a) HTML annotation via footnote/link back; b) 19th-century exchange/inflation-adjusted rates

November 9, 2007 8:45 AM

Posted for sushiwiththejury: I have to annotate some texts from the 19th century. The final product must be an HTML-only webpage. I have two major issues:

1. I'm working through the texts from start to finish, i.e, when I come across a word/proper noun/etc to annotate, I do so and move on (by annotate I mean link to a footnote/web resource). For every word/text that I annotate, I insert an anchor (to linkback) and a link to the resource.

The problem with this: updating the anchors later on will be a major headache, because I'm positive I'll come up with many more links as I go through multiple readings of the texts.

Is there a simpler solution?

2. Is there a resource for converting various 19th century European/American/African currencies to current inflation-adjusted values?

I've seen some of the annotation software out there, and it isn't what I want.

I'm on Ubuntu using Mozilla Composer, could use Windows if needed.

(via)

posted by DevilsAdvocate to computers & internet (4 comments total)

You could use Movable Type to generate static HTML files, and use the Bö plugin (with Markdown) to manage the footnotes.
posted by kirkaracha at 9:04 AM on November 9, 2007

I think you're taking the wrong approach. If I understand your goals correctly, what would work better would be to A) compile a list of annotated words/phrases, and B) assemble an array of annotated words/phrases paired with their annotation.

Once you've done that, a very simple script in perl/python/php would iterate over the array, find the first (or every) use of the annotated word/phrase, and mark it up appropriately. A simple version of this script would probably be about 10 lines long; a fancier one might be 30. My own script-fu is weak, but I can imagine roughly how it would look.
posted by adamrice at 10:52 AM on November 9, 2007

The approach suggested by adamrice is really the correct thing to do. Marking as you go is sure to turn into a nightmare if each keyword is going to come up more than once.

I don't think a program to do this, good enough for real-world purposes, would fit in 30 lines: You would have to use a fair amount of code to make sure the input files are parsed correctly, and if you have more than a few keywords it would be really inefficient to scan the entire input for every keyword. The problem is that one usually massively underestimates the amount of time needed to make it work correctly, and proof the results to make sure your code is not doing something stupid. I'm not really a programmer (I'm guessing you aren't either), but it would probably take me at least a couple of hours to get right.

If you haven't done something like this before, I suggest using index cards and find/replace.
posted by the number 17 at 11:40 AM on November 9, 2007

Check out Roy Davies's page of links on estimating historical currency values.

I don't understand what you mean by "updating the anchors later on." Do you mean renumbering notes? If so, consider using OpenOffice.org for your annotating -- it handles note numbering in HTML perfectly, and generates relatively clean HTML over all. I often use this for editing text before bringing it over to an HTML editor for cleanup and formatting.
posted by gum at 12:55 PM on November 9, 2007