RegEx question (again)

phossler · 01-19-2013, 06:44 PM

Cleaning a book that someone has converted through Calibre a number of times.

Mostly a lot of grunt work, but I've come across the following many times, not only in this one but many others. The chapter number is H1, so the TOC just consists of a string of numbers

So if I can save the F&R in Tools, I can use it whenever I need it

Looks like this ..

Code:

<body>
  <h1>TWO</h1>

  <p>TITLE OF CHAPTER</p>

After manually making changes (lots of trial and error), this seems to work best

Code:

<body>
  <h1>TWO<br />
  TITLE OF CHAPTER</h1>

or if I could be even more clever

Code:

<body>
  <h1>Two<br />
  Title Of Chapter</h1>

Paul

cybmole · 01-19-2013, 09:19 PM

is there an actual question in there, somewhere ?

theducks · 01-19-2013, 09:35 PM

His TOC will look like mush

There needs to be a   before the break to prevent:
ONETitle of Chapter

DaleDe · 01-20-2013, 01:05 AM

Actually <h1>Two<br /> Title Of Chapter</h1> will likely work just fine.

Doitsu · 01-20-2013, 03:40 AM

The following quick and dirty regex will replace:

Code:

  <h1>TWO</h1>

  <p>TITLE OF CHAPTER</p>

with:

Code:

     <h1 title="TITLE OF CHAPTER">TWO<br />
TITLE OF CHAPTER</h1>

Find: <h1>(.*?)</h1>\s+<p>(.*?)</p>
Replace: <h1 title="\2">\1<br />\n\2</h1>

It's not perfect, but should save you some editing. (Sigil uses the title attribute for the toc.)

Sunlite · 01-20-2013, 06:07 AM

I have a solution that is nearly the same as Doitsu, but also starts with the upper case to lower case conversion.

search for:

Code:

<h1>(.)(.*?)</h1>\s+<p>(.)(.*?)</p>

replace with:

Code:

<h1>\1\L\2\E<br /> \3\L\4\E</h1>

This will change

Code:

  <h1>TWO</h1>

  <p>TITLE OF CHAPTER</p>

into

Code:

<h1>Two<br /> Title of chapter</h1>

I can't think of a way to get title case for the title of the chapter without a second regex. Maybe someone else can.

theducks · 01-20-2013, 07:49 AM

Quote:

Originally Posted by DaleDe

Actually <h1>Two<br /> Title Of Chapter</h1> will likely work just fine.

I don't use the Leading space solution on the title because it throws off the display when Left justified or (sometimes obvious) when centered

Code:

Two
 Title of Chapter

ElMiko · 01-20-2013, 11:25 AM

Just for the record:

Code:

<h1>TWO<br />
Chapter Title</h1>

won't result in "TWOChapter Title" as the TOC value. It will display as "TWO Chapter Title". There is no real reason to add either leading or trailing spaces. Of course, as has already been stated, you can specify the desired title value by using the title attribute. Alternately you could just edit it the values directly in the .ncx file after generating the TOC.

theducks · 01-20-2013, 12:47 PM

Quote:

Originally Posted by ElMiko

Just for the record:

Code:

<h1>TWO<br />
Chapter Title</h1>

won't result in "TWOChapter Title" as the TOC value. It will display as "TWO Chapter Title". There is no real reason to add either leading or trailing spaces. Of course, as has already been stated, you can specify the desired title value by using the title attribute. Alternately you could just edit it the values directly in the .ncx file after generating the TOC.

It runs together for me if I don't include some sort of space

phossler · 01-20-2013, 01:08 PM

Thanks to everyone for taking time to offer comments and make suggestions

Based on what I learned, I did change what I thought I wanted to do

I know it's not perfect, but this is what I ended up with

Find:

Code:

<h1>(.)(.*?)</h1>\s+<p>(.)(.*?)</p>

Replace:

Code:

<h1 title="\1\2&nbsp;&ndash;&nbsp;\3\4">\1\2<br />\3\4</h1>

I don't quite grasp all the Find 'tokens' ( esp. that pesky ?) but I'm working through it

Thanks again

Paul

DiapDealer · 01-20-2013, 01:09 PM

Quote:

Originally Posted by theducks

It runs together for me if I don't include some sort of space

Probably all depend on your Clean Source Settings. If it's turned completely off and your code starts life as one line:

Code:

<h2>Chapter One<br />The Title</h2>

Then the entry will be created in the NCX file (when clicking "Generate ToC") with no space...

Quote:

Chapter OneThe Title

If, however, you have at least Pretty Print turned on, then

Code:

<h2>Chapter One<br />The Title</h2>

becomes

Code:

<h2>Chapter One<br />
The Title</h2>

and the entry will be created in the NCX file with a space when clicking Generate ToC...

Quote:

Chapter One The Title

That's the way it works for me anyway.

ElMiko · 01-20-2013, 01:10 PM

EDIT:
Refer to DD's post

theducks · 01-20-2013, 02:37 PM

Quote:

Originally Posted by DiapDealer

Probably all depend on your Clean Source Settings. If it's turned completely off and your code starts life as one line:

Code:

<h2>Chapter One<br />The Title</h2>

Then the entry will be created in the NCX file (when clicking "Generate ToC") with no space...

If, however, you have at least Pretty Print turned on, then

Code:

<h2>Chapter One<br />The Title</h2>

becomes

Code:

<h2>Chapter One<br />
The Title</h2>

and the entry will be created in the NCX file with a space when clicking Generate ToC...

That's the way it works for me anyway.

I have Pretty Print ON, so maybe I just started doing this with an older version and did not change (and no harm seems to be done, so I never noticed

)

01-19-2013, 06:44 PM	#1
phossler Connoisseur Posts: 82 Karma: 10 Join Date: Jan 2009 Location: Valley Forge, PA, USA Device: kindle	RegEx question (again) Cleaning a book that someone has converted through Calibre a number of times. Mostly a lot of grunt work, but I've come across the following many times, not only in this one but many others. The chapter number is H1, so the TOC just consists of a string of numbers So if I can save the F&R in Tools, I can use it whenever I need it Looks like this .. Code: <body> <h1>TWO</h1> <p>TITLE OF CHAPTER</p> After manually making changes (lots of trial and error), this seems to work best Code: <body> <h1>TWO<br /> TITLE OF CHAPTER</h1> or if I could be even more clever Code: <body> <h1>Two<br /> Title Of Chapter</h1> Paul

01-19-2013, 09:35 PM	#3
theducks Staff to 4 Cats Posts: 10,707 Karma: 2485850 Join Date: Aug 2009 Location: The (original) Silicon Valley, USA Device: Galaxy Tab 2,Black Astak PEz, K4NT(now Wifes)	His TOC will look like mush There needs to be a   before the break to prevent: ONETitle of Chapter __________________ Using: Ubuntu(32 bit):Oneric,Precise and XPpro SP3, W7HP(64)- - Libre Office w/Writer2EPUB

01-20-2013, 01:05 AM	#4
DaleDe Grand Sorcerer Posts: 8,695 Karma: 3644259 Join Date: Aug 2007 Location: Grass Valley, CA Device: EB 1150, EZ Reader, Literati, iPad 2	Actually <h1>Two<br /> Title Of Chapter</h1> will likely work just fine. __________________ Dale DePriest http://pages.suddenlink.net/dalede or http://daledepriest.wikispaces.com currently using an EZ Reader or a Literati or my iPad.

01-20-2013, 03:40 AM	#5
Doitsu Wizard Posts: 1,080 Karma: 2677223 Join Date: Dec 2010 Device: Kindle 3	The following quick and dirty regex will replace: Code: <h1>TWO</h1> <p>TITLE OF CHAPTER</p> with: Code: <h1 title="TITLE OF CHAPTER">TWO<br /> TITLE OF CHAPTER</h1> Find: *<h1>(.?)</h1>\s+<p>(.?)</p>* Replace: <h1 title="\2">\1<br />\n\2</h1> It's not perfect, but should save you some editing. (Sigil uses the title attribute for the toc.)

01-20-2013, 06:07 AM	#6
Sunlite Connoisseur Posts: 91 Karma: 8474 Join Date: Mar 2008 Location: Berlin, Germany Device: PRS-T1, PB602, CyBook Gen3	I have a solution that is nearly the same as Doitsu, but also starts with the upper case to lower case conversion. search for: Code: <h1>(.)(.?)</h1>\s+<p>(.)(.?)</p> replace with: Code: <h1>\1\L\2\E<br /> \3\L\4\E</h1> This will change Code: <h1>TWO</h1> <p>TITLE OF CHAPTER</p> into Code: <h1>Two<br /> Title of chapter</h1> I can't think of a way to get title case for the title of the chapter without a second regex. Maybe someone else can.

01-19-2013, 09:19 PM	#2
cybmole Wizard Posts: 1,686 Karma: 772746 Join Date: Sep 2010 Device: Sony PRS 650, Kindle for PC	is there an actual question in there, somewhere ?

01-20-2013, 11:25 AM	#8
ElMiko Addict Posts: 286 Karma: 56788 Join Date: Jun 2011 Device: Kindle	Just for the record: Code: <h1>TWO<br /> Chapter Title</h1> won't result in "TWOChapter Title" as the TOC value. It will display as "TWO Chapter Title". There is no real reason to add either leading or trailing spaces. Of course, as has already been stated, you can specify the desired title value by using the title attribute. Alternately you could just edit it the values directly in the .ncx file after generating the TOC. __________________ When some wild-eyed, eight-foot-tall maniac grabs your neck, taps the back of your favorite head up against the barroom wall, and he looks you crooked in the eye and he asks you if you've paid your dues, well, you just stare that big sucker right back in the eye, and you remember what ol' Jack Burton always says at a time like that: "Have ya paid your dues, Jack?" "Yessir, the check is in the mail." Last edited by ElMiko; 01-20-2013 at 12:17 PM.

01-20-2013, 01:08 PM	#10
phossler Connoisseur Posts: 82 Karma: 10 Join Date: Jan 2009 Location: Valley Forge, PA, USA Device: kindle	Thanks to everyone for taking time to offer comments and make suggestions Based on what I learned, I did change what I thought I wanted to do I know it's not perfect, but this is what I ended up with Find: Code: <h1>(.)(.?)</h1>\s+<p>(.)(.?)</p> Replace: Code: <h1 title="\1\2 – \3\4">\1\2<br />\3\4</h1> I don't quite grasp all the Find 'tokens' ( esp. that pesky ?) but I'm working through it Thanks again Paul

01-20-2013, 01:10 PM	#12
ElMiko Addict Posts: 286 Karma: 56788 Join Date: Jun 2011 Device: Kindle	EDIT: Refer to DD's post __________________ When some wild-eyed, eight-foot-tall maniac grabs your neck, taps the back of your favorite head up against the barroom wall, and he looks you crooked in the eye and he asks you if you've paid your dues, well, you just stare that big sucker right back in the eye, and you remember what ol' Jack Burton always says at a time like that: "Have ya paid your dues, Jack?" "Yessir, the check is in the mail." Last edited by ElMiko; 01-20-2013 at 01:14 PM.

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
A regex question	PatNY	Sigil	30	06-03-2012 02:03 PM
Quick Regex Question	cptsmidge	Sigil	6	03-06-2012 04:20 AM
Yet another regex question	Jabby	Sigil	8	01-30-2012 08:41 PM
Regex question and maybe some help	crutledge	Sigil	9	03-10-2011 04:37 PM
Regex Question	Archon	Conversion	11	02-05-2011 10:13 AM


Advertisement Enthusiast