begin
require 'htmlentities'
rescue LoadError
- gems = nil
- begin
- gems = require 'rubygems'
- rescue LoadError
- gems = false
- end
- if gems
- retry
- else
module ::Irc
module Utils
UNESCAPE_TABLE = {
'raquo' => '»',
'quot' => '"',
'apos' => '\'',
+ 'deg' => '°',
'micro' => 'µ',
'copy' => '©',
'trade' => '™',
'gt' => '>',
'hellip' => '…',
'nbsp' => ' ',
+ 'ndash' => '–',
'Agrave' => 'À',
'Aacute' => 'Á',
'Acirc' => 'Â',
}
end
end
- end
end
begin
end
end
rescue LoadError
- gems = nil
- begin
- gems = require 'rubygems'
- rescue LoadError
- gems = false
- end
- if gems
- retry
- else
module ::Irc
module Utils
# Some regular expressions to manage HTML data
# Some blogging and forum platforms use spans or divs with a 'body' or 'message' or 'text' in their class
# to mark actual text
- AFTER_PAR1_REGEX = /<\w+\s+[^>]*(?:body|message|text)[^>]*>.*?<\/?(?:p|div|html|body|table|td|tr)(?:\s+[^>]*)?>/im
+ AFTER_PAR1_REGEX = /<\w+\s+[^>]*(?:body|message|text|post)[^>]*>.*?<\/?(?:p|div|html|body|table|td|tr)(?:\s+[^>]*)?>/im
# At worst, we can try stuff which is comprised between two <br>
AFTER_PAR2_REGEX = /<br(?:\s+[^>]*)?\/?>.*?<\/?(?:br|p|div|html|body|table|td|tr)(?:\s+[^>]*)?\/?>/im
end
end
- end
end
module ::Irc
def Utils.bot=(b)
debug "initializing utils"
@@bot = b
- @@safe_save_dir = "#{@@bot.botclass}/safe_save"
+ @@safe_save_dir = @@bot.path('safe_save')
end
_("%{m} minutes") % { :m => secs/SEC_PER_MIN }
when secs > 1
_("%{m} seconds") % { :m => secs }
- else
+ else
_("one second")
end
end
# Execute an external program, returning a String obtained by redirecting
- # the program's standards errors and output
+ # the program's standards errors and output
#
+ # TODO: find a way to expose some common errors (e.g. Errno::NOENT)
+ # to the caller
def Utils.safe_exec(command, *args)
- IO.popen("-") { |p|
+ output = IO.popen("-") { |p|
if p
- return p.readlines.join("\n")
+ break p.readlines.join("\n")
else
begin
$stderr.reopen($stdout)
exec(command, *args)
rescue Exception => e
- puts "exec of #{command} led to exception: #{e.pretty_inspect}"
- Kernel::exit! 0
+ puts "exception #{e.pretty_inspect} trying to run #{command}"
+ Kernel::exit! 1
end
puts "exec of #{command} failed"
- Kernel::exit! 0
+ Kernel::exit! 1
end
}
+ raise "safe execution of #{command} returned #{$?}" unless $?.success?
+ return output
end
+ # Try executing an external program, returning true if the run was successful
+ # and false otherwise
+ def Utils.try_exec(command, *args)
+ IO.popen("-") { |p|
+ if p.nil?
+ begin
+ $stderr.reopen($stdout)
+ exec(command, *args)
+ rescue Exception => e
+ Kernel::exit! 1
+ end
+ Kernel::exit! 1
+ else
+ debug p.readlines
+ end
+ }
+ debug $?
+ return $?.success?
+ end
# Safely (atomically) save to _file_, by passing a tempfile to the block
# and then moving the tempfile to its final location when done.
# Decode HTML entities in the String _str_, using HTMLEntities if the
# package was found, or UNESCAPE_TABLE otherwise.
#
- def Utils.decode_html_entities(str)
- if defined? ::HTMLEntities
- return HTMLEntities.decode_entities(str)
+
+ if defined? ::HTMLEntities
+ if ::HTMLEntities.respond_to? :decode_entities
+ def Utils.decode_html_entities(str)
+ return HTMLEntities.decode_entities(str)
+ end
else
- str.gsub(/(&(.+?);)/) {
+ @@html_entities = HTMLEntities.new
+ def Utils.decode_html_entities(str)
+ return @@html_entities.decode str
+ end
+ end
+ else
+ def Utils.decode_html_entities(str)
+ return str.gsub(/(&(.+?);)/) {
symbol = $2
# remove the 0-paddng from unicode integers
- if symbol =~ /^#(\d+)$/
+ case symbol
+ when /^#x([0-9a-fA-F]+)$/
+ symbol = $1.to_i(16).to_s
+ when /^#(\d+)$/
symbol = $1.to_i.to_s
end
# HTML first par grabber without hpricot
def Utils.ircify_first_html_par_woh(xml_org, opts={})
- xml = xml_org.gsub(/<!--.*?-->/m, '').gsub(/<script(?:\s+[^>]*)?>.*?<\/script>/im, "").gsub(/<style(?:\s+[^>]*)?>.*?<\/style>/im, "")
+ xml = xml_org.gsub(/<!--.*?-->/m,
+ "").gsub(/<script(?:\s+[^>]*)?>.*?<\/script>/im,
+ "").gsub(/<style(?:\s+[^>]*)?>.*?<\/style>/im,
+ "").gsub(/<select(?:\s+[^>]*)?>.*?<\/select>/im,
+ "")
strip = opts[:strip]
strip = Regexp.new(/^#{Regexp.escape(strip)}/) if strip.kind_of?(String)
# returns non-nil, its results are merged in _ds_ and returned. Otherwise
# nil is returned.
#
- # The input DataStream shuold have the downloaded HTML as primary key
+ # The input DataStream should have the downloaded HTML as primary key
# (:text) and possibly a :headers key holding the resonse headers.
#
def Utils.try_htmlinfo_filters(ds)
cur = nil
# TODO filter priority
filters.each { |n|
- debug "testing filter #{n}"
+ debug "testing htmlinfo filter #{n}"
cur = @@bot.filter(@@bot.global_filter_name(n, :htmlinfo), ds)
debug "returned #{cur.pretty_inspect}"
break if cur
return retval
end
+ # Returns a comma separated list except for the last element
+ # which is joined in with specified conjunction
+ #
+ def Utils.comma_list(words, options={})
+ defaults = { :join_with => ", ", :join_last_with => _(" and ") }
+ opts = defaults.merge(options)
+
+ if words.size < 2
+ words.last
+ else
+ [words[0..-2].join(opts[:join_with]), words.last].join(opts[:join_last_with])
+ end
+ end
+
end
end
projects / user / henk / code / ruby / rbot.git / blobdiff