cbeta 0.0.1
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/lib/cbeta.rb +11 -0
- data/lib/cbeta/bm_to_text.rb +94 -0
- data/lib/cbeta/gaiji.json +126188 -0
- data/lib/cbeta/gaiji.rb +39 -0
- data/lib/cbeta/html_to_text.rb +151 -0
- data/lib/cbeta/p5a_to_html.rb +664 -0
- data/lib/cbeta/unicode-1.1.json +1 -0
- metadata +50 -0
data/lib/cbeta/gaiji.rb
ADDED
@@ -0,0 +1,39 @@
|
|
1
|
+
require 'json'
|
2
|
+
|
3
|
+
# 存取 CBETA 缺字資料庫
|
4
|
+
class CBETA::Gaiji
|
5
|
+
# 載入 CBETA 缺字資料庫
|
6
|
+
def initialize()
|
7
|
+
fn = File.join(File.dirname(__FILE__), 'gaiji.json')
|
8
|
+
@gaijis = JSON.parse(File.read(fn))
|
9
|
+
end
|
10
|
+
|
11
|
+
# 傳入缺字 CB 碼,傳回 hash 缺字資訊
|
12
|
+
#
|
13
|
+
# 例如:
|
14
|
+
#
|
15
|
+
# g = Cbeta::Gaiji.new
|
16
|
+
# g["CB01002"]
|
17
|
+
#
|
18
|
+
# 回傳:
|
19
|
+
# {
|
20
|
+
# "zzs": "[得-彳]",
|
21
|
+
# "unicode": "3775",
|
22
|
+
# "unicode-char": "㝵",
|
23
|
+
# "zhuyin": [ "ㄉㄜˊ", "ㄞˋ" ]
|
24
|
+
# }
|
25
|
+
def [](cb)
|
26
|
+
@gaijis[cb]
|
27
|
+
end
|
28
|
+
|
29
|
+
# 傳入缺字 CB 碼,傳回注音 array
|
30
|
+
#
|
31
|
+
# 例如:
|
32
|
+
#
|
33
|
+
# g = Cbeta::Gaiji.new
|
34
|
+
# g.zhuyin("CB00023") # return [ "ㄍㄢˇ", "ㄍㄢ", "ㄧㄤˊ", "ㄇㄧˇ", "ㄇㄧㄝ", "ㄒㄧㄤˊ" ]
|
35
|
+
def zhuyin(cb)
|
36
|
+
return nil unless @gaijis.key? cb
|
37
|
+
@gaijis[cb]['zhuyin']
|
38
|
+
end
|
39
|
+
end
|
@@ -0,0 +1,151 @@
|
|
1
|
+
require 'fileutils'
|
2
|
+
require 'nokogiri'
|
3
|
+
|
4
|
+
# 將 CBETA HTML 轉為 純文字(含行首資訊)
|
5
|
+
#
|
6
|
+
# example:
|
7
|
+
#
|
8
|
+
# h2t = CBETA::HTMLToText.new('/temp/cbeta-html', '/temp/cbeta-text')
|
9
|
+
# h2t.convert("T01") # 轉換大正藏第一冊
|
10
|
+
class CBETA::HTMLToText
|
11
|
+
# html_root:: 來源 HTML 路徑
|
12
|
+
# out_root:: 輸出路徑
|
13
|
+
def initialize(html_root, out_root)
|
14
|
+
@html_root = html_root
|
15
|
+
@out_root = out_root
|
16
|
+
end
|
17
|
+
|
18
|
+
# 例如執行大正藏第一冊:
|
19
|
+
#
|
20
|
+
# convert("T01")
|
21
|
+
def convert(arg)
|
22
|
+
@dirty = false
|
23
|
+
@vol = arg.upcase
|
24
|
+
@corpus = @vol[0]
|
25
|
+
handle_vol
|
26
|
+
end
|
27
|
+
|
28
|
+
private
|
29
|
+
|
30
|
+
def traverse(e)
|
31
|
+
r = ''
|
32
|
+
e.children.each { |c|
|
33
|
+
r += handle_node(c)
|
34
|
+
}
|
35
|
+
r.gsub(' ', '')
|
36
|
+
end
|
37
|
+
|
38
|
+
def handle_text(e)
|
39
|
+
s = e.content().chomp
|
40
|
+
return '' if s.empty?
|
41
|
+
s.gsub(/[\n,、—!。:「]/, '')
|
42
|
+
end
|
43
|
+
|
44
|
+
def handle_span(e)
|
45
|
+
r = ''
|
46
|
+
case e['class']
|
47
|
+
when 'doube-line-note'
|
48
|
+
r = traverse(e)
|
49
|
+
unless r.start_with? '('
|
50
|
+
r = "(#{r})"
|
51
|
+
end
|
52
|
+
when 'lb'
|
53
|
+
if @dirty
|
54
|
+
r += "\n"
|
55
|
+
else
|
56
|
+
@dirty = true
|
57
|
+
end
|
58
|
+
# 行首資訊 T05n0220a 改為 T05n0220
|
59
|
+
lb = e['id'].sub(/^(T0\dn0220)[a-z](.*)$/, '\1\2')
|
60
|
+
r += lb + '║'
|
61
|
+
when 'lineInfo'
|
62
|
+
when 'ranja'
|
63
|
+
r = '【◇】'
|
64
|
+
when 'siddam'
|
65
|
+
r = '【◇】'
|
66
|
+
when 'star'
|
67
|
+
else
|
68
|
+
r = traverse(e)
|
69
|
+
end
|
70
|
+
r
|
71
|
+
end
|
72
|
+
|
73
|
+
def handle_node(e)
|
74
|
+
return '' if e.comment?
|
75
|
+
return handle_text(e) if e.text?
|
76
|
+
r = ''
|
77
|
+
case e.name
|
78
|
+
when 'a'
|
79
|
+
if e['class'] == 'gaijiAnchor'
|
80
|
+
id = e['href'][1..-1]
|
81
|
+
r = @gaiji[id]
|
82
|
+
else
|
83
|
+
r = traverse(e)
|
84
|
+
end
|
85
|
+
when 'div'
|
86
|
+
if e['id'] != 'back'
|
87
|
+
r = traverse(e)
|
88
|
+
end
|
89
|
+
when 'head'
|
90
|
+
when 'p'
|
91
|
+
if e['class'] == 'figure'
|
92
|
+
r = '【圖】'
|
93
|
+
else
|
94
|
+
r = traverse(e)
|
95
|
+
end
|
96
|
+
when 'span'
|
97
|
+
r = handle_span(e)
|
98
|
+
else
|
99
|
+
r = traverse(e)
|
100
|
+
end
|
101
|
+
r
|
102
|
+
end
|
103
|
+
|
104
|
+
def prepare_folder()
|
105
|
+
folder = File.join(@out_root, @corpus, @vol)
|
106
|
+
FileUtils.remove_dir(folder, force=true)
|
107
|
+
FileUtils.mkdir_p(folder)
|
108
|
+
folder
|
109
|
+
end
|
110
|
+
|
111
|
+
def handle_file(path)
|
112
|
+
sutra = File.basename(path, ".*")
|
113
|
+
sutra.sub!(/^(.*)_.*$/, '\1')
|
114
|
+
sutra.sub!(/(T\d\dn0220).*$/, '\1') # T0220 BM 沒有分 a, b, c...
|
115
|
+
|
116
|
+
if sutra != @last_sutra
|
117
|
+
txt_fn = sutra + '.txt'
|
118
|
+
txt_path = File.join(@folder_out, txt_fn)
|
119
|
+
puts "h2t #{txt_path}"
|
120
|
+
@fo = File.open(txt_path, 'w')
|
121
|
+
@last_sutra = sutra
|
122
|
+
@dirty = false
|
123
|
+
end
|
124
|
+
|
125
|
+
f = File.open(path)
|
126
|
+
doc = Nokogiri::HTML(f)
|
127
|
+
f.close
|
128
|
+
|
129
|
+
@gaiji = {}
|
130
|
+
doc.css("span.gaijiInfo").each { |e|
|
131
|
+
@gaiji[e['id']] = e['zzs']
|
132
|
+
}
|
133
|
+
|
134
|
+
text = traverse(doc.root)
|
135
|
+
|
136
|
+
# 悉曇字
|
137
|
+
text.gsub!(/(\((【◇】)+\)|(【◇】)|【◇】)+/, '【◇】')
|
138
|
+
|
139
|
+
@fo.write(text)
|
140
|
+
end
|
141
|
+
|
142
|
+
def handle_vol()
|
143
|
+
folder_in = File.join(@html_root, @corpus, @vol)
|
144
|
+
@folder_out = prepare_folder
|
145
|
+
@last_sutra = ''
|
146
|
+
Dir["#{folder_in}/*"].each { |f|
|
147
|
+
handle_file(f)
|
148
|
+
}
|
149
|
+
end
|
150
|
+
|
151
|
+
end
|
@@ -0,0 +1,664 @@
|
|
1
|
+
require 'cgi'
|
2
|
+
require 'date'
|
3
|
+
require 'fileutils'
|
4
|
+
require 'json'
|
5
|
+
require 'nokogiri'
|
6
|
+
require 'set'
|
7
|
+
|
8
|
+
# 內容不輸出的元素
|
9
|
+
PASS=['back', 'teiHeader']
|
10
|
+
|
11
|
+
# 某版用字缺的符號
|
12
|
+
MISSING = '-'
|
13
|
+
|
14
|
+
# 處理 CBETA XML P5a
|
15
|
+
#
|
16
|
+
# CBETA XML P5a 可由此取得: https://github.com/cbeta-git/xml-p5a
|
17
|
+
#
|
18
|
+
# 轉檔規則請參考: http://wiki.ddbc.edu.tw/pages/CBETA_XML_P5a_轉_HTML
|
19
|
+
class CBETA::P5aToHTML
|
20
|
+
|
21
|
+
# xml_root:: 來源 CBETA XML P5a 路徑
|
22
|
+
# out_root:: 輸出 HTML 路徑
|
23
|
+
def initialize(xml_root, out_root)
|
24
|
+
@xml_root = xml_root
|
25
|
+
@out_root = out_root
|
26
|
+
@gaijis = CBETA::Gaiji.new
|
27
|
+
|
28
|
+
# 載入 unicode 1.1 字集列表
|
29
|
+
fn = File.join(File.dirname(__FILE__), 'unicode-1.1.json')
|
30
|
+
json = File.read(fn)
|
31
|
+
@unicode1 = JSON.parse(json)
|
32
|
+
end
|
33
|
+
|
34
|
+
# 將 CBETA XML P5a 轉為 HTML
|
35
|
+
#
|
36
|
+
# 例如 轉出大正藏第一冊
|
37
|
+
#
|
38
|
+
# x2h = CBETA::P5aToHTML.new('/PATH/TO/CBETA/XML/P5a', '/OUTPUT/FOLDER')
|
39
|
+
# x2h.convert('T01')
|
40
|
+
#
|
41
|
+
# 例如 轉出大正藏全部
|
42
|
+
#
|
43
|
+
# x2h = CBETA::P5aToHTML.new('/PATH/TO/CBETA/XML/P5a', '/OUTPUT/FOLDER')
|
44
|
+
# x2h.convert('T')
|
45
|
+
#
|
46
|
+
# T 是大正藏的 ID, CBETA 的藏經 ID 系統請參考: http://www.cbeta.org/format/id.php
|
47
|
+
def convert(arg=nil)
|
48
|
+
return convert_all if arg.nil?
|
49
|
+
|
50
|
+
arg.upcase!
|
51
|
+
if arg.size == 1
|
52
|
+
handle_collection(arg)
|
53
|
+
else
|
54
|
+
if arg.include? '..'
|
55
|
+
arg.match(/^([^\.]+?)\.\.([^\.]+)$/) {
|
56
|
+
handle_vols($1, $2)
|
57
|
+
}
|
58
|
+
else
|
59
|
+
handle_vol(arg)
|
60
|
+
end
|
61
|
+
end
|
62
|
+
end
|
63
|
+
|
64
|
+
private
|
65
|
+
|
66
|
+
def convert_all
|
67
|
+
Dir.foreach(@xml_root) { |c|
|
68
|
+
next unless c.match(/^[A-Z]$/)
|
69
|
+
handle_collection(c)
|
70
|
+
}
|
71
|
+
end
|
72
|
+
|
73
|
+
def handle_anchor(e)
|
74
|
+
id = e['id']
|
75
|
+
if e.has_attribute?('id')
|
76
|
+
if id.start_with?('nkr_note_orig')
|
77
|
+
note = @notes[id]
|
78
|
+
note_text = traverse(note)
|
79
|
+
n = id[/^nkr_note_orig_(.*)$/, 1]
|
80
|
+
@back[@juan] += "<span class='footnote' id='n#{n}'>#{note_text}</span>\n"
|
81
|
+
return "<a class='noteAnchor' href='#n#{n}'></a>"
|
82
|
+
elsif id.start_with? 'fx'
|
83
|
+
return "<span class='star'>[*]</span>"
|
84
|
+
end
|
85
|
+
end
|
86
|
+
|
87
|
+
if e.has_attribute?('type')
|
88
|
+
if e['type'] == 'circle'
|
89
|
+
return '◎'
|
90
|
+
end
|
91
|
+
end
|
92
|
+
|
93
|
+
''
|
94
|
+
end
|
95
|
+
|
96
|
+
def handle_app(e)
|
97
|
+
r = ''
|
98
|
+
if e['type'] == 'star'
|
99
|
+
c = e['corresp'][1..-1]
|
100
|
+
r = "<a class='noteAnchor star' href='#n#{c}'></a>"
|
101
|
+
end
|
102
|
+
r + traverse(e)
|
103
|
+
end
|
104
|
+
|
105
|
+
def handle_byline(e)
|
106
|
+
r = '<p class="byline">'
|
107
|
+
r += "<span class='lineInfo'>#{@lb}</span>"
|
108
|
+
r += traverse(e)
|
109
|
+
r + '</p>'
|
110
|
+
end
|
111
|
+
|
112
|
+
def handle_collection(c)
|
113
|
+
@series = c
|
114
|
+
puts 'handle_collection ' + c
|
115
|
+
folder = File.join(@xml_root, @series)
|
116
|
+
Dir.foreach(folder) { |vol|
|
117
|
+
next if ['.', '..', '.DS_Store'].include? vol
|
118
|
+
handle_vol(vol)
|
119
|
+
}
|
120
|
+
end
|
121
|
+
|
122
|
+
def handle_corr(e)
|
123
|
+
r = ''
|
124
|
+
if e.parent.name == 'choice'
|
125
|
+
sic = e.parent.at_xpath('sic')
|
126
|
+
unless sic.nil?
|
127
|
+
@dila_note += 1
|
128
|
+
r = "<a class='noteAnchor dila' href='#dila_note#{@dila_note}'></a>"
|
129
|
+
|
130
|
+
note = @orig
|
131
|
+
sic_text = traverse(sic, 'back')
|
132
|
+
if sic_text.empty?
|
133
|
+
note += MISSING
|
134
|
+
else
|
135
|
+
note += sic_text
|
136
|
+
end
|
137
|
+
@back[@juan] += "<span class='footnote_dila' id='dila_note#{@dila_note}'>#{note}</span>\n"
|
138
|
+
end
|
139
|
+
end
|
140
|
+
r + "<span class='cbeta'>%s</span>" % traverse(e)
|
141
|
+
end
|
142
|
+
|
143
|
+
def handle_div(e)
|
144
|
+
@div_count += 1
|
145
|
+
n = @div_count
|
146
|
+
if e.has_attribute? 'type'
|
147
|
+
@open_divs << e
|
148
|
+
r = traverse(e)
|
149
|
+
@open_divs.pop
|
150
|
+
return "<!-- begin div#{n}--><div class='div-#{e['type']}'>#{r}</div><!-- end of div#{n} -->"
|
151
|
+
else
|
152
|
+
return traverse(e)
|
153
|
+
end
|
154
|
+
end
|
155
|
+
|
156
|
+
def handle_figure(e)
|
157
|
+
"<p class='figure'>%s</p>" % traverse(e)
|
158
|
+
end
|
159
|
+
|
160
|
+
def handle_g(e, mode)
|
161
|
+
# if 有 <mapping type="unicode">
|
162
|
+
# if 在 unicode 1.1 範圍裡
|
163
|
+
# 直接採用
|
164
|
+
# else
|
165
|
+
# 預設呈現 unicode, 但仍包缺字資訊,供點選開 popup
|
166
|
+
# else if 有 <mapping type="normal_unicode">
|
167
|
+
# 預設呈現 normal_unicode, 但仍包缺字資訊,供點選開 popup
|
168
|
+
# else if 有 normalized form
|
169
|
+
# 預設呈現 normalized form, 但仍包缺字資訊,供點選開 popup
|
170
|
+
# else
|
171
|
+
# 預設呈現組字式, 但仍包缺字資訊,供點選開 popup
|
172
|
+
gid = e['ref'][1..-1]
|
173
|
+
g = @gaijis[gid]
|
174
|
+
abort "Line:#{__LINE__} 無缺字資料:#{gid}" if g.nil?
|
175
|
+
zzs = g['zzs']
|
176
|
+
|
177
|
+
if mode == 'txt'
|
178
|
+
return g['roman'] if gid.start_with?('SD')
|
179
|
+
if zzs.nil?
|
180
|
+
abort "缺組字式:#{g}"
|
181
|
+
else
|
182
|
+
return zzs
|
183
|
+
end
|
184
|
+
end
|
185
|
+
|
186
|
+
@char_count += 1
|
187
|
+
|
188
|
+
if gid.start_with?('SD')
|
189
|
+
case gid
|
190
|
+
when 'SD-E35A'
|
191
|
+
return '('
|
192
|
+
when 'SD-E35B'
|
193
|
+
return ')'
|
194
|
+
else
|
195
|
+
return "<span class='siddam' roman='#{g['roman']}' code='#{gid}' char='#{g['sd-char']}'/>"
|
196
|
+
end
|
197
|
+
end
|
198
|
+
|
199
|
+
if gid.start_with?('RJ')
|
200
|
+
return "<span class='ranja' roman='#{g['roman']}' code='#{gid}' char='#{g['rj-char']}'/>"
|
201
|
+
end
|
202
|
+
|
203
|
+
default = ''
|
204
|
+
if g.has_key?('unicode')
|
205
|
+
if @unicode1.include?(g['unicode'])
|
206
|
+
return g['unicode-char'] # unicode 1.1 直接用
|
207
|
+
else
|
208
|
+
default = g['unicode-char']
|
209
|
+
end
|
210
|
+
end
|
211
|
+
|
212
|
+
nor = ''
|
213
|
+
if g.has_key?('normal_unicode')
|
214
|
+
nor = g['normal_unicode']
|
215
|
+
default = nor if default.empty?
|
216
|
+
end
|
217
|
+
|
218
|
+
if g.has_key?('normal')
|
219
|
+
nor += ', ' unless nor==''
|
220
|
+
nor += g['normal']
|
221
|
+
default = g['normal'] if default.empty?
|
222
|
+
end
|
223
|
+
|
224
|
+
default = zzs if default.empty?
|
225
|
+
|
226
|
+
href = 'http://dict.cbeta.org/dict_word/gaiji-cb/%s/%s.gif' % [gid[2, 2], gid]
|
227
|
+
unless @back[@juan].include?(href)
|
228
|
+
@back[@juan] += "<span id='#{gid}' class='gaijiInfo' figure_url='#{href}' zzs='#{zzs}' nor='#{nor}'>#{default}</span>\n"
|
229
|
+
end
|
230
|
+
"<a class='gaijiAnchor' href='##{gid}'>#{default}</a>"
|
231
|
+
end
|
232
|
+
|
233
|
+
def handle_graphic(e)
|
234
|
+
url = File.basename(e['url'])
|
235
|
+
"<span imgsrc='#{url}' class='graphic'></span>"
|
236
|
+
end
|
237
|
+
|
238
|
+
def handle_head(e)
|
239
|
+
r = ''
|
240
|
+
unless e['type'] == 'added'
|
241
|
+
r = "<p class='head'>%s</p>" % traverse(e)
|
242
|
+
end
|
243
|
+
r
|
244
|
+
end
|
245
|
+
|
246
|
+
def handle_item(e)
|
247
|
+
"<li>%s</li>\n" % traverse(e)
|
248
|
+
end
|
249
|
+
|
250
|
+
def handle_juan(e)
|
251
|
+
"<p class='juan'>%s</p>" % traverse(e)
|
252
|
+
end
|
253
|
+
|
254
|
+
def handle_l(e)
|
255
|
+
if @lg_type == 'abnormal'
|
256
|
+
return traverse(e)
|
257
|
+
end
|
258
|
+
|
259
|
+
@in_l = true
|
260
|
+
|
261
|
+
doc = Nokogiri::XML::Document.new
|
262
|
+
cell = doc.create_element('div')
|
263
|
+
cell['class'] = 'lg-cell'
|
264
|
+
cell.inner_html = traverse(e)
|
265
|
+
|
266
|
+
if @first_l
|
267
|
+
parent = e.parent()
|
268
|
+
if parent.has_attribute?('rend')
|
269
|
+
indent = parent['rend'].scan(/text-indent:[^:]*/)
|
270
|
+
unless indent.empty?
|
271
|
+
cell['style'] = indent[0]
|
272
|
+
end
|
273
|
+
end
|
274
|
+
@first_l = false
|
275
|
+
end
|
276
|
+
r = cell.to_s
|
277
|
+
|
278
|
+
unless @lg_row_open
|
279
|
+
r = "\n<div class='lg-row'>" + r
|
280
|
+
@lg_row_open = true
|
281
|
+
end
|
282
|
+
@in_l = false
|
283
|
+
r
|
284
|
+
end
|
285
|
+
|
286
|
+
def handle_lb(e)
|
287
|
+
@char_count = 1
|
288
|
+
@lb = e['n']
|
289
|
+
line_head = @sutra_no + '_p' + e['n']
|
290
|
+
r = ''
|
291
|
+
#if e.parent.name == 'lg' and $lg_row_open
|
292
|
+
if @lg_row_open && !@in_l
|
293
|
+
# 每行偈頌放在一個 lg-row 裡面
|
294
|
+
# T46n1937, p. 914a01, l 包雙行夾註跨行
|
295
|
+
# T20n1092, 337c16, lb 在 l 中間,不結束 lg-row
|
296
|
+
r += "</div><!-- end of lg-row -->"
|
297
|
+
@lg_row_open = false
|
298
|
+
end
|
299
|
+
r + "<span class='lb' \nid='#{line_head}'>#{line_head}</span>"
|
300
|
+
end
|
301
|
+
|
302
|
+
def handle_lem(e)
|
303
|
+
r = ''
|
304
|
+
w = e['wit']
|
305
|
+
if w.include? 'CBETA' and not w.include? @orig
|
306
|
+
@dila_note += 1
|
307
|
+
r = "<a class='noteAnchor dila' href='#dila_note#{@dila_note}'></a>"
|
308
|
+
r += "<span class='cbeta'>%s</span>" % traverse(e)
|
309
|
+
|
310
|
+
note = lem_note_cf(e)
|
311
|
+
note += lem_note_rdg(e)
|
312
|
+
@back[@juan] += "<span class='footnote_dila' id='dila_note#{@dila_note}'>#{note}</span>\n"
|
313
|
+
else
|
314
|
+
r = traverse(e)
|
315
|
+
end
|
316
|
+
r
|
317
|
+
end
|
318
|
+
|
319
|
+
def handle_lg(e)
|
320
|
+
r = ''
|
321
|
+
@lg_type = e['type']
|
322
|
+
if @lg_type == 'abnormal'
|
323
|
+
r = "<p class='lg-abnormal'>" + traverse(e) + "</p>"
|
324
|
+
else
|
325
|
+
@first_l = true
|
326
|
+
doc = Nokogiri::XML::Document.new
|
327
|
+
node = doc.create_element('div')
|
328
|
+
node['class'] = 'lg'
|
329
|
+
if e.has_attribute?('rend')
|
330
|
+
rend = e['rend'].gsub(/text-indent:[^:]*/, '')
|
331
|
+
node['style'] = rend
|
332
|
+
end
|
333
|
+
@lg_row_open = false
|
334
|
+
node.inner_html = traverse(e)
|
335
|
+
if @lg_row_open
|
336
|
+
node.inner_html += '</div><!-- end of lg -->'
|
337
|
+
@lg_row_open = false
|
338
|
+
end
|
339
|
+
r = "\n" + node.to_s
|
340
|
+
end
|
341
|
+
r
|
342
|
+
end
|
343
|
+
|
344
|
+
def handle_list(e)
|
345
|
+
"<ul>%s</ul>" % traverse(e)
|
346
|
+
end
|
347
|
+
|
348
|
+
def handle_milestone(e)
|
349
|
+
r = ''
|
350
|
+
if e['unit'] == 'juan'
|
351
|
+
|
352
|
+
r += "</div>" * @open_divs.size # 如果有 div 跨卷,要先結束, ex: T55n2154, p. 680a29, 跨 19, 20 兩卷
|
353
|
+
@juan = e['n'].to_i
|
354
|
+
@back[@juan] = @back[0]
|
355
|
+
r += "<juan #{@juan}>"
|
356
|
+
@open_divs.each { |d|
|
357
|
+
r += "<div class='#{d['type']}'>"
|
358
|
+
}
|
359
|
+
end
|
360
|
+
r
|
361
|
+
end
|
362
|
+
|
363
|
+
def handle_mulu(e)
|
364
|
+
r = ''
|
365
|
+
if e['type'] == '品'
|
366
|
+
@pass << false
|
367
|
+
r = "<mulu class='pin' s='%s'/>" % traverse(e, 'txt')
|
368
|
+
@pass.pop
|
369
|
+
end
|
370
|
+
r
|
371
|
+
end
|
372
|
+
|
373
|
+
def handle_node(e, mode)
|
374
|
+
return '' if e.comment?
|
375
|
+
return handle_text(e, mode) if e.text?
|
376
|
+
return '' if PASS.include?(e.name)
|
377
|
+
r = case e.name
|
378
|
+
when 'anchor' then handle_anchor(e)
|
379
|
+
when 'app' then handle_app(e)
|
380
|
+
when 'byline' then handle_byline(e)
|
381
|
+
when 'corr' then handle_corr(e)
|
382
|
+
when 'div' then handle_div(e)
|
383
|
+
when 'figure' then handle_figure(e)
|
384
|
+
when 'foreign' then ''
|
385
|
+
when 'g' then handle_g(e, mode)
|
386
|
+
when 'graphic' then handle_graphic(e)
|
387
|
+
when 'head' then handle_head(e)
|
388
|
+
when 'item' then handle_item(e)
|
389
|
+
when 'juan' then handle_juan(e)
|
390
|
+
when 'l' then handle_l(e)
|
391
|
+
when 'lb' then handle_lb(e)
|
392
|
+
when 'lem' then handle_lem(e)
|
393
|
+
when 'lg' then handle_lg(e)
|
394
|
+
when 'list' then handle_list(e)
|
395
|
+
when 'mulu' then handle_mulu(e)
|
396
|
+
when 'note' then handle_note(e)
|
397
|
+
when 'milestone' then handle_milestone(e)
|
398
|
+
when 'p' then handle_p(e)
|
399
|
+
when 'rdg' then ''
|
400
|
+
when 'reg' then ''
|
401
|
+
when 'sic' then ''
|
402
|
+
when 'sg' then handle_sg(e)
|
403
|
+
when 't' then handle_t(e)
|
404
|
+
else traverse(e)
|
405
|
+
end
|
406
|
+
r
|
407
|
+
end
|
408
|
+
|
409
|
+
|
410
|
+
def handle_note(e)
|
411
|
+
n = e['n']
|
412
|
+
if e.has_attribute?('type')
|
413
|
+
t = e['type']
|
414
|
+
case t
|
415
|
+
when 'equivalent'
|
416
|
+
return ''
|
417
|
+
when 'orig'
|
418
|
+
@pass << false
|
419
|
+
s = traverse(e)
|
420
|
+
@pass.pop
|
421
|
+
@back[@juan] += "<span class='footnote_orig' id='n#{n}'>#{s}</span>\n"
|
422
|
+
|
423
|
+
if @mod_notes.include? n
|
424
|
+
return ''
|
425
|
+
else
|
426
|
+
return "<a class='noteAnchor' href='#n#{n}'></a>"
|
427
|
+
end
|
428
|
+
when 'mod'
|
429
|
+
@pass << false
|
430
|
+
s = traverse(e)
|
431
|
+
@pass.pop
|
432
|
+
@back[@juan] += "<span class='footnote_cb' id='n#{n}'>#{s}</span>\n"
|
433
|
+
return "<a class='noteAnchor' href='#n#{n}'></a>"
|
434
|
+
when 'rest'
|
435
|
+
return ''
|
436
|
+
else
|
437
|
+
return '' if t.start_with?('cf')
|
438
|
+
end
|
439
|
+
end
|
440
|
+
|
441
|
+
if e.has_attribute?('resp')
|
442
|
+
return '' if e['resp'].start_with? 'CBETA'
|
443
|
+
end
|
444
|
+
|
445
|
+
if e.has_attribute?('place') && e['place']=='inline'
|
446
|
+
r = traverse(e)
|
447
|
+
return "<span class='doube-line-note'>#{r}</span>"
|
448
|
+
else
|
449
|
+
return traverse(e)
|
450
|
+
end
|
451
|
+
end
|
452
|
+
|
453
|
+
def handle_p(e)
|
454
|
+
r = '<p>'
|
455
|
+
r += "<span class='lineInfo'>#{@lb}</span>"
|
456
|
+
r += traverse(e)
|
457
|
+
r + '</p>'
|
458
|
+
end
|
459
|
+
|
460
|
+
def handle_sg(e)
|
461
|
+
'(' + traverse(e) + ')'
|
462
|
+
end
|
463
|
+
|
464
|
+
def handle_sutra(xml_fn)
|
465
|
+
puts "handle sutra #{xml_fn}"
|
466
|
+
@back = { 0 => '' }
|
467
|
+
@char_count = 1
|
468
|
+
@dila_note = 0
|
469
|
+
@div_count = 0
|
470
|
+
@in_l = false
|
471
|
+
@juan = 0
|
472
|
+
@lg_row_open = false
|
473
|
+
@mod_notes = Set.new
|
474
|
+
@open_divs = []
|
475
|
+
@sutra_no = File.basename(xml_fn, ".xml")
|
476
|
+
|
477
|
+
text = parse_xml(xml_fn)
|
478
|
+
|
479
|
+
# 註標移到 lg-cell 裡面,不然以 table 呈現 lg 會有問題
|
480
|
+
text.gsub!(/(<a class='noteAnchor'[^>]*><\/a>)(<div class="lg-cell"[^>]*>)/, '\2\1')
|
481
|
+
|
482
|
+
juans = text.split(/(<juan \d+>)/)
|
483
|
+
open = false
|
484
|
+
fo = nil
|
485
|
+
juan_no = nil
|
486
|
+
fn = ''
|
487
|
+
buf = ''
|
488
|
+
# 一卷一檔
|
489
|
+
juans.each { |j|
|
490
|
+
if j =~ /<juan (\d+)>$/
|
491
|
+
juan_no = $1.to_i
|
492
|
+
fn = "#{@sutra_no}_%03d.htm" % juan_no
|
493
|
+
output_path = File.join(@out_folder, fn)
|
494
|
+
fo = File.open(output_path, 'w')
|
495
|
+
open = true
|
496
|
+
s = <<eos
|
497
|
+
<html>
|
498
|
+
<head>
|
499
|
+
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
|
500
|
+
<meta name="filename" content="#{fn}" />
|
501
|
+
<title>#{@title}</title>
|
502
|
+
</head>
|
503
|
+
<body>
|
504
|
+
<!--
|
505
|
+
來源 XML CBETA P5a: https://github.com/cbeta-org/xml-p5a.git
|
506
|
+
轉檔程式: Dropbox/DILA-DA/cbeta-html/bin/x2h.rb Version #{Date.today}
|
507
|
+
說明文件: http://wiki.ddbc.edu.tw/pages/CBETA_XML_P5a_%E8%BD%89_HTML
|
508
|
+
-->
|
509
|
+
<div id='body'>
|
510
|
+
eos
|
511
|
+
fo.write(s)
|
512
|
+
fo.write(buf)
|
513
|
+
buf = ''
|
514
|
+
elsif open
|
515
|
+
fo.write(j + "\n</div><!-- end of div[@id='body'] -->\n")
|
516
|
+
fo.write("<div id='back'>\n" + @back[juan_no] + "</div>\n")
|
517
|
+
fo.write('</body></html>')
|
518
|
+
fo.close
|
519
|
+
else
|
520
|
+
buf = j
|
521
|
+
end
|
522
|
+
}
|
523
|
+
end
|
524
|
+
|
525
|
+
def handle_t(e)
|
526
|
+
if e.has_attribute? 'place'
|
527
|
+
return '' if e['place'].include? 'foot'
|
528
|
+
end
|
529
|
+
traverse(e)
|
530
|
+
end
|
531
|
+
|
532
|
+
def handle_text(e, mode)
|
533
|
+
s = e.content().chomp
|
534
|
+
return '' if s.empty?
|
535
|
+
return '' if e.parent.name == 'app'
|
536
|
+
|
537
|
+
# cbeta xml 文字之間會有多餘的換行
|
538
|
+
r = s.gsub(/[\n\r]/, '')
|
539
|
+
|
540
|
+
# 把 & 轉為 &
|
541
|
+
r = CGI.escapeHTML(r)
|
542
|
+
|
543
|
+
# 正文區的文字外面要包 span
|
544
|
+
if @pass.last and mode=='html'
|
545
|
+
r = "<span class='t' l='#{@lb}' w='#{@char_count}'>#{r}</span>"
|
546
|
+
@char_count += r.size
|
547
|
+
end
|
548
|
+
r
|
549
|
+
end
|
550
|
+
|
551
|
+
def handle_vol(vol)
|
552
|
+
puts 'x2h ' + vol
|
553
|
+
if vol.start_with? 'T'
|
554
|
+
@orig = "【大】"
|
555
|
+
else
|
556
|
+
abort "未處理底本"
|
557
|
+
end
|
558
|
+
@vol = vol
|
559
|
+
@series = vol[0]
|
560
|
+
@out_folder = File.join(@out_root, @series, vol)
|
561
|
+
FileUtils.remove_dir(@out_folder, force=true)
|
562
|
+
FileUtils::mkdir_p @out_folder
|
563
|
+
|
564
|
+
source = File.join(@xml_root, @series, vol)
|
565
|
+
Dir[source+"/*"].each { |f|
|
566
|
+
handle_sutra(f)
|
567
|
+
}
|
568
|
+
end
|
569
|
+
|
570
|
+
def handle_vols(v1, v2)
|
571
|
+
@series = v1[0]
|
572
|
+
folder = File.join(IN, @series)
|
573
|
+
Dir.foreach(folder) { |vol|
|
574
|
+
next if vol < v1
|
575
|
+
next if vol > v2
|
576
|
+
handle_vol(vol)
|
577
|
+
}
|
578
|
+
end
|
579
|
+
|
580
|
+
def lem_note_cf(e)
|
581
|
+
# ex: T32n1670A.xml, p. 703a16
|
582
|
+
# <note type="cf1">K30n1002_p0257a01-a23</note>
|
583
|
+
refs = []
|
584
|
+
e.xpath('./note').each { |n|
|
585
|
+
if n.key?('type') and n['type'].start_with? 'cf'
|
586
|
+
refs << n.content
|
587
|
+
end
|
588
|
+
}
|
589
|
+
if refs.empty?
|
590
|
+
''
|
591
|
+
else
|
592
|
+
'修訂依據:' + refs.join(';') + '。'
|
593
|
+
end
|
594
|
+
end
|
595
|
+
|
596
|
+
def lem_note_rdg(lem)
|
597
|
+
r = ''
|
598
|
+
app = lem.parent
|
599
|
+
@pass << false
|
600
|
+
app.xpath('rdg').each { |rdg|
|
601
|
+
if rdg['wit'].include? @orig
|
602
|
+
s = traverse(rdg, 'back')
|
603
|
+
s = MISSING if s.empty?
|
604
|
+
r += @orig + s
|
605
|
+
end
|
606
|
+
}
|
607
|
+
@pass.pop
|
608
|
+
r += '。' unless r.empty?
|
609
|
+
r
|
610
|
+
end
|
611
|
+
|
612
|
+
def open_xml(fn)
|
613
|
+
s = File.read(fn)
|
614
|
+
|
615
|
+
if fn.include? 'T16n0657'
|
616
|
+
# 這個地方 雙行夾註 跨兩行偈頌
|
617
|
+
# 把 lb 移到 note 結束之前
|
618
|
+
# 讓 lg-row 先結束,再結束雙行夾註
|
619
|
+
s.sub!(/(<\/note>)(\n<lb n="0206b29" ed="T"\/>)/, '\2\1')
|
620
|
+
end
|
621
|
+
|
622
|
+
# <milestone unit="juan"> 前面的 lb 屬於新的這一卷
|
623
|
+
s.gsub!(/((?:<pb [^>]+>\n?)?<lb [^>]+>\n?)(<milestone [^>]*unit="juan"\/>)/, '\2\1')
|
624
|
+
|
625
|
+
doc = Nokogiri::XML(s)
|
626
|
+
doc.remove_namespaces!()
|
627
|
+
doc
|
628
|
+
end
|
629
|
+
|
630
|
+
def read_mod_notes(doc)
|
631
|
+
doc.xpath("//note[@type='mod']").each { |e|
|
632
|
+
@mod_notes << e['n']
|
633
|
+
}
|
634
|
+
end
|
635
|
+
|
636
|
+
def parse_xml(xml_fn)
|
637
|
+
@pass = [false]
|
638
|
+
|
639
|
+
doc = open_xml(xml_fn)
|
640
|
+
|
641
|
+
e = doc.xpath("//titleStmt/title")[0]
|
642
|
+
@title = traverse(e, 'txt')
|
643
|
+
@title = @title.split()[-1]
|
644
|
+
|
645
|
+
read_mod_notes(doc)
|
646
|
+
|
647
|
+
root = doc.root()
|
648
|
+
body = root.xpath("text/body")[0]
|
649
|
+
@pass = [true]
|
650
|
+
|
651
|
+
text = traverse(body)
|
652
|
+
text
|
653
|
+
end
|
654
|
+
|
655
|
+
def traverse(e, mode='html')
|
656
|
+
r = ''
|
657
|
+
e.children.each { |c|
|
658
|
+
s = handle_node(c, mode)
|
659
|
+
r += s
|
660
|
+
}
|
661
|
+
r
|
662
|
+
end
|
663
|
+
|
664
|
+
end
|