sunflower 0.4.3 → 0.4.4

data/README

@@ -1,4 +1,4 @@
-Version: 0.4.3 alpha
+Version: 0.4.4 alpha
 
 >>> English:
 

data/lib/sunflower/commontasks.rb

@@ -2,7 +2,6 @@
 # extends Page with some methods letting easily perform common tasks
 
 class Page
-	def execute commands
 	# executes methods on self
 	# "commands" is array of arrays
 	# page.execute([
@@ -18,6 +17,7 @@ class Page
 	# oi:module, only-if:module
 	# !oi:module, only-if-not:module
 	# s:append to summary, summary:append to summary
+	def execute commands
 		originalText = self.text.dup
 	
 		commands.each do |cmd|
@@ -75,9 +75,9 @@ class Page
 	
 	
 	
-	def replace from, to, once=false
 	# replaces "from" with "to" in page text
 	# "from" may be regex
+	def replace from, to, once=false
 		self.text = self.text.send( (once ? 'sub' : 'gsub'), from, to )
 	end
 	def gsub from, to
@@ -87,31 +87,23 @@ class Page
 		self.replace from, to, true
 	end
 	
-	def append txt, newlines=2
 	# appends newlines and text
 	# by default - 2 newlines
+	def append txt, newlines=2
 		self.text = self.text.rstrip + ("\n"*newlines) + txt
 	end
 	
-	def prepend txt, newlines=2
 	# prepends text and newlines
 	# by default - 2 newlines
+	def prepend txt, newlines=2
 		self.text = txt + ("\n"*newlines) + self.text.lstrip
 	end
 	
-	def code_cleanup
-	# simple, safe code cleanup
-	# use Sunflower.always_do_code_cleanup=true to do it automatically just before saving page
+	# plwiki-specific cleanup routines.
 	# based on Nux's cleaner: http://pl.wikipedia.org/wiki/Wikipedysta:Nux/wp_sk.js
-		str=self.text.gsub(/\r\n/,"\n")
+	def code_cleanup_plwiki str
+		str = str.dup
 		
-		str.gsub!(/\{\{\s*([^|{}]+ |uni|)stub2?(\|[^{}]+)?\}\}/i){
-			if $1=='sekcja '
-				'{{sekcja stub}}'
-			else
-				'{{stub}}'
-			end
-		}
 		str.gsub!(/\{\{\{(?:poprzednik|następca|pop|nast|lata|info|lang)\|(.+?)\}\}\}/i,'\1')
 		str.gsub!(/(={1,5})\s*Przypisy\s*\1\s*<references\s?\/>/i){
 			if $1=='=' || $1=='=='
@@ -121,21 +113,6 @@ class Page
 			end
 		}
 		
-		str.gsub!(/\[\[([^\|#\]]*)([^\|\]]*)(\||\]\])/){
-			name, anchor, _end = $1, $2, $3
-			
-			begin
-				name = name.gsub(/((?:%[0-9a-fA-F]{2})+)/){ [$1.delete('%')].pack('H*') }
-				anchor = (anchor||'').gsub(/\.([0-9A-F]{2})/, '%\1').gsub(/((?:%[0-9a-fA-F]{2})+)/){ [$1.delete('%')].pack('H*') }
-				a='[['+name+anchor+(_end||'')
-				a=a.gsub '_', ' '
-			rescue
-				a=('[['+name+(anchor||'')+(_end||'')).gsub '_', ' '
-			end
-			
-			a
-		}
-		
 		# sklejanie skrótów linkowych
 		str.gsub!(/m\.? ?\[\[n\.? ?p\.? ?m\.?\]\]/, 'm [[n.p.m.]]');
 
@@ -165,102 +142,46 @@ class Page
 		str.gsub!(/[ \n\t]*\n'''? *((Zewnętrzn[ey] )?(Linki?|Łącza|Stron[ay]|Zobacz w (internecie|sieci))( zewn[eę]trzn[aey])?):* *'''?[ \n\t]*/i, "\n\n== Linki zewnętrzne ==\n");
 		str.gsub!(/[ \n\t]*\n(=+) *((Zewnętrzn[ey] )?(Linki?|Łącza|Stron[ay]|Zobacz w (internecie|sieci))( zewn[eę]trzn[aey])?):* *=+[ \n\t]*/i, "\n\n\\1 Linki zewnętrzne \\1\n");
 
-		# nagłówki
+		return str
+	end
+	
+	# simple, safe code cleanup
+	# use Sunflower.always_do_code_cleanup=true to do it automatically just before saving page
+	def code_cleanup
+		str = self.text.gsub /\r\n/, "\n"
+		
+		str.gsub!(/\[\[([^\|\]]+)(\||\]\])/){
+			name, rest = $1, $2
+			"[[#{self.sunflower.cleanup_title name}#{rest}"
+		}
+		
+		# headings
 		str.gsub!(/(^|\n)(=+) *([^=\n]*[^ :=\n])[ :]*=/, '\1\2 \3 ='); # =a= > = a =, =a:= > = a =
-		str.gsub!(/(^|\n)(=+[^=\n]+=+)[\n]{2,}/, "\\1\\2\n");	# jeden \n
+		str.gsub!(/(^|\n)(=+[^=\n]+=+)[\n]{2,}/, "\\1\\2\n"); # one newline
 
-		# listy ze spacjami
+		# spaced lists
 		str.gsub!(/(\n[#*:;]+)([^ \t\n#*:;{])/, '\1 \2');
 		
-		# poprawa nazw przestrzeni i drobne okoliczne
-		str.gsub!(/\[\[(:?) *(image|grafika|file|plik) *: *([^ ])/i){'[['+$1+'Plik:'+$3.upcase}
-		str.gsub!(/\[\[(:?) *(category|kategoria) *: *([^ ])/i){'[['+$1+'Kategoria:'+$3.upcase}
-		str.gsub!(/\[\[ *(:?) *(template|szablon) *: *([^ ])/i){'[['+'Szablon:'+$3.upcase}
-		str.gsub!(/\[\[ *(:?) *(special|specjalna) *: *([^ ])/i){'[['+'Specjalna:'+$3.upcase}
-		
-		3.times { str.gsub!('{{stub}}{{stub}}', '{{stub}}') }
+		if wikiid = self.sunflower.siteinfo['general']['wikiid']
+			if self.respond_to? :"code_cleanup_#{wikiid}"
+				str = self.call :"code_cleanup_#{wikiid}", str
+			end
+		end
 		
 		self.text = str
 	end
-	
-	def friendly_infobox
-	# cleans up infoboxes
-	# might make mistakes! use at your own risk!
-		def makeFriendly(nazwa,zaw)
-			zaw.gsub!(/<!--.+?-->/,'')
-			nazwa=nazwa.gsub('_',' ').strip
-			
-			#escapowanie parametrów
-			zaw.gsub!(/<<<(#+)>>>/,"<<<#\\1>>>")
-			#wewnętrzne szablony
-			while zaw=~/\{\{[^}]+\|[^}]+\}\}/
-				zaw.gsub!($&,$&.gsub(/\|/,'<<<#>>>'))
-			end
-			#wewnętrzne linki
-			while zaw=~/\[\[[^\]]+\|[^\]]+\]\]/
-				zaw.gsub!($&,$&.gsub(/\|/,'<<<#>>>'))
-			end
-			
-			zaw.sub!(/\A\s*\|\s*/,'') #usunięcie pierwszego pipe'a
-			lines=zaw.split('|')
-			
-			# te tablice przechowują odpowiednio nazwy i wartości kolejnych parametrów
-			names=[]
-			values=[]
-				
-			for line in lines
-				line.gsub!(/<<<#>>>/,'|')
-				line.gsub!(/<<<#(#+)>>>/,"<<<\\1>>>") #odescapowanie
-				
-				line=~/\A\s*(.+?)\s*=\s*([\s\S]*?)\s*\Z/
-				if $&==nil
-					next
-				end
-				name=$1.strip
-				value=$2.strip
-				
-				names<<name
-				values<<value
-			end
-			
-			zaw=''
-			names.each_index{|i|
-				zaw+=' | '+names[i]+' = '+values[i]+"\n"
-			}
-			
-			# grupowane koordynaty
-			zaw.gsub!(/\s*\| minut/, ' | minut')
-			zaw.gsub!(/\s*\| sekund/, ' | sekund')
-			
-			return '{{'+nazwa[0,1].upcase+nazwa[1,999]+"\n"+zaw+'}}'+"\n"
-		end
 
-		nstr=''
-		while str!=''
-			str=~/(\s*)\{\{([^|}]+[ _]infobo[^|}]+|[wW]ładca)((?:[^{}]|[^{}][{}][^{}]|\{\{(?:[^{}]|[^{}][{}][^{}]|\{\{[^{}]+\}\})+\}\})+)\}\}(?:\s*)/
-			
-			spaces=($1!='' ? "\n" : '')
-			before=($`==nil ? '' : $`)
-			name=$2
-			inner=$3
-			match=$&
-			if match!=nil
-				result=makeFriendly(name,inner)
-				nstr+=before+spaces+result
-			else
-				nstr+=str
-				break
-			end
-			
-			str=str.sub(before+match,'')
-		end
-	
-		self.text = nstr
-	end
-	
+	# Replace the category from with category to in page wikitext.
+	# 
+	# Inputs can be either with the Category: prefix (or localised version) or without.
 	def change_category from, to
-		from=from.sub(/\A\s*([cC]ategory|[kK]ategoria):/, '').strip
-		to=to.sub(/\A\s*([cC]ategory|[kK]ategoria):/, '').strip
-		self.text = self.text.gsub(/\[\[ *(?:[cC]ategory|[kK]ategoria) *: *#{Regexp.escape from} *(\|[^\]]*|)\]\]/){'[[Kategoria:'+to+($1=='| ' ? $1 : $1.rstrip)+']]'}
+		cat_regex = self.sunflower.ns_regex_for 'Category'
+		from = self.sunflower.cleanup_title(from).sub(/^#{cat_regex}:/, '')
+		to   = self.sunflower.cleanup_title(to  ).sub(/^#{cat_regex}:/, '')
+		
+		self.text.gsub!(/\[\[ *#{cat_regex} *: *#{Regexp.escape from} *(\||\]\])/){
+			rest = $1
+			"[[#{self.sunflower.ns_local_for 'Category'}:#{to}#{rest}"
+		}
 	end
-end
+end

data/lib/sunflower/core.rb

@@ -19,7 +19,7 @@ class SunflowerError < StandardError; end
 #
 # You can use multiple Sunflowers at once, to work on multiple wikis.
 class Sunflower
-	VERSION = '0.4.3'
+	VERSION = '0.4.4'
 	
 	INVALID_CHARS = %w(# < > [ ] | { })
 	INVALID_CHARS_REGEX = Regexp.union *INVALID_CHARS
@@ -29,14 +29,24 @@ class Sunflower
 		File.join(ENV['HOME'], 'sunflower-userdata')
 	end
 
-	# Options for this Sunflower.
-	attr_accessor :summary, :always_do_code_cleanup
-	attr_accessor :cookie, :headers, :wikiURL
+	# Summary used when saving edits with this Sunflower.
+	attr_accessor :summary
+	# Whether to run #code_cleanup when calling #save.
+	attr_accessor :always_do_code_cleanup
+	# The URL this Sunflower works on, as provided as argument to #initialize.
+	attr_reader :wikiURL
+	# Siteinfo, as returned by API call.
+	attr_accessor :siteinfo
 	
+	# Whether this user (if logged in) has bot rights.
 	def is_bot?; @is_bot; end
 	
-	attr_writer :warnings, :log
+	# Whether to output warning messages (using Kernel#warn). Defaults to true.
+	attr_writer :warnings
 	def warnings?; @warnings; end
+	
+	# Whether to output log messages (to a file named log.txt in current directory). Defaults to false.
+	attr_writer :log
 	def log?; @log; end
 	
 	# Initialize a new Sunflower working on a wiki with given URL, for ex. "pl.wikipedia.org".
@@ -62,6 +72,32 @@ class Sunflower
 		@wikiURL=url
 		
 		@loggedin=false
+		
+		siprop = 'general|namespaces|namespacealiases|specialpagealiases|magicwords|interwikimap|dbrepllag|statistics|usergroups|extensions|fileextensions|rightsinfo|languages|skins|extensiontags|functionhooks|showhooks|variables'
+		@siteinfo = self.API(action: 'query', meta: 'siteinfo', siprop: siprop)['query']
+		
+		_build_ns_map
+	end
+	
+	# Private. Massages data from siteinfo to be used for recognizing namespaces.
+	def _build_ns_map
+		@namespace_to_id = {} # all keys lowercase
+		@namespace_id_to_canon = {}
+		@namespace_id_to_local = {}
+		
+		@siteinfo['namespaces'].each_value do |h|
+			next if h['content']
+			
+			id = h['id'].to_i
+			@namespace_id_to_canon[id] = h['canonical']
+			@namespace_id_to_local[id] = h['*']
+			
+			@namespace_to_id[ h['canonical'].downcase ] = id
+			@namespace_to_id[ h['*'].downcase ] = id
+		end
+		@siteinfo['namespacealiases'].each do |h|
+			@namespace_to_id[ h['*'].downcase ] = h['id'].to_i
+		end
 	end
 	
 	# Call the API. Returns a hash of JSON response. Request can be a HTTP request string or a hash.
@@ -69,7 +105,7 @@ class Sunflower
 		if request.is_a? String
 			request += '&format=json'
 		elsif request.is_a? Hash
-			request.merge!({format:'json'})
+			request = request.merge({format:'json'})
 		end
 		
 		resp = RestClient.post(
@@ -145,33 +181,88 @@ class Sunflower
 		return self
 	end
 	
-	def log t
-		File.open('log.txt','a'){|f| f.puts t} if @log
+	# Log message to a file named log.txt in current directory, if logging is enabled. See #log= / #log?.
+	def log message
+		File.open('log.txt','a'){|f| f.puts message} if @log
+	end
+	
+	# Cleans up underscores, percent-encoding and title-casing in title (with optional anchor).
+	def cleanup_title title
+		name, anchor = title.split '#', 2
+		
+		# CGI.unescape also changes pluses to spaces; code borrowed from there
+		unescape = lambda{|a| a.gsub(/((?:%[0-9a-fA-F]{2})+)/){ [$1.delete('%')].pack('H*') } }
+		
+		ns = nil
+		name = unescape.call(name).gsub(/[ _]+/, ' ').strip
+		anchor = unescape.call(anchor.gsub(/\.([0-9a-fA-F]{2})/, '%\1')).gsub(/[ _]+/, ' ').strip if anchor
+		
+		# FIXME unicode? downcase, upcase
+		
+		if name.include? ':'
+			maybe_ns, part_name = name.split ':', 2
+			if ns_id = @namespace_to_id[maybe_ns.strip.downcase]
+				ns, name = @namespace_id_to_local[ns_id], part_name.strip
+			end
+		end
+		
+		name[0] = name[0].upcase if @siteinfo["general"]["case"] == "first-letter"
+		
+		return [ns ? "#{ns}:" : nil,  name,  anchor ? "##{anchor}" : nil].join ''
+	end
+	
+	# Returns the localized namespace name for ns, which may be namespace number, canonical name, or any namespace alias.
+	# 
+	# Returns nil if passed an invalid namespace.
+	def ns_local_for ns
+		case ns
+		when Numeric
+			@namespace_id_to_local[ns.to_i]
+		when String
+			@namespace_id_to_local[ @namespace_to_id[cleanup_title(ns).downcase] ]
+		end
+	end
+	
+	# Like #ns_local_for, but returns canonical (English) name.
+	def ns_canon_for ns
+		case ns
+		when Numeric
+			@namespace_id_to_canon[ns.to_i]
+		when String
+			@namespace_id_to_canon[ @namespace_to_id[cleanup_title(ns).downcase] ]
+		end
+	end
+	
+	# Returns a regular expression that will match given namespace. Rules for input like #ns_local_for.
+	# 
+	# Does NOT handle percent-encoding and underscores. Use #cleanup_title to canonicalize the namespace first.
+	def ns_regex_for ns
+		id = ns.is_a?(Numeric) ? ns.to_i : @namespace_to_id[cleanup_title(ns).downcase]
+		return nil if !id
+		
+		/#{@namespace_to_id.to_a.select{|a| a[1] == id }.map{|a| Regexp.escape a[0] }.join '|' }/i
 	end
 end
 
-# Class representng single Wiki page. To load specified page, use #new/#get/#load method.
-#
-# When calling Page.new, at first only the text will be loaded - attributes and edit token will be loaded when needed, or when you call #preload_attrs.
-#
-# If you are using multiple Sunflowers, you have to specify which wiki this page belongs to using second argument of function; you can pass whole URL (same as when creating new Sunflower) or just language code.
-#
-# To save page, use #save/#put method. Optional argument is new title page, if ommited, page is saved at old title. Summary can be passed as second parameter. If it's ommited, s.summary is used. If it's empty too, error is raised.
-#
-# To get Sunflower instance which this page belongs to, use #sunflower of #belongs_to.
+# Class representing a single Wiki page. To load specified page, use #new. To save it back, use #save.
 class Page
+	# Characters which MediaWiki does not permit in page title.
 	INVALID_CHARS = %w(# < > [ ] | { })
+	# Regex matching characters which MediaWiki does not permit in page title.
 	INVALID_CHARS_REGEX = Regexp.union *INVALID_CHARS
 	
+	# The current text of the page.
 	attr_accessor :text
+	# The text of the page, as of when it was loaded.
 	attr_reader :orig_text
 	
+	# The Sunflower instance this page belongs to.
 	attr_reader :sunflower
-	alias :belongs_to :sunflower
 	
 	# this is only for RDoc. wrapped in "if false" to avoid warnings when running with ruby -w
 	if false
-	attr_reader :pageid, :ns, :title, :touched, :lastrevid, :counter, :length, :starttimestamp, :edittoken, :protection #prop=info
+	# Return value of given attribute, as returned by API call prop=info for this page. Lazy-loaded.
+	attr_reader :pageid, :ns, :title, :touched, :lastrevid, :counter, :length, :starttimestamp, :edittoken, :protection
 	end
 	
 	# calling any of these accessors will fetch the data.
@@ -182,15 +273,15 @@ class Page
 		end
 	end
 	
+	# Load the specified page. Only the text will be immediately loaded - attributes and edit token will be loaded when needed, or when you call #preload_attrs.
+	#
+	# If you are using multiple Sunflowers, you have to specify which wiki this page belongs to using the second argument of function; you can pass whole URL (same as when creating new Sunflower) or just the language code.
 	def initialize title='', wiki=''
 		raise SunflowerError, 'title invalid: '+title if title =~ INVALID_CHARS_REGEX
 		
 		@modulesExecd=[] #used by sunflower-commontasks.rb
 		@summaryAppend=[] #used by sunflower-commontasks.rb
 		
-		@title=title
-		wiki=wiki+'.wikipedia.org' if wiki.index('.')==nil && wiki!=''
-		
 		if wiki=='' 
 			count=ObjectSpace.each_object(Sunflower){|o| @sunflower=o}
 			raise SunflowerError, 'you must pass wiki name if using multiple Sunflowers at once!' if count>1
@@ -198,6 +289,9 @@ class Page
 			ObjectSpace.each_object(Sunflower){|o| @sunflower=o if o.wikiURL==wiki}
 		end
 		
+		@title = self.sunflower.cleanup_title title
+		wiki = wiki+'.wikipedia.org' if wiki.index('.')==nil && wiki!=''
+		
 		if title==''
 			@text=''
 			@orig_text=''
@@ -207,6 +301,7 @@ class Page
 		preload_text
 	end
 	
+	# Load the text of this page. Semi-private.
 	def preload_text
 		r = @sunflower.API('action=query&prop=revisions&rvprop=content&titles='+CGI.escape(@title))
 		r = r['query']['pages'].first
@@ -223,6 +318,7 @@ class Page
 		@preloaded_text = true
 	end
 	
+	# Load the metadata associated with this page. Semi-private.
 	def preload_attrs
 		r = @sunflower.API('action=query&prop=info&inprop=protection&intoken=edit&titles='+CGI.escape(@title))
 		r = r['query']['pages'].first
@@ -233,6 +329,7 @@ class Page
 		@preloaded_attrs = true
 	end
 	
+	# Save the current text of this page to file (which can be either a filename or an IO).
 	def dump_to file
 		if file.respond_to? :write #probably file or IO
 			file.write @text
@@ -241,10 +338,18 @@ class Page
 		end
 	end
 	
+	# Save the current text of this page to a file whose name is based on page title, with non-alphanumeric characters stripped.
 	def dump
 		self.dump_to @title.gsub(/[^a-zA-Z0-9\-]/,'_')+'.txt'
 	end
 	
+	# Save the modifications to this page, possibly under a different title. Default summary is this page's Sunflower's summary (see Sunflower#summary=). Default title is the current title.
+	# 
+	# Will not perform API request if no changes were made.
+	# 
+	# Will call #code_cleanup if Sunflower#always_do_code_cleanup is set.
+	# 
+	# Returns the JSON result of API call or nil when API call was not made.
 	def save title=@title, summary=nil
 		preload_attrs unless @preloaded_attrs
 		

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: sunflower
 version: !ruby/object:Gem::Version
-  version: 0.4.3
+  version: 0.4.4
   prerelease: 
 platform: ruby
 authors:
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-08-01 00:00:00.000000000 Z
+date: 2012-08-03 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: json
@@ -95,7 +95,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 1.8.23
+rubygems_version: 1.8.24
 signing_key: 
 specification_version: 3
 summary: Sunflower is a lightweight library to provide access to MediaWiki API from