[Scribus] Index

John R. Culleton john
Mon Jul 23 22:09:12 CEST 2007

On Monday 23 July 2007, John Jason Jordan wrote:
> On Mon, 23 Jul 2007 11:10:42 -0400
> "John R. Culleton" <john at wexfordpress.com> dijo:
> > > Might it be possible to link makeindex or xindy, or some other
> > > open source indexing utility, to Scribus somehow? It would be
> > > cool to have an indexing utility for Scribus.
> >
> > Yes it would. But makeindex is effectively a standalone program.
> > One pass thorugh Tex generates the IDX file from embedded tags ,
> > then makeindex is run creating the ind file, and the next run of
> > TeX finds the ind file and imports it  into the document. This
> > three run process can be hidden somewhat by a driver program but
> > that would have no applicability to Scribus.
> >
> > The best implementation I can see is a four step process:
> > 1. Run scribus and create a final pdf.
> > 2. Run a pdf reader program and manually transfer the index items
> > and their page numbers into an .idx file, perhaps using a little
> > Tcl/Tk program to help ease the process (spinner for page number
> > etc.) 3. Run makeindex,
> > 4. Flow the finished index file into scribus as text.
> >
> > I have asked for help in copying/ pasting words from Scibus
> > itself, thus eliminating step 1, modifying step 2, and
> > incorporating step 3 as part of the original Scribus run. 
> > Haven't read the replies yet.
> I have thought of a differrent approach - OpenOffice.org. Suppose
> we had a 500-page book to index. What if we could select the text
> (in linked frames, presumably) in the Scribus document, copy and
> paste into Writer, with a manual page break at the points where the
> pages break in the Scribus document. Then the user can just use
> OOo's wonderful indexing utility to create the index. Having done
> so, copy and paste the index back into the Scribus document. The
> advantage is that OOo's indexing facility is easily the equal of
> InDesign's with a great and easy to understand GUI interface.
> The problem with this is the page breaks. For our hypothetical
> 500-page book you'd have to hit Ctrl-Enter 500 times, while
> checking each one against the Scribus document. If that could be
> automated somehow, it would work perfectly.
> Another approach might be to see if OOo's indexing code could be
> adapted for and integrated into Scribus to give us a built-in
> indexing tool.
> Personally, I need Scribus to have all the integration with OOo
> that it can. I'd love to have copy/paste working reliably between
> the two, and when pasting into Scribus it would be great to have
> Scribus import styles as well. But further discussion of that needs
> its own thread. _______________________________________________
> Scribus mailing list
> Scribus at nashi.altmuehlnet.de
> http://nashi.altmuehlnet.de/mailman/listinfo/scribus

IMO too many cooks spoil the broth and too many layout programs in 
train in a single job cause confusion. That is why typesetters ask 
that  authors who use MSWord do no or minimal formatting.  Most 
professional indexers work from a finished pdf or even paper copy of 
the final book.  OTOH MSWord, OO, and TeX allow for embedded 
indexing.  I like the TeX approach using a separate program best, 
obviously.  But here is how I really work when indexing for pay:

1. I receive a pdf file of the finished book with final pagination 
2. Using a utility I convert the  pdf to ascii, preserving the page 
3. I massage this file a bit in Vim, making it into a TeX file with
 the page breaks converted to \vfil\break  TeX tags. 
4. I go through the file with Vim, marking items as indexed following 
the eplain.tex tags.  I do mass changes, enter "see" and "see also" 
tags, change sort order of items or subitems were necessary etc. The 
eplain tags allow for all this. 
5. I run a dummy plain tex run to create the idx file
6. I run makeindex on the idx file.
7. I run TeX again to get pages with marginal notations of where items 
are indexed (an eplain feature.)   This is handy when look for page 
sequences with no or very few index items.  I can also check the 
appearance ot the index. 
8. Finally I produce the index from makeindex but this time in plain 
ascii form for shipping to the customer. 

Most of my time is spent on step 4. Since the pagination is automatic 
I can concentrate on what is indexed, not where it occurs. 

I don't know what faciities OO has fo indexing. If it doesn't provide 
for "see" and "see also", altered sort order where required,mass 
tagging with accept reject on each individual tag  etc. I would 
consider it suboptimal. 

John Culleton
ATTN Publishers/authors:
If you don't read you don't succeed.
Free short list of publishing/marketing books. 

Need personalized email and website? Look no further. It's easy
with Doteasy $0 Web Hosting! Learn more at www.doteasy.com

More information about the scribus mailing list