Format and sort Zotero HTML bibliographies

format-zotero.rb

#encoding: UTF-8
require 'nokogiri'

# This script is intended to be used with the custom CSL at https://gist.github.com/sheepeeh/dbb7b02973644d397378
# as it relies on sorting by call number to work.
# Takes a directory of HTML bibliographies exported by Zotero and makes it a little nicer 
# for display on an Omeka Simple Page.

# Expects files to be named CollectionNumber-BoxNumber.html

Dir.glob("./*.html").each do |page|
	name = File.basename(page,".html")
	puts "Processing #{name}"
	coll,box = name.split("-")

	doc = Nokogiri::HTML(open("#{page}"))

	# Array to make sure folders are not printed twice.
	folders = []

	# Open output file.
	f = File.open("./#{name}_folders.html","a")

	# Adds an H3 with the box number and a link to a box search. 
	# (Requires an Item Type Metadata element for box.)
	# CHANGE TO FIT YOUR LINK STRUCTURE
	# Also starts csl-bib-body DIV.
	f.puts "<h3><a href='/exhibits/merrigan/items/browse?search=&advanced[0][element_id]=53&advanced[0][type]=contains&advanced[0][terms]=#{name}' title='View other items in box #{box}'>Box #{box}</a></h3>\n<div style='line-height: 1.35; padding-left: 2em; text-indent:-2em;' class='csl-bib-body'>"


	doc.xpath("//div[@class='csl-entry']").each do |entry|
		
		text = entry.content
		# This regex is based on our local identifier construction. 
		# CHANGE TO MATCH YOUR LOCAL IDENTIFIER STRUCTURE
		id = text.match(/(455(-([0-9]{3})){3})/)
		id = id.to_s
		coll,box,folder,item = id.split("-")
		
		folder = "not listed" if folder.nil?


		
		# Adds an H4 with the folder number and a link to a folder search
		# unless one has already been created. 
		# (Requires an Item Type Metadata element for folder.)
		# CHANGE TO FIT YOUR LINK STRUCTURE
		f.puts "\n<h4><a href='/exhibits/merrigan/items/browse?search=&advanced[0][element_id]=54&advanced[0][type]=contains&advanced[0][terms]=455-#{box}-#{folder}'>Folder #{folder}</a></h4>" unless folders.include?(folder)
		f.puts entry
		
		folders << folder unless folder.nil?
	end

	# Outpus COINS spans
	doc.xpath("//span[@class='Z3988']").each do |coins|
		f.puts coins
	end

	# Close csl-bib-body DIV.
	f.puts "</div>"

	f.close

	# Replace all of the weirdness introduced by Nokogiri
	text = File.read("./#{name}_folders.html")
	text.force_encoding 'utf-8'
	    replacements = [["\u201C","&quot;"],
	    		   ["\u201D","&quot;"], 
	    		   ['&#13;', '' ],
	    		   [" :",":"],
	    		   ["———","&#8212;"],
	                   [ '&#151;', '&#8212;' ],
	                   [ '&#146;', '&#8217;' ],
	                   [ '&#145;', '&#8216;' ],
	                   [ '&#148;', '&#8221;' ],
	                   [ '&#147;', '&#8220;' ],
	                   [ '&nbsp;', ' ' ],
	                   [ 'Â<A0>', ' '],
	                   [ '—', "—" ],
	                   [ '–', "—" ],
	                   [ '…', '…'],
	                   [ 'é', 'é'],
	                   [ '“', '“'],
	                   [ '‘', '‘'],
	                   [ '’', '’'],
	                   [ 'â€', '”'],
	                   [ ' ', ' '], 
	                   ["–",'&#45;'],
	                   ["‘","'"],
	                   ["’","'"]
	               ]
	                 
	replacements.each { |replacement| text.gsub!(replacement[0], replacement[1]) }
	File.open("./#{name}_folders.html", "w") { |file| file.puts text }

	# Add/modify some formatting.
	text = File.read("./#{name}_folders.html")
	replacements = [['<div class="csl-block">NAL Notes:', '<div class="csl-block"><strong>NAL Notes:</strong>'],
		        ['<div class="csl-block">Abstract:', '<div class="csl-block"><strong>Abstract:</strong>'],
	        	["NAL identifier:","<strong>NAL identifier:</strong>"],
	        	["<i>","<em>"],
	        	["</i>","</em>"]
        	]

	replacements.each { |replacement| text.gsub!(replacement[0], replacement[1]) }
	File.open("#{name}_folders.html", "w") { |file| file.puts text }
end

puts "Done."

Within Omeka, I add the following CSS

.csl-bib-body {
    margin-left:1em;
    margin-bottom:2.5em;
}

.csl-entry {
    margin-bottom:1em;
}

.csl-block {
    margin-left:2em;
}