clutterbuck-router 0.0.1
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/.gitignore +4 -0
- data/.yardopts +1 -0
- data/LICENCE +674 -0
- data/README.md +103 -0
- data/clutterbuck-router.gemspec +33 -0
- data/config.ru +4 -0
- data/lib/.gitkeep +0 -0
- data/lib/clutterbuck-router.rb +1 -0
- data/lib/clutterbuck/router.rb +269 -0
- data/lib/clutterbuck/router/route.rb +59 -0
- metadata +200 -0
data/README.md
ADDED
@@ -0,0 +1,103 @@
|
|
1
|
+
This is a fairly minimal, unobtrusive request routing library, part of the
|
2
|
+
Clutterbuck Web Application Construction Kit.
|
3
|
+
|
4
|
+
|
5
|
+
# Installation
|
6
|
+
|
7
|
+
It's a gem:
|
8
|
+
|
9
|
+
gem install clutterbuck-router
|
10
|
+
|
11
|
+
There's also the wonders of [the Gemfile](http://bundler.io):
|
12
|
+
|
13
|
+
gem 'clutterbuck-router'
|
14
|
+
|
15
|
+
If you're the sturdy type that likes to run from git:
|
16
|
+
|
17
|
+
rake install
|
18
|
+
|
19
|
+
Or, if you've eschewed the convenience of Rubygems entirely, then you
|
20
|
+
presumably know what to do already.
|
21
|
+
|
22
|
+
|
23
|
+
# Usage
|
24
|
+
|
25
|
+
Load the code:
|
26
|
+
|
27
|
+
require 'clutterbuck-router'
|
28
|
+
|
29
|
+
Then include {Clutterbuck::Router} in any class you wish to be a Rack
|
30
|
+
application using the Clutterbuck router:
|
31
|
+
|
32
|
+
class ExampleApp
|
33
|
+
include Clutterbuck::Router
|
34
|
+
end
|
35
|
+
|
36
|
+
Now, you define the routes you wish the application to respond to:
|
37
|
+
|
38
|
+
class ExampleApp
|
39
|
+
include Clutterbuck::Router
|
40
|
+
|
41
|
+
get '/' do
|
42
|
+
[200, [["Content-Type", "text/plain"]], ["Ohai!"]]
|
43
|
+
end
|
44
|
+
|
45
|
+
post %r{^/mail/([^/]+)$} do |path_opt|
|
46
|
+
[200, [["Content-Type", "text/plain"]],
|
47
|
+
["You posted #{env['rack.input'].read} to #{path_opt}"]
|
48
|
+
]
|
49
|
+
end
|
50
|
+
end
|
51
|
+
|
52
|
+
The above example pretty much demonstrates all the features that
|
53
|
+
{Clutterbuck::Router} provides. You define routes by means of methods named
|
54
|
+
after the HTTP verb to respond to, and the path is either a string or a
|
55
|
+
regex. If you specify a string, then the path provided must match *exactly*
|
56
|
+
the request path. If you specify a regex, then the first route which
|
57
|
+
matches the request path and method gets run, with any captured
|
58
|
+
subexpressions (ie "the bits in the parentheses") get passed as arguments to
|
59
|
+
the block.
|
60
|
+
|
61
|
+
The `env` method returns the request's Rack environment; apart from that,
|
62
|
+
you're on your own as far as interacting with Rack itself -- you have to
|
63
|
+
parse out the query params and request body, and return your own response
|
64
|
+
array in the Rack-compatible format. If that all sounds like too much work,
|
65
|
+
you might want to look at `clutterbuck-request` and/or
|
66
|
+
`clutterbuck-response` to get syntactic sugar to help with those parts of
|
67
|
+
your app.
|
68
|
+
|
69
|
+
Once your app is crafted to your liking, you can put it into a `config.ru`:
|
70
|
+
|
71
|
+
require 'example_app'
|
72
|
+
|
73
|
+
use ExampleApp
|
74
|
+
|
75
|
+
Fire that up via `rackup`, and you're off and running.
|
76
|
+
|
77
|
+
|
78
|
+
# Contributing
|
79
|
+
|
80
|
+
Bug reports should be sent to the [Github issue
|
81
|
+
tracker](https://github.com/mpalmer/clutterbuck-router/issues), or
|
82
|
+
[e-mailed](mailto:theshed+clutterbuck@hezmatt.org). Patches can be sent as a
|
83
|
+
Github pull request, or [e-mailed](mailto:theshed+clutterbuck@hezmatt.org).
|
84
|
+
|
85
|
+
|
86
|
+
# Licence
|
87
|
+
|
88
|
+
Unless otherwise stated, everything in this repo is covered by the following
|
89
|
+
copyright notice:
|
90
|
+
|
91
|
+
Copyright (C) 2015 Matt Palmer <matt@hezmatt.org>
|
92
|
+
|
93
|
+
This program is free software: you can redistribute it and/or modify it
|
94
|
+
under the terms of the GNU General Public License version 3, as
|
95
|
+
published by the Free Software Foundation.
|
96
|
+
|
97
|
+
This program is distributed in the hope that it will be useful,
|
98
|
+
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
99
|
+
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
100
|
+
GNU General Public License for more details.
|
101
|
+
|
102
|
+
You should have received a copy of the GNU General Public License
|
103
|
+
along with this program. If not, see <http://www.gnu.org/licenses/>.
|
@@ -0,0 +1,33 @@
|
|
1
|
+
require 'git-version-bump' rescue nil
|
2
|
+
|
3
|
+
Gem::Specification.new do |s|
|
4
|
+
s.name = "clutterbuck-router"
|
5
|
+
|
6
|
+
s.version = GVB.version rescue "0.0.0.1.NOGVB"
|
7
|
+
s.date = GVB.date rescue Time.now.strftime("%Y-%m-%d")
|
8
|
+
|
9
|
+
s.platform = Gem::Platform::RUBY
|
10
|
+
|
11
|
+
s.summary = "Rack-based minimal request router"
|
12
|
+
|
13
|
+
s.authors = ["Matt Palmer"]
|
14
|
+
s.email = ["theshed+clutterbuck@hezmatt.org"]
|
15
|
+
s.homepage = "http://theshed.hezmatt.org/clutterbuck"
|
16
|
+
|
17
|
+
s.files = `git ls-files -z`.split("\0").reject { |f| f =~ /^(G|spec|Rakefile)/ }
|
18
|
+
|
19
|
+
s.required_ruby_version = ">= 1.9.3"
|
20
|
+
|
21
|
+
s.add_runtime_dependency "rack", "~> 1.5"
|
22
|
+
|
23
|
+
s.add_development_dependency 'bundler'
|
24
|
+
s.add_development_dependency 'github-release'
|
25
|
+
s.add_development_dependency 'guard-spork'
|
26
|
+
s.add_development_dependency 'guard-rspec'
|
27
|
+
s.add_development_dependency 'rake', '~> 10.4', '>= 10.4.2'
|
28
|
+
# Needed for guard
|
29
|
+
s.add_development_dependency 'rb-inotify', '~> 0.9'
|
30
|
+
s.add_development_dependency 'redcarpet'
|
31
|
+
s.add_development_dependency 'rspec'
|
32
|
+
s.add_development_dependency 'yard'
|
33
|
+
end
|
data/config.ru
ADDED
data/lib/.gitkeep
ADDED
File without changes
|
@@ -0,0 +1 @@
|
|
1
|
+
require 'clutterbuck/router'
|
@@ -0,0 +1,269 @@
|
|
1
|
+
require 'clutterbuck/router/route'
|
2
|
+
|
3
|
+
#:nodoc:
|
4
|
+
module Clutterbuck; end
|
5
|
+
|
6
|
+
# A minimal router for Rack-compatible web applications.
|
7
|
+
#
|
8
|
+
# Does request routing, and *only* request routing. You specify
|
9
|
+
# routes as strings (exact match) or regexes, and then when the
|
10
|
+
# app gets a request, the appropriate route gets executed.
|
11
|
+
#
|
12
|
+
module Clutterbuck::Router
|
13
|
+
#:nodoc:
|
14
|
+
# Signals that the router got a 404. Should never escape the app call.
|
15
|
+
#
|
16
|
+
class NotFoundError < StandardError; end
|
17
|
+
|
18
|
+
#:nodoc:
|
19
|
+
# Signals that the router got a 405. Should never escape the app call.
|
20
|
+
#
|
21
|
+
class MethodNotAllowedError < StandardError; end
|
22
|
+
|
23
|
+
# All of the methods to define the app's routing behaviour are defined
|
24
|
+
# on the class, because that's where the config lives. Instances of the
|
25
|
+
# app class are created to handle requests.
|
26
|
+
#
|
27
|
+
module ClassMethods
|
28
|
+
# Process a request from the app.
|
29
|
+
#
|
30
|
+
def call(env)
|
31
|
+
self.new(env).call
|
32
|
+
end
|
33
|
+
|
34
|
+
# @!macro handler_args
|
35
|
+
# See {.add_handler} for all the gory details.
|
36
|
+
#
|
37
|
+
# @param path [String, Regexp]
|
38
|
+
#
|
39
|
+
# @param block [Proc]
|
40
|
+
#
|
41
|
+
# @return void
|
42
|
+
#
|
43
|
+
|
44
|
+
# Define a handler for `GET <path>` and `HEAD <path>` requests.
|
45
|
+
#
|
46
|
+
# @macro handler_args
|
47
|
+
#
|
48
|
+
def get(path, &block)
|
49
|
+
add_handler('GET', path, &block)
|
50
|
+
add_handler('HEAD', path, &block)
|
51
|
+
end
|
52
|
+
|
53
|
+
# Define a handler for `PUT <path>` requests.
|
54
|
+
#
|
55
|
+
# @macro handler_args
|
56
|
+
#
|
57
|
+
def put(path, &block)
|
58
|
+
add_handler('PUT', path, &block)
|
59
|
+
end
|
60
|
+
|
61
|
+
# Define a handler for `POST <path>` requests.
|
62
|
+
#
|
63
|
+
# @macro handler_args
|
64
|
+
#
|
65
|
+
def post(path, &block)
|
66
|
+
add_handler('POST', path, &block)
|
67
|
+
end
|
68
|
+
|
69
|
+
# Define a handler for `DELETE <path>` requests.
|
70
|
+
#
|
71
|
+
# @macro handler_args
|
72
|
+
#
|
73
|
+
def delete(path, &block)
|
74
|
+
add_handler('DELETE', path, &block)
|
75
|
+
end
|
76
|
+
|
77
|
+
# Define a handler for `PATCH <path>` requests.
|
78
|
+
#
|
79
|
+
# @macro handler_args
|
80
|
+
#
|
81
|
+
def patch(path, &block)
|
82
|
+
add_handler('PATCH', path, &block)
|
83
|
+
end
|
84
|
+
|
85
|
+
# Define a handler for an arbitrary HTTP method and path.
|
86
|
+
#
|
87
|
+
# If more than one handler for a given `verb` and `path` match a URL,
|
88
|
+
# then the first handler defined will be called.
|
89
|
+
#
|
90
|
+
# If no handler matches given `path`, then `404 Not Found` will be
|
91
|
+
# returned. If a handler exists for `path`, but does not support the
|
92
|
+
# method of the request, then `405 Method Not Allowed` will be
|
93
|
+
# returned to the client.
|
94
|
+
#
|
95
|
+
# The return value of the `block` can be one of a number of different
|
96
|
+
# things. If you set the `Content-Type` response header (by calling
|
97
|
+
# `set_header 'Content-Type', <something>`), then no special
|
98
|
+
# processing is done and your content is sent to the client
|
99
|
+
# more-or-less as-is, with only `Content-Length` calculation and
|
100
|
+
# wrapping your returned object in an array (if it doesn't already
|
101
|
+
# respond to `#each`, as required by Rack). This allows your API to
|
102
|
+
# return anything at all if it feels like it.
|
103
|
+
#
|
104
|
+
# However, by far the most common response will be a JSON document.
|
105
|
+
# If you don't set a `Content-Type` header in your handler, then we
|
106
|
+
# try quite hard to turn what you return from your handler into either
|
107
|
+
# JSON (if possible), or `text/plain`.
|
108
|
+
#
|
109
|
+
# For starters, if what you send back responds to `#each`, then we
|
110
|
+
# assume that you know what you're doing and will be sending back
|
111
|
+
# strings that will come together to be valid JSON. We'll set
|
112
|
+
# `Content-Type` and `Link` headers to match with the handler's
|
113
|
+
# schema, and that is that.
|
114
|
+
#
|
115
|
+
# If what you send back doesn't respond to `#each`, then we try to
|
116
|
+
# determine if its valid JSON by parsing it as JSON (if it's a string)
|
117
|
+
# or trying to call `#to_json` on it. If either of those work, then
|
118
|
+
# `Content-Type` and `Link` are set to match the schema for your
|
119
|
+
# handler, and all is well. Otherwise, we're kinda out of options and
|
120
|
+
# we'll send back the content as `text/plain` and hope the client
|
121
|
+
# knows what to do with it.
|
122
|
+
#
|
123
|
+
# @param verb [String] the **case sensitive** HTTP method which you
|
124
|
+
# wish this handler to respond for. If you're using the
|
125
|
+
# `get`/`post/`put`/etc wrappers for `add_handler`, this argument is
|
126
|
+
# taken care of for you. It's only if you want to define your own
|
127
|
+
# custom HTTP verbs that you'd ever need to worry about this
|
128
|
+
# argument.
|
129
|
+
#
|
130
|
+
# @param path [String, Regexp] defines the path which is to be handled
|
131
|
+
# by this handler. The path is defined relative to the root of the
|
132
|
+
# application; that is, there may be path components in the request
|
133
|
+
# URI which won't be matched against, because they're handled by the
|
134
|
+
# webserver or Rack itself (if the app is routed to via a `map`
|
135
|
+
# block, for example).
|
136
|
+
#
|
137
|
+
# If `path` is a string, the matching logic is very simple: if the
|
138
|
+
# path of the request matches exactly with `path`, then we run this
|
139
|
+
# handler. If not, we skip it.
|
140
|
+
#
|
141
|
+
# If `path` is a regex, things are ever so slightly more
|
142
|
+
# complicated. In that instance, we'll run the handler if the given
|
143
|
+
# regex matches the path of the request. In addition, any capturing
|
144
|
+
# subexpressions (aka "the bits in parentheses") in the regular
|
145
|
+
# expression will be passed as arguments to the handler block.
|
146
|
+
#
|
147
|
+
# In almost all cases, you'll want to anchor your regular
|
148
|
+
# expressions (surround them in `^` and `$`); while it's very
|
149
|
+
# unlikely that you'll want to handle a URL with `/foo` anywhere in
|
150
|
+
# the path, we don't want to *forbid* you from doing so, so there's
|
151
|
+
# no automatic anchoring of regexes.
|
152
|
+
#
|
153
|
+
# @param block [Proc] the code to execute when this handler is
|
154
|
+
# invoked. An arbitrary number of arguments may be passed to this
|
155
|
+
# block, if `path` is a regex which contains capturing
|
156
|
+
# subexpressions (that is, parts of the regular expression
|
157
|
+
# surrounded by unescaped parentheses). This is useful to capture
|
158
|
+
# portions of the URL, such as resource IDs, and feed them into your
|
159
|
+
# handler as arguments.
|
160
|
+
#
|
161
|
+
# @raise [ArgumentError] if you don't pass in a block, or you pass an
|
162
|
+
# invalid type for `path`.
|
163
|
+
#
|
164
|
+
def add_handler(verb, path, &block)
|
165
|
+
unless block_given?
|
166
|
+
raise ArgumentError,
|
167
|
+
"Must pass a block"
|
168
|
+
end
|
169
|
+
|
170
|
+
@routes ||= []
|
171
|
+
@routes << Route.new(verb, path, method_for(verb, path, block))
|
172
|
+
end
|
173
|
+
|
174
|
+
# :nodoc:
|
175
|
+
# Grovel through the list of routes, looking for a match.
|
176
|
+
#
|
177
|
+
# @raise [Clutterbuck::Router::NotFoundError] if no route matched the
|
178
|
+
# given path.
|
179
|
+
#
|
180
|
+
# @raise [Clutterbuck::Router::MethodNotAllowedError] if a route matched
|
181
|
+
# the given path, but it didn't have the right verb.
|
182
|
+
#
|
183
|
+
def find_route(verb, path)
|
184
|
+
if @routes.nil?
|
185
|
+
raise Clutterbuck::Router::NotFoundError,
|
186
|
+
path
|
187
|
+
end
|
188
|
+
|
189
|
+
candidates = @routes.select { |r| r.handles?(path) }
|
190
|
+
|
191
|
+
if candidates.empty?
|
192
|
+
raise Clutterbuck::Router::NotFoundError,
|
193
|
+
path
|
194
|
+
end
|
195
|
+
|
196
|
+
candidates = candidates.select { |r| r.verb == verb }
|
197
|
+
|
198
|
+
if candidates.empty?
|
199
|
+
raise Clutterbuck::Router::MethodNotAllowedError,
|
200
|
+
"#{verb} not permitted on #{path}"
|
201
|
+
end
|
202
|
+
|
203
|
+
candidates.first
|
204
|
+
end
|
205
|
+
|
206
|
+
private
|
207
|
+
|
208
|
+
# Get an unbound instance method for the given verb/path/block.
|
209
|
+
#
|
210
|
+
# @param verb [String]
|
211
|
+
#
|
212
|
+
# @param path [String, Regexp]
|
213
|
+
#
|
214
|
+
# @param block [Proc]
|
215
|
+
#
|
216
|
+
# @return [UnboundMethod]
|
217
|
+
#
|
218
|
+
def method_for(verb, path, block)
|
219
|
+
name = "#{verb} #{path}".to_sym
|
220
|
+
define_method(name, &block)
|
221
|
+
instance_method(name).tap do |m|
|
222
|
+
remove_method name
|
223
|
+
end
|
224
|
+
end
|
225
|
+
end
|
226
|
+
|
227
|
+
#:nodoc:
|
228
|
+
# Add in the class-level methods to the app.
|
229
|
+
#
|
230
|
+
def self.included(mod)
|
231
|
+
mod.extend(ClassMethods)
|
232
|
+
end
|
233
|
+
|
234
|
+
# Create a new instance of the app.
|
235
|
+
#
|
236
|
+
# @param env [Hash] The Rack environment for this request.
|
237
|
+
#
|
238
|
+
def initialize(env)
|
239
|
+
@env = env
|
240
|
+
end
|
241
|
+
|
242
|
+
# Handle the request.
|
243
|
+
#
|
244
|
+
# Find the route, run it, and return the result.
|
245
|
+
#
|
246
|
+
def call
|
247
|
+
path = env["PATH_INFO"].empty? ? "/" : env["PATH_INFO"]
|
248
|
+
|
249
|
+
begin
|
250
|
+
route = self.class.find_route(env["REQUEST_METHOD"], path)
|
251
|
+
rescue Clutterbuck::Router::NotFoundError
|
252
|
+
return [404, [["Content-Type", "text/plain"]], ["Not found"]]
|
253
|
+
rescue Clutterbuck::Router::MethodNotAllowedError => ex
|
254
|
+
return [405, [["Content-Type", "text/plain"]], [ex]]
|
255
|
+
end
|
256
|
+
|
257
|
+
route.run(self, path).tap do |response|
|
258
|
+
if env["REQUEST_METHOD"] == "HEAD"
|
259
|
+
response[2] = []
|
260
|
+
end
|
261
|
+
end
|
262
|
+
end
|
263
|
+
|
264
|
+
protected
|
265
|
+
# Return the Rack environment for this request.
|
266
|
+
def env
|
267
|
+
@env
|
268
|
+
end
|
269
|
+
end
|
@@ -0,0 +1,59 @@
|
|
1
|
+
#:nodoc:
|
2
|
+
module Clutterbuck; end
|
3
|
+
#:nodoc:
|
4
|
+
module Clutterbuck::Router; end
|
5
|
+
|
6
|
+
# A route within the application.
|
7
|
+
#
|
8
|
+
# Not something you should ever have to deal with yourself directly, it's
|
9
|
+
# part of the internal plumbing of {Clutterbuck::Router}.
|
10
|
+
#
|
11
|
+
class Clutterbuck::Router::Route
|
12
|
+
attr_accessor :verb, :path_match
|
13
|
+
|
14
|
+
# @param verb [String]
|
15
|
+
#
|
16
|
+
# @param path_match [String, Regexp]
|
17
|
+
#
|
18
|
+
# @param method [UnboundMethod] is a method to call on an instance of the
|
19
|
+
# app class this route is defined on, which will do whatever is needed
|
20
|
+
# to handle the route. Because you can't just create an
|
21
|
+
# `UnboundMethod` out of thin air, the `UnboundMethod` needs to be
|
22
|
+
# created in the class, and then passed into here. Ugly.
|
23
|
+
#
|
24
|
+
def initialize(verb, path_match, method)
|
25
|
+
unless path_match.is_a?(String) or path_match.is_a?(Regexp)
|
26
|
+
raise ArgumentError,
|
27
|
+
"path must be either a string or a regexp"
|
28
|
+
end
|
29
|
+
|
30
|
+
@verb, @path_match, @method = verb, path_match, method
|
31
|
+
end
|
32
|
+
|
33
|
+
# Can this route handle a request for the specified path?
|
34
|
+
#
|
35
|
+
# @param path [String]
|
36
|
+
#
|
37
|
+
# @return Boolean
|
38
|
+
#
|
39
|
+
def handles?(path)
|
40
|
+
!!case @path_match
|
41
|
+
when String
|
42
|
+
@path_match == path
|
43
|
+
when Regexp
|
44
|
+
@path_match =~ path
|
45
|
+
end
|
46
|
+
end
|
47
|
+
|
48
|
+
# Execute the handler for the route
|
49
|
+
#
|
50
|
+
# Run the method
|
51
|
+
def run(obj, path)
|
52
|
+
args = case @path_match
|
53
|
+
when String then []
|
54
|
+
when Regexp then @path_match.match(path)[1..-1]
|
55
|
+
end
|
56
|
+
|
57
|
+
@method.bind(obj).call(*args)
|
58
|
+
end
|
59
|
+
end
|