Text to HTML (or any e-book format, really) program that detects chapters?

[JeremyR]

I have many old text files that are broken up into chapters by simple labels. Chapter 1 (or Chapter I) and so forth.

I've been just loading them into Sigil, but then I have to go through it and mark each chapter by hand.

While this isn't that big a deal, only takes like 10 minutes each, it does get kind of old. It seems like there must be a way to automate it, and then be able to save it as a html file with the chapters labeled so I can import that into Sigil.

Any ideas? (Sadly, ten years ago I could probably write a program to do this, but I don't even have a programming language installed on my computer anymore...)

[Paulinafrica]

Hello Jeremy, Would it be possible for you to use Calibre to do this for you?

[frabjous]

Regular expression find and replace?

It would depend on the particular system/editor. In Vim I would search for:

^\s*Chapter\s*[0-9XVI]*\s*$

That breaks down like this:
^ = start of line
\s* = any amount (inluding none) of whitespace (spaces, tabs, etc.)
Chapter = the word Chapter
\s* = any amount of whitespace again
[0-9XVI]* = a string of any length consisting only of 0-9 X and V and I
\s* = any amount of whitespace again
$ = the end of the line
And replace with

<h1>&<\/h1>

which is the search pattern but with <h1> HTML tags around it.

Not sure if Sigil supports something like that (I don't use it), but it might. If not, you can use any advanced Text editor.

[JeremyR]

Originally Posted by Paulinafrica

Hello Jeremy, Would it be possible for you to use Calibre to do this for you?

I actually thought Calibre did this at first (it's what gave me the idea). But it apparently only detects chapters that have the h1/h2 tags around it

[JeremyR]

Originally Posted by frabjous

Regular expression find and replace?

It would depend on the particular system/editor. In Vim I would search for:

^\s*Chapter\s*[0-9XVI]*\s*$

That breaks down like this:
^ = start of line
\s* = any amount (inluding none) of whitespace (spaces, tabs, etc.)
Chapter = the word Chapter
\s* = any amount of whitespace again
[0-9XVI]* = a string of any length consisting only of 0-9 X and V and I
\s* = any amount of whitespace again
$ = the end of the line
And replace with

<h1>&<\/h1>

which is the search pattern but with <h1> HTML tags around it.

Not sure if Sigil supports something like that (I don't use it), but it might. If not, you can use any advanced Text editor.

Thanks! I will probably end up having to try that. I was hoping for a simpler (and more automated) answer, but this is probably as good as it gets.

[DaleDe]

Originally Posted by JeremyR

I actually thought Calibre did this at first (it's what gave me the idea). But it apparently only detects chapters that have the h1/h2 tags around it

Calibre lets you define your own search criteria.

Dale

[user_none]

calibre has an entire heuristic processing section. One of the heuristics is for detecting and applying the appropriate style to chapter headings. TXT input defaults will try to auto detect the paragraph type and the formatting used. If Textile or Markdown are not detected the default is to enable the majority of the heuristic processing options.

[DMcCunney]

Originally Posted by JeremyR

Any ideas? (Sadly, ten years ago I could probably write a program to do this, but I don't even have a programming language installed on my computer anymore...)

You might want to install Perl (a/k/a Practical Extraction and Report Language.) It's designed for manipulating text files and performing programmed edits on the content. You use regular expressions to locate what you want to change and take actions based on the results.

It's cross-platform, free, and open source. The best Windows version is probably Active State's: http://www.activestate.com/activeperl/downloads

They sell supported commercial versions, but a free Community edition is available, with source.

You can also look at Text2HTML, an open source conversion program built on Perl.
_______
Dennis

[JeremyR]

Originally Posted by DaleDe

Calibre lets you define your own search criteria.

Dale

Unless I am reading it wrong (and that is quite possible), it only searches things that have been tagged. But my problem is, as it's plain text, it hasn't been tagged with anything.

Originally Posted by user_none

calibre has an entire heuristic processing section. One of the heuristics is for detecting and applying the appropriate style to chapter headings. TXT input defaults will try to auto detect the paragraph type and the formatting used. If Textile or Markdown are not detected the default is to enable the majority of the heuristic processing options.

As impressive as Calibre is as a whole, all I can say is, it just doesn't work for the ones I've tried it on..

Indeed, not only does it not detect chapters, it runs together most of the paragraphs making the end result unreadable.

So really, Calibre doesn't enter the picture into this process. I am just looking for something to apply before loading the file into Sigil (which detects the paragraphs just fine).

[JeremyR]

Originally Posted by DMcCunney

You might want to install Perl (a/k/a Practical Extraction and Report Language.) It's designed for manipulating text files and performing programmed edits on the content. You use regular expressions to locate what you want to change and take actions based on the results.

Thanks for the suggestion, but I think I am well past learning any new programming languages (since I was never great at it to begin with). I think I will just look for my old Visual Basic cds and see what I can do with that...

[DaleDe]

Originally Posted by JeremyR

Unless I am reading it wrong (and that is quite possible), it only searches things that have been tagged. But my problem is, as it's plain text, it hasn't been tagged with anything.



As impressive as Calibre is as a whole, all I can say is, it just doesn't work for the ones I've tried it on..

Indeed, not only does it not detect chapters, it runs together most of the paragraphs making the end result unreadable.

So really, Calibre doesn't enter the picture into this process. I am just looking for something to apply before loading the file into Sigil (which detects the paragraphs just fine).

Calibre could search for something like the word Chapter I think if all your chapters start with that word.

Have you just tried pasting the text onto an empty Sigil page?

Dale

[JeremyR]

Thanks for the help everyone (esp frabjous ), I've got it now.

Sigil will do it, sort of. It mostly supports REGEX

So I searched for

CHAPTER [0-9XVI]+

and then replaced with

<h3>\0</h3>

or if you want it to add chapter break marks

<hr class="sigilChapterBreak" /><h3>\0</h3>