Yet Another Gutenberg Book/HTML converter

[FangornUK]

More updates: Small fix for the BODY tag not being picked up correctly and causing crashes.
Thanks silkpag for the bug report.

[FangornUK]

Update: Added option to specify LRF output filename

[igorsk]

Say, would it be too hard to adapt it for Baen's HTML books?

[FangornUK]

Would be hard to do, the Baen books aren't formatted very well in HTML and gutlrf only really works with single HTML files (theirs are split). The RTF versions look to have the best formatting though.

[NatCh]

For Baen's offerings, I've been quite happy with taking the RTFs and bumping the font size up. I also change the font face, but that's just preference.

[igorsk]

Well, their RTFs are definitely readable, but it would be nice to have a TOC...

[HarryT]

Not to mention that many of them have useful pictures (maps, etc) but I guess that's out of the question at the moment if we're generating RTF.

[grebki]

Here is a script adapted from FangornUK's nice scripts (OK, shamelessly copied.) It is for HTML's that are already split up and need no additional processing.

In order to use it, you must create a file list (dir > filelist.txt, then edit text file to clean it up into a clean list.)

The script reads the file list and calls HTML2LRF.exe with the list of files. It's easier than manually typing in a super-long list of files on the command line (30, 40, 50 chapters...)

Each HTML file is listed in a working TOC as well (HTML2LRF does this automagically.) You can also specify multiple tags for Chapter titles (for example, I had a book that used H3 for a tag along the lines of "Chapter 1" and then an H4 tag with the actual name of the chapter -- something like "Fastidious Incompetence." ) Anyway, hope it helps -- feel free to modify.

A few of notes:
* Needs to be run in the directory of HTML2LRF.EXE
* You can either specify a base directory with -d option -or- specify complete path-names for the files in the filelist.txt file. HTML2LRF needs to get complete path names or it doesn't work.
* If no chapter tags are specified, then the <title> <\title> tag will be used by HTML2LRF for the TOC entry (which is fine sometimes) otherwise, the script will replace <title> <\title> with the contents of the specified tags ... unless it can't find them, in which case it will leave the title unchanged.

To Do:
* Allow chapter headings by regular expression instead of HTML tags (i.e. "<p>Chapter something<\p>" or ALL CAPS)
* Pull in text files, split on regexp, remove arbitrary line-breaks, and convert to rudimentary HTML before combining into a BBeB (just to get that nice text-flow and TOC)

-G

[FangornUK]

New fix "Handle Gutenberg ZIPs with missing subdirectories".

I still use gutlrf even though, Book Designer is a much better solution, to prepare a Gutenberg HTML book for Book Designer. It (gutlrf.pl) will retrieve, extract, and clean up the HTML book (like removing page numbers) which can then be loaded into Book Designer to generate a LRF.

[UncleDuke]

Great stuff but you're forcing me to learn Perl!

[FangornUK]

You don't need to learn it, just run it!

[gr8tech4me]

FangornUK, this tool is awesome, I love it.

My only hope is that there is a way to send huge files into HTML2LRF. I want to put a dictionary on my Reader even if I need to use it the old fashioned way(or add bookmarks to it). The Gutenberg offers the Unabridged Websters but it is over 1 MB per volume or 15 MB in total. This crashes HTML2LRF everytime, any ideas would be great!

[FangornUK]

The Websters appears to be in some strange XML format and is not HTML.

[FangornUK]

Update:

• gutlrf.pl now uses kovidgoyal's wonderful libprs500 library instead of the old HTML2LRF (utility is still called HTML2LRF in libprs500). Don't forget to delete the old HTML2LRF.exe in the gutlrf.pl directory.

• Added text based books conversion (with help from Gutenmark)

• Removed splitbook.pl as the new HTML2LRF no longer requires this

[kovidgoyal]

Nice you should see about getting manybooks.net to use this for the LRFs.