properly, because it just isn’t set up that way. Imported files (my understanding) become .rtf format in the drafts; I think what you’re seeing in the binder from the split names is a good clue; the document structure is changed. Remember, markdown files are simply text files with Markdown codes. They’ll convert to .rtf in Scrivener, but they’ll be used as text files in Markdown conversions–not as .rtf. I’ve also used pandoc, but always with a Markdown text file, with suffix [name].md.txt, to produce an html or pdf or odt format file.

I do a lot of posting to an online story site. My master markdown drafts are kept in Scrivener; and most times I simply copy the Markdown content of a chapter, paste it into a text editor, and upload it as the normal [name].md.txt file to the online site. If for some reason I make edits in the text editor, I then copy that content, and paste it back into the Scrivener draft binder to replace the chapter content. In all cases, the Scrivener .rtf internal format is immaterial when copying & using the chapter file as [name].md.txt .

Since Markdown is so versatile, I rarely use the “compile” feature of Scrivener… but it’s always there and it works very well with the markdown compile filter. It’s a one-way feature. Scriv outputs a markdown compiled product, but it won’t import it properly as a structured document. That’s why I use a copy-paste method with selected chapters to keep Scriv drafts updated for my master copies. It’s really much simpler than it sounds, and incredibly flexible for output.

Graybyrd, this is a very helpful answer, and I appreciate it. It’s a pain to do, but I don’t mind going through the reformat process. I was just concerned that I was missing something, and your reply tells me I don’t. Your additional comments are helpful, as well. I’ve been using plain text (via Editorial on the iPad and SublimeText on Windows and Linux) to write lately, and have my Pandoc output pretty much the way I like it. But I’ve wanted a portable Scrivener, and a bit of reformatting is a small price to pay.

Thanks again.

I’ll volunteer a word about methods & practices. On Linux (a Debian variant distro) I’ve found Scrivener for Linux (the last upgraded version) to be useful; paired with Pandoc and LibreOffice, it’s a powerful suite of writing and publishing tools. Either Scriv or Pandoc can output various formats, including ePub; and Calibre then becomes most useful for eBook tweaking and conversions. With all this, there’s little reason to look beyond Linux for useful tools. Behind it all, no matter a web page, a printed document, or a published novel in ePub format… my master archive of markdown files resides in the Scrivener project, safe and always handy. Everything reverts back to a master .md.txt file in Scriv.

That said, I continue a long-standing love affair with the early-days system of Mac with OS-X 10.4.11 “Tiger” and its insanely useful suite of tools. Scrivener 2.5 is superb; I’ve never lost a byte and it just… works. The only format I use in Scriv is markdown; it compiles beautiful output. But more usually, I simply copy a chapter’s content into a text editor: either TextWranger or BBEdit or TextMate. Why? Well, TW and BBE have internal linkage to Markdown conversion scripts similar to Pandoc… outputting HTML, etc. BBE is a superb HTML editor for my web work, using Scriv-drafted stuff. TextMate is more powerful with its output choices, with powerful “bundle” scripts. It’s a great alternative to Pandoc. Each editor has its own keyboard and commands ‘personality;’ I use whichever suits.

The on-screen interface and immediate access of OSX, integrated with so many utilities and features, is simply unbeatable. I did find a simple “work space” utility that gives me Linux-like access to multiple desktop workspaces, so I can have research, drafting, file sorting, and publishing output all laid out seamlessly. Windows is hopelessly splintered and useless in this regard; Linux still clings to its CLI roots and without GUI standards between the apps and the OS, it’s hopelessly clumsy. As old as it is, Mac OSX “Tiger” remains light-years ahead.

I’m sure the newer versions of Mac and OSX, with TextWrangler (free) or BBEdit or TextMate are even better; personally, I’ve found the Mac integrated system for writing and publishing to be unsurpassed. Combined with Scriv and DevonThink (document database) there’s no equivalent anywhere in Windows or Linux. There just isn’t. Linux is workable as a writer’s toolkit, but it’s still light-years behind even my old “Tiger” suite of software.

Just one ol’ guy’s opinion from the geriatric wing with the sun-facing rec room at the Restorium. 8)

Thanks for your advice! I love learning how other people work.

Unfortunately, much of the Mac stuff you say is lost on me; though I was introduced to Scrivener on a Mac, I haven’t had one to use for years. I’m on Linux, Windows and now iOS. I’ve been away from Scrivener for a while because of the lack of portability on mobile devices, so I need to find my own workflow. I’m a big fan of Pandoc and Calibre as well.