Regex s&r for dashes used where emdashes should be

[alansplace]

I need a pair of regex s&r strings for dashes used where emdashes should have been used?

[Jellby]

There's no such thing, without actually understanding the language.

I don't want to go up-stairs (it used to be written with a hyphen in "old" times)
I don't want to go up—stairs or not (maybe not the best sentence, but you get my point)

At most, you can maybe catch some common mistakes, like a hyphen at the end of the sentence.

[Sabardeyn]

Not to mention that in many cases in previously edited ebooks many users will change dashes/en dashes/em dashes to comply with their own display preferences. For instance, word—word is frequently changed to word - word (note the added spaces). While grammatically incorrect (per Chicago Manual of Style), it is frequently changed to something similar to allow for better line breaks, or as a form of "soft hyphening".

[PeterT]

@saberdyn: I'm not sure I would say "grammatically incorrect"; "stylistically incorrect" maybe....

[Tex2002ans]

Also, if you want help with Regex, please post samples to test. Sometimes it is very hard to figure out exactly what is wanted.

Are we talking:

#1:

It is immaterial whether the restriction of a recipe's use is made possible by institutional conditions--such as patents and copyright laws--or by the fact that a formula is kept secret and other people fail to guess it.

#2:

It is immaterial whether the restriction of a recipe's use is made possible by institutional conditions -- such as patents and copyright laws -- or by the fact that a formula is kept secret and other people fail to guess it.

#3:

It is immaterial whether the restriction of a recipe's use is made possible by institutional conditions-such as patents and copyright laws-or by the fact that a formula is kept secret and other people fail to guess it.

After #1:

It is immaterial whether the restriction of a recipe's use is made possible by institutional conditions—such as patents and copyright laws—or by the fact that a formula is kept secret and other people fail to guess it.

After #2:

It is immaterial whether the restriction of a recipe's use is made possible by institutional conditions — such as patents and copyright laws — or by the fact that a formula is kept secret and other people fail to guess it.

If it is #1 or #2, those can be done with a simple search and replace.

#3 will involve checking nearly every hyphenation in the book, and may or may not be worth it. I would recommend using the Sigil Spellchecker for that (Tools - Spellcheck - Spellcheck (Alt+Q)), and in the search box typing a hyphen. This will give you every single word with a hyphen in it (see attached image).

I recommend at least two passes when fixing hyphenation this way. First with "Show All Words" off, and then one pass with "Show All Words" on.

[cybmole]

what I do , sometimes, is set up a search to find dash and replace with em or endash but then manually step thru the text- hit the find button to find 1st instance then either hit replace+find to change it or hit find again to leave it as-is & move on. This is better done AFTER reading the book if you don't want to see plot spoilers! you can visually make a context-sensitive decision in less than a second so it does not take too long.

you can automate the special case of a dash immediately before a closing quote if that bugs you

[alansplace]

Thanks to Jellby, Sabardeyn, PeterT, Tex2002ans and cybmole for their replys. I'll report here if any of your suggestions work out for me!

[phossler]

Originally Posted by alansplace

I need a pair of regex s&r strings for dashes used where emdashes should have been used?

What are your rules for 'should have been used'?

I've seem many house rules and personal preferences for those darn things, all the way from "They're old fashioned--don't use them" up to "No space before or after unless ..."

For the dead trees versions, in MS Word I have a macro that does (something like) replace variations of 'space+double dash+space' with an an em-dash and 'number+dash+number' with an en-dash. But again, that's personal style. Sometimes depending on the typeface, I'll do it differently


Paul

[theducks]

Paul
Personal style wins every time (unless the customer is PAYING you to do it their way)

[BetterRed]

Originally Posted by theducks

Paul
Personal style wins every time (unless the customer is PAYING you to do it their way)

"In matters of style, swim with the current; in matters of principle, stand like a rock." Thomas Jefferson

BR

[alansplace]

Originally Posted by phossler

What are your rules for 'should have been used'?

I've seem many house rules and personal preferences for those darn things, all the way from "They're old fashioned--don't use them" up to "No space before or after unless ..."

For the dead trees versions, in MS Word I have a macro that does (something like) replace variations of 'space+double dash+space' with an an em-dash and 'number+dash+number' with an en-dash. But again, that's personal style. Sometimes depending on the typeface, I'll do it differently


Paul

Was looking for a way/method to change the character being used where a dash was being used that was not in a standardly hyphenated, or compound word but instead was being used to set off a phrase or indicate a pause.

[Tex2002ans]

There is also this regex that I use, but be warned again, this will find (nearly) every single hyphen... and takes forever to click through the entire book one by one.

This is why I initially recommended the Sigil Spellcheck method.

Search:

([A-Za-z])-([A-Za-z])

Replace:

\1—\2

This will work on hyphens like in my Case #3 above.

And again, I like to stress... SAMPLES SAMPLES SAMPLES. They really help everyone figure out what you mean, and allows us to help better.

As a side note, I use a variation of this regex to handle adding en dashes between years/page numbers:


Search:

([0-9])-([0-9])

Replace:

\1–\2

[alansplace]

Originally Posted by Tex2002ans

There is also this regex that I use, but be warned again, this will find (nearly) every single hyphen... and takes forever to click through the entire book one by one.

This is why I initially recommended the Sigil Spellcheck method.

Search:

([A-Za-z])-([A-Za-z])

Replace:

\1—\2

This will work on hyphens like in my Case #3 above.

Oh thanks for that. I'd already tried that method before I started this thread and since it had almost 900 hits I decided not to proceed.

Concerning the spell checker in the Sigil that I use, 0.6.2, doesn't work like what you posted. I suppose you're using version 7.

[Tex2002ans]

Indeed. The spellcheck list was added in version ~0.7.0 due to my recommendation.

In 0.6, I was hackishly using the Index Editor to try to find every hyphenated word that was misspelled (I was probably the only person who exists who used it like that?).

The spellcheck list is infinitely better than anything else I have used. I have zero clue why it isn't in your typical word processor program. Now I can't live without it, and it saves me hours of time fixing typos.

[BetterRed]

Originally Posted by Tex2002ans

I have zero clue why it isn't in your typical word processor program. Now I can't live without it, and it saves me hours of time fixing typos.

I've had such a thing in Word since they first supported custom dictionaries, I think I filched from one of Xerox's boat-anchors.

BR