Top Level Namespace

Defined Under Namespace

Classes: Server, SiteData

Constant Summary collapse

TMPFILE =

content cache #

Dir.tmpdir + "/w3m-autopagerize.tmp.html"
DOCTYPE =

BUG: libxml2 cannot handle id() function without doctype. labs.gmo.jp/blog/ku/2008/07/libxmlhtmlxpathid.html

'<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">'

Instance Method Summary collapse

Instance Method Details

#addstring(url_or_pattern, string) ⇒ Object 



131
132
133
134
135
136
 
# File 'lib/w3m-autopagerize-server.rb', line 131

def addstring(url_or_pattern, string)
  defnext_(url_or_pattern) {|u,m|
    $logger.info "Use addstring for #{url_or_pattern}"
    u.to_s + string
  } 
end

#defnext(url_or_pattern, nexturl = nil, &b) ⇒ Object 



113
114
115
116
117
118
 
# File 'lib/w3m-autopagerize-server.rb', line 113

def defnext(url_or_pattern, nexturl=nil, &b)
  defnext_ url_or_pattern, nexturl do |u,m|
    $logger.info "Use defnext for #{url_or_pattern}"
    instance_exec(u, m, &b)
  end
end

#defnext_(url_or_pattern, nexturl = nil, &block) ⇒ Object 



120
121
122
123
124
125
126
127
128
 
# File 'lib/w3m-autopagerize-server.rb', line 120

def defnext_(url_or_pattern, nexturl=nil, &block)
  sd = SiteData.new 
  if nexturl
    sd.block = lambda{|u,m| nexturl }
  else
    sd.block = block
  end
  $SITEINFO << [ url_or_pattern, sd ]
end

#get_content(url) ⇒ Object 



224
225
226
227
228
229
230
 
# File 'lib/w3m-autopagerize-server.rb', line 224

def get_content(url)
  if url =~ /^https?:/
    get_header_and_content(url)[1]
  else                          # local file
    File.zread(File.expand_path(url)).force_encoding "ASCII-8BIT"
  end
end

#get_header_and_content(url) ⇒ Object

Utilities #



214
215
216
217
218
219
220
221
222
 
# File 'lib/w3m-autopagerize-server.rb', line 214

def get_header_and_content(url)
  output = `w3m #$W3M_EXTRA_OPTIONS -dump_both -o accept_encoding='gzip' "#{url}"`
  header, source = output.force_encoding("ASCII-8BIT").split(/\n\n/, 2)
  open(TMPFILE,"wb"){|f| f.write source}
  source = File.zread(TMPFILE).force_encoding "ASCII-8BIT"
  [header, source]
ensure
  File.unlink TMPFILE if File.exist? TMPFILE
end

#increment(url_or_pattern, n = 1) ⇒ Object 



138
139
140
141
142
143
144
145
146
147
148
149
150
 
# File 'lib/w3m-autopagerize-server.rb', line 138

def increment(url_or_pattern, n=1)
  defnext_(url_or_pattern) {|u,m|
    $logger.info "Use increment for #{url_or_pattern}"
    url=u.to_s
    nextvar = m[1].to_i + n
    url[ m.begin(1) ... m.end(1) ] = if m[1] =~ /^0/
                                       format("%0#{m[1].length}d", nextvar)
                                     else
                                       nextvar.to_s
                                     end
    url
  }
end

#normalize_charset(charset) ⇒ Object 



232
233
234
235
236
 
# File 'lib/w3m-autopagerize-server.rb', line 232

def normalize_charset(charset)
  charset = charset.to_s
  # FIXME I do not know other charsets than Japanese.
  charset.downcase == "shift_jis" ? "cp932" : charset
end

#reinit ⇒ Object 



239
240
241
 
# File 'lib/w3m-autopagerize-server.rb', line 239

def reinit
  $SITEINFO = []
end

#w3mctl(*strings) ⇒ Object 



152
153
154
155
156
157
158
159
160
161
162
163
 
# File 'lib/w3m-autopagerize-server.rb', line 152

def w3mctl(*strings)
  strings.each do |str|
    if str
      if str==true
        $client.print "\r\n\r\n"
      else
        $client.print "W3m-Control: #{str}\r\n"
      end
    end
  end
  nil
end