wishlist: import with separators

I’m using the trial of scrivener, love it, will buy it.

I particularly like the command-k split feature since I collect ideas in a long text file and then scrivener allows me to split up the ideas easily and classify them in folders easily. (and organize them into a narrative easily). All that convenience revolves around the great command-k idea which makes the splitting up of the long text file so easy. And this basic process (collect, divide, organize, present) is so fundamental to the writing process itself.

Its along these lines that I had an idea for the wishlist. Scrivener of course can import a text file. But why not allow the user to define a delimiter – a tab or a series of dashes or whatever – within the text file, so that as Scrivener imports it, it creates index-card breakpoint at those delimiters automatically?

You see the huge advantage of this, I’m sure. So a person could bring in a file they’ve been working on, and just put (for example) a series of dashes between each chunk or each idea. Scrivener imports the file but automatically imports the chunks as separate entries (based on the defined delimiter).

Hours of work saved. And scrivener saves the day again.

What do you think? Isnt that a fairly simple (and pretty intuitive) feature for the import function in scrivener?

I’m fairly sure I saw it announced that this would be a feature in the next version, precisely the way you described it.

For the time being, are you aware of MultiMarkdown? I’m not suggesting you adopt the system, but if you are using text files you aren’t worried about losing formatting anyway. To have a document automatically split up, you need only insert a single hash mark in front of the titles. If you already have a document split up with dashes or something, you could do a search and replace. For example, search for:

----------------------------------------------------------------

Replace with:

# Chapter break

Then import the text file as a MultiMarkdown file (it’s in the File menu).

If you want to get more fancy you can even import entire hierarchies. The number of hashes you put in front indicates the depth. So:

[code]# Level one of the Binder

Bunch of text…

Level two of the Binder

Bunch of text[/code]

Would create two text files, the second one beneath the first.

you cant see me right now, but i’m doing handstands of joy. :slight_smile: I hope fervently that what you say is true.

Meanwhile, I’ll look into multimarkdown… thanks very much for the tips and pointers.

Yes, exactly this feature is coming with 2.0.
All the best,
Keith

hi guys - so I finally had time to try this, and ran into a problem.

So I have an rtf file (called test.rtf).

In it, I just have the following: (without the dashed lines of course)


note 1

text of note 1

sub note 1

text of subnote 1

When I try to import it as a multimarkdown file, scrivener says: “nothing to import, file is not a valid multimarkdown file”. What am I missing?
thanks!

I don’t use multimarkdown, but I have a feeling it has to be a .txt not a .rtf … I am willing to be told I’m wrong though!

Mark

hmmm, i tried txt, no luck!
i’ll try html too, but I think i must be doing something very obviously wrong.

MultiMarkdown files should indeed be plain text and not RTF - it definitely wouldn’t work as RTF.

Also, in 1.x versions of Scrivener (this is fixed for 2.0) you need a line break between titles and text:

# note 1

text of note 1

## sub note 1

text of subnote 1

Otherwise the files will import as empty documents. I just tried File > Import > MultiMarkdown File… and imported a .txt file containing the above and it worked fine.

All the best,
Keith

hi kb - your example file worked for me too. It imported into scrivener correctly as an MM import.

Strangely, my test file is as straightforward too, but mine didnt work until I re-typed the second line (see below, the word “todo” is the second line). I re-typed it exactly as is, and then it worked, but if I dont retype the second line, it does not work.

Basically I’m exporting from a pc from an outlining program called ‘the guide’ (on which I have extensive notes from before I got my mac) which exports in rtf. I bring that over to the mac, and open the rtf file in textedit. I convert it (‘make plain text’) in textedit to a txt plain text format. Then I bring it into scriv as a MM import.

The entire text file is reproduced below (without the dashed lines of course). I’ve also attached it to this post.

title 1

todo

title 2

this is second one

Is this an encoding issue? I dont suppose you know what encoding I ought to have for the text file? (ie, utf8? 16? windows latin? mac roman? etc)
Untitled2.txt (51 Bytes)

There seem to be some gremlin characters in there and not a proper return after “todo”. It looks fine when you open it in TextEdit, but if you open it in TextWrangler (free from Bare Bones Software) you will see what I mean. Because there’s not a proper return in there, it doesn’t get recognised properly in Scrivener.

All the best,
Keith

Thanks KB. I’ll try a different conversion process. (And downloading textwrangler :wink:

KB, I’d like to just add a note here – as it says in the scrivener wiki, scrivener is “a conduit” for the writing workflow, and import/export to/from word processors is something scrivener recognizes as important. I think thats great because for me as well scrivener is a part of the workflow and not the entire workflow. I begin my projects in either textfile notes or in word outlines, then ideally I would do the actual composition in scriv (because scriv works really well for that part of the process), and then get the final product back into word.

So for the most part I’m dealing with either TXT/RTF or DOC/DOCX. But I need a way to get ‘discrete index card sized information, in a set heirarchy’, into scriv (in other words, ‘outlines’). Thats where I keep running into hurdles with scriv.

So far I’ve been unable to:
-get a word (doc/docx) outline into scriv (ie, divided automatically into files and folders based on the word outline) (or for that matter out of scriv with the heirarchies intact). As per the discussion above in this thread, I gather thats not going to be something I can do ever. Thats frustrating. I know it cant be helped, but its frustrating. It blocks my ability to “plug” scriv into this workflow.
-get a text (rtf/txt) file outline into scriv. I know MM is supposed to allow that - i’ve had problems (not resolved yet and it seems like some obscure issue which probably isnt worth anyone’s time to really pursue). (Additionally, the MM route is not very convenient even if it did work - since it involves for me 2 or 3 levels of format conversions each time, and since my text tends to be filled with *'s and --'s and other symbols which probably would interfere with MM’s markup, requiring me to rewrite most of my outlines if they’re going to be importable as MM markup). So the MM route for me doesnt seem convenient either, even if I got it to work.
I know the next version of scriv may recognize separators inherently and bypass MM in that regard, and thats great (and I hope it can recognize heirarchy as well as separators!). But thats not due out till end of year-- so once again, frustrating :wink:

So at the moment, despite my admiration for scriv’s UI, and how much it makes the work of composition much easier than ever before, I’m still unable to plug it into my workflow. Its like being offered a bottle of water in the desert and not being given a bottle opener :slight_smile:

Anyway, just venting. None of this is anyone’s fault. I just wish MS Word integration was easier to do and outlines were supported; and I just wish the MM import wasnt so picky about file formats. I wish 2.0 was out already (and that it bypasses MM entirely by natively supporting heirarchy/separator markup in the text, no matter the format of the file itself!). I wish Word had an MM exporter. I wish copying a file from a PC to a Mac wouldnt introduce ghost characters. I wish a lot of things… :slight_smile:

TextWranger ought to have a “Straighten quotes” feature available somewhere, which will fix all of those typographic apostrophes and quotation marks. That said, have you tried just leaving them in? If your MMD workflow stops at import, it really doesn’t matter a whole lot. Those cautions are for further down the MMD chain. Some issues can result in the LaTeX files without a few code tweaks.

I know everyone has their own method, but part of the problem you might be facing is that Scrivener wasn’t really designed to be a plugged somewhere in the middle of a workflow, for the most part, but rather at the source of the stream—where text originates. It has very good import features, including OPML, MMD, and lots of ordinary un-structured formats (and yes it will get better in the future), but really it is meant to be the start of the chain.

I used to work heavily with the outliner in Word, until I found Scrivener. Outlining is so easy, just create a folder, then a document, type a little, hit return, and go to the next point in your outline. You can start with the Corkboard and jot out ideas, scenes, or just fragments of thought. Move them around to create patterns. Select a chunk of text and drag it to the Binder, and it becomes a new document. This way of working is far more intuitive than the top-down progression in Word. As Amber suggests, if you start the writing process in Scrivener, it’s far easier to draft and then later export out to Word.

hi guys - I think what you say is true. If one uses scriv as the starting point, then the hurdles of importing outlines/notes into scriv becomes less of an issue.

However I guess I’d just point out two things in passing, on that point.

One is that scriv, in an ideal world, would be better off if it were truly “plug-in-able” into any workflow. Since every author has their own needs and tools which are usually quite customized based on what kind of research and writing they’re doing. The more versatile scriv is in that regard, the better a product it is for writers. For instance I already use extensive database-type programs in which to collect research long before any writing begins. And that means that whatever tool I use to write, it needs an efficient import function. Even on the mac, for instance, people using devonthink or evernote or tinderbox or etc are sooner or later going to face these hurdles. And lets face it, these are all extraordinarily popular and useful programs for writers and researchers.
I think its good that scriv focus on its main strong points – ease of composition – and that it should dominate that product niche. But then it will need to address the import issue (specifically preservation of heirarchy and separators, and an efficient (one step) way to achieve that).

The other is that, for me specifically, I live in a world that includes both pc’s and macs (a situation that, I think, going forward, is only going to become more common out there). I need the PC not only for legacy purposes, but because in some cases for the academic research I do, its ‘required’ in labs and classrooms. Much of my research collected over the last 10 years is on PCs. So for many reasons I (and many others who own both pc’s and macs) will always need to bring files that originated on PC’s into the mac universe and then export them out again. The pc and mac universes are less and less insular with each passing year.
So again it comes back to versatility and flexibility; the more scriv is able to be be ‘plugged in’ to different workflows, the more useful and useable it will be as a writing tool; the more it embraces cosmopolitanism rather than insularity, the more useful it will be. Just as a general principle.

I’m not saying scriv needs to have 2 dozen importers to achieve this - not at all. Having ONE standard importing format (that preserves heirarchy and separators) would accomplish this. I’m saying tho, that at the moment, scriv does NOT have even one standard importing format that preserves heirarchy and separators (and also does not have an exporting format that does).

(MM, For reasons I’ve stated above, while promising in that regard, is less of a streamlined solution than I initially thought.)

OPML is platform independent, and seems to be the de facto standard for exchanging outlines these days. The problem with your workflow seems to me to be more that Word doesn’t export outlines to OPML. It might be possible to solve the problem by outlining in Word, exporting as rtf, then somehow converting the rtf outline to OPML and importing that into Scrivener. I don’t know of an application that can import rtf and export OPML, but someone here might…?

OmniOutliner? It says it can, though whether it would import a Word outline as a hierarchical tree or a flat list, I don’t know. However, I agree OPML is now in the process of becoming a de facto standard.

H

Unfortunately, I don’t think it works. I just tested Omnioutliner 3.8 and it seems to import rtf (produced by Word) as a series of flat paragraphs, ignoring styles called ‘Heading 1’, ‘Heading 2’ (and which are set as heading styles in Word).

well exporting an outline from word and importing it into scriv, is one challenge. It would be wonderful if that can be achieved, though I’m perfectly willing to settle though just for scriv to import a ‘marked up’ text file.

(MM will work if I type a simple file up from scratch; but there are several reasons why this remains problematic for me (1. saving as text file on pc, copying to mac, seems to be introducing ghost characters; 2. my existing notes have symbols like asterisks and dashes all over them, seem to be messing up mm import, etc; 3. even if I imported it, scriv cant export an outline back to word, so there’s more work on the export side too)).

So unless anyone has other suggestions, I’m pinning my hopes (for importing anyway) on scriv 2.0, that it can bypass MM and define its own markup requirements (and can read in both separators and heirarchy), and thus read a txt or rtf file as an outline and import it as an outline… the export side isnt AS crucial since I can resign myself to exporting finished narrative (as opposed to exporting an outline) back to word. (Though obviously an export-as-word-outline would be the bees knees, i’m not holding my breath on that one :wink:

'Fraid there are no plans for anything quite as radical as that. :frowning: