An OpenURL is a specially formatted URL containing bibliographic information. It allows you to link to a specialized search engine, called a resolver, at a library. The resolver gathers information from the OpenURL, then searches the library's resources to find out whether or not the library has a copy of that document available. The resolver, which is most commonly provided by Ex Libris' SFX software, searches not just the catalog, but also the library's proprietary database subscriptions. Then it presents the user with the results:

• If the document is available electronically, they get a link.
• If the document is available in the stacks, they get a call number.
• If the document is not available, they get a “sorry” message and maybe a link to Inter-Library Loan services.

OpenURL is also known as Z39.88 after the name of the formal standard established by the National Information Standards Organization.

OpenURL systems are fairly expensive to create and maintain, and so they are not very widespread yet. Many universities have them, and a few large public libraries.

“COinS” stands for “ContextObject in Span.” “ContextObject” is OpenURL jargon for the part of an OpenURL carrying the bibliographic information - basically, it's a query string, everything that would follow the ? in a regular URL. And “span” refers here to the HTML tag <span>. It's the ContextObject that carries the bibliographic information.

The ContextObject is very flexible, and can carry everything from a simple title and author to more obscure things like a Serial Item and Contribution Identifier (SICI).

The upshot of which is that if you stick COinS in your web page, Zotero can pull all that data out again.

The basic structure of a COinS reference consists of a span tag of the class “Z3988” and a title attribute with the ContextObject in it. Here's what a simple one looks like:

<span class=“Z3988” title=“ContextObject goes here”></span>

Let's look at the ContextObject:

1. ctx_ver=Z39.88-2004 specifies the version of the OpenURL standard we're using.
2. &amp;rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook says that the document is a book.
3. &amp;rft.btitle=The+Wind+in+the+Willows is the title of the book.
4. &amp;rft.au=Grahame,+Kenneth is the author of the book.

Putting all of those values on one line in the title attribute makes it a COinS reference. Here's a complete version:

<span class="Z3988" title="ctx_ver=Z39.88-2004&amp;rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&amp;rft.btitle=The+Wind+in+the+Willows&amp;rft.au=Grahame,+Kenneth"></span>

Apologies for the sideways scrolling.

There are a great many different fields you can use in COinS references. For a more complete reference, please see OpenURL COinS: A Convention to Embed Bibliographic Metadata in HTML, especially the two “brief guide” sections listed in the table of contents.

Here is a brief function I worked out for making COinS references. I hereby consign it to the public domain - take it, use it, modify it, print it out and sleep on it, whatever. Read the comments before the function begins for complete documentation. You'll need to modify it to suit your own project.

<?php
/*
OpenURL() constructs an NISO Z39.88 compliant ContextObject for use in OpenURL links and COinS.  It returns 
the proper query string, which you must embed in a <span></span> thus:
 
<span class="Z3988" title="<?php print OpenURL($Document, $People) ?>">Content of your choice goes here</span>
 
This span will work with Zotero. You can also use the output of OpenURL() to link to your library's OpenURL resolver, thus:
 
<a href="http://www.lib.utexas.edu:9003/sfx_local?<?php print OpenURL($Document, $People); ?>" title="Search for a copy of this document in UT's libraries">Find it at UT!</a>
 
Replace "http://www.lib.utexas.edu:9003/sfx_local?" with the correct resolver for your library.
 
OpenURL() takes two arguments.
 
$Document - a document object, having an array (fields) with the following properties:
	$Document->fields["DocType"]
		1 = Article
		2 = Book Item (e.g. a chapter, section, etc)
		3 = Book
		4 = Unpublished MA thesis
		5 = Unpublished PhD thesis
 
	$Document->fields["DocTitle"] - Title of the document.
	$Document->fields["JournalTitle"] - Title of the journal/magazine the article was published in, or false if this is not an article.
 
	$Document->fields["BookTitle"] - Title of the book in which this item was published, or false if this is not a book item.
 
	$Document->fields["Volume"] - The volume of the journal this article was published in as an integer, or false if this is not an article.  Optional.
	$Document->fields["JournalIssue"] - The issue of the journal this article was published in as an integer, or false if this is not an article.  Optional.
	$Document->fields["JournalSeason"] Optional.
		The season of the journal this article was published in, as a string, where:
			Spring
			Summer
			Fall
			Winter
			false = not applicable
	$Document->fields["JournalQuarter"] - The quarter of the journal this article was published in as an integer between 1 and 4, or false. Optional.
	$Document->fields["ISSN"] - The volume of the journal this article was published in, or false.  Optional.
 
 
	$Document->fields["BookPublisher"] - The publisher of the book, or false. Optional.
	$Document->fields["PubPlace"] - The publication place, or false.  Optional.
	$Document->fields["ISBN"] - The ISBN of the book.  Optional but highly recommended.
 
	$Document->fields["StartPage"] - Start page for the article or item, or false if this is a complete book.
	$Document->fields["EndPage"] - End page for the article or item, or false if this is a complete book.
 
$Document->fields["DocYear"] - The year in which this document was published.
 
$People - An array of person objects, each having an array, fields, with these properties:
	$People->fields["DocRelationship"]
		An integer indicating what kind of relationship the person has to this document.
		0 = author
		1 = editor
		2 = translator
	$People->fields["FirstName"] - The person's first name.
	$People->fields["LastName"] - The person's last name.
*/
function OpenURL($Document, $People){
	$DocType = $Document->fields["DocType"];
	if($DocType > 2){ return false; }
 
	// Base of the OpenURL specifying which version of the standard we're using.
	$URL = "ctx_ver=Z39.88-2004";
 
	// Metadata format - e.g. article or book.
	if($DocType == 0){ $URL .= "&amp;rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal"; }
	if($DocType > 0){ $URL .= "&amp;rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook"; }
 
	// An ID for your application.  Replace yoursite.com and specify a name for your application.
	$URL .= "&amp;rfr_id=info%3Asid%2Fyoursite.com%3AYour+Name+Here+Using+Plus+Signs+For+Spaces";
 
	// Document Genre
	if($DocType == 0){ $URL .= "&amp;rft.genre=article"; }
	if($DocType == 1){ $URL .= "&amp;rft.genre=bookitem"; }
	if($DocType == 2){ $URL .= "&amp;rft.genre=book"; }
 
	// Document Title
	if($DocType < 2){ $URL .= "&amp;rft.atitle=".urlencode($Document->fields["DocTitle"]); }
	if($DocType == 2){ $URL .= "&amp;rft.btitle=".urlencode($Document->fields["DocTitle"]); }
 
	// Publication Title
	if($DocType == 0){ $URL .= "&amp;rft.jtitle=".urlencode($Document->fields["JournalTitle"]); }
	if($DocType == 1){ $URL .= "&amp;rft.btitle=".urlencode($Document->fields["BookTitle"]); }
 
	// Volume, Issue, Season, Quarter, and ISSN (for journals)
	if($DocType == 0){
		if($Document->fields["Volume"]){ $URL .= "&amp;rft.volume=".urlencode($Document->fields["Volume"]); }
		if($Document->fields["JournalIssue"]){ $URL .= "&amp;rft.issue=".urlencode($Document->fields["JournalIssue"]); }
		if($Document->fields["JournalSeason"]){ $URL .= "&amp;rft.ssn=".urlencode($Document->fields["JournalSeason"]); }
		if($Document->fields["JournalQuarter"]){ $URL .= "&amp;rft.quarter=".urlencode($Document->fields["JournalQuarter"]); }
		if($Document->fields["JournalQuarter"]){ $URL .= "&amp;rft.quarter=".urlencode($Document->fields["ISSN"]); }
	}
 
	// Publisher, Publication Place, and ISBN (for books)
	if($DocType > 0){
		$URL .= "&amp;rft.pub=".urlencode($Document->fields["BookPublisher"]);
		$URL .= "&amp;rft.place=".urlencode($Document->fields["PubPlace"]);
		$URL .= "&amp;rft.isbn=".urlencode($Document->fields["ISBN"]);
	}
 
	// Start page and end page (for journals and book articles)
	if($DocType < 2){
		$URL .= "&amp;rft.spage=".urlencode($Document->fields["StartPage"]);
		$URL .= "&amp;rft.epage=".urlencode($Document->fields["EndPage"]);
	}
 
	// Publication year.
	$URL .= "&amp;rft.date=".$Document->fields["DocYear"];
 
	// Authors
	$i = 0;
	while($People[$i]){
		if($People[$i]->fields["DocRelationship"] == 0){
			$URL .= "&amp;rft.au=".urlencode($People[$i]->fields["LastName"]).",+".urlencode($People[$i]->fields["FirstName"]);
		}
		$i++;
	}
 
	return $URL;
}
?>