How are characters indexed in the indices values of entities? My guess would be 
that they are indexed as Unicode code points--not bytes—and that the indexes 
refer to the text before entity expansion (“&” -> “&”) is done. Is that 


Like I mentioned in the previous thread, it would be very useful if we could 
use the entities construct when posting. First, it would help to mitigate 
problems when replying to somebody that has changed their username since the 
tweet you are replying to. Also, it would allow an application to clarify what 
is exactly in a URL. For example, “Check out” should 
have the URL parsed without the trailing period, but currently Twitter 
considers the period to be part of the URL.


>From other observations it seems like Twitter has already built a short-link 
>expansion service that is used internally. Will expanded links be exposed via 
>this entities mechanism anytime soon?





From: Taylor Singletary


Raffi's already introduced the concept of entities to you in a previous post:

You can now retrieve entities for tweets by specifying a include_entities=true 
parameter to statuses/home_timeline, statuses/user_timeline, 
statuses/friends_timeline, and statuses/mentions API calls to receive 
additional per-tweet payloads dissecting parse-able elements from the tweet 
body like @mentions, links, and hashtags. It's really cool! Some examples of 
how entities are represented can be found here: 

Reply via email to