How Could this have happened?

I have a project in Scrivener, with externally (that is, required) numbering as shown:

Yes, I wanted to use Markdown instead of paragraph styles. Primarily to have alternatives when weird things happen after compiling.

So I compiled to Word. How is this incorrect numbering even possible? There were 44 instances. How can I prevent this from reoccurring/what is the cause?

I think the issue is not actually the numbering itself, but the indent causing it.

Have you checked the compatibility checkboxes in your compile format?

You’ll note that in the original, the paragraph is numbered:

After compiling to Word, it’s

x1.
2.

Which is incorrect.

I just did a straight, out of the box compile to Markdown. No changes.

The

x1

here is because the forum’s software assumes this is a list and “corrects” it creating new errors.
m

Could be Word doing that?
What if you test a PDF compile?

The Word file cam out of Scrivener, so I would think the answer is “no,” except that “normal.dot” has a mind of its own.

Could be your Word template?

The issue is the indent, I’m pretty sure.
That’s what needs to be fixed.

That’s all the “help” I can provide.

Not mine. I sent it to a Service Bureau for creation of epubs with multiple functional ToC’s.

Which is wrong, 27 or 44? Is 44 nothing but hard-coded text, or something else?

This is all plain text. Or at least, that’s how it started life in the Scrivener project.

You are compiling as Markdown [using Markdown], so the numeration in your

1.
2.
3.

is not meaningful. Markdown [processing] just uses the digits to register that you want to make a list. It actually handles the numeration. So, it would give you the same result if you gave it

1.
1.
1.

That explains half your mystery.

The other half regards why you are getting a sublist when it appears that you have not asked for one. I think Vincent is on to the right line here – if you look at how Markdown works and how sublists are created, they are triggered from plain text by indentation (a couple of initial spaces followed by a number). So, the question is, how is it that when the Markdown interpreter gets invoked during compile, your subsequent paragraphs are triggering the sublist routine?

I wonder if the difference between the way the first numbered paragraph is treated and the subsequent lines (as sublist items) is down to something in your compile settings treating the first paragraph as different. Likely, it is the setting to not indent paragraphs that are preceded by whitespace. Subsequent paragraphs are set to have an indent. Prolly first-line indent is converted to spaces on the way to markdown land. Armchair result: The last three sentences, if correct would explain the phenomenon you are seeing. Presumably what you want to do is tell your compile format that your no-style paragraphs should have no indent added.

One way or another it would appear the compile format you are using is introducing some initial spaces (or some first line indent) to the subsequent paragraphs and this is getting interpreted by the markdown interpreter as an indication of sublist.

If any copy of Word that opened the file was configured to automatically recognize and format lists – which I believe is Word’s out-of-the-box default – that is exactly what it would have done.

Except I’m not compiling to Markdown, I’m writing in Markdown. I’m compiling to Word, because this is how the Service Bureau wanted the file.

Thank you, no doubt I mispoke. I understand you are writing using markdown notation. But when you compile, the markdown interpreter is being invoked. It is that which makes my diagnostic idea pertinent, all mispeaking aside.

In the first line of my post, I ought to have said ‘using Markdown’ instead of ‘compiling as Markdown’. The rest reads correctly as is.

The Markdown interpreter is invoked when compiling straight to Word?

Yes.

How else is Scrivener supposed to figure out that things like ## Chapter 4 and ### Article 44 should be interpreted as heading styles? Those aren’t Word commands.

No, they’re not.
I expected the Word document post-compile to contain/display these commands, not implement them.
The solution then, is to compile to plaint text. This does not invoke any Markdown processor.
Then, the plain text file can be opened in Word and then saved as a Word file. This will neither invoke a Markdown processor nor will Word do anything with those commands other than display them.

Whether Scrivener’s Compile command interprets embedded Markdown is configurable, via the general settings (“gear”) tab on the main Compile screen.

2 Likes