Wishlist item - word count rules configurable

It would be good to be able to have as a configurable item in the options so the user can toggle options on or off for word counting.
I would like to be able to have a hyphenated word counted as one word.
Others may have different preferences for other word count rules, hence the idea that it might be a configurable options for word count.

In this regard, we are limited by what the frameworks we use consider to be a word, and things would get slow if we tried to intervene and override that code. You’ll find that all word processors vary in this regard.
All the best,
Keith

I’m going to open this up again. Many scientific papers have word count limits, and all ones I’ve tested seem to treat hyphenated pairs as a single word (they have online submission forms to count words). Scientific subject specific terminology words are often hyphenated word pairs which is why I think they all treat hyphenated words as singular. This means word counts in MS Word are more accurate for us than Scrivener’s a-word-is-a-word-goddamit method. Has anything changed in the frameworks you use regarding this — a new cool .count-hyphenations(‘no’) method? :smiley:

A quick fudge, but if you load the document in scrivenings and then search and replace a hyphen with a hyphen (thereby not changing the text at all), you’ll get a note of the number of hyphens in the document. Assuming they are only used in hyphenated words, you can reduce your word count by the number of hyphens. If a project of 10,000 words contains 58 hyphens, the count would be 9,942 words.

Thanks Briar, that will have to be the way to do it in the meantime, but it is quite a pain when each section has a different word limit and you are editing down and switching documents to try to reach your word limits. So I still want Keith to tell me Apple in their wisdom have added a scientific-jargon count mode to the Sierra APIs (rather than all the bugs they seem to have added to PDF Kit!)

I’m afraid not - if Apple had changed this, you would see the results in Scrivener. When it comes to hyphenated words, Scrivener’s word counts match Pages and Nisus Writer, because it uses Apple’s standard methods for counting words.

All the best,
Keith

Agree, not an ideal solution, but reasonably swift.

Does replacing your hyphens with a “non-breaking hyphen” make any difference? You can get to it with the keyboard pallet, I believe. It’s Unicode 2011 (see discussions.apple.com/thread/1089908?tstart=0 for details). If it works, a project replace would fix what exists, and setting up an auto-substitution would work for newly-typed hyphens.

Non-breaking hyphens don’t make any difference to how words are counted.

That’s too bad. But surely there is a unicode character that could stand in for a hyphen (even if it doesn’t look like a hyphen) inside Scrivener that it counts as just another character in a word.

Aye. One could even write greatXZXZgrandfather for great-grandfather to fool the word count and then set up a suitable replacement for XZXZ in compile…but not very elegant.

Ha, a unicode letter lookalike is a great idea. The Apple mechanism does know if the unicode character is from a letter or symbol range so for example maths symbols like ⊸ (multimap) or other technical symbols like ⌙ don’t fool the count mechanism. Even hyphens from minority languages like ᠆ (mongolian todo soft hyphen) are counted as punctuation thus don’t unify words. So you need a “letter” that looks hyphenish, or at least minimal. Lots of hyphenish symbols are actually combining marks and thus cannot be used. Also lots of letters in e.g. hebrew or arabic cause baseline shifts or font substitution issues.

Here are some characters that work:

ʹ = greek numeral
ͺ = greek YPOGEGRAMMENI
ʭ = latin bidental percussive
؀ = arabic number sign
؃ = arabic sanah
٠ = arabic zero
ᅳ = Hangul jungseuong
ᐨ = Canadian aboriginal short horizontal

Some of these will triggger a font substitution and look different in different substituted fonts. I’m partial to the arabic zero myself and it is correctly replaced with – in compile. But the issue is using it, even with a text replacement utility it is more effort than just hitting - and the problem with an auto-substitution is that as Scrivener uses the Apple substitution system it would affect everything not just Scrivener :cry:

There is a character that does work that is easy to access: _ (lowline or ascii underscore, unicode U+005F) but this is used by MMD and can have other uses so makes it not ideal as a hyphen replacement.

So here’s a dumb idea for Keith: how hard would it be to have Scrivener perform the character-replacement hijinks behind the scenes before performing the word count? If it’s not that hard or terrible on performance, then we could have the configuration option for default textkit behavior or the opposite, and Scrivener could (behind the scenes) create a copy of the text, perform the replace with the chosen Unicode character, and send that copy to the count function – and then throw that copy away. Of course, this would require the word count to be a discrete action the user performs – it would be unworkable for the on-screen counts.

I am new to Scrivener and this message board. But I have two wishes for the software. I apologize if these topics have already been discussed. If they have, could someone send me the link?

  1. When I move items sometimes, I notice if try to drag and drop to create a new space, the item instead falls into the category before or after the empty space between them. If I drag and hold in the space between two items, after about 3 seconds, it will want to drop into one or the other. This forces me to open the item and replace the item again. Is this a bug?

  2. Is there a word count option that shows how many words were typed on a daily basis. A list that shows word counts for each day of a project would be helpful. Setting a date and dividing those days into total words is helpful for planning, but I’d like something that shows me how I did every day.

I’m going to revive this request.

I write in Portuguese. We’ve got plenty of words that have hyphens and always count as one word. To make it worse, we can have two hypens (dar-lhe-ia) in a single word so the work around above doesn’t work.

I’ve got the first volume of a novel almost finished, divided into a ton of chapters and scenes. To make it even worse, I’ve got a character with a stutter (naturally marked as ‘I n-need t-to c-c-count hyphenat-ted words as ONE word’).

Please, tell me you can find a way to count these as one. It’s so frustrating! I can’t even keep track of daily word counts because it’s all inflated.

I know the stutter is a ‘minority’ detail, as another member put it, but Portuguese is not a minority language. And there are more languages out there that have hyphenated words counting as one. So this is not a minority request.

Forgive my venting but it’s so frustrating (especially because keeping word count stats is very important for me for which I’ve got a dedicated excel file) that I sometimes wonder if it’s worth it and I shouldn’t just give it all up and make do with Word. Even if I prefer Scrivener in everything else. Please try to give this non-minority request some serious thought.