Perl's excellent Unicode support
on
Ask Larry Wall
·
· Score: 1
I've been impressed by your commitment to implementing Unicode support in Perl. It looks like you've played/worked/studied hard to get it right. Any comments about the highlights of this learning process --- the good, the bad, the culturally provocative?
I just finished using docbook-xml 4.1.2 - docbook-dsssl 1.71(contains a print and an HTML sytlesheet for use with RTF and TeX backends) - openjade 1.4devel1 (which uses the stylesheets just mentioned in its RTF and TeX backends) and pdfjadetex to produce a 140 page technical manual in.pdf format. I used psgml mode with Emacs 20.7 as my editor. (I was looking at the interesting group of programs called psgmlx but didn't have time to actually try them.)
In short, for editing, Emacs/psgml mode worked fine, but the complexity of keyboard commands, my aging brain, and deadline pressure meant that I could only learn to use a small chunk of the possible features, especially of psgml mode which, unlike Emacs, was new to me. I opened one Emacs frame on one page on a 4 x 7 desktop for each chapter and then skipped from desktop page to desktop page with a full screen for each. Hint: Make a tags table so that you can then do search and replace operations across all files that make up your book. (See 22.16 Tags Tables in the latest GNU Emacs Manual (14th edition).)
Using docbook-xml, I found that tables were difficult, too difficult. I was unable to make a simple 3 colum x 5 row table within the hour I'd alloted and finally gave up. Ended up using the tag to get literal text. Fortunately, only one or two tables were desired. Checking the archives on the docbook-related mailing lists, I was not able to put Norm Walsh's advice (given to someone with a similar table-related problem) to go read the info on the new table model to practical use within my limited timeframe.
Also, index generation was a bitch. I used collateindex.pl in conjunction with openjade. There must be a better way, but I couldn't find it in the limited time I had.
No graphics were included, but some of the traffic regarding the problems of others gave me a headsup that I should allow plenty of time for experimentation before actual work could begin, learning how to insert both vector and bitmap graphic files in a document that would be turned into a.pdf.
Incidently, I was also able to produce MIF files for FrameMaker from the same XML source with openjade, a client requirement.
Slashdot Mirror
I've been impressed by your commitment to implementing Unicode support in Perl. It looks like you've played/worked/studied hard to get it right. Any comments about the highlights of this learning process --- the good, the bad, the culturally provocative?
Thanks.
Jon Babcock
I just finished using docbook-xml 4.1.2 - docbook-dsssl 1.71(contains a print and an HTML sytlesheet for use with RTF and TeX backends) - openjade 1.4devel1 (which uses the stylesheets just mentioned in its RTF and TeX backends) and pdfjadetex to produce a 140 page technical manual in .pdf format. I used psgml mode with Emacs 20.7 as my editor. (I was looking at the interesting group of programs called psgmlx but didn't have time to actually try them.)
.pdf.
In short, for editing, Emacs/psgml mode worked fine, but the complexity of keyboard commands, my aging brain, and deadline pressure meant that I could only learn to use a small chunk of the possible features, especially of psgml mode which, unlike Emacs, was new to me. I opened one Emacs frame on one page on a 4 x 7 desktop for each chapter and then skipped from desktop page to desktop page with a full screen for each. Hint: Make a tags table so that you can then do search and replace operations across all files that make up your book. (See 22.16 Tags Tables in the latest GNU Emacs Manual (14th edition).)
Using docbook-xml, I found that tables were difficult, too difficult. I was unable to make a simple 3 colum x 5 row table within the hour I'd alloted and finally gave up. Ended up using the tag to get literal text. Fortunately, only one or two tables were desired. Checking the archives on the docbook-related mailing lists, I was not able to put Norm Walsh's advice (given to someone with a similar table-related problem) to go read the info on the new table model to practical use within my limited timeframe.
Also, index generation was a bitch. I used collateindex.pl in conjunction with openjade. There must be a better way, but I couldn't find it in the limited time I had.
No graphics were included, but some of the traffic regarding the problems of others gave me a headsup that I should allow plenty of time for experimentation before actual work could begin, learning how to insert both vector and bitmap graphic files in a document that would be turned into a
Incidently, I was also able to produce MIF files for FrameMaker from the same XML source with openjade, a client requirement.
YMMV.