From 0a56871c84a00f77352e9b7c08a16d201dd8b51f Mon Sep 17 00:00:00 2001
From: dbs <dbs@dcc99617-32d9-48b4-a31d-7c20da2025e4>
Date: Wed, 27 Apr 2011 15:54:48 +0000
Subject: [PATCH] Teach OpenLibrary.pm how to extract ebook links from
 OpenLibrary results

If the Internet Archive has an available ebook, then OpenLibrary will
provide links to it in various formats. For now, we return the results
as a plain set of <div> elements with class attributes denoting what
kind of ebook was returned; i18n will have to take care of providing
an actual label. Oh, and I guess we want to turn these into actual
<a> links...

Signed-off-by: Dan Scott <dscott@laurentian.ca>


git-svn-id: svn://svn.open-ils.org/ILS/trunk@20344 dcc99617-32d9-48b4-a31d-7c20da2025e4
---
 .../OpenILS/WWW/AddedContent/OpenLibrary.pm   | 86 ++++++++++++++++---
 1 file changed, 75 insertions(+), 11 deletions(-)

diff --git a/Open-ILS/src/perlmods/lib/OpenILS/WWW/AddedContent/OpenLibrary.pm b/Open-ILS/src/perlmods/lib/OpenILS/WWW/AddedContent/OpenLibrary.pm
index fab9ea9aed..c3d6c07da9 100644
--- a/Open-ILS/src/perlmods/lib/OpenILS/WWW/AddedContent/OpenLibrary.pm
+++ b/Open-ILS/src/perlmods/lib/OpenILS/WWW/AddedContent/OpenLibrary.pm
@@ -1,6 +1,6 @@
 # ---------------------------------------------------------------
 # Copyright (C) 2009 David Christensen <david.a.christensen@gmail.com>
-# Copyright (C) 2009 Dan Scott <dscott@laurentian.ca>
+# Copyright (C) 2009-2011 Dan Scott <dscott@laurentian.ca>
 #
 # This program is free software; you can redistribute it and/or
 # modify it under the terms of the GNU General Public License
@@ -30,7 +30,13 @@ my $AC = 'OpenILS::WWW::AddedContent';
 
 # These URLs are always the same for OpenLibrary, so there's no advantage to
 # pulling from opensrf.xml; we hardcode them here
-my $base_url = 'http://openlibrary.org/api/books?details=true&bibkeys=ISBN:';
+
+# jscmd=details is unstable but includes goodies such as Table of Contents
+my $base_url_details = 'http://openlibrary.org/api/books?format=json&jscmd=details&bibkeys=ISBN:';
+
+# jscmd=data is stable and contains links to ebooks, excerpts, etc
+my $base_url_data = 'http://openlibrary.org/api/books?format=json&jscmd=data&bibkeys=ISBN:';
+
 my $cover_base_url = 'http://covers.openlibrary.org/b/isbn/';
 
 sub new {
@@ -60,6 +66,62 @@ sub jacket_large {
 
 # --------------------------------------------------------------------------
 
+sub ebooks_html {
+    my( $self, $key ) = @_;
+    my $book_data_json = $self->fetch_data_response($key)->content();
+
+    $logger->debug("$key: " . $book_data_json);
+
+    my $ebook_html;
+    
+    my $book_data = OpenSRF::Utils::JSON->JSON2perl($book_data_json);
+    my $book_key = (keys %$book_data)[0];
+
+    # We didn't find a matching book; short-circuit our response
+    if (!$book_key) {
+        $logger->debug("$key: no found book");
+        return 0;
+    }
+
+    my $ebooks_json = $book_data->{$book_key}->{ebooks};
+
+    # No ebooks are available for this book; short-circuit
+    if (!$ebooks_json or !scalar(@$ebooks_json)) {
+        $logger->debug("$key: no ebooks");
+        return 0;
+    }
+
+    # Check the availability of the ebooks
+    my $available = $ebooks_json->[0]->{'availability'} || '';
+    if (!$available) {
+        $logger->debug("$key: no available ebooks");
+        return 0;
+    }
+
+    # Build a basic list of available ebook types and their URLs
+    # ebooks appears to be an array containing one element - a hash
+
+    # First element of the hash is 'read_url' which is a URL to
+    # Internet Archive online reader
+    my $stream_url = $ebooks_json->[0]->{'read_url'} || '';
+    if ($stream_url) {
+        $ebook_html .= "<div class='ebook_stream'>$stream_url</div>\n";
+        $logger->debug("$key: stream URL = $stream_url");
+    }
+
+    my $ebook_formats = $ebooks_json->[0]->{'formats'} || '';
+    # Next elements are various ebook formats that are available
+    foreach my $ebook (keys %{$ebook_formats}) {
+        if ($ebook_formats->{$ebook} eq 'read_url') {
+            next;
+        }
+        $ebook_html .= "<div class='$ebook'>" . 
+            $ebook_formats->{$ebook}->{'url'} . "</div>\n";
+    }
+
+    $logger->debug("$key: $ebook_html");
+    $self->send_html("<div class='ebooks'>$ebook_html</div>");
+}
 =head1
 
 OpenLibrary returns a JSON hash of zero or more book responses matching our
@@ -74,12 +136,7 @@ HTML table.
 
 sub toc_html {
     my( $self, $key ) = @_;
-    my $book_details_json = $self->fetch_response($key)->content();
-
-
-    # Trim the "var _OlBookInfo = " declaration that makes this
-    # invalid JSON
-    $book_details_json =~ s/^.+?({.*?});$/$1/s;
+    my $book_details_json = $self->fetch_details_response($key)->content();
 
     $logger->debug("$key: " . $book_details_json);
 
@@ -127,7 +184,7 @@ sub toc_html {
 sub toc_json {
     my( $self, $key ) = @_;
     my $toc = $self->send_json(
-        $self->fetch_response($key)
+        $self->fetch_details_response($key)
     );
 }
 
@@ -167,9 +224,16 @@ sub send_html {
 }
 
 # returns the HTTP response object from the URL fetch
-sub fetch_response {
+sub fetch_data_response {
+    my( $self, $key ) = @_;
+    my $url = $base_url_data . "$key";
+    my $response = $AC->get_url($url);
+    return $response;
+}
+# returns the HTTP response object from the URL fetch
+sub fetch_details_response {
     my( $self, $key ) = @_;
-    my $url = $base_url . "$key";
+    my $url = $base_url_details . "$key";
     my $response = $AC->get_url($url);
     return $response;
 }
-- 
2.43.2