From Scrivener to any WP, then back again


I’m currently working with two other writers, one using OpenOffice, and the other Microsoft Word.

I send them an RTF file ‘compiled’ from Scrivener, and receive a DOC document back.

I want to do my edit back in Scrivener, but it seems that this means that I have to go through the whole ‘Split at selection as title’ and ‘Rename document’ process each time.

Is there any way to exchange text with users of other wordprocessors, and have it read by Scrivener back as a tokenized document with all the original titles at back at their place?



No, there’s no way of doing this. I played with the idea of something like this for 2.0, but it would be incredibly difficult, and dependent upon other writers keeping certain things intact in the document.

All the best,

Hi Keith,

Just a last try. Assuming all participants to the collaborative work are aware of the special text to be left untouched, can’t reserved tags be used for this?

If, when importing an RTF file, Scrivener would find something like this:

<$title:This is the name of my doc no. 1>

couldn’t it consider the beginning of a new document in the Binder, with the name “This is the name of my doc no. 1”?

At the same time, couldn’t the tags (optionally) be saved when compiling into an RTF file?

Best regards,

I am not a Multi-Markdown user, but wouldn’t MM be the perfect solution for this? May the MM goddess can weigh in with a suggestions.

Paolo - things are nowhere near as simple, I’m afraid. Documents use internal tags, then you have notes, synopses and so on to think about too. If the order is changed in the document it could be next-to-impossible for Scrivener to translate that back into a decent binder order, especially if some items have ID numbers removed or inadvertently changed, or parent sections removed and new sections added… I spent a couple of weeks trying to come up with something that worked and everything hit major problems or would require a couple of months of coding. Thus, it is on the back-burner, though not entirely abandoned.
All the best,


This would work if we didn’t use other RTF features, like text highlighters/colors, footnotes, and various formattings.

Could we make our text formatting simpler? Let me see:

  • Text highlighters/colors are used to show which author wrote a passage or notes. Tags would not be as effective when browsing through a document. Maybe there is an easy way to convert MM to RTF, and viceversa?

  • Footnotes are the easiest way to separate narative text from notes. Notes should be footnotes, since jumping from the narrative to the endnotes would not be very practical while revising a text. I’m starting to use RTF comments more and more, since they are equivalent to footnotes in Scrivener, and are way more handy in Nisus Writer; shame OpenOffice and Word seem to understand them in a different way.

  • Various font faces are something my director seems not to be able to stay away from. I don’t know how to treat him. Ya 'now, when they write biographies of film directors, there is always that too over-precise screenwriter to blame of, as opposed to the director’s uncontainable vitality…


I’m sure this misses some nuances Keith mentioned, but it might be quite a useful start anyway.

  • Export binder structure into folder structure.

  • Prefix number to the sub-folders (e.g. prefix to their name) based on their order
    01 - This is my Section 1.1
    02 - This is my Section 1.2

  • Document content dependent on export choices e.g. text, mmd, rtf, …

  • Docs may be edited, including structure using folder structure + numbering

  • Import back following same rules

I know I would love to get the structure/order part of this when I do MMD round-trips.

  • Sophie

I don’t know if this will be of any help, but here goes…

One of the strengths of Word (it does have some) is that it will Find/Exchange styles or formatting. It is therefore possible to make Word find a section heading or title that has been formatted in a particular way and insert text before it while retaining the original text. Perhaps this might go some way to solving the problem. I certainly used a similar technique to take an OmniOutliner file and convert it into a MultiMarkdown file which I imported into Scrivener.

Best of luck with it.

Martin BB.

Hi Martin,

May you elaborate a little on your idea? I would love to know how to use this formatting strategiy to make Scrivener split the imported RTF file into separate documents in the Binder. Is there a particular sequence of conversions involved?


Molto volontieri.

It is a bit long, and I don’t know how you might deal with footnotes. You may have to find a work-round. Highlighting you may be able to do – you will need to look at the documentation for MultiMarkdown. If you decide to use the procedure, you could record it as a macro in Word 2004, which I believe you can’t do in Word 2008, but you might be able to use AppleScript instead. I’m using Word 2004 on an English system, but I trust you will be able to find equivalents on an Italian system. The (example) procedure is:

Create test document in MS Word.

Type and format headings. (You can use any formatting or fonts, as Word will Find/Replace these, but for the sake of example I will use the built-in heading styles. Easiest way to use these is to select your first heading and press Cmd-Opt-1. If you still have default settings this formats the text with the style for Heading 1. Cmd-Opt-2 gives Heading 2, Cmd-Opt-3 gives Heading 3. This only works for the first three heading styles. Other heading styles have to be selected via the menu.)

Type in body text using Normal style, or whatever other style you want.

Picture 1.png
Save test document as normal MS Word document (might as well save it, in case of crashes or other mishaps).

Go to the “Find and Replace” dialogue (via Edit menu on the English system, or via Cmd-Shift-h – no idea what it is on an Italian system [Cerca?]).

Do not type any text in the first slot (“Find what:” on the UK system).

Go to bottom of dialogue window and click on “Format”.

Select “Style…”.

Select “Heading 1”

“Find what:” slot should now have “Format: Style: Heading 1” or some such under it.

Move to “Replace with:” slot.

Type "# " (that is: hash followed by space).

At bottom of dialogue box, click on “Special”.

Select “Find What Text” (alternatively, you can just type “^&” [that is: caret followed by ampersand] into the “Replace with:” slot).

Go back to “Special” and choose “Paragraph Mark” (or type “^p” – caret followed by letter p). It is important to have a blank line after every heading or the import into Scrivener won’t work properly.

Picture 5.png
Click on “Replace All” (unless you are feeling nervous and want to do it one at a time).

Word will prepend a hash and a space to every piece of text formatted as “Heading 1”, and append a paragraph break.

Repeat this process by “Finding” text formatted as Heading 2 and using two hashes, etc., etc.

Body text can be left as it is.

The resulting file should be saved as a plain text file.

The problem now is that Scrivener seems to want the text file to be in UTF-8 encoding with Unix line endings. I get this by opening the file in TextWrangler (free and very useful at times), going to the Edit menu and choosing “Document Options”. You can then choose UTF-8 and Unix line endings, and Save.

In Scrivener, import as MultiMarkdown file. Result as below:

Picture 8.png
I hope this may be of some help.

Best wishes,

Martin BB.

PS: un saluto a Leopardi.

PPS: Someone posted a script on another thread which does the same sort of thing more quickly (see download/file.php?id=307) but I think it only looks for the Heading styles. You would need to adapt it if you need to Find/Replace text with other sorts of formatting, so the procedure I have described above may help you to work out how to do it.

Could you elaborate on this, please? Structure should remain intact when importing from an MMD text file (or are you specifically talking about multiple files?). You should be able to compile an MMD file from Scrivener, then drag it back in to build the same structure of files and containers.

All the best,

Specifically multiple files.

The ordering within the binder structure gets lost on the round-trip. If you pre-fix a serial number on the export, and parse that serial number on the import, perhaps round trip could become loss-free?