Protocol for modifying existing styles into a new style

fshic

I recently needed a csl for the journal Brain Research; the easiest match I could find was the journal Cell. I made a not-so-great series of modification, but it seems as if those modifications will get me through the day. I wanted to know if 1) I should somehow note that this style was derived from another somewhere in the csl, and 2) if modifying the Cell csl directly wasn't even the right way to go about creating a Brain Research csl.

Part 1 of this is just wanting to "give credit where credit is due"; Part 2 is wanting to know if there is an easier way to derive from other styles.

Also, I found finding a "best matching" csl somewhat tedious, since I had to walk through the style list. I wonder if someone might recommend a better process for the future, e.g. where I could get a giant document with all styles and examples. Then I could write a script to go find a reasonable match rather than wading through them all.

adamsmith

1) Leave the original authors and contributors as contributors in the <info> section of the style - see e.g. the APA style for an example
2) Yes, modifying an old style is usually the best strategy, there is no better strategy.

About downloading all csls - you can get an SVN client such as Tortoise SVN on WIndows, Rabbit VCS on Linux and I forgot what it's called on Mac - you can then checkout the entire styles from
https://www.zotero.org/svn/csl
I'm pretty sure you can check out a copy anonymously, if you want to submit you'll need trac access
https://www.zotero.org/dev/trac_access

I believe a more structured repository is planned, but I think the milestone is Zotero 3.0

noksagt

I wanted to know if 1) I should somehow note that this style was derived from another somewhere in the csl

I would place the following in the file:<link href="http://www.zotero.org/styles/cell.csl" rel="template"/>To me, leaving contributors in is redundant to this: the old style is readily available.

if modifying the Cell csl directly wasn't even the right way to go about creating a Brain Research csl.

If it works, it works. A few "easy" ways to find probable style matches are to look at journals from the same publisher and from the same field. Often, you can google the example citation listed in the author instructions and will find other journals that use the same formatting.

Quite a while back, I downloaded all of the examples of formatted citations from the Zotero webpage & had a python script that calculated the edit distance between my (manually edited) version of what the new style should generate & all of the other styles. It is not userfriendly (somewhat prone to input errors & also sensitive to ordering of markup tags), and the shortest edit distance does not equate to the simplest style edits (it finds a "good" style to change, but not the best one). You're welcome to it & something like this should be added to future online editors/repositories.

It has been a while since I used the script. The first time that I did, I found several redundant styles. I have since found more manually, but I'm sure that there are more that could be found if I ran it again.