pg2html Detailed Information

zip pg2html-0.17.zip
(DOS/Windows executable, C source code and documentation) 2004-05-14

This utility converts an HTML-ready or PG-ASCII formatted text file to HTML 4.01 Transitional or XHTML 1.0 Transitional.

General usage: pg2html inputfile >outputfile

For: Windows (98, XP, 2000); Linux.
For other platforms, the C source should compile without issues.

This software does not do all the work for you. What is does do is take a huge amount of manual effort off the post-processor's hands to quickly mark up most of the essential elements required in creating an HTML version of an extext. The resulting output should be examined and edited with the same care given to producing an etext in any format. At minimum, you should manually review the complete text for formatting, links, TOC, and images. In practice, many post-processors have used it on a reasonably broad selection of texts with very good results; any editing required is usually minimal, and of no trouble to anyone with average experience with HTML.

In fact, Tonya says:

Ooh, lovely! I am very pleased. ... This is very nice, it is saving me a lot of time. Thank you for your work in developing this tool! (And the best part is that when you make it available to everyone, I won't be doing everyone's HTML for them anymore!!!)

Contexts:

Auto-detection of Author and Title, placed in <title> element.
Marks title of work as <h1>.
Marks as <h2> and generates <a name=""> anchors for the following contexts: Chapter, Book, Part, Index, Preface, Introduction, Conclusion, Contents, Prologue, Epilogue, Acknowledgments, References, Appendices, Bibliography, Summary, Glossary, Endnotes, Footnotes, Line Notes, Notes.
Marks lines with roman numerals only as <h3>.
Lines in all-caps not matching an above context are given <center> tags.
Wraps lead-spaced sequences with <pre> tags.
Everything else is treated as a <p> paragraph.
Detects [Illustration:] markup and creates <img> tags with default values for height, width, and src, and converts the text to an alt="" value.
Images are always wrapped in a <center> tag, and an HTML comment line written above which explains how to otherwise align the images. (For left/right flowed text.)

Converts:

* * * * * rules to <hr> tags.
"--" and "&" to "—" and "&" named entities.
_italic_ markup to <i></i> tags. (Reports and partially fixes mismatches.)
Be aware that the program strips DOS newlines ('\r') to simplify processing.

TOC Generation:

A navigable (linked) TOC is generated for all <h2> and <img> elements in order as two sections and inserted just above the existing table of contents, or just above the first <h2> element if the script cannot determine the location of an existing table of contents.
Bonus: If the links can be associated with lines in the existing Table of Contents, the existing TOC is modified to use these links.

Footnotes:

Footnote markers of the form “[xxxx]” get generated links; the footnotes themselves get a named anchor tag as long as they match the form “[n|t]e xxxx:” (where ‘xxxx’ is 1–4 numeric characters).
This really only works well for out-of-line footnotes which have been renumbered in ascending numeric order and optionally moved to a separate section.

Bugs/Known Issues:

It's not perfect yet. ;-)
Does not completely handle non-lead-spaced poetry well.
Section and <h3> link/anchor generation not fully implemented.
<hr> in <pre> sequences sometimes gets removed.
TOC generation sometimes fails. In particular, texts which use roman numerals only to indicate chapters will usually have no links generated.
Patterns for languages other than English are not yet implemented, but should be easy to add.
No provision is made at this time for any user preferences or customization of the output.

Donovan