Hello Richard
Many scientific papers have a tendency to provide a bibliography layout file
in endnote format and never a .bst file. Perhaps it would be possible to
convert endnote->bst config?
Now, apart from this, I support your idea to make a simplified version of
makebst (which is a very useful tool, but I often have to redo the process
several times to get everything in order - and I am constantly confused in
this process as well).
First thing is to realize that a bibtex entry consists of a number of blocks
and that the blocks used in an bibtex entry depends on the entry type.
Second, the order of blocks should be configurable (title block before date
block or visa verse). Thirdly, each block should be configurable in all the
details (et al in italics or not, etc.).
Keeping these three things in mind, I think it is important to sort out the
question about which blocks should be used for a given entry type, then what
order the blocks should be in for this entry type, and finally the layout
for the blocks. Keeping these three levels separate should also simplify the
syntax.
some pseudo syntax:
<entry=article>
<author block>
<" ">
<title block>
<". ">
<journal block>
<". ">
<pages block>
<", ">
<date block>
</entry>
<block=author>
<last name>
<"., ">
<first name>
</block>
<block=title>
<title, font-style="italics">
</block>
And most importantly. A small set of test references should be available so
that you can test the outcome of you bibliography layout with the click of a
button. That is a problem with makebst, that you have to go through a long
process before you can test the layout. Actually, this is one spot where I
would like to see some WYSIWYG.
Best regards,
Martin
On 6/27/06, Richard Heck <[EMAIL PROTECTED]> wrote:
In a recent thread, there was some discussion about inadequacies of
standard BibTeX styles, especially as regards work in the humanities.
Having a little time free, I decided to educate myself as much as
possible about BibTeX styles and see what I could do. To this point, I
have a working TRANSLATOR field; I can use both author and editor in
books (e.g., "Hume, David (1978). /A Treatise on Human Nature/, ed. by
L. A. Selby-Bigge", etc);/ / I've manged to overcome this problem with
how makebst handles dashed versions of repeated authors:
Heck, R. (1993). ...
----- (1995). ...
Heck, R. (ed). (1997). ...
Heck, R. (2001). ...
----- (2003). ....
and, most importantly to me, I've figured out how to allow the use of
abbreviations for some references (i.e., instead of (Hume, 1978, p. 23),
I can have (/Treatise/, p. 23), or whatever, but only for specific
works); and to do all of that using natbib, so that I don't have to give
up LyX's support for natbib or deal with the complexities of jurabib.
There's more work to do. In particular, I'm working on an alternative to
makebst that is intended to be more flexible but not much harder to use,
especially for those of us who need more customization than the question
and answer format offers. The idea here is that formats for various
types of entries can be written in a sort of format-by-example syntax, e.g
.:
NAME.FORMAT=last-first
BIBTYPE ARTICLE
NAME=AUTHOR
BIBFORMAT
$NAME ($CITATION): `$TITLE', \emph{$JOURNAL} $VOLUME<|($NUMBER)|>: %
<|IF<|$PAGES.MULTI|><|pp. |><|pg. |>|>$PAGES.
ENDFORMAT
ENDTYPE
BIBTYPE BOOK
NAME=AUTHOR, EDITOR
EDITOR.FORMAT=first-last-etal
BIBFORMAT
$NAME ($CITATION): $TITLE.emph<|IF<|$AUTHOR|><||><|, ed. by
$EDITOR|>|>. %
$ADDRESS, $PUBLISHER.
ENDFORMAT
ENDTYPE
That's an example of the intended syntax. I'm guessing most of you will
be able to figure out what it should do. (It does nothing yet, as the
program hasn't been written.) Comments appreciated.
That leads to my question. While I'm at this, I'd like to have some
sense what it is that other people wish their BibTeX styles would do. So
what DO people wish BibTeX would do? Other than make it easier to create
formats for specific journals and the like: That's what the above will
do. Are there other frustrations? as, for me, were the absence of a
TRANSLATOR field, the inability to use author and editor at the same
time, the inability to define citation abbreviations, and the like?
Best,
Richard
