4 # :title: rbot HTTP provider
6 # Author:: Tom Gilbert <tom@linuxbrit.co.uk>
7 # Author:: Giuseppe "Oblomov" Bilotta <giuseppe.bilotta@gmail.com>
8 # Author:: Dmitry "jsn" Kim <dmitry point kim at gmail point com>
10 # Copyright:: (C) 2002-2005 Tom Gilbert
11 # Copyright:: (C) 2006 Tom Gilbert, Giuseppe Bilotta
12 # Copyright:: (C) 2007 Giuseppe Bilotta, Dmitry Kim
20 error "Couldn't load 'net/https': #{e.inspect}"
21 error "Secured HTTP connections will fail"
26 attr_accessor :no_cache
27 if !instance_methods.include?('raw_body')
31 def body_charset(str=self.raw_body)
32 ctype = self['content-type'] || 'text/html'
33 return nil unless ctype =~ /^text/i || ctype =~ /x(ht)?ml/i
35 charset = 'latin1' # should be in config
37 if self['content-type'].match(/charset=["']?([^\s"']+)["']?/i)
39 debug "charset #{charset} set from header"
43 when /<\?xml\s[^>]*encoding=['"]([^\s"'>]+)["'][^>]*\?>/i
45 debug "xml charset #{charset} set from xml pi"
46 when /<(meta\s[^>]*http-equiv=["']?Content-Type["']?[^>]*)>/i
48 if meta =~ /charset=['"]?([^\s'";]+)['"]?/
50 debug "html charset #{charset} set from meta"
57 charset = self.body_charset(str) or return str
60 return Iconv.iconv('utf-8//ignore', charset, str).first
62 debug "conversion failed"
68 return self.body_to_utf(self.raw_body)
71 # Read chunks from the body until we have at least _size_ bytes, yielding
72 # the partial text at each chunk. Return the partial body.
73 def partial_body(size=0, &block)
78 self.read_body { |chunk|
80 yield self.body_to_utf(partial) if block_given?
81 break if size and size > 0 and partial.length >= size
84 return self.body_to_utf(partial)
94 # class for making http requests easier (mainly for plugins to use)
95 # this class can check the bot proxy configuration to determine if a proxy
96 # needs to be used, which includes support for per-url proxy configuration.
98 BotConfig.register BotConfigBooleanValue.new('http.use_proxy',
99 :default => false, :desc => "should a proxy be used for HTTP requests?")
100 BotConfig.register BotConfigStringValue.new('http.proxy_uri', :default => false,
101 :desc => "Proxy server to use for HTTP requests (URI, e.g http://proxy.host:port)")
102 BotConfig.register BotConfigStringValue.new('http.proxy_user',
104 :desc => "User for authenticating with the http proxy (if required)")
105 BotConfig.register BotConfigStringValue.new('http.proxy_pass',
107 :desc => "Password for authenticating with the http proxy (if required)")
108 BotConfig.register BotConfigArrayValue.new('http.proxy_include',
110 :desc => "List of regexps to check against a URI's hostname/ip to see if we should use the proxy to access this URI. All URIs are proxied by default if the proxy is set, so this is only required to re-include URIs that might have been excluded by the exclude list. e.g. exclude /.*\.foo\.com/, include bar\.foo\.com")
111 BotConfig.register BotConfigArrayValue.new('http.proxy_exclude',
113 :desc => "List of regexps to check against a URI's hostname/ip to see if we should use avoid the proxy to access this URI and access it directly")
114 BotConfig.register BotConfigIntegerValue.new('http.max_redir',
116 :desc => "Maximum number of redirections to be used when getting a document")
117 BotConfig.register BotConfigIntegerValue.new('http.expire_time',
119 :desc => "After how many minutes since last use a cached document is considered to be expired")
120 BotConfig.register BotConfigIntegerValue.new('http.max_cache_time',
122 :desc => "After how many minutes since first use a cached document is considered to be expired")
123 BotConfig.register BotConfigIntegerValue.new('http.no_expire_cache',
125 :desc => "Set this to true if you want the bot to never expire the cached pages")
126 BotConfig.register BotConfigIntegerValue.new('http.info_bytes',
128 :desc => "How many bytes to download from a web page to find some information. Set to 0 to let the bot download the whole page.")
131 attr_accessor :response, :last_used, :first_used, :count, :expires, :date
133 def self.maybe_new(resp)
134 debug "maybe new #{resp}"
135 return nil if resp.no_cache
136 return nil unless Net::HTTPOK === resp ||
137 Net::HTTPMovedPermanently === resp ||
138 Net::HTTPFound === resp ||
139 Net::HTTPPartialContent === resp
141 cc = resp['cache-control']
142 return nil if cc && (cc =~ /no-cache/i)
146 date = Time.httpdate(d)
149 return nil if resp['expires'] && (Time.httpdate(resp['expires']) < date)
151 debug "creating cache obj"
158 @first_used = now if @count == 0
164 debug "checking expired?"
165 if cc = self.response['cache-control'] && cc =~ /must-revalidate/
168 return self.expires < Time.now
171 def setup_headers(hdr)
172 hdr['if-modified-since'] = self.date.rfc2822
174 debug "ims == #{hdr['if-modified-since']}"
176 if etag = self.response['etag']
177 hdr['if-none-match'] = etag
178 debug "etag: #{etag}"
182 def revalidate(resp = self.response)
185 self.date = resp.key?('date') ? Time.httpdate(resp['date']) : Time.now
187 cc = resp['cache-control']
188 if cc && (cc =~ /max-age=(\d+)/)
189 self.expires = self.date + $1.to_i
190 elsif resp.key?('expires')
191 self.expires = Time.httpdate(resp['expires'])
192 elsif lm = resp['last-modified']
193 delta = self.date - Time.httpdate(lm)
194 delta = 10 if delta <= 0
196 self.expires = self.date + delta
198 self.expires = self.date + 300
200 # self.expires = Time.now + 10 # DEBUG
201 debug "expires on #{self.expires}"
211 self.response.raw_body
212 rescue Exception => e
214 error e.backtrace.join("\n")
224 'Accept-Charset' => 'utf-8;q=1.0, *;q=0.8',
226 "rbot http util #{$version} (http://linuxbrit.co.uk/rbot/)"
228 debug "starting http cache cleanup timer"
229 @timer = @bot.timer.add(300) {
230 self.remove_stale_cache unless @bot.config['http.no_expire_cache']
235 debug 'stopping http cache cleanup timer'
236 @bot.timer.remove(@timer)
239 # if http_proxy_include or http_proxy_exclude are set, then examine the
240 # uri to see if this is a proxied uri
241 # the in/excludes are a list of regexps, and each regexp is checked against
242 # the server name, and its IP addresses
243 def proxy_required(uri)
245 if @bot.config["http.proxy_exclude"].empty? && @bot.config["http.proxy_include"].empty?
251 list.concat Resolv.getaddresses(uri.host)
252 rescue StandardError => err
253 warning "couldn't resolve host uri.host"
256 unless @bot.config["http.proxy_exclude"].empty?
257 re = @bot.config["http.proxy_exclude"].collect{|r| Regexp.new(r)}
267 unless @bot.config["http.proxy_include"].empty?
268 re = @bot.config["http.proxy_include"].collect{|r| Regexp.new(r)}
278 debug "using proxy for uri #{uri}?: #{use_proxy}"
282 # uri:: Uri to create a proxy for
284 # return a net/http Proxy object, which is configured correctly for
285 # proxying based on the bot's proxy configuration.
286 # This will include per-url proxy configuration based on the bot config
287 # +http_proxy_include/exclude+ options.
289 def get_proxy(uri, options = {})
301 if @bot.config["http.use_proxy"]
302 if (ENV['http_proxy'])
303 proxy = URI.parse ENV['http_proxy'] rescue nil
305 if (@bot.config["http.proxy_uri"])
306 proxy = URI.parse @bot.config["http.proxy_uri"] rescue nil
309 debug "proxy is set to #{proxy.host} port #{proxy.port}"
310 if proxy_required(uri)
311 proxy_host = proxy.host
312 proxy_port = proxy.port
313 proxy_user = @bot.config["http.proxy_user"]
314 proxy_pass = @bot.config["http.proxy_pass"]
319 h = Net::HTTP.new(uri.host, uri.port, proxy_host, proxy_port, proxy_user, proxy_port)
320 h.use_ssl = true if uri.scheme == "https"
322 h.read_timeout = opts[:read_timeout]
323 h.open_timeout = opts[:open_timeout]
327 def handle_response(uri, resp, opts, &block)
328 if Net::HTTPRedirection === resp && opts[:max_redir] >= 0
329 if resp.key?('location')
330 raise 'Too many redirections' if opts[:max_redir] <= 0
331 yield resp if opts[:yield] == :all && block_given?
332 loc = resp['location']
333 new_loc = URI.join(uri.to_s, loc) rescue URI.parse(loc)
335 new_opts[:max_redir] -= 1
336 case opts[:method].to_s.downcase.intern
337 when :post, :"net::http::post"
338 new_opts[:method] = :get
340 debug "following the redirect to #{new_loc}"
341 return get_response(new_loc, new_opts, &block)
343 warning ":| redirect w/o location?"
349 # Net::HTTP wants us to read the whole body here
355 # uri:: uri to query (Uri object or String)
356 # opts:: options. Currently used:
357 # :method:: request method [:get (default), :post or :head]
358 # :open_timeout:: open timeout for the proxy
359 # :read_timeout:: read timeout for the proxy
360 # :cache:: should we cache results?
361 # :yield:: if :final [default], call &block for the response object
362 # if :all, call &block for all intermediate redirects, too
363 # :max_redir:: how many redirects to follow before raising the exception
364 # if -1, don't follow redirects, just return them
365 # :range:: make a ranged request (usually GET). accepts a string
366 # for HTTP/1.1 "Range:" header (i.e. "bytes=0-1000")
367 # :body:: request body (usually for POST requests)
369 # Generic http transaction method
371 # It will return a HTTP::Response object or raise an exception
373 # If a block is given, it will yield the response (see :yield option)
375 def get_response(uri_or_s, options = {}, &block)
376 uri = uri_or_s.kind_of?(URI) ? uri_or_s : URI.parse(uri_or_s.to_s)
378 :max_redir => @bot.config['http.max_redir'],
387 req_class = case opts[:method].to_s.downcase.intern
388 when :head, :"net::http::head"
389 opts[:max_redir] = -1
391 when :get, :"net::http::get"
393 when :post, :"net::http::post"
395 opts[:body] or raise 'post request w/o a body?'
396 warning "refusing to cache POST request" if options[:cache]
399 warning "unsupported method #{opts[:method]}, doing GET"
403 if req_class != Net::HTTP::Get && opts[:range]
404 warning "can't request ranges for #{req_class}"
408 cache_key = "#{opts[:range]}|#{req_class}|#{uri.to_s}"
410 if req_class != Net::HTTP::Get && req_class != Net::HTTP::Head
412 warning "can't cache #{req_class.inspect} requests, working w/o cache"
417 debug "get_response(#{uri}, #{opts.inspect})"
419 if opts[:cache] && cached = @cache[cache_key]
424 return handle_response(uri, cached.response, opts, &block)
428 headers = @headers.dup.merge(opts[:headers] || {})
429 headers['Range'] = opts[:range] if opts[:range]
431 cached.setup_headers(headers) if cached && (req_class == Net::HTTP::Get)
432 req = req_class.new(uri.request_uri, headers)
433 req.basic_auth(uri.user, uri.password) if uri.user && uri.password
434 req.body = opts[:body] if req_class == Net::HTTP::Post
435 debug "prepared request: #{req.to_hash.inspect}"
437 get_proxy(uri, opts).start do |http|
438 http.request(req) do |resp|
439 if Net::HTTPNotModified === resp
442 cached.revalidate(resp)
443 rescue Exception => e
445 error e.backtrace.join("\n")
447 debug "reusing cached"
448 resp = cached.response
449 elsif Net::HTTPServerError === resp || Net::HTTPClientError === resp
450 debug "http error, deleting cached obj" if cached
451 @cache.delete(cache_key)
454 return handle_response(uri, resp, opts, &block)
456 if cached = CachedObject.maybe_new(resp) rescue nil
457 debug "storing to cache"
458 @cache[cache_key] = cached
463 return handle_response(uri, resp, opts, &block)
468 # uri:: uri to query (Uri object)
470 # simple get request, returns (if possible) response body following redirs
471 # and caching if requested
472 def get(uri, opts = {}, &block)
474 resp = get_response(uri, opts, &block)
475 raise "http error: #{resp}" unless Net::HTTPOK === resp ||
476 Net::HTTPPartialContent === resp
478 rescue Exception => e
480 error e.backtrace.join("\n")
485 def head(uri, options = {}, &block)
486 opts = {:method => :head}.merge(options)
488 resp = get_response(uri, opts, &block)
489 raise "http error #{resp}" if Net::HTTPClientError === resp ||
490 Net::HTTPServerError == resp
492 rescue Exception => e
494 error e.backtrace.join("\n")
499 def post(uri, data, options = {}, &block)
500 opts = {:method => :post, :body => data, :cache => false}.merge(options)
502 resp = get_response(uri, opts, &block)
503 raise 'http error' unless Net::HTTPOK === resp
505 rescue Exception => e
507 error e.backtrace.join("\n")
512 def get_partial(uri, nbytes = @bot.config['http.info_bytes'], options = {}, &block)
513 opts = {:range => "bytes=0-#{nbytes}"}.merge(options)
514 return get(uri, opts, &block)
517 def remove_stale_cache
518 debug "Removing stale cache"
520 max_last = @bot.config['http.expire_time'] * 60
521 max_first = @bot.config['http.max_cache_time'] * 60
522 debug "#{@cache.size} pages before"
524 @cache.reject! { |k, val|
525 (now - val.last_used > max_last) || (now - val.first_used > max_first)
528 error "Failed to remove stale cache: #{e.inspect}"
530 debug "#{@cache.size} pages after"
537 class HttpUtilPlugin < CoreBotModule
540 debug 'initializing httputil'
541 @bot.httputil = Irc::Utils::HttpUtil.new(@bot)
545 debug 'shutting down httputil'
546 @bot.httputil.cleanup