| - new tests for the gruber functionality - 001_gruber_rb is just 000_testcases.rb with simple_find changed to gruber_find - 003_gruber_switch.rb tests the global preference switch - TODO 003 will break if URI.extract ever finds bit.ly/whatever as a URI |
|
More...
|
over 15 years ago
|
| - allow use of the liberal regexp matcher from mr gruber - http://daringfireball.net/2010/07/improved_regex_for_matching_urls - handy for finding things like bit.ly, etc. |
|
More...
|
over 15 years ago
|
| - flickr_static: new plugin to work out flickr details from a static picture or picture page |
|
More...
|
over 16 years ago
|
| - newsbbc: different sections have different title formats, it seems - just look for a non-word character separating the parts instead of a pipe - if we cannot find anything, default to returning the original title instead of crashing |
|
More...
|
over 16 years ago
|
| Merge branch 'master' of [email protected]:rjp/linkylinky |
|
More...
|
over 16 years ago
|
| - amazon: they like to put the <title> 120k into the body of HTML so we have to use fetch_all |
|
More...
|
over 16 years ago
|
| - some sites send compressed content even without asking (foreignpolicy.com) - which means we have to deal with Content-Encoding headers - if we have one and it matches /deflate/gzip/, assuming Zlib::Inflate will work |
|
More...
|
over 16 years ago
|
| Merge branch 'master' of github.com:rjp/linkylinky |
|
More...
|
over 16 years ago
|
| - meta-title: comments for clarity - meta-title: TODO proper rdoc to describe plugins etc. |
|
More...
|
over 16 years ago
|
| - uri-find: add a new simple_find method that just returns uri.normalize.to_s - uri-find: clean up tests, add some more, etc. |
|
More...
|
over 16 years ago
|
| - comments for great ohloh justice |
|
More...
|
over 16 years ago
|
| - return the suppress_domain functionality |
|
More...
|
over 16 years ago
|
| - fetch: PNG can be broken by fetching the last 16k so special-case an exception for that |
|
More...
|
over 16 years ago
|
| - twitter: expanding shorturls for andym - TODO limit the expansion to things in the longurl list of shorteners |
|
More...
|
over 16 years ago
|
| Merge branch 'master' of [email protected]:rjp/linkylinky |
|
More...
|
over 16 years ago
|
| - twitter: shorturl expansion for andym |
|
More...
|
over 16 years ago
|
| - default: base plugin for last resort |
|
More...
|
over 16 years ago
|
| Merge branch 'master' of [email protected]:rjp/linkylinky |
|
More...
|
over 16 years ago
|
| - remove debugging output from linkyplugin - wrap curl fetching in error handling - remove leading/trailing whitespace from title - twitter: andym wants long URLs expanding so let's try that |
|
More...
|
over 16 years ago
|
| Merge branch 'master' of [email protected]:rjp/ribot |
|
More...
|
over 16 years ago
|
| - bot: topic changes arrive as groupchat but with no body |
|
More...
|
over 16 years ago
|
| - fix up whitespace by removing tabs and deleting trailing - make the queues global because it's just easier that way - add an outgoing queue so things like deli_tags don't need to know bot internals - deli: wrap deli_tags in exception handling to avoid breaking everything on error - deli: show top 8 tags and ... if there's more - deli: reformatted message as per blech's fc015019 suggestion - bot: new thread solely for fetching outgoing messages to send - bot: message handling wrapped in exception handling to avoid hanging on error - TODO report errors in URL fetching to the outgoing stream? |
|
More...
|
over 16 years ago
|
| - reorganisation and rationalisation of unit tests |
|
More...
|
over 16 years ago
|
| - flickr: rationalise the unit tests |
|
More...
|
over 16 years ago
|
| - twitter: rationalise the unit tests |
|
More...
|
over 16 years ago
|
| - .fetch returns everything we pass into .title for easier chaining - also allows an internal url rewriting step if we want to use it |
|
More...
|
over 16 years ago
|
| - flickr: update unit tests - TODO consider returning the uri from the .fetch method so you can pass that straight to the .title method - TODO also allows internal uri rewriting (static flickr image => flickr photo page for title?) |
|
More...
|
over 16 years ago
|
| - massive refactoring to treat plugins as objects - everything happens in a fetch -> title -> postfilter sequence - new method, fetch_all, does a full fetch |
|
More...
|
over 16 years ago
|
| - plan for rationalised plugin handling |
|
More...
|
over 16 years ago
|
| - newsbbc: sometimes the title only has A|B|C so our previous matching would fail - newsbbc: change the output from "B | C" to "C ((B))" to emphasise the title over category |
|
More...
|
over 16 years ago
|
