ruby.codefetch.com offers API and books search

[cosmo]

Greetings!

On Oct. 1 we pushed out a new version of Codefetch that supports search
of the Ruby API as well as the publically available source code from a
number of Ruby books. Try it out at http://ruby.codefetch.com

With Codefetch you can search using punctuation as part of your query
and you can use some handy regular expressions, such as .*, \b, and \s.

Enjoy!

 

[James Britt]

Greetings!

On Oct. 1 we pushed out a new version of Codefetch that supports search
of the Ruby API as well as the publically available source code from a
number of Ruby books. Try it out at http://ruby.codefetch.com

With Codefetch you can search using punctuation as part of your query
and you can use some handy regular expressions, such as .*, \b, and \s.

Enjoy!

Nice. But why does it say "Ruby Rails", rather than simply "Ruby"?

And no API docs are returned when I search, for example, on 'REXML'

James


--

http://www.ruby-doc.org - The Ruby Documentation Site
http://www.rubyxml.com - News, Articles, and Listings for Ruby & XML
http://www.rubystuff.com - The Ruby Store for Ruby Stuff
http://www.jamesbritt.com - Playing with Better Toys

 

[cosmo]

But why does it say "Ruby Rails", rather than simply "Ruby"?

Really, it should say Ruby / Rails, and HTML CSS should be HTML / CSS.
We mean to say that we include books on Rails in the index. We're
working on adding the slash.

Right now we only index the core classes, not the stdlib. Sounds like
we should include stdlib, eh? Shouldn't be a big deal-- I'll see what
I can do.

- cosmo

 

[cosmo]

Right now we only index the core classes, not the stdlib.

It looks like indexing the stdlib would flummox our scanner because
there are class names that are identical to core class names. Am I
missing something, or are some classes simply documented twice, once in
stdlib and once in core? An example of this would be the class
"abbrev".

- cosmo

 

[Austin Ziegler]

It looks like indexing the stdlib would flummox our scanner because
there are class names that are identical to core class names. Am I
missing something, or are some classes simply documented twice, once in
stdlib and once in core? An example of this would be the class
"abbrev".

How are you trying to index? You *do* realize that classes are
completely open in Ruby, don't you, allowing stdlib to add new
methods, etc. that aren't supported/in the C-based core Ruby?

And, please, as James said, it isn't "Ruby Rails". It's just "Ruby".
Rails is just an application framework on top of Ruby. Ruby makes
Rails possible, not the other way around.

-austin

 

[Gene Tani]

maybe run the ruby C source through your tokenizer/indexer and then we
can compare to gonzui

 

[cosmo]

You *do* realize that classes are
No I didn't, (I've haven't yet had the pleasure of using Ruby), and in
fact I didn't understand what you meant until right now. Hmmm. I'll
have to see how our indexer can handle this situation.

In the meantime, the core Ruby api is indexed as is the Ruby on Rails
API and also, as an experiment, the REXML API. Also, the format of api
search results has been adjusted to better accommodate long class
names.

The category is now Ruby / Rails to indicate that our search includes
info on pure Ruby as well as Ruby on Rails.

These changes went live Friday morning, Oct 7, at
http://ruby.codefetch.com

- cosmo