1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
|
require 'rubygems'
require 'mongrel'
require 'inline'
READER = "WP, the free offline Wikipedia reader thing"
MAXRES = 40
USAGE = %Q[
<strong>Supported requests:</strong>
<ul>
<li><tt>/wiki/Foo</tt> - parsed Wikipedia article on Foo</li>
<li><tt>/raw/Foo</tt> - raw text of article</li>
<li><tt>/search?s=foo</tt> - article titles containing foo</li>
</ul>
]
SEARCH_BOX=%Q{<div width="100" style="float: right;">
<form action="/search" method="get">
<input type="text" name="s" accesskey="s" /></form></div>}
class String
def titleize
self[0..0].upcase + self[1..-1]
end
end
class Parser
# Parser.php is 5k lines, and has a billion dependencies, transclusion logic,
# magic-word support, langauge processing, templating code (a literally
# Turing-complete sublanguage), and stuff... so let's just use regexps
def parse_infobox(str)
str.gsub(/^\|([^=]+)[:space:]*=[:space:]*(.*)$/) { "<b>#{$1}</b> #{$2}<br>" }
end
def parse_template(str)
str.gsub(/\{\{(([^[:space:]]+) (\w+)$)?([^\}\{]+)\}\}/) {
if $2 == "Infobox"
%Q[ <div style="width: 200px; background-color: #eee; border: 1px solid #bbb; float: right">
<h2>#{$3}</h2>
#{parse_infobox($4)} </div> ]
else
""
end
}
end
def parse(str)
# Lists; this barely works in even the simplest cases
str = str.gsub(/^\*(.+)$/) { "<span style=\"display: block\">* #{$1}</span>" }
# Bold
str = str.gsub(/'''(.+?)'''/) { "<b>#{$1}</b>" }
# Italic
str = str.gsub(/''(.+?)''/) { "<i>#{$1}</i>" }
# Headings
str = str.gsub(/(={2,})([^=]+)\1/) {"<h#{$1.size}>#{$2}</h#{$1.size}>"}
# Interwiki links
str = str.gsub(/\[\[([^\|\]]+)\|([^\]]+)\]\](\w*)/) { Article.link_to($1, $2 + $3) }
str = str.gsub(/\[\[(.*?)\]\](\w*)/) { Article.link_to($1, $1 + $2) }
# Strip refs
str = str.gsub(/<ref>.*?<\/ref>/, '')
# Templates, which may be nested, and which we need to parse from innermost first
while (new = parse_template(str)) != str
str = new
end
# External links
str = str.gsub(/\[([^\][:space:]]+) ([^\]]+)\]/) {"<a href=\"#{$1}\">#{$2}</a>"}
str = str.gsub(/\[([^\][:space:]]+)\]/) {"<a href=\"#{$1}\">#{$1}</a>"}
end
end
class Article
attr_accessor :text
attr_accessor :block
attr_accessor :title
def parsed_text
Parser.new.parse(self.text)
end
def as_html
%Q[<html><head><title>#{title} - #{READER}</title></head>
<body>
#{SEARCH_BOX}
<h1>#{title}</h1><h5>#{READER}</h5><p><small>#{text.size} bytes from block #{block}
(<a href="/raw/#{CGI::escape(title.titleize)}">raw</a>)</small></p>
#{parsed_text}
</body></html>]
end
def self.link_to(name, text)
"<a href=\"/wiki/#{CGI::escape(name.titleize)}\">#{text}</a>"
end
end
class SearchResult
attr_accessor :results
attr_accessor :needle
def as_html
%Q[<html><head><title>Search: #{needle} - #{READER}</title></head>
<body>
#{SEARCH_BOX}
<h1>Search: #{needle}</h1>
<ul>#{results.map{|r| "<li>" + Article.link_to(r, r) + "</li>"}.join}</ul>
<p><small>Searches return up to #{MAXRES} articles containing the search string anywhere in their title.
Results are case-insensitive. Exact matches appear first, followed by prefix matches, and, lastly, substring matches.
Press ^S to quickly jump to search.</small></p>
</body>
</html>]
end
end
class WPArticleReader
inline(:C) do |builder|
builder.add_compile_flags "-I../c -I. -lbz2 -DDEBUG"
builder.add_compile_flags "../c/bzipreader.c"
builder.add_compile_flags "../c/wp.c"
builder.add_compile_flags "../c/lsearcher.c"
builder.add_compile_flags "../c/safe.c"
builder.add_compile_flags "../c/blocks.c"
builder.prefix %Q$
#include "wp.h"
#define MAXRES #{MAXRES}
#define MAXSTR 1024
wp_dump d = {0};
wp_article a = {0};
char results[MAXRES][MAXSTR];
int nresults;
bool __handle_result(char *s) {
strncpy(results[nresults], s, MAXSTR);
results[nresults][MAXSTR - 1] = \'\\0\';
char *end = strrchr(results[nresults], \' \');
if(end) {
*(end - 1) = \'\\0\';
nresults++;
}
return nresults < MAXRES;
}
$
builder.c 'void __load_dump(char *dump, char *loc, char *ploc, char *blocks) {
load_dump(&d, dump, loc, ploc, blocks);
init_article(&a);
}'
builder.c 'char *__load_article(char *name) {
a.block = 0;
a.text[0] = \'\0\';
load_article(&d, name, &a);
return a.text;
}'
builder.c 'int __article_block() {
return a.block;
}'
builder.c 'int __article_size() {
return strlen(a.text);
}'
builder.c 'int __search(char *needle) {
nresults = 0;
search(&d.index, needle, __handle_result, NULL, true, true);
return nresults;
}'
builder.c 'char *__result(int n) {
return results[n];
}'
end
def initialize(opts)
@locatedb = opts[:locatedb]
@prefixdb = opts[:prefixdb]
@blockdb = opts[:blockdb]
@dump = opts[:dump]
__load_dump(@dump, @locatedb, @prefixdb, @blockdb)
end
def fetch(name)
text = __load_article(name)
a = Article.new
a.text = text
a.block = __article_block
a.title = name
a
end
def find(name)
n = __search(name)
r = SearchResult.new
r.needle = name
r.results = (0..n - 1).map {|n| __result(n)}
r
end
end
class WPHandler < Mongrel::HttpHandler
def initialize(base)
$stderr.puts "Using base #{base}"
@reader = WPArticleReader.new(:locatedb => "#{base}.locate.db",
:prefixdb => "#{base}.locate.prefixdb",
:blockdb => "#{base}.blocks.db",
:dump => "#{base}.processed")
@parser = Parser.new
end
def path(req)
CGI::unescape(req.params["REQUEST_URI"])
end
def notfound(resp, str)
respond(resp, true, 404) { "Couldn't find #{str}. <p>#{USAGE}</p>" }
end
def respond(resp, html=true, status=200)
resp.start(200) do |h, o|
h["Content-type"] = (html ? 'text/html' : 'text/plain') + '; charset=utf-8'
o.write yield
end
end
def process(req, resp)
if path(req) =~ /^\/(wiki|raw)\/(.+)$/
article = @reader.fetch($2)
if article.text.empty?
notfound(resp, $2)
else
if $1 == "wiki"
respond(resp) { article.as_html }
elsif $1 == "raw"
respond(resp, false) { article.text }
end
end
elsif path(req) =~ /^\/search\?s=(.+)$/
respond(resp) { @reader.find($1).as_html }
else
notfound(resp, path(req))
end
end
end
class WPServer
def self.start_on(port)
self.new(:port => port).run
end
def initialize(opts={})
@port = opts[:port] || 9000
@host = opts[:host] || '0.0.0.0'
end
def run
$stderr.puts "Binding to #{@host}:#{@port}"
conf = Mongrel::Configurator.new(:port => @port,
:host => @host) {
listener do
uri '/', :handler => WPHandler.new(ARGV.first), :in_front => true
end
trap("INT") { stop }
run
}
conf.join
end
end
|
