1 Revision history for Perl extension WWW::OpenSearch
4 - Avoid hash randomization problem in test (RT #82796)
7 - Fix tests for Test::More 0.95_02
10 - repackaged with a new version of Module::Install
13 - Handle broken LWP 5.810 in Makefile.PL
16 - Fix pod coverage (Ian Beckwith)
19 - switch from ||= to a "defined" idiom for some params
20 - switch to Module::Install
23 - added a simple OSD parsing test
26 - fix Request.pm to clone the extra attributes we've added (Marc Tobias)
29 - use URI::Template for parsing OSD uri templates
30 - added opensearch-specific Request and Agent classes
31 - un-break get_best_url()
32 - you can now pass a WWW::OpenSearch::Url to WWW::OpenSearch's search()
34 - added ns() (namespace) field to Url.pm
35 - re-worked paging to use the new Request object
37 [THINGS THAT MAY BREAK YOUR CODE]
38 - using URI::Template means some methods are now proxied to that class
39 - removed agent argument in new() in OpenSearch.pm
40 - handling of POST requests in prepare_query() in Url.pm now returns data
41 suitable for passing to HTTP::Request
42 - un-link Response.pm and the parent WWW::OpenSearch object -- a Response
43 is now in the context of whatever WWW::OpenSearch::Url was used
46 - fix link fetching from atom feeds
49 - fix optional attributes for Image
50 - added strict to Url object
53 - inflate query and image fields to objects for OSD 1.1
58 - updated available description info
59 - try to handle POST urls better
60 - update XML::Atom calls to latest release API
61 - added option to override the useragent from new()
63 - turn utf8 flag off for queries
66 - Complete re-write to handle OpenSearch 1.1/Atom
67 - Now uses XML::Feed to abstract RSS and Atom
68 - Improved OpenSearch Description/Url handling
69 - Encapsulate returns in a Response object
72 - Added partial support of OpenSearch 1.1. There's no API change and old
73 elements just work as before, with 1.1 OpenSearch document as well. See
74 http://opensearch.a9.com/docs/upgrading10.jsp (Thanks to Masaaki Hirose)
77 - Drop off XML::Simple for efficiency. LibXMLify everything. Now you need
81 - Hacked current_page so you can specify page number correctly (Thanks to
85 - Added accessors for attributes