tidy-ext 0.1.7

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (65) hide show
  1. data/.gitignore +4 -0
  2. data/LICENSE +50 -0
  3. data/README +12 -0
  4. data/Rakefile +60 -0
  5. data/VERSION +1 -0
  6. data/ext/tidy/access.c +3310 -0
  7. data/ext/tidy/access.h +279 -0
  8. data/ext/tidy/alloc.c +107 -0
  9. data/ext/tidy/attrask.c +209 -0
  10. data/ext/tidy/attrdict.c +2398 -0
  11. data/ext/tidy/attrdict.h +122 -0
  12. data/ext/tidy/attrget.c +213 -0
  13. data/ext/tidy/attrs.c +1911 -0
  14. data/ext/tidy/attrs.h +374 -0
  15. data/ext/tidy/buffio.c +232 -0
  16. data/ext/tidy/buffio.h +118 -0
  17. data/ext/tidy/charsets.c +1032 -0
  18. data/ext/tidy/charsets.h +14 -0
  19. data/ext/tidy/clean.c +2674 -0
  20. data/ext/tidy/clean.h +87 -0
  21. data/ext/tidy/config.c +1746 -0
  22. data/ext/tidy/config.h +153 -0
  23. data/ext/tidy/entities.c +419 -0
  24. data/ext/tidy/entities.h +24 -0
  25. data/ext/tidy/extconf.rb +5 -0
  26. data/ext/tidy/fileio.c +106 -0
  27. data/ext/tidy/fileio.h +46 -0
  28. data/ext/tidy/forward.h +69 -0
  29. data/ext/tidy/iconvtc.c +105 -0
  30. data/ext/tidy/iconvtc.h +15 -0
  31. data/ext/tidy/istack.c +373 -0
  32. data/ext/tidy/lexer.c +3825 -0
  33. data/ext/tidy/lexer.h +617 -0
  34. data/ext/tidy/localize.c +1882 -0
  35. data/ext/tidy/mappedio.c +329 -0
  36. data/ext/tidy/mappedio.h +16 -0
  37. data/ext/tidy/message.h +207 -0
  38. data/ext/tidy/parser.c +4408 -0
  39. data/ext/tidy/parser.h +76 -0
  40. data/ext/tidy/platform.h +636 -0
  41. data/ext/tidy/pprint.c +2276 -0
  42. data/ext/tidy/pprint.h +93 -0
  43. data/ext/tidy/ruby-tidy.c +195 -0
  44. data/ext/tidy/streamio.c +1407 -0
  45. data/ext/tidy/streamio.h +222 -0
  46. data/ext/tidy/tagask.c +286 -0
  47. data/ext/tidy/tags.c +955 -0
  48. data/ext/tidy/tags.h +235 -0
  49. data/ext/tidy/tidy-int.h +129 -0
  50. data/ext/tidy/tidy.h +1097 -0
  51. data/ext/tidy/tidyenum.h +622 -0
  52. data/ext/tidy/tidylib.c +1751 -0
  53. data/ext/tidy/tmbstr.c +306 -0
  54. data/ext/tidy/tmbstr.h +92 -0
  55. data/ext/tidy/utf8.c +539 -0
  56. data/ext/tidy/utf8.h +52 -0
  57. data/ext/tidy/version.h +14 -0
  58. data/ext/tidy/win32tc.c +795 -0
  59. data/ext/tidy/win32tc.h +19 -0
  60. data/spec/spec_helper.rb +5 -0
  61. data/spec/tidy/compat_spec.rb +44 -0
  62. data/spec/tidy/remote_uri_spec.rb +14 -0
  63. data/spec/tidy/test1.html +5 -0
  64. data/spec/tidy/tidy_spec.rb +34 -0
  65. metadata +125 -0
data/ext/tidy/pprint.c ADDED
@@ -0,0 +1,2276 @@
1
+ /*
2
+ pprint.c -- pretty print parse tree
3
+
4
+ (c) 1998-2007 (W3C) MIT, ERCIM, Keio University
5
+ See tidy.h for the copyright notice.
6
+
7
+ CVS Info :
8
+
9
+ $Author: arnaud02 $
10
+ $Date: 2008/03/22 20:23:37 $
11
+ $Revision: 1.119 $
12
+
13
+ */
14
+
15
+ #include <stdio.h>
16
+ #include <stdlib.h>
17
+ #include <string.h>
18
+
19
+ #include "pprint.h"
20
+ #include "tidy-int.h"
21
+ #include "parser.h"
22
+ #include "entities.h"
23
+ #include "tmbstr.h"
24
+ #include "utf8.h"
25
+
26
+ /*
27
+ Block-level and unknown elements are printed on
28
+ new lines and their contents indented 2 spaces
29
+
30
+ Inline elements are printed inline.
31
+
32
+ Inline content is wrapped on spaces (except in
33
+ attribute values or preformatted text, after
34
+ start tags and before end tags
35
+ */
36
+
37
+ static void PPrintAsp( TidyDocImpl* doc, uint indent, Node* node );
38
+ static void PPrintJste( TidyDocImpl* doc, uint indent, Node* node );
39
+ static void PPrintPhp( TidyDocImpl* doc, uint indent, Node* node );
40
+ static int TextEndsWithNewline( Lexer *lexer, Node *node, uint mode );
41
+ static int TextStartsWithWhitespace( Lexer *lexer, Node *node, uint start, uint mode );
42
+ static Bool InsideHead( TidyDocImpl* doc, Node *node );
43
+ static Bool ShouldIndent( TidyDocImpl* doc, Node *node );
44
+
45
+ #if SUPPORT_ASIAN_ENCODINGS
46
+ /* #431953 - start RJ Wraplen adjusted for smooth international ride */
47
+
48
+ #if 0
49
+ uint CWrapLen( TidyDocImpl* doc, uint ind )
50
+ {
51
+ ctmbstr lang = cfgStr( doc, TidyLanguage );
52
+ uint wraplen = cfg( doc, TidyWrapLen );
53
+
54
+ if ( !TY_(tmbstrcasecmp)(lang, "zh") )
55
+ /* Chinese characters take two positions on a fixed-width screen */
56
+ /* It would be more accurate to keep a parallel linelen and wraphere
57
+ incremented by 2 for Chinese characters and 1 otherwise, but this
58
+ is way simpler.
59
+ */
60
+ return (ind + (( wraplen - ind ) / 2)) ;
61
+
62
+ if ( !TY_(tmbstrcasecmp)(lang, "ja") )
63
+ /* average Japanese text is 30% kanji */
64
+ return (ind + ((( wraplen - ind ) * 7) / 10)) ;
65
+
66
+ return wraplen;
67
+ }
68
+ #endif
69
+
70
+ typedef enum
71
+ {
72
+ UC00, /* None */
73
+ UCPC, /* Punctuation, Connector */
74
+ UCPD, /* Punctuation, Dash */
75
+ UCPE, /* Punctuation, Close */
76
+ UCPS, /* Punctuation, Open */
77
+ UCPI, /* Punctuation, Initial quote */
78
+ UCPF, /* Punctuation, Final quote */
79
+ UCPO, /* Punctuation, Other */
80
+ UCZS, /* Separator, Space */
81
+ UCZL, /* Separator, Line */
82
+ UCZP /* Separator, Paragraph */
83
+ } UnicodeCategory;
84
+
85
+ /*
86
+ From the original code, the following characters are removed:
87
+
88
+ U+2011 (non-breaking hyphen)
89
+ U+202F (narrow non-break space)
90
+ U+2044 (fraction slash)
91
+ U+200B (zero width space)
92
+ ...... (bidi formatting control characters)
93
+
94
+ U+2011 and U+202F are non-breaking, U+2044 is a Sm character,
95
+ U+200B is a non-visible space, wrapping after it would make
96
+ this space visible, bidi should be done using HTML features
97
+ and the characters are neither Px or Zx.
98
+
99
+ The following Unicode 3.0 punctuation characters are added:
100
+
101
+ U+2048 (question exclamation mark)
102
+ U+2049 (exclamation question mark)
103
+ U+204A (tironian sign et)
104
+ U+204B (reversed pilcrow sign)
105
+ U+204C (black leftwards bullet)
106
+ U+204D (black rightwards bullet)
107
+ U+3030 (wavy dash)
108
+ U+30FB (katakana middle dot)
109
+ U+FE63 (small hyphen-minus)
110
+ U+FE68 (small reverse solidus)
111
+ U+FF3F (fullwidth low line)
112
+ U+FF5B (fullwidth left curly bracket)
113
+ U+FF5D (fullwidth right curly bracket)
114
+
115
+ Other additional characters were not included in Unicode 3.0.
116
+ The table is based on Unicode 4.0. It must include only those
117
+ characters marking a wrapping point, "before" if the general
118
+ category is UCPS or UCPI, otherwise "after".
119
+ */
120
+ static struct _unicode4cat
121
+ {
122
+ unsigned long code;
123
+ UnicodeCategory category;
124
+ } const unicode4cat[] =
125
+ {
126
+ #if 0
127
+ { 0x037E, UCPO }, { 0x0387, UCPO }, { 0x055A, UCPO }, { 0x055B, UCPO },
128
+ { 0x055C, UCPO }, { 0x055D, UCPO }, { 0x055E, UCPO }, { 0x055F, UCPO },
129
+ { 0x0589, UCPO }, { 0x058A, UCPD }, { 0x05BE, UCPO }, { 0x05C0, UCPO },
130
+ { 0x05C3, UCPO }, { 0x05F3, UCPO }, { 0x05F4, UCPO }, { 0x060C, UCPO },
131
+ { 0x060D, UCPO }, { 0x061B, UCPO }, { 0x061F, UCPO }, { 0x066A, UCPO },
132
+ { 0x066B, UCPO }, { 0x066C, UCPO }, { 0x066D, UCPO }, { 0x06D4, UCPO },
133
+ { 0x0700, UCPO }, { 0x0701, UCPO }, { 0x0702, UCPO }, { 0x0703, UCPO },
134
+ { 0x0704, UCPO }, { 0x0705, UCPO }, { 0x0706, UCPO }, { 0x0707, UCPO },
135
+ { 0x0708, UCPO }, { 0x0709, UCPO }, { 0x070A, UCPO }, { 0x070B, UCPO },
136
+ { 0x070C, UCPO }, { 0x070D, UCPO }, { 0x0964, UCPO }, { 0x0965, UCPO },
137
+ { 0x0970, UCPO }, { 0x0DF4, UCPO }, { 0x0E4F, UCPO }, { 0x0E5A, UCPO },
138
+ { 0x0E5B, UCPO }, { 0x0F04, UCPO }, { 0x0F05, UCPO }, { 0x0F06, UCPO },
139
+ { 0x0F07, UCPO }, { 0x0F08, UCPO }, { 0x0F09, UCPO }, { 0x0F0A, UCPO },
140
+ { 0x0F0B, UCPO }, { 0x0F0D, UCPO }, { 0x0F0E, UCPO }, { 0x0F0F, UCPO },
141
+ { 0x0F10, UCPO }, { 0x0F11, UCPO }, { 0x0F12, UCPO }, { 0x0F3A, UCPS },
142
+ { 0x0F3B, UCPE }, { 0x0F3C, UCPS }, { 0x0F3D, UCPE }, { 0x0F85, UCPO },
143
+ { 0x104A, UCPO }, { 0x104B, UCPO }, { 0x104C, UCPO }, { 0x104D, UCPO },
144
+ { 0x104E, UCPO }, { 0x104F, UCPO }, { 0x10FB, UCPO }, { 0x1361, UCPO },
145
+ { 0x1362, UCPO }, { 0x1363, UCPO }, { 0x1364, UCPO }, { 0x1365, UCPO },
146
+ { 0x1366, UCPO }, { 0x1367, UCPO }, { 0x1368, UCPO }, { 0x166D, UCPO },
147
+ { 0x166E, UCPO }, { 0x1680, UCZS }, { 0x169B, UCPS }, { 0x169C, UCPE },
148
+ { 0x16EB, UCPO }, { 0x16EC, UCPO }, { 0x16ED, UCPO }, { 0x1735, UCPO },
149
+ { 0x1736, UCPO }, { 0x17D4, UCPO }, { 0x17D5, UCPO }, { 0x17D6, UCPO },
150
+ { 0x17D8, UCPO }, { 0x17D9, UCPO }, { 0x17DA, UCPO }, { 0x1800, UCPO },
151
+ { 0x1801, UCPO }, { 0x1802, UCPO }, { 0x1803, UCPO }, { 0x1804, UCPO },
152
+ { 0x1805, UCPO }, { 0x1806, UCPD }, { 0x1807, UCPO }, { 0x1808, UCPO },
153
+ { 0x1809, UCPO }, { 0x180A, UCPO }, { 0x180E, UCZS }, { 0x1944, UCPO },
154
+ { 0x1945, UCPO },
155
+ #endif
156
+ { 0x2000, UCZS }, { 0x2001, UCZS }, { 0x2002, UCZS }, { 0x2003, UCZS },
157
+ { 0x2004, UCZS }, { 0x2005, UCZS }, { 0x2006, UCZS }, { 0x2008, UCZS },
158
+ { 0x2009, UCZS }, { 0x200A, UCZS }, { 0x2010, UCPD }, { 0x2012, UCPD },
159
+ { 0x2013, UCPD }, { 0x2014, UCPD }, { 0x2015, UCPD }, { 0x2016, UCPO },
160
+ { 0x2017, UCPO }, { 0x2018, UCPI }, { 0x2019, UCPF }, { 0x201A, UCPS },
161
+ { 0x201B, UCPI }, { 0x201C, UCPI }, { 0x201D, UCPF }, { 0x201E, UCPS },
162
+ { 0x201F, UCPI }, { 0x2020, UCPO }, { 0x2021, UCPO }, { 0x2022, UCPO },
163
+ { 0x2023, UCPO }, { 0x2024, UCPO }, { 0x2025, UCPO }, { 0x2026, UCPO },
164
+ { 0x2027, UCPO }, { 0x2028, UCZL }, { 0x2029, UCZP }, { 0x2030, UCPO },
165
+ { 0x2031, UCPO }, { 0x2032, UCPO }, { 0x2033, UCPO }, { 0x2034, UCPO },
166
+ { 0x2035, UCPO }, { 0x2036, UCPO }, { 0x2037, UCPO }, { 0x2038, UCPO },
167
+ { 0x2039, UCPI }, { 0x203A, UCPF }, { 0x203B, UCPO }, { 0x203C, UCPO },
168
+ { 0x203D, UCPO }, { 0x203E, UCPO }, { 0x203F, UCPC }, { 0x2040, UCPC },
169
+ { 0x2041, UCPO }, { 0x2042, UCPO }, { 0x2043, UCPO }, { 0x2045, UCPS },
170
+ { 0x2046, UCPE }, { 0x2047, UCPO }, { 0x2048, UCPO }, { 0x2049, UCPO },
171
+ { 0x204A, UCPO }, { 0x204B, UCPO }, { 0x204C, UCPO }, { 0x204D, UCPO },
172
+ { 0x204E, UCPO }, { 0x204F, UCPO }, { 0x2050, UCPO }, { 0x2051, UCPO },
173
+ { 0x2053, UCPO }, { 0x2054, UCPC }, { 0x2057, UCPO }, { 0x205F, UCZS },
174
+ { 0x207D, UCPS }, { 0x207E, UCPE }, { 0x208D, UCPS }, { 0x208E, UCPE },
175
+ { 0x2329, UCPS }, { 0x232A, UCPE }, { 0x23B4, UCPS }, { 0x23B5, UCPE },
176
+ { 0x23B6, UCPO }, { 0x2768, UCPS }, { 0x2769, UCPE }, { 0x276A, UCPS },
177
+ { 0x276B, UCPE }, { 0x276C, UCPS }, { 0x276D, UCPE }, { 0x276E, UCPS },
178
+ { 0x276F, UCPE }, { 0x2770, UCPS }, { 0x2771, UCPE }, { 0x2772, UCPS },
179
+ { 0x2773, UCPE }, { 0x2774, UCPS }, { 0x2775, UCPE }, { 0x27E6, UCPS },
180
+ { 0x27E7, UCPE }, { 0x27E8, UCPS }, { 0x27E9, UCPE }, { 0x27EA, UCPS },
181
+ { 0x27EB, UCPE }, { 0x2983, UCPS }, { 0x2984, UCPE }, { 0x2985, UCPS },
182
+ { 0x2986, UCPE }, { 0x2987, UCPS }, { 0x2988, UCPE }, { 0x2989, UCPS },
183
+ { 0x298A, UCPE }, { 0x298B, UCPS }, { 0x298C, UCPE }, { 0x298D, UCPS },
184
+ { 0x298E, UCPE }, { 0x298F, UCPS }, { 0x2990, UCPE }, { 0x2991, UCPS },
185
+ { 0x2992, UCPE }, { 0x2993, UCPS }, { 0x2994, UCPE }, { 0x2995, UCPS },
186
+ { 0x2996, UCPE }, { 0x2997, UCPS }, { 0x2998, UCPE }, { 0x29D8, UCPS },
187
+ { 0x29D9, UCPE }, { 0x29DA, UCPS }, { 0x29DB, UCPE }, { 0x29FC, UCPS },
188
+ { 0x29FD, UCPE }, { 0x3001, UCPO }, { 0x3002, UCPO }, { 0x3003, UCPO },
189
+ { 0x3008, UCPS }, { 0x3009, UCPE }, { 0x300A, UCPS }, { 0x300B, UCPE },
190
+ { 0x300C, UCPS }, { 0x300D, UCPE }, { 0x300E, UCPS }, { 0x300F, UCPE },
191
+ { 0x3010, UCPS }, { 0x3011, UCPE }, { 0x3014, UCPS }, { 0x3015, UCPE },
192
+ { 0x3016, UCPS }, { 0x3017, UCPE }, { 0x3018, UCPS }, { 0x3019, UCPE },
193
+ { 0x301A, UCPS }, { 0x301B, UCPE }, { 0x301C, UCPD }, { 0x301D, UCPS },
194
+ { 0x301E, UCPE }, { 0x301F, UCPE }, { 0x3030, UCPD }, { 0x303D, UCPO },
195
+ { 0x30A0, UCPD }, { 0x30FB, UCPC }, { 0xFD3E, UCPS }, { 0xFD3F, UCPE },
196
+ { 0xFE30, UCPO }, { 0xFE31, UCPD }, { 0xFE32, UCPD }, { 0xFE33, UCPC },
197
+ { 0xFE34, UCPC }, { 0xFE35, UCPS }, { 0xFE36, UCPE }, { 0xFE37, UCPS },
198
+ { 0xFE38, UCPE }, { 0xFE39, UCPS }, { 0xFE3A, UCPE }, { 0xFE3B, UCPS },
199
+ { 0xFE3C, UCPE }, { 0xFE3D, UCPS }, { 0xFE3E, UCPE }, { 0xFE3F, UCPS },
200
+ { 0xFE40, UCPE }, { 0xFE41, UCPS }, { 0xFE42, UCPE }, { 0xFE43, UCPS },
201
+ { 0xFE44, UCPE }, { 0xFE45, UCPO }, { 0xFE46, UCPO }, { 0xFE47, UCPS },
202
+ { 0xFE48, UCPE }, { 0xFE49, UCPO }, { 0xFE4A, UCPO }, { 0xFE4B, UCPO },
203
+ { 0xFE4C, UCPO }, { 0xFE4D, UCPC }, { 0xFE4E, UCPC }, { 0xFE4F, UCPC },
204
+ { 0xFE50, UCPO }, { 0xFE51, UCPO }, { 0xFE52, UCPO }, { 0xFE54, UCPO },
205
+ { 0xFE55, UCPO }, { 0xFE56, UCPO }, { 0xFE57, UCPO }, { 0xFE58, UCPD },
206
+ { 0xFE59, UCPS }, { 0xFE5A, UCPE }, { 0xFE5B, UCPS }, { 0xFE5C, UCPE },
207
+ { 0xFE5D, UCPS }, { 0xFE5E, UCPE }, { 0xFE5F, UCPO }, { 0xFE60, UCPO },
208
+ { 0xFE61, UCPO }, { 0xFE63, UCPD }, { 0xFE68, UCPO }, { 0xFE6A, UCPO },
209
+ { 0xFE6B, UCPO }, { 0xFF01, UCPO }, { 0xFF02, UCPO }, { 0xFF03, UCPO },
210
+ { 0xFF05, UCPO }, { 0xFF06, UCPO }, { 0xFF07, UCPO }, { 0xFF08, UCPS },
211
+ { 0xFF09, UCPE }, { 0xFF0A, UCPO }, { 0xFF0C, UCPO }, { 0xFF0D, UCPD },
212
+ { 0xFF0E, UCPO }, { 0xFF0F, UCPO }, { 0xFF1A, UCPO }, { 0xFF1B, UCPO },
213
+ { 0xFF1F, UCPO }, { 0xFF20, UCPO }, { 0xFF3B, UCPS }, { 0xFF3C, UCPO },
214
+ { 0xFF3D, UCPE }, { 0xFF3F, UCPC }, { 0xFF5B, UCPS }, { 0xFF5D, UCPE },
215
+ { 0xFF5F, UCPS }, { 0xFF60, UCPE }, { 0xFF61, UCPO }, { 0xFF62, UCPS },
216
+ { 0xFF63, UCPE }, { 0xFF64, UCPO }, { 0xFF65, UCPC }, { 0x10100,UCPO },
217
+ { 0x10101,UCPO }, { 0x1039F,UCPO },
218
+
219
+ /* final entry */
220
+ { 0x0000, UC00 }
221
+ };
222
+
223
+ typedef enum
224
+ {
225
+ NoWrapPoint,
226
+ WrapBefore,
227
+ WrapAfter
228
+ } WrapPoint;
229
+
230
+ /*
231
+ If long lines of text have no white space as defined in HTML 4
232
+ (U+0009, U+000A, U+000D, U+000C, U+0020) other characters could
233
+ be used to determine a wrap point. Since user agents would
234
+ normalize the inserted newline character to a space character,
235
+ this wrapping behaviour would insert visual whitespace into the
236
+ document.
237
+
238
+ Characters of the General Category Pi and Ps in the Unicode
239
+ character database (opening punctuation and intial quote
240
+ characters) mark a wrapping point before the character, other
241
+ punctuation characters (Pc, Pd, Pe, Pf, and Po), breakable
242
+ space characters (Zs), and paragraph and line separators
243
+ (Zl, Zp) mark a wrap point after the character. Using this
244
+ function Tidy can for example pretty print
245
+
246
+ <p>....................&ldquo;...quote...&rdquo;...</p>
247
+ as
248
+ <p>....................\n&ldquo;...quote...&rdquo;...</p>
249
+ or
250
+ <p>....................&ldquo;...quote...&rdquo;\n...</p>
251
+
252
+ if the next normal wrapping point would exceed the user
253
+ chosen wrapping column.
254
+ */
255
+ static WrapPoint CharacterWrapPoint(tchar c)
256
+ {
257
+ int i;
258
+ for (i = 0; unicode4cat[i].code && unicode4cat[i].code <= c; ++i)
259
+ if (unicode4cat[i].code == c)
260
+ {
261
+ /* wrapping before opening punctuation and initial quotes */
262
+ if (unicode4cat[i].category == UCPS ||
263
+ unicode4cat[i].category == UCPI)
264
+ return WrapBefore;
265
+ /* else wrapping after this character */
266
+ else
267
+ return WrapAfter;
268
+ }
269
+ /* character has no effect on line wrapping */
270
+ return NoWrapPoint;
271
+ }
272
+
273
+ static WrapPoint Big5WrapPoint(tchar c)
274
+ {
275
+ if ((c & 0xFF00) == 0xA100)
276
+ {
277
+ /* opening brackets have odd codes: break before them */
278
+ if ( c > 0xA15C && c < 0xA1AD && (c & 1) == 1 )
279
+ return WrapBefore;
280
+ return WrapAfter;
281
+ }
282
+ return NoWrapPoint;
283
+ }
284
+
285
+ #endif /* SUPPORT_ASIAN_ENCODINGS */
286
+
287
+ static void InitIndent( TidyIndent* ind )
288
+ {
289
+ ind->spaces = -1;
290
+ ind->attrValStart = -1;
291
+ ind->attrStringStart = -1;
292
+ }
293
+
294
+ void TY_(InitPrintBuf)( TidyDocImpl* doc )
295
+ {
296
+ TidyClearMemory( &doc->pprint, sizeof(TidyPrintImpl) );
297
+ InitIndent( &doc->pprint.indent[0] );
298
+ InitIndent( &doc->pprint.indent[1] );
299
+ doc->pprint.allocator = doc->allocator;
300
+ }
301
+
302
+ void TY_(FreePrintBuf)( TidyDocImpl* doc )
303
+ {
304
+ TidyDocFree( doc, doc->pprint.linebuf );
305
+ TY_(InitPrintBuf)( doc );
306
+ }
307
+
308
+ static void expand( TidyPrintImpl* pprint, uint len )
309
+ {
310
+ uint* ip;
311
+ uint buflen = pprint->lbufsize;
312
+
313
+ if ( buflen == 0 )
314
+ buflen = 256;
315
+ while ( len >= buflen )
316
+ buflen *= 2;
317
+
318
+ ip = (uint*) TidyRealloc( pprint->allocator, pprint->linebuf, buflen*sizeof(uint) );
319
+ if ( ip )
320
+ {
321
+ TidyClearMemory( ip+pprint->lbufsize,
322
+ (buflen-pprint->lbufsize)*sizeof(uint) );
323
+ pprint->lbufsize = buflen;
324
+ pprint->linebuf = ip;
325
+ }
326
+ }
327
+
328
+ static uint GetSpaces( TidyPrintImpl* pprint )
329
+ {
330
+ int spaces = pprint->indent[ 0 ].spaces;
331
+ return ( spaces < 0 ? 0U : (uint) spaces );
332
+ }
333
+ static int ClearInString( TidyPrintImpl* pprint )
334
+ {
335
+ TidyIndent *ind = pprint->indent + pprint->ixInd;
336
+ return ind->attrStringStart = -1;
337
+ }
338
+ static int ToggleInString( TidyPrintImpl* pprint )
339
+ {
340
+ TidyIndent *ind = pprint->indent + pprint->ixInd;
341
+ Bool inString = ( ind->attrStringStart >= 0 );
342
+ return ind->attrStringStart = ( inString ? -1 : (int) pprint->linelen );
343
+ }
344
+ static Bool IsInString( TidyPrintImpl* pprint )
345
+ {
346
+ TidyIndent *ind = pprint->indent + 0; /* Always 1st */
347
+ return ( ind->attrStringStart >= 0 &&
348
+ ind->attrStringStart < (int) pprint->linelen );
349
+ }
350
+ static Bool IsWrapInString( TidyPrintImpl* pprint )
351
+ {
352
+ TidyIndent *ind = pprint->indent + 0; /* Always 1st */
353
+ int wrap = (int) pprint->wraphere;
354
+ return ( ind->attrStringStart == 0 ||
355
+ (ind->attrStringStart > 0 && ind->attrStringStart < wrap) );
356
+ }
357
+
358
+ static Bool HasMixedContent (Node *element)
359
+ {
360
+ Node * node;
361
+
362
+ if (!element)
363
+ return no;
364
+
365
+ for (node = element->content; node; node = node->next)
366
+ if ( TY_(nodeIsText)(node) )
367
+ return yes;
368
+
369
+ return no;
370
+ }
371
+
372
+ static void ClearInAttrVal( TidyPrintImpl* pprint )
373
+ {
374
+ TidyIndent *ind = pprint->indent + pprint->ixInd;
375
+ ind->attrValStart = -1;
376
+ }
377
+ static int SetInAttrVal( TidyPrintImpl* pprint )
378
+ {
379
+ TidyIndent *ind = pprint->indent + pprint->ixInd;
380
+ return ind->attrValStart = (int) pprint->linelen;
381
+ }
382
+ static Bool IsWrapInAttrVal( TidyPrintImpl* pprint )
383
+ {
384
+ TidyIndent *ind = pprint->indent + 0; /* Always 1st */
385
+ int wrap = (int) pprint->wraphere;
386
+ return ( ind->attrValStart == 0 ||
387
+ (ind->attrValStart > 0 && ind->attrValStart < wrap) );
388
+ }
389
+
390
+ static Bool WantIndent( TidyDocImpl* doc )
391
+ {
392
+ TidyPrintImpl* pprint = &doc->pprint;
393
+ Bool wantIt = GetSpaces(pprint) > 0;
394
+ if ( wantIt )
395
+ {
396
+ Bool indentAttrs = cfgBool( doc, TidyIndentAttributes );
397
+ wantIt = ( ( !IsWrapInAttrVal(pprint) || indentAttrs ) &&
398
+ !IsWrapInString(pprint) );
399
+ }
400
+ return wantIt;
401
+ }
402
+
403
+
404
+ static uint WrapOff( TidyDocImpl* doc )
405
+ {
406
+ uint saveWrap = cfg( doc, TidyWrapLen );
407
+ TY_(SetOptionInt)( doc, TidyWrapLen, 0xFFFFFFFF ); /* very large number */
408
+ return saveWrap;
409
+ }
410
+
411
+ static void WrapOn( TidyDocImpl* doc, uint saveWrap )
412
+ {
413
+ TY_(SetOptionInt)( doc, TidyWrapLen, saveWrap );
414
+ }
415
+
416
+ static uint WrapOffCond( TidyDocImpl* doc, Bool onoff )
417
+ {
418
+ if ( onoff )
419
+ return WrapOff( doc );
420
+ return cfg( doc, TidyWrapLen );
421
+ }
422
+
423
+
424
+ static void AddC( TidyPrintImpl* pprint, uint c, uint string_index)
425
+ {
426
+ if ( string_index + 1 >= pprint->lbufsize )
427
+ expand( pprint, string_index + 1 );
428
+ pprint->linebuf[string_index] = c;
429
+ }
430
+
431
+ static uint AddChar( TidyPrintImpl* pprint, uint c )
432
+ {
433
+ AddC( pprint, c, pprint->linelen );
434
+ return ++pprint->linelen;
435
+ }
436
+
437
+ static uint AddAsciiString( TidyPrintImpl* pprint, ctmbstr str, uint string_index )
438
+ {
439
+ uint ix, len = TY_(tmbstrlen)( str );
440
+ if ( string_index + len >= pprint->lbufsize )
441
+ expand( pprint, string_index + len );
442
+
443
+ for ( ix=0; ix<len; ++ix )
444
+ pprint->linebuf[string_index + ix] = str[ ix ];
445
+ return string_index + len;
446
+ }
447
+
448
+ static uint AddString( TidyPrintImpl* pprint, ctmbstr str )
449
+ {
450
+ return pprint->linelen = AddAsciiString( pprint, str, pprint->linelen );
451
+ }
452
+
453
+ /* Saves current output point as the wrap point,
454
+ ** but only if indentation would NOT overflow
455
+ ** the current line. Otherwise keep previous wrap point.
456
+ */
457
+ static Bool SetWrap( TidyDocImpl* doc, uint indent )
458
+ {
459
+ TidyPrintImpl* pprint = &doc->pprint;
460
+ Bool wrap = ( indent + pprint->linelen < cfg(doc, TidyWrapLen) );
461
+ if ( wrap )
462
+ {
463
+ if ( pprint->indent[0].spaces < 0 )
464
+ pprint->indent[0].spaces = indent;
465
+ pprint->wraphere = pprint->linelen;
466
+ }
467
+ else if ( pprint->ixInd == 0 )
468
+ {
469
+ /* Save indent 1st time we pass the the wrap line */
470
+ pprint->indent[ 1 ].spaces = indent;
471
+ pprint->ixInd = 1;
472
+ }
473
+ return wrap;
474
+ }
475
+
476
+ static void CarryOver( int* valTo, int* valFrom, uint wrapPoint )
477
+ {
478
+ if ( *valFrom > (int) wrapPoint )
479
+ {
480
+ *valTo = *valFrom - wrapPoint;
481
+ *valFrom = -1;
482
+ }
483
+ }
484
+
485
+
486
+ static Bool SetWrapAttr( TidyDocImpl* doc,
487
+ uint indent, int attrStart, int strStart )
488
+ {
489
+ TidyPrintImpl* pprint = &doc->pprint;
490
+ TidyIndent *ind = pprint->indent + 0;
491
+
492
+ Bool wrap = ( indent + pprint->linelen < cfg(doc, TidyWrapLen) );
493
+ if ( wrap )
494
+ {
495
+ if ( ind[0].spaces < 0 )
496
+ ind[0].spaces = indent;
497
+ pprint->wraphere = pprint->linelen;
498
+ }
499
+ else if ( pprint->ixInd == 0 )
500
+ {
501
+ /* Save indent 1st time we pass the the wrap line */
502
+ pprint->indent[ 1 ].spaces = indent;
503
+ pprint->ixInd = 1;
504
+
505
+ /* Carry over string state */
506
+ CarryOver( &ind[1].attrStringStart, &ind[0].attrStringStart, pprint->wraphere );
507
+ CarryOver( &ind[1].attrValStart, &ind[0].attrValStart, pprint->wraphere );
508
+ }
509
+ ind += doc->pprint.ixInd;
510
+ ind->attrValStart = attrStart;
511
+ ind->attrStringStart = strStart;
512
+ return wrap;
513
+ }
514
+
515
+
516
+ /* Reset indent state after flushing a new line
517
+ */
518
+ static void ResetLine( TidyPrintImpl* pprint )
519
+ {
520
+ TidyIndent* ind = pprint->indent + 0;
521
+ if ( pprint->ixInd > 0 )
522
+ {
523
+ ind[0] = ind[1];
524
+ InitIndent( &ind[1] );
525
+ }
526
+
527
+ if ( pprint->wraphere > 0 )
528
+ {
529
+ int wrap = (int) pprint->wraphere;
530
+ if ( ind[0].attrStringStart > wrap )
531
+ ind[0].attrStringStart -= wrap;
532
+ if ( ind[0].attrValStart > wrap )
533
+ ind[0].attrValStart -= wrap;
534
+ }
535
+ else
536
+ {
537
+ if ( ind[0].attrStringStart > 0 )
538
+ ind[0].attrStringStart = 0;
539
+ if ( ind[0].attrValStart > 0 )
540
+ ind[0].attrValStart = 0;
541
+ }
542
+ pprint->wraphere = pprint->ixInd = 0;
543
+ }
544
+
545
+ /* Shift text after wrap point to
546
+ ** beginning of next line.
547
+ */
548
+ static void ResetLineAfterWrap( TidyPrintImpl* pprint )
549
+ {
550
+ if ( pprint->linelen > pprint->wraphere )
551
+ {
552
+ uint *p = pprint->linebuf;
553
+ uint *q = p + pprint->wraphere;
554
+ uint *end = p + pprint->linelen;
555
+
556
+ if ( ! IsWrapInAttrVal(pprint) )
557
+ {
558
+ while ( q < end && *q == ' ' )
559
+ ++q, ++pprint->wraphere;
560
+ }
561
+
562
+ while ( q < end )
563
+ *p++ = *q++;
564
+
565
+ pprint->linelen -= pprint->wraphere;
566
+ }
567
+ else
568
+ {
569
+ pprint->linelen = 0;
570
+ }
571
+
572
+ ResetLine( pprint );
573
+ }
574
+
575
+ /* Goes ahead with writing current line up to
576
+ ** previously saved wrap point. Shifts unwritten
577
+ ** text in output buffer to beginning of next line.
578
+ */
579
+ static void WrapLine( TidyDocImpl* doc )
580
+ {
581
+ TidyPrintImpl* pprint = &doc->pprint;
582
+ uint i;
583
+
584
+ if ( pprint->wraphere == 0 )
585
+ return;
586
+
587
+ if ( WantIndent(doc) )
588
+ {
589
+ uint spaces = GetSpaces( pprint );
590
+ for ( i = 0; i < spaces; ++i )
591
+ TY_(WriteChar)( ' ', doc->docOut );
592
+ }
593
+
594
+ for ( i = 0; i < pprint->wraphere; ++i )
595
+ TY_(WriteChar)( pprint->linebuf[i], doc->docOut );
596
+
597
+ if ( IsWrapInString(pprint) )
598
+ TY_(WriteChar)( '\\', doc->docOut );
599
+
600
+ TY_(WriteChar)( '\n', doc->docOut );
601
+ ResetLineAfterWrap( pprint );
602
+ }
603
+
604
+ /* Checks current output line length along with current indent.
605
+ ** If combined they overflow output line length, go ahead
606
+ ** and flush output up to the current wrap point.
607
+ */
608
+ static Bool CheckWrapLine( TidyDocImpl* doc )
609
+ {
610
+ TidyPrintImpl* pprint = &doc->pprint;
611
+ if ( GetSpaces(pprint) + pprint->linelen >= cfg(doc, TidyWrapLen) )
612
+ {
613
+ WrapLine( doc );
614
+ return yes;
615
+ }
616
+ return no;
617
+ }
618
+
619
+ static Bool CheckWrapIndent( TidyDocImpl* doc, uint indent )
620
+ {
621
+ TidyPrintImpl* pprint = &doc->pprint;
622
+ if ( GetSpaces(pprint) + pprint->linelen >= cfg(doc, TidyWrapLen) )
623
+ {
624
+ WrapLine( doc );
625
+ if ( pprint->indent[ 0 ].spaces < 0 )
626
+ pprint->indent[ 0 ].spaces = indent;
627
+ return yes;
628
+ }
629
+ return no;
630
+ }
631
+
632
+ static void WrapAttrVal( TidyDocImpl* doc )
633
+ {
634
+ TidyPrintImpl* pprint = &doc->pprint;
635
+ uint i;
636
+
637
+ /* assert( IsWrapInAttrVal(pprint) ); */
638
+ if ( WantIndent(doc) )
639
+ {
640
+ uint spaces = GetSpaces( pprint );
641
+ for ( i = 0; i < spaces; ++i )
642
+ TY_(WriteChar)( ' ', doc->docOut );
643
+ }
644
+
645
+ for ( i = 0; i < pprint->wraphere; ++i )
646
+ TY_(WriteChar)( pprint->linebuf[i], doc->docOut );
647
+
648
+ if ( IsWrapInString(pprint) )
649
+ TY_(WriteChar)( '\\', doc->docOut );
650
+ else
651
+ TY_(WriteChar)( ' ', doc->docOut );
652
+
653
+ TY_(WriteChar)( '\n', doc->docOut );
654
+ ResetLineAfterWrap( pprint );
655
+ }
656
+
657
+ static void PFlushLineImpl( TidyDocImpl* doc )
658
+ {
659
+ TidyPrintImpl* pprint = &doc->pprint;
660
+
661
+ uint i;
662
+
663
+ CheckWrapLine( doc );
664
+
665
+ if ( WantIndent(doc) )
666
+ {
667
+ uint spaces = GetSpaces( pprint );
668
+ for ( i = 0; i < spaces; ++i )
669
+ TY_(WriteChar)( ' ', doc->docOut );
670
+ }
671
+
672
+ for ( i = 0; i < pprint->linelen; ++i )
673
+ TY_(WriteChar)( pprint->linebuf[i], doc->docOut );
674
+
675
+ if ( IsInString(pprint) )
676
+ TY_(WriteChar)( '\\', doc->docOut );
677
+ ResetLine( pprint );
678
+ pprint->linelen = 0;
679
+ }
680
+
681
+ void TY_(PFlushLine)( TidyDocImpl* doc, uint indent )
682
+ {
683
+ TidyPrintImpl* pprint = &doc->pprint;
684
+
685
+ if ( pprint->linelen > 0 )
686
+ PFlushLineImpl( doc );
687
+
688
+ TY_(WriteChar)( '\n', doc->docOut );
689
+ pprint->indent[ 0 ].spaces = indent;
690
+ }
691
+
692
+ static void PCondFlushLine( TidyDocImpl* doc, uint indent )
693
+ {
694
+ TidyPrintImpl* pprint = &doc->pprint;
695
+
696
+ if ( pprint->linelen > 0 )
697
+ {
698
+ PFlushLineImpl( doc );
699
+
700
+ TY_(WriteChar)( '\n', doc->docOut );
701
+ pprint->indent[ 0 ].spaces = indent;
702
+ }
703
+ }
704
+
705
+ static void PPrintChar( TidyDocImpl* doc, uint c, uint mode )
706
+ {
707
+ tmbchar entity[128];
708
+ ctmbstr p;
709
+ TidyPrintImpl* pprint = &doc->pprint;
710
+ uint outenc = cfg( doc, TidyOutCharEncoding );
711
+ Bool qmark = cfgBool( doc, TidyQuoteMarks );
712
+
713
+ if ( c == ' ' && !(mode & (PREFORMATTED | COMMENT | ATTRIBVALUE | CDATA)))
714
+ {
715
+ /* coerce a space character to a non-breaking space */
716
+ if (mode & NOWRAP)
717
+ {
718
+ ctmbstr ent = "&nbsp;";
719
+ /* by default XML doesn't define &nbsp; */
720
+ if ( cfgBool(doc, TidyNumEntities) || cfgBool(doc, TidyXmlTags) )
721
+ ent = "&#160;";
722
+ AddString( pprint, ent );
723
+ return;
724
+ }
725
+ else
726
+ pprint->wraphere = pprint->linelen;
727
+ }
728
+
729
+ /* comment characters are passed raw */
730
+ if ( mode & (COMMENT | CDATA) )
731
+ {
732
+ AddChar( pprint, c );
733
+ return;
734
+ }
735
+
736
+ /* except in CDATA map < to &lt; etc. */
737
+ if ( !(mode & CDATA) )
738
+ {
739
+ if ( c == '<')
740
+ {
741
+ AddString( pprint, "&lt;" );
742
+ return;
743
+ }
744
+
745
+ if ( c == '>')
746
+ {
747
+ AddString( pprint, "&gt;" );
748
+ return;
749
+ }
750
+
751
+ /*
752
+ naked '&' chars can be left alone or
753
+ quoted as &amp; The latter is required
754
+ for XML where naked '&' are illegal.
755
+ */
756
+ if ( c == '&' && cfgBool(doc, TidyQuoteAmpersand)
757
+ && !cfgBool(doc, TidyPreserveEntities) )
758
+ {
759
+ AddString( pprint, "&amp;" );
760
+ return;
761
+ }
762
+
763
+ if ( c == '"' && qmark )
764
+ {
765
+ AddString( pprint, "&quot;" );
766
+ return;
767
+ }
768
+
769
+ if ( c == '\'' && qmark )
770
+ {
771
+ AddString( pprint, "&#39;" );
772
+ return;
773
+ }
774
+
775
+ if ( c == 160 && outenc != RAW )
776
+ {
777
+ if ( cfgBool(doc, TidyQuoteNbsp) )
778
+ {
779
+ if ( cfgBool(doc, TidyNumEntities) ||
780
+ cfgBool(doc, TidyXmlTags) )
781
+ AddString( pprint, "&#160;" );
782
+ else
783
+ AddString( pprint, "&nbsp;" );
784
+ }
785
+ else
786
+ AddChar( pprint, c );
787
+ return;
788
+ }
789
+ }
790
+
791
+ #if SUPPORT_ASIAN_ENCODINGS
792
+
793
+ /* #431953 - start RJ */
794
+ /* Handle encoding-specific issues */
795
+ switch ( outenc )
796
+ {
797
+ case UTF8:
798
+ #if SUPPORT_UTF16_ENCODINGS
799
+ case UTF16:
800
+ case UTF16LE:
801
+ case UTF16BE:
802
+ #endif
803
+ if (!(mode & PREFORMATTED) && cfg(doc, TidyPunctWrap))
804
+ {
805
+ WrapPoint wp = CharacterWrapPoint(c);
806
+ if (wp == WrapBefore)
807
+ pprint->wraphere = pprint->linelen;
808
+ else if (wp == WrapAfter)
809
+ pprint->wraphere = pprint->linelen + 1;
810
+ }
811
+ break;
812
+
813
+ case BIG5:
814
+ /* Allow linebreak at Chinese punctuation characters */
815
+ /* There are not many spaces in Chinese */
816
+ AddChar( pprint, c );
817
+ if (!(mode & PREFORMATTED) && cfg(doc, TidyPunctWrap))
818
+ {
819
+ WrapPoint wp = Big5WrapPoint(c);
820
+ if (wp == WrapBefore)
821
+ pprint->wraphere = pprint->linelen;
822
+ else if (wp == WrapAfter)
823
+ pprint->wraphere = pprint->linelen + 1;
824
+ }
825
+ return;
826
+
827
+ case SHIFTJIS:
828
+ #ifndef NO_NATIVE_ISO2022_SUPPORT
829
+ case ISO2022: /* ISO 2022 characters are passed raw */
830
+ #endif
831
+ case RAW:
832
+ AddChar( pprint, c );
833
+ return;
834
+ }
835
+ /* #431953 - end RJ */
836
+
837
+ #else /* SUPPORT_ASIAN_ENCODINGS */
838
+
839
+ /* otherwise ISO 2022 characters are passed raw */
840
+ if (
841
+ #ifndef NO_NATIVE_ISO2022_SUPPORT
842
+ outenc == ISO2022 ||
843
+ #endif
844
+ outenc == RAW )
845
+ {
846
+ AddChar( pprint, c );
847
+ return;
848
+ }
849
+
850
+ #endif /* SUPPORT_ASIAN_ENCODINGS */
851
+
852
+ /* don't map latin-1 chars to entities */
853
+ if ( outenc == LATIN1 )
854
+ {
855
+ if (c > 255) /* multi byte chars */
856
+ {
857
+ uint vers = TY_(HTMLVersion)( doc );
858
+ if ( !cfgBool(doc, TidyNumEntities) && (p = TY_(EntityName)(c, vers)) )
859
+ TY_(tmbsnprintf)(entity, sizeof(entity), "&%s;", p);
860
+ else
861
+ TY_(tmbsnprintf)(entity, sizeof(entity), "&#%u;", c);
862
+
863
+ AddString( pprint, entity );
864
+ return;
865
+ }
866
+
867
+ if (c > 126 && c < 160)
868
+ {
869
+ TY_(tmbsnprintf)(entity, sizeof(entity), "&#%u;", c);
870
+ AddString( pprint, entity );
871
+ return;
872
+ }
873
+
874
+ AddChar( pprint, c );
875
+ return;
876
+ }
877
+
878
+ /* don't map UTF-8 chars to entities */
879
+ if ( outenc == UTF8 )
880
+ {
881
+ AddChar( pprint, c );
882
+ return;
883
+ }
884
+
885
+ #if SUPPORT_UTF16_ENCODINGS
886
+ /* don't map UTF-16 chars to entities */
887
+ if ( outenc == UTF16 || outenc == UTF16LE || outenc == UTF16BE )
888
+ {
889
+ AddChar( pprint, c );
890
+ return;
891
+ }
892
+ #endif
893
+
894
+ /* use numeric entities only for XML */
895
+ if ( cfgBool(doc, TidyXmlTags) )
896
+ {
897
+ /* if ASCII use numeric entities for chars > 127 */
898
+ if ( c > 127 && outenc == ASCII )
899
+ {
900
+ TY_(tmbsnprintf)(entity, sizeof(entity), "&#%u;", c);
901
+ AddString( pprint, entity );
902
+ return;
903
+ }
904
+
905
+ /* otherwise output char raw */
906
+ AddChar( pprint, c );
907
+ return;
908
+ }
909
+
910
+ /* default treatment for ASCII */
911
+ if ( outenc == ASCII && (c > 126 || (c < ' ' && c != '\t')) )
912
+ {
913
+ uint vers = TY_(HTMLVersion)( doc );
914
+ if (!cfgBool(doc, TidyNumEntities) && (p = TY_(EntityName)(c, vers)) )
915
+ TY_(tmbsnprintf)(entity, sizeof(entity), "&%s;", p);
916
+ else
917
+ TY_(tmbsnprintf)(entity, sizeof(entity), "&#%u;", c);
918
+
919
+ AddString( pprint, entity );
920
+ return;
921
+ }
922
+
923
+ AddChar( pprint, c );
924
+ }
925
+
926
+ static uint IncrWS( uint start, uint end, uint indent, int ixWS )
927
+ {
928
+ if ( ixWS > 0 )
929
+ {
930
+ uint st = start + MIN( (uint)ixWS, indent );
931
+ start = MIN( st, end );
932
+ }
933
+ return start;
934
+ }
935
+ /*
936
+ The line buffer is uint not char so we can
937
+ hold Unicode values unencoded. The translation
938
+ to UTF-8 is deferred to the TY_(WriteChar)() routine called
939
+ to flush the line buffer.
940
+ */
941
+ static void PPrintText( TidyDocImpl* doc, uint mode, uint indent,
942
+ Node* node )
943
+ {
944
+ uint start = node->start;
945
+ uint end = node->end;
946
+ uint ix, c = 0;
947
+ int ixNL = TextEndsWithNewline( doc->lexer, node, mode );
948
+ int ixWS = TextStartsWithWhitespace( doc->lexer, node, start, mode );
949
+ if ( ixNL > 0 )
950
+ end -= ixNL;
951
+ start = IncrWS( start, end, indent, ixWS );
952
+
953
+ for ( ix = start; ix < end; ++ix )
954
+ {
955
+ CheckWrapIndent( doc, indent );
956
+ /*
957
+ if ( CheckWrapIndent(doc, indent) )
958
+ {
959
+ ixWS = TextStartsWithWhitespace( doc->lexer, node, ix );
960
+ ix = IncrWS( ix, end, indent, ixWS );
961
+ }
962
+ */
963
+ c = (byte) doc->lexer->lexbuf[ix];
964
+
965
+ /* look for UTF-8 multibyte character */
966
+ if ( c > 0x7F )
967
+ ix += TY_(GetUTF8)( doc->lexer->lexbuf + ix, &c );
968
+
969
+ if ( c == '\n' )
970
+ {
971
+ TY_(PFlushLine)( doc, indent );
972
+ ixWS = TextStartsWithWhitespace( doc->lexer, node, ix+1, mode );
973
+ ix = IncrWS( ix, end, indent, ixWS );
974
+ }
975
+ else
976
+ {
977
+ PPrintChar( doc, c, mode );
978
+ }
979
+ }
980
+ }
981
+
982
+ #if 0
983
+ static void PPrintString( TidyDocImpl* doc, uint indent, ctmbstr str )
984
+ {
985
+ while ( *str != '\0' )
986
+ AddChar( &doc->pprint, *str++ );
987
+ }
988
+ #endif /* 0 */
989
+
990
+
991
+ static void PPrintAttrValue( TidyDocImpl* doc, uint indent,
992
+ ctmbstr value, uint delim, Bool wrappable, Bool scriptAttr )
993
+ {
994
+ TidyPrintImpl* pprint = &doc->pprint;
995
+ Bool scriptlets = cfgBool(doc, TidyWrapScriptlets);
996
+
997
+ uint mode = PREFORMATTED | ATTRIBVALUE;
998
+ if ( wrappable )
999
+ mode = NORMAL | ATTRIBVALUE;
1000
+
1001
+ /* look for ASP, Tango or PHP instructions for computed attribute value */
1002
+ if ( value && value[0] == '<' )
1003
+ {
1004
+ if ( value[1] == '%' || value[1] == '@'||
1005
+ TY_(tmbstrncmp)(value, "<?php", 5) == 0 )
1006
+ mode |= CDATA;
1007
+ }
1008
+
1009
+ if ( delim == 0 )
1010
+ delim = '"';
1011
+
1012
+ AddChar( pprint, '=' );
1013
+
1014
+ /* don't wrap after "=" for xml documents */
1015
+ if ( !cfgBool(doc, TidyXmlOut) || cfgBool(doc, TidyXhtmlOut) )
1016
+ {
1017
+ SetWrap( doc, indent );
1018
+ CheckWrapIndent( doc, indent );
1019
+ /*
1020
+ if ( !SetWrap(doc, indent) )
1021
+ PCondFlushLine( doc, indent );
1022
+ */
1023
+ }
1024
+
1025
+ AddChar( pprint, delim );
1026
+
1027
+ if ( value )
1028
+ {
1029
+ uint wraplen = cfg( doc, TidyWrapLen );
1030
+ int attrStart = SetInAttrVal( pprint );
1031
+ int strStart = ClearInString( pprint );
1032
+
1033
+ while (*value != '\0')
1034
+ {
1035
+ uint c = *value;
1036
+
1037
+ if ( wrappable && c == ' ' )
1038
+ SetWrapAttr( doc, indent, attrStart, strStart );
1039
+
1040
+ if ( wrappable && pprint->wraphere > 0 &&
1041
+ GetSpaces(pprint) + pprint->linelen >= wraplen )
1042
+ WrapAttrVal( doc );
1043
+
1044
+ if ( c == delim )
1045
+ {
1046
+ ctmbstr entity = (c == '"' ? "&quot;" : "&#39;");
1047
+ AddString( pprint, entity );
1048
+ ++value;
1049
+ continue;
1050
+ }
1051
+ else if (c == '"')
1052
+ {
1053
+ if ( cfgBool(doc, TidyQuoteMarks) )
1054
+ AddString( pprint, "&quot;" );
1055
+ else
1056
+ AddChar( pprint, c );
1057
+
1058
+ if ( delim == '\'' && scriptAttr && scriptlets )
1059
+ strStart = ToggleInString( pprint );
1060
+
1061
+ ++value;
1062
+ continue;
1063
+ }
1064
+ else if ( c == '\'' )
1065
+ {
1066
+ if ( cfgBool(doc, TidyQuoteMarks) )
1067
+ AddString( pprint, "&#39;" );
1068
+ else
1069
+ AddChar( pprint, c );
1070
+
1071
+ if ( delim == '"' && scriptAttr && scriptlets )
1072
+ strStart = ToggleInString( pprint );
1073
+
1074
+ ++value;
1075
+ continue;
1076
+ }
1077
+
1078
+ /* look for UTF-8 multibyte character */
1079
+ if ( c > 0x7F )
1080
+ value += TY_(GetUTF8)( value, &c );
1081
+ ++value;
1082
+
1083
+ if ( c == '\n' )
1084
+ {
1085
+ /* No indent inside Javascript literals */
1086
+ TY_(PFlushLine)( doc, (strStart < 0
1087
+ && !cfgBool(doc, TidyLiteralAttribs) ?
1088
+ indent : 0) );
1089
+ continue;
1090
+ }
1091
+ PPrintChar( doc, c, mode );
1092
+ }
1093
+ ClearInAttrVal( pprint );
1094
+ ClearInString( pprint );
1095
+ }
1096
+ AddChar( pprint, delim );
1097
+ }
1098
+
1099
+ static uint AttrIndent( TidyDocImpl* doc, Node* node, AttVal* ARG_UNUSED(attr) )
1100
+ {
1101
+ uint spaces = cfg( doc, TidyIndentSpaces );
1102
+ uint xtra = 2; /* 1 for the '<', another for the ' ' */
1103
+ if ( node->element == NULL )
1104
+ return spaces;
1105
+
1106
+ if ( !TY_(nodeHasCM)(node, CM_INLINE) ||
1107
+ !ShouldIndent(doc, node->parent ? node->parent: node) )
1108
+ return xtra + TY_(tmbstrlen)( node->element );
1109
+
1110
+ if ( NULL != (node = TY_(FindContainer)(node)) )
1111
+ return xtra + TY_(tmbstrlen)( node->element );
1112
+ return spaces;
1113
+ }
1114
+
1115
+ static Bool AttrNoIndentFirst( /*TidyDocImpl* doc,*/ Node* node, AttVal* attr )
1116
+ {
1117
+ return ( attr==node->attributes );
1118
+
1119
+ /*&&
1120
+ ( InsideHead(doc, node) ||
1121
+ !TY_(nodeHasCM)(node, CM_INLINE) ) );
1122
+ */
1123
+ }
1124
+
1125
+ static void PPrintAttribute( TidyDocImpl* doc, uint indent,
1126
+ Node *node, AttVal *attr )
1127
+ {
1128
+ TidyPrintImpl* pprint = &doc->pprint;
1129
+ Bool xmlOut = cfgBool( doc, TidyXmlOut );
1130
+ Bool xhtmlOut = cfgBool( doc, TidyXhtmlOut );
1131
+ Bool wrapAttrs = cfgBool( doc, TidyWrapAttVals );
1132
+ Bool ucAttrs = cfgBool( doc, TidyUpperCaseAttrs );
1133
+ Bool indAttrs = cfgBool( doc, TidyIndentAttributes );
1134
+ uint xtra = AttrIndent( doc, node, attr );
1135
+ Bool first = AttrNoIndentFirst( /*doc,*/ node, attr );
1136
+ tmbstr name = attr->attribute;
1137
+ Bool wrappable = no;
1138
+ tchar c;
1139
+
1140
+ /* fix for odd attribute indentation bug triggered by long values */
1141
+ if (!indAttrs)
1142
+ xtra = 0;
1143
+
1144
+ if ( indAttrs )
1145
+ {
1146
+ if ( TY_(nodeIsElement)(node) && !first )
1147
+ {
1148
+ indent += xtra;
1149
+ PCondFlushLine( doc, indent );
1150
+ }
1151
+ else
1152
+ indAttrs = no;
1153
+ }
1154
+
1155
+ CheckWrapIndent( doc, indent );
1156
+
1157
+ if ( !xmlOut && !xhtmlOut && attr->dict )
1158
+ {
1159
+ if ( TY_(IsScript)(doc, name) )
1160
+ wrappable = cfgBool( doc, TidyWrapScriptlets );
1161
+ else if (!(attrIsCONTENT(attr) || attrIsVALUE(attr) || attrIsALT(attr)) && wrapAttrs )
1162
+ wrappable = yes;
1163
+ }
1164
+
1165
+ if ( !first && !SetWrap(doc, indent) )
1166
+ {
1167
+ TY_(PFlushLine)( doc, indent+xtra ); /* Put it on next line */
1168
+ }
1169
+ else if ( pprint->linelen > 0 )
1170
+ {
1171
+ AddChar( pprint, ' ' );
1172
+ }
1173
+
1174
+ /* Attribute name */
1175
+ while (*name)
1176
+ {
1177
+ c = (unsigned char)*name;
1178
+
1179
+ if (c > 0x7F)
1180
+ name += TY_(GetUTF8)(name, &c);
1181
+ else if (ucAttrs)
1182
+ c = TY_(ToUpper)(c);
1183
+
1184
+ AddChar(pprint, c);
1185
+ ++name;
1186
+ }
1187
+
1188
+ /* fix for bug 732038 */
1189
+ #if 0
1190
+ /* If not indenting attributes, bump up indent for
1191
+ ** value after putting out name.
1192
+ */
1193
+ if ( !indAttrs )
1194
+ indent += xtra;
1195
+ #endif
1196
+
1197
+ CheckWrapIndent( doc, indent );
1198
+
1199
+ if ( attr->value == NULL )
1200
+ {
1201
+ Bool isB = TY_(IsBoolAttribute)(attr);
1202
+ Bool scriptAttr = TY_(attrIsEvent)(attr);
1203
+
1204
+ if ( xmlOut )
1205
+ PPrintAttrValue( doc, indent, isB ? attr->attribute : NULLSTR,
1206
+ attr->delim, no, scriptAttr );
1207
+
1208
+ else if ( !isB && !TY_(IsNewNode)(node) )
1209
+ PPrintAttrValue( doc, indent, "", attr->delim, yes, scriptAttr );
1210
+
1211
+ else
1212
+ SetWrap( doc, indent );
1213
+ }
1214
+ else
1215
+ PPrintAttrValue( doc, indent, attr->value, attr->delim, wrappable, no );
1216
+ }
1217
+
1218
+ static void PPrintAttrs( TidyDocImpl* doc, uint indent, Node *node )
1219
+ {
1220
+ TidyPrintImpl* pprint = &doc->pprint;
1221
+ AttVal* av;
1222
+
1223
+ /* add xml:space attribute to pre and other elements */
1224
+ if ( cfgBool(doc, TidyXmlOut) && cfgBool(doc, TidyXmlSpace) &&
1225
+ !TY_(GetAttrByName)(node, "xml:space") &&
1226
+ TY_(XMLPreserveWhiteSpace)(doc, node) )
1227
+ {
1228
+ TY_(AddAttribute)( doc, node, "xml:space", "preserve" );
1229
+ }
1230
+
1231
+ for ( av = node->attributes; av; av = av->next )
1232
+ {
1233
+ if ( av->attribute != NULL )
1234
+ {
1235
+ PPrintAttribute( doc, indent, node, av );
1236
+ }
1237
+ else if ( av->asp != NULL )
1238
+ {
1239
+ AddChar( pprint, ' ' );
1240
+ PPrintAsp( doc, indent, av->asp );
1241
+ }
1242
+ else if ( av->php != NULL )
1243
+ {
1244
+ AddChar( pprint, ' ' );
1245
+ PPrintPhp( doc, indent, av->php );
1246
+ }
1247
+ }
1248
+ }
1249
+
1250
+ Bool TY_(TextNodeEndWithSpace)( Lexer *lexer, Node *node )
1251
+ {
1252
+ if (TY_(nodeIsText)(node) && node->end > node->start)
1253
+ {
1254
+ uint i, c = '\0'; /* initialised to avoid warnings */
1255
+ for (i = node->start; i < node->end; ++i)
1256
+ {
1257
+ c = (byte) lexer->lexbuf[i];
1258
+ if ( c > 0x7F )
1259
+ i += TY_(GetUTF8)( lexer->lexbuf + i, &c );
1260
+ }
1261
+
1262
+ if ( c == ' ' || c == '\n' )
1263
+ return yes;
1264
+ }
1265
+ return no;
1266
+ }
1267
+
1268
+ /*
1269
+ Line can be wrapped immediately after inline start tag provided
1270
+ if follows a text node ending in a space, or it follows a <br>,
1271
+ or its parent is an inline element that that rule applies to.
1272
+ This behaviour was reverse engineered from Netscape 3.0.
1273
+
1274
+ Line wrapping can occur if an element is not empty and before a block
1275
+ level. For instance:
1276
+ <p><span>
1277
+ x</span>y</p>
1278
+ will display properly. Whereas
1279
+ <p><img />
1280
+ x</p> won't.
1281
+ */
1282
+ static Bool AfterSpaceImp(Lexer *lexer, Node *node, Bool isEmpty)
1283
+ {
1284
+ Node *prev;
1285
+
1286
+ if ( !TY_(nodeCMIsInline)(node) )
1287
+ return yes;
1288
+
1289
+ prev = node->prev;
1290
+ if (prev)
1291
+ {
1292
+ if (TY_(nodeIsText)(prev))
1293
+ return TY_(TextNodeEndWithSpace)( lexer, prev );
1294
+ else if (nodeIsBR(prev))
1295
+ return yes;
1296
+
1297
+ return no;
1298
+ }
1299
+
1300
+ if ( isEmpty && !TY_(nodeCMIsInline)(node->parent) )
1301
+ return no;
1302
+
1303
+ return AfterSpaceImp(lexer, node->parent, isEmpty);
1304
+ }
1305
+
1306
+ static Bool AfterSpace(Lexer *lexer, Node *node)
1307
+ {
1308
+ return AfterSpaceImp(lexer, node, TY_(nodeCMIsEmpty)(node));
1309
+ }
1310
+
1311
+ static void PPrintTag( TidyDocImpl* doc,
1312
+ uint mode, uint indent, Node *node )
1313
+ {
1314
+ TidyPrintImpl* pprint = &doc->pprint;
1315
+ Bool uc = cfgBool( doc, TidyUpperCaseTags );
1316
+ Bool xhtmlOut = cfgBool( doc, TidyXhtmlOut );
1317
+ Bool xmlOut = cfgBool( doc, TidyXmlOut );
1318
+ tchar c;
1319
+ tmbstr s = node->element;
1320
+
1321
+ AddChar( pprint, '<' );
1322
+
1323
+ if ( node->type == EndTag )
1324
+ AddChar( pprint, '/' );
1325
+
1326
+ if (s)
1327
+ {
1328
+ while (*s)
1329
+ {
1330
+ c = (unsigned char)*s;
1331
+
1332
+ if (c > 0x7F)
1333
+ s += TY_(GetUTF8)(s, &c);
1334
+ else if (uc)
1335
+ c = TY_(ToUpper)(c);
1336
+
1337
+ AddChar(pprint, c);
1338
+ ++s;
1339
+ }
1340
+ }
1341
+
1342
+ PPrintAttrs( doc, indent, node );
1343
+
1344
+ if ( (xmlOut || xhtmlOut) &&
1345
+ (node->type == StartEndTag || TY_(nodeCMIsEmpty)(node)) )
1346
+ {
1347
+ AddChar( pprint, ' ' ); /* Space is NS compatibility hack <br /> */
1348
+ AddChar( pprint, '/' ); /* Required end tag marker */
1349
+ }
1350
+
1351
+ AddChar( pprint, '>' );
1352
+
1353
+ if ( (node->type != StartEndTag || xhtmlOut) && !(mode & PREFORMATTED) )
1354
+ {
1355
+ uint wraplen = cfg( doc, TidyWrapLen );
1356
+ CheckWrapIndent( doc, indent );
1357
+
1358
+ if ( indent + pprint->linelen < wraplen )
1359
+ {
1360
+ /* wrap after start tag if is <br/> or if it's not inline.
1361
+ Technically, it would be safe to call only AfterSpace.
1362
+ However, it would disrupt the existing algorithm. So let's
1363
+ leave as is. Note that AfterSpace returns true for non inline
1364
+ elements but can still be false for some <br>. So it has to
1365
+ stay as well. */
1366
+ if (!(mode & NOWRAP)
1367
+ && (!TY_(nodeCMIsInline)(node) || nodeIsBR(node))
1368
+ && AfterSpace(doc->lexer, node))
1369
+ {
1370
+ pprint->wraphere = pprint->linelen;
1371
+ }
1372
+ }
1373
+ /* flush the current buffer only if it is known to be safe,
1374
+ i.e. it will not introduce some spurious white spaces.
1375
+ See bug #996484 */
1376
+ else if ( mode & NOWRAP ||
1377
+ nodeIsBR(node) || AfterSpace(doc->lexer, node))
1378
+ PCondFlushLine( doc, indent );
1379
+ }
1380
+ }
1381
+
1382
+ static void PPrintEndTag( TidyDocImpl* doc, uint ARG_UNUSED(mode),
1383
+ uint ARG_UNUSED(indent), Node *node )
1384
+ {
1385
+ TidyPrintImpl* pprint = &doc->pprint;
1386
+ Bool uc = cfgBool( doc, TidyUpperCaseTags );
1387
+ tmbstr s = node->element;
1388
+ tchar c;
1389
+
1390
+ /*
1391
+ Netscape ignores SGML standard by not ignoring a
1392
+ line break before </A> or </U> etc. To avoid rendering
1393
+ this as an underlined space, I disable line wrapping
1394
+ before inline end tags by the #if 0 ... #endif
1395
+ */
1396
+ #if 0
1397
+ if ( !(mode & NOWRAP) )
1398
+ SetWrap( doc, indent );
1399
+ #endif
1400
+
1401
+ AddString( pprint, "</" );
1402
+
1403
+ if (s)
1404
+ {
1405
+ while (*s)
1406
+ {
1407
+ c = (unsigned char)*s;
1408
+
1409
+ if (c > 0x7F)
1410
+ s += TY_(GetUTF8)(s, &c);
1411
+ else if (uc)
1412
+ c = TY_(ToUpper)(c);
1413
+
1414
+ AddChar(pprint, c);
1415
+ ++s;
1416
+ }
1417
+ }
1418
+
1419
+ AddChar( pprint, '>' );
1420
+ }
1421
+
1422
+ static void PPrintComment( TidyDocImpl* doc, uint indent, Node* node )
1423
+ {
1424
+ TidyPrintImpl* pprint = &doc->pprint;
1425
+
1426
+ SetWrap( doc, indent );
1427
+ AddString( pprint, "<!--" );
1428
+
1429
+ #if 0
1430
+ SetWrap( doc, indent );
1431
+ #endif
1432
+
1433
+ PPrintText(doc, COMMENT, 0, node);
1434
+
1435
+ #if 0
1436
+ SetWrap( doc, indent );
1437
+ AddString( pprint, "--" );
1438
+ #endif
1439
+
1440
+ AddString(pprint, "--");
1441
+ AddChar( pprint, '>' );
1442
+ if ( node->linebreak && node->next )
1443
+ TY_(PFlushLine)( doc, indent );
1444
+ }
1445
+
1446
+ static void PPrintDocType( TidyDocImpl* doc, uint indent, Node *node )
1447
+ {
1448
+ TidyPrintImpl* pprint = &doc->pprint;
1449
+ uint wraplen = cfg( doc, TidyWrapLen );
1450
+ uint spaces = cfg( doc, TidyIndentSpaces );
1451
+ AttVal* fpi = TY_(GetAttrByName)(node, "PUBLIC");
1452
+ AttVal* sys = TY_(GetAttrByName)(node, "SYSTEM");
1453
+
1454
+ /* todo: handle non-ASCII characters in FPI / SI / node->element */
1455
+
1456
+ SetWrap( doc, indent );
1457
+ PCondFlushLine( doc, indent );
1458
+
1459
+ AddString( pprint, "<!DOCTYPE " );
1460
+ SetWrap( doc, indent );
1461
+ if (node->element)
1462
+ {
1463
+ AddString(pprint, node->element);
1464
+ }
1465
+
1466
+ if (fpi && fpi->value)
1467
+ {
1468
+ AddString(pprint, " PUBLIC ");
1469
+ AddChar(pprint, fpi->delim);
1470
+ AddString(pprint, fpi->value);
1471
+ AddChar(pprint, fpi->delim);
1472
+ }
1473
+
1474
+ if (fpi && fpi->value && sys && sys->value)
1475
+ {
1476
+ uint i = pprint->linelen - (TY_(tmbstrlen)(sys->value) + 2) - 1;
1477
+ if (!(i>0&&TY_(tmbstrlen)(sys->value)+2+i<wraplen&&i<=(spaces?spaces:2)*2))
1478
+ i = 0;
1479
+
1480
+ PCondFlushLine(doc, i);
1481
+ if (pprint->linelen)
1482
+ AddChar(pprint, ' ');
1483
+ }
1484
+ else if (sys && sys->value)
1485
+ {
1486
+ AddString(pprint, " SYSTEM ");
1487
+ }
1488
+
1489
+ if (sys && sys->value)
1490
+ {
1491
+ AddChar(pprint, sys->delim);
1492
+ AddString(pprint, sys->value);
1493
+ AddChar(pprint, sys->delim);
1494
+ }
1495
+
1496
+ if (node->content)
1497
+ {
1498
+ PCondFlushLine(doc, indent);
1499
+ AddChar(pprint, '[');
1500
+ PPrintText(doc, CDATA, 0, node->content);
1501
+ AddChar(pprint, ']');
1502
+ }
1503
+
1504
+ SetWrap( doc, 0 );
1505
+ AddChar( pprint, '>' );
1506
+ PCondFlushLine( doc, indent );
1507
+ }
1508
+
1509
+ static void PPrintPI( TidyDocImpl* doc, uint indent, Node *node )
1510
+ {
1511
+ TidyPrintImpl* pprint = &doc->pprint;
1512
+ tchar c;
1513
+ tmbstr s;
1514
+
1515
+ SetWrap( doc, indent );
1516
+ AddString( pprint, "<?" );
1517
+
1518
+ s = node->element;
1519
+
1520
+ while (s && *s)
1521
+ {
1522
+ c = (unsigned char)*s;
1523
+ if (c > 0x7F)
1524
+ s += TY_(GetUTF8)(s, &c);
1525
+ AddChar(pprint, c);
1526
+ ++s;
1527
+ }
1528
+
1529
+ /* set CDATA to pass < and > unescaped */
1530
+ PPrintText( doc, CDATA, indent, node );
1531
+
1532
+ if (cfgBool(doc, TidyXmlOut) ||
1533
+ cfgBool(doc, TidyXhtmlOut) || node->closed)
1534
+ AddChar( pprint, '?' );
1535
+
1536
+ AddChar( pprint, '>' );
1537
+ PCondFlushLine( doc, indent );
1538
+ }
1539
+
1540
+ static void PPrintXmlDecl( TidyDocImpl* doc, uint indent, Node *node )
1541
+ {
1542
+ AttVal* att;
1543
+ uint saveWrap;
1544
+ TidyPrintImpl* pprint = &doc->pprint;
1545
+ Bool ucAttrs;
1546
+ SetWrap( doc, indent );
1547
+ saveWrap = WrapOff( doc );
1548
+
1549
+ /* no case translation for XML declaration pseudo attributes */
1550
+ ucAttrs = cfgBool(doc, TidyUpperCaseAttrs);
1551
+ TY_(SetOptionBool)(doc, TidyUpperCaseAttrs, no);
1552
+
1553
+ AddString( pprint, "<?xml" );
1554
+
1555
+ /* Force order of XML declaration attributes */
1556
+ /* PPrintAttrs( doc, indent, node ); */
1557
+ if ( NULL != (att = TY_(AttrGetById)(node, TidyAttr_VERSION)) )
1558
+ PPrintAttribute( doc, indent, node, att );
1559
+ if ( NULL != (att = TY_(AttrGetById)(node, TidyAttr_ENCODING)) )
1560
+ PPrintAttribute( doc, indent, node, att );
1561
+ if ( NULL != (att = TY_(GetAttrByName)(node, "standalone")) )
1562
+ PPrintAttribute( doc, indent, node, att );
1563
+
1564
+ /* restore old config value */
1565
+ TY_(SetOptionBool)(doc, TidyUpperCaseAttrs, ucAttrs);
1566
+
1567
+ if ( node->end <= 0 || doc->lexer->lexbuf[node->end - 1] != '?' )
1568
+ AddChar( pprint, '?' );
1569
+ AddChar( pprint, '>' );
1570
+ WrapOn( doc, saveWrap );
1571
+ TY_(PFlushLine)( doc, indent );
1572
+ }
1573
+
1574
+ /* note ASP and JSTE share <% ... %> syntax */
1575
+ static void PPrintAsp( TidyDocImpl* doc, uint indent, Node *node )
1576
+ {
1577
+ TidyPrintImpl* pprint = &doc->pprint;
1578
+ Bool wrapAsp = cfgBool( doc, TidyWrapAsp );
1579
+ Bool wrapJste = cfgBool( doc, TidyWrapJste );
1580
+ uint saveWrap = WrapOffCond( doc, !wrapAsp || !wrapJste );
1581
+
1582
+ #if 0
1583
+ SetWrap( doc, indent );
1584
+ #endif
1585
+ AddString( pprint, "<%" );
1586
+ PPrintText( doc, (wrapAsp ? CDATA : COMMENT), indent, node );
1587
+ AddString( pprint, "%>" );
1588
+
1589
+ /* PCondFlushLine( doc, indent ); */
1590
+ WrapOn( doc, saveWrap );
1591
+ }
1592
+
1593
+ /* JSTE also supports <# ... #> syntax */
1594
+ static void PPrintJste( TidyDocImpl* doc, uint indent, Node *node )
1595
+ {
1596
+ TidyPrintImpl* pprint = &doc->pprint;
1597
+ Bool wrapAsp = cfgBool( doc, TidyWrapAsp );
1598
+ uint saveWrap = WrapOffCond( doc, !wrapAsp );
1599
+
1600
+ AddString( pprint, "<#" );
1601
+ PPrintText( doc, (cfgBool(doc, TidyWrapJste) ? CDATA : COMMENT),
1602
+ indent, node );
1603
+ AddString( pprint, "#>" );
1604
+
1605
+ /* PCondFlushLine( doc, indent ); */
1606
+ WrapOn( doc, saveWrap );
1607
+ }
1608
+
1609
+ /* PHP is based on XML processing instructions */
1610
+ static void PPrintPhp( TidyDocImpl* doc, uint indent, Node *node )
1611
+ {
1612
+ TidyPrintImpl* pprint = &doc->pprint;
1613
+ Bool wrapPhp = cfgBool( doc, TidyWrapPhp );
1614
+ uint saveWrap = WrapOffCond( doc, !wrapPhp );
1615
+ #if 0
1616
+ SetWrap( doc, indent );
1617
+ #endif
1618
+
1619
+ AddString( pprint, "<?" );
1620
+ PPrintText( doc, (wrapPhp ? CDATA : COMMENT),
1621
+ indent, node );
1622
+ AddString( pprint, "?>" );
1623
+
1624
+ /* PCondFlushLine( doc, indent ); */
1625
+ WrapOn( doc, saveWrap );
1626
+ }
1627
+
1628
+ static void PPrintCDATA( TidyDocImpl* doc, uint indent, Node *node )
1629
+ {
1630
+ uint saveWrap;
1631
+ TidyPrintImpl* pprint = &doc->pprint;
1632
+ Bool indentCData = cfgBool( doc, TidyIndentCdata );
1633
+ if ( !indentCData )
1634
+ indent = 0;
1635
+
1636
+ PCondFlushLine( doc, indent );
1637
+ saveWrap = WrapOff( doc ); /* disable wrapping */
1638
+
1639
+ AddString( pprint, "<![CDATA[" );
1640
+ PPrintText( doc, COMMENT, indent, node );
1641
+ AddString( pprint, "]]>" );
1642
+
1643
+ PCondFlushLine( doc, indent );
1644
+ WrapOn( doc, saveWrap ); /* restore wrapping */
1645
+ }
1646
+
1647
+ static void PPrintSection( TidyDocImpl* doc, uint indent, Node *node )
1648
+ {
1649
+ TidyPrintImpl* pprint = &doc->pprint;
1650
+ Bool wrapSect = cfgBool( doc, TidyWrapSection );
1651
+ uint saveWrap = WrapOffCond( doc, !wrapSect );
1652
+ #if 0
1653
+ SetWrap( doc, indent );
1654
+ #endif
1655
+
1656
+ AddString( pprint, "<![" );
1657
+ PPrintText( doc, (wrapSect ? CDATA : COMMENT),
1658
+ indent, node );
1659
+ AddString( pprint, "]>" );
1660
+
1661
+ /* PCondFlushLine( doc, indent ); */
1662
+ WrapOn( doc, saveWrap );
1663
+ }
1664
+
1665
+
1666
+ #if 0
1667
+ /*
1668
+ ** Print script and style elements. For XHTML, wrap the content as follows:
1669
+ **
1670
+ ** JavaScript:
1671
+ ** //<![CDATA[
1672
+ ** content
1673
+ ** //]]>
1674
+ ** VBScript:
1675
+ ** '<![CDATA[
1676
+ ** content
1677
+ ** ']]>
1678
+ ** CSS:
1679
+ ** / *<![CDATA[* / Extra spaces to keep compiler happy
1680
+ ** content
1681
+ ** / *]]>* /
1682
+ ** other:
1683
+ ** <![CDATA[
1684
+ ** content
1685
+ ** ]]>
1686
+ */
1687
+ #endif
1688
+
1689
+ static ctmbstr CDATA_START = "<![CDATA[";
1690
+ static ctmbstr CDATA_END = "]]>";
1691
+ static ctmbstr JS_COMMENT_START = "//";
1692
+ static ctmbstr JS_COMMENT_END = "";
1693
+ static ctmbstr VB_COMMENT_START = "\'";
1694
+ static ctmbstr VB_COMMENT_END = "";
1695
+ static ctmbstr CSS_COMMENT_START = "/*";
1696
+ static ctmbstr CSS_COMMENT_END = "*/";
1697
+ static ctmbstr DEFAULT_COMMENT_START = "";
1698
+ static ctmbstr DEFAULT_COMMENT_END = "";
1699
+
1700
+ static Bool InsideHead( TidyDocImpl* doc, Node *node )
1701
+ {
1702
+ if ( nodeIsHEAD(node) )
1703
+ return yes;
1704
+
1705
+ if ( node->parent != NULL )
1706
+ return InsideHead( doc, node->parent );
1707
+
1708
+ return no;
1709
+ }
1710
+
1711
+ /* Is text node and already ends w/ a newline?
1712
+
1713
+ Used to pretty print CDATA/PRE text content.
1714
+ If it already ends on a newline, it is not
1715
+ necessary to print another before printing end tag.
1716
+ */
1717
+ static int TextEndsWithNewline(Lexer *lexer, Node *node, uint mode )
1718
+ {
1719
+ if ( (mode & (CDATA|COMMENT)) && TY_(nodeIsText)(node) && node->end > node->start )
1720
+ {
1721
+ uint ch, ix = node->end - 1;
1722
+ /* Skip non-newline whitespace. */
1723
+ while ( ix >= node->start && (ch = (lexer->lexbuf[ix] & 0xff))
1724
+ && ( ch == ' ' || ch == '\t' || ch == '\r' ) )
1725
+ --ix;
1726
+
1727
+ if ( lexer->lexbuf[ ix ] == '\n' )
1728
+ return node->end - ix - 1; /* #543262 tidy eats all memory */
1729
+ }
1730
+ return -1;
1731
+ }
1732
+
1733
+ static int TextStartsWithWhitespace( Lexer *lexer, Node *node, uint start, uint mode )
1734
+ {
1735
+ assert( node != NULL );
1736
+ if ( (mode & (CDATA|COMMENT)) && TY_(nodeIsText)(node) && node->end > node->start && start >= node->start )
1737
+ {
1738
+ uint ch, ix = start;
1739
+ /* Skip whitespace. */
1740
+ while ( ix < node->end && (ch = (lexer->lexbuf[ix] & 0xff))
1741
+ && ( ch==' ' || ch=='\t' || ch=='\r' ) )
1742
+ ++ix;
1743
+
1744
+ if ( ix > start )
1745
+ return ix - start;
1746
+ }
1747
+ return -1;
1748
+ }
1749
+
1750
+ static Bool HasCDATA( Lexer* lexer, Node* node )
1751
+ {
1752
+ /* Scan forward through the textarray. Since the characters we're
1753
+ ** looking for are < 0x7f, we don't have to do any UTF-8 decoding.
1754
+ */
1755
+ ctmbstr start = lexer->lexbuf + node->start;
1756
+ int len = node->end - node->start + 1;
1757
+
1758
+ if ( node->type != TextNode )
1759
+ return no;
1760
+
1761
+ return ( NULL != TY_(tmbsubstrn)( start, len, CDATA_START ));
1762
+ }
1763
+
1764
+
1765
+ static
1766
+ void PPrintScriptStyle( TidyDocImpl* doc, uint mode, uint indent, Node *node )
1767
+ {
1768
+ TidyPrintImpl* pprint = &doc->pprint;
1769
+ Node* content;
1770
+ ctmbstr commentStart = DEFAULT_COMMENT_START;
1771
+ ctmbstr commentEnd = DEFAULT_COMMENT_END;
1772
+ Bool hasCData = no;
1773
+ int contentIndent = -1;
1774
+ Bool xhtmlOut = cfgBool( doc, TidyXhtmlOut );
1775
+
1776
+ if ( InsideHead(doc, node) )
1777
+ TY_(PFlushLine)( doc, indent );
1778
+
1779
+ PPrintTag( doc, mode, indent, node );
1780
+
1781
+ /* use zero indent here, see http://tidy.sf.net/bug/729972 */
1782
+ TY_(PFlushLine)(doc, 0);
1783
+
1784
+ if ( xhtmlOut && node->content != NULL )
1785
+ {
1786
+ AttVal* type = attrGetTYPE(node);
1787
+
1788
+ if (AttrValueIs(type, "text/javascript"))
1789
+ {
1790
+ commentStart = JS_COMMENT_START;
1791
+ commentEnd = JS_COMMENT_END;
1792
+ }
1793
+ else if (AttrValueIs(type, "text/css"))
1794
+ {
1795
+ commentStart = CSS_COMMENT_START;
1796
+ commentEnd = CSS_COMMENT_END;
1797
+ }
1798
+ else if (AttrValueIs(type, "text/vbscript"))
1799
+ {
1800
+ commentStart = VB_COMMENT_START;
1801
+ commentEnd = VB_COMMENT_END;
1802
+ }
1803
+
1804
+ hasCData = HasCDATA(doc->lexer, node->content);
1805
+
1806
+ if (!hasCData)
1807
+ {
1808
+ uint saveWrap = WrapOff( doc );
1809
+
1810
+ AddString( pprint, commentStart );
1811
+ AddString( pprint, CDATA_START );
1812
+ AddString( pprint, commentEnd );
1813
+ PCondFlushLine( doc, indent );
1814
+
1815
+ WrapOn( doc, saveWrap );
1816
+ }
1817
+ }
1818
+
1819
+ for ( content = node->content;
1820
+ content != NULL;
1821
+ content = content->next )
1822
+ {
1823
+ /*
1824
+ This is a bit odd, with the current code there can only
1825
+ be one child and the only caller of this function defines
1826
+ all these modes already...
1827
+ */
1828
+ TY_(PPrintTree)( doc, (mode | PREFORMATTED | NOWRAP | CDATA),
1829
+ indent, content );
1830
+
1831
+ if ( content == node->last )
1832
+ contentIndent = TextEndsWithNewline( doc->lexer, content, CDATA );
1833
+ }
1834
+
1835
+ if ( contentIndent < 0 )
1836
+ {
1837
+ PCondFlushLine( doc, indent );
1838
+ contentIndent = 0;
1839
+ }
1840
+
1841
+ if ( xhtmlOut && node->content != NULL )
1842
+ {
1843
+ if ( ! hasCData )
1844
+ {
1845
+ uint saveWrap = WrapOff( doc );
1846
+
1847
+ AddString( pprint, commentStart );
1848
+ AddString( pprint, CDATA_END );
1849
+ AddString( pprint, commentEnd );
1850
+
1851
+ WrapOn( doc, saveWrap );
1852
+ PCondFlushLine( doc, indent );
1853
+ }
1854
+ }
1855
+
1856
+ if ( node->content && pprint->indent[ 0 ].spaces != (int)indent )
1857
+ {
1858
+ pprint->indent[ 0 ].spaces = indent;
1859
+ }
1860
+ PPrintEndTag( doc, mode, indent, node );
1861
+ if ( cfgAutoBool(doc, TidyIndentContent) == TidyNoState
1862
+ && node->next != NULL &&
1863
+ !( TY_(nodeHasCM)(node, CM_INLINE) || TY_(nodeIsText)(node) ) )
1864
+ TY_(PFlushLine)( doc, indent );
1865
+ }
1866
+
1867
+
1868
+
1869
+ static Bool ShouldIndent( TidyDocImpl* doc, Node *node )
1870
+ {
1871
+ TidyTriState indentContent = cfgAutoBool( doc, TidyIndentContent );
1872
+ if ( indentContent == TidyNoState )
1873
+ return no;
1874
+
1875
+ if ( nodeIsTEXTAREA(node) )
1876
+ return no;
1877
+
1878
+ if ( indentContent == TidyAutoState )
1879
+ {
1880
+ if ( node->content && TY_(nodeHasCM)(node, CM_NO_INDENT) )
1881
+ {
1882
+ for ( node = node->content; node; node = node->next )
1883
+ if ( TY_(nodeHasCM)(node, CM_BLOCK) )
1884
+ return yes;
1885
+ return no;
1886
+ }
1887
+
1888
+ if ( TY_(nodeHasCM)(node, CM_HEADING) )
1889
+ return no;
1890
+
1891
+ if ( nodeIsHTML(node) )
1892
+ return no;
1893
+
1894
+ if ( nodeIsP(node) )
1895
+ return no;
1896
+
1897
+ if ( nodeIsTITLE(node) )
1898
+ return no;
1899
+
1900
+ /* http://tidy.sf.net/issue/1610888
1901
+ Indenting <div><img /></div> produces spurious lines with IE 6.x */
1902
+ if ( nodeIsDIV(node) && node->last && nodeIsIMG(node->last) )
1903
+ return no;
1904
+ }
1905
+
1906
+ if ( TY_(nodeHasCM)(node, CM_FIELD | CM_OBJECT) )
1907
+ return yes;
1908
+
1909
+ if ( nodeIsMAP(node) )
1910
+ return yes;
1911
+
1912
+ return ( !TY_(nodeHasCM)( node, CM_INLINE ) && node->content );
1913
+ }
1914
+
1915
+ /*
1916
+ Feature request #434940 - fix by Dave Raggett/Ignacio Vazquez-Abrams 21 Jun 01
1917
+ print just the content of the body element.
1918
+ useful when you want to reuse material from
1919
+ other documents.
1920
+
1921
+ -- Sebastiano Vigna <vigna@dsi.unimi.it>
1922
+ */
1923
+ void TY_(PrintBody)( TidyDocImpl* doc )
1924
+ {
1925
+ Node *node = TY_(FindBody)( doc );
1926
+
1927
+ if ( node )
1928
+ {
1929
+ for ( node = node->content; node != NULL; node = node->next )
1930
+ TY_(PPrintTree)( doc, NORMAL, 0, node );
1931
+ }
1932
+ }
1933
+
1934
+ void TY_(PPrintTree)( TidyDocImpl* doc, uint mode, uint indent, Node *node )
1935
+ {
1936
+ Node *content, *last;
1937
+ uint spaces = cfg( doc, TidyIndentSpaces );
1938
+ Bool xhtml = cfgBool( doc, TidyXhtmlOut );
1939
+
1940
+ if ( node == NULL )
1941
+ return;
1942
+
1943
+ if (node->type == TextNode)
1944
+ {
1945
+ PPrintText( doc, mode, indent, node );
1946
+ }
1947
+ else if ( node->type == CommentTag )
1948
+ {
1949
+ PPrintComment( doc, indent, node );
1950
+ }
1951
+ else if ( node->type == RootNode )
1952
+ {
1953
+ for ( content = node->content; content; content = content->next )
1954
+ TY_(PPrintTree)( doc, mode, indent, content );
1955
+ }
1956
+ else if ( node->type == DocTypeTag )
1957
+ PPrintDocType( doc, indent, node );
1958
+ else if ( node->type == ProcInsTag)
1959
+ PPrintPI( doc, indent, node );
1960
+ else if ( node->type == XmlDecl)
1961
+ PPrintXmlDecl( doc, indent, node );
1962
+ else if ( node->type == CDATATag)
1963
+ PPrintCDATA( doc, indent, node );
1964
+ else if ( node->type == SectionTag)
1965
+ PPrintSection( doc, indent, node );
1966
+ else if ( node->type == AspTag)
1967
+ PPrintAsp( doc, indent, node );
1968
+ else if ( node->type == JsteTag)
1969
+ PPrintJste( doc, indent, node );
1970
+ else if ( node->type == PhpTag)
1971
+ PPrintPhp( doc, indent, node );
1972
+ else if ( TY_(nodeCMIsEmpty)(node) ||
1973
+ (node->type == StartEndTag && !xhtml) )
1974
+ {
1975
+ if ( ! TY_(nodeHasCM)(node, CM_INLINE) )
1976
+ PCondFlushLine( doc, indent );
1977
+
1978
+ if ( nodeIsBR(node) && node->prev &&
1979
+ !(nodeIsBR(node->prev) || (mode & PREFORMATTED)) &&
1980
+ cfgBool(doc, TidyBreakBeforeBR) )
1981
+ TY_(PFlushLine)( doc, indent );
1982
+
1983
+ if ( nodeIsHR(node) )
1984
+ {
1985
+ /* insert extra newline for classic formatting */
1986
+ Bool classic = cfgBool( doc, TidyVertSpace );
1987
+ if (classic && node->parent && node->parent->content != node)
1988
+ {
1989
+ TY_(PFlushLine)( doc, indent );
1990
+ }
1991
+ }
1992
+
1993
+ PPrintTag( doc, mode, indent, node );
1994
+
1995
+ if (node->next)
1996
+ {
1997
+ if (nodeIsPARAM(node) || nodeIsAREA(node))
1998
+ PCondFlushLine(doc, indent);
1999
+ else if ((nodeIsBR(node) && !(mode & PREFORMATTED))
2000
+ || nodeIsHR(node))
2001
+ TY_(PFlushLine)(doc, indent);
2002
+ }
2003
+ }
2004
+ else /* some kind of container element */
2005
+ {
2006
+ if ( node->type == StartEndTag )
2007
+ node->type = StartTag;
2008
+
2009
+ if ( node->tag &&
2010
+ (node->tag->parser == TY_(ParsePre) || nodeIsTEXTAREA(node)) )
2011
+ {
2012
+ Bool classic = cfgBool( doc, TidyVertSpace );
2013
+ uint indprev = indent;
2014
+ PCondFlushLine( doc, indent );
2015
+
2016
+ PCondFlushLine( doc, indent );
2017
+
2018
+ /* insert extra newline for classic formatting */
2019
+ if (classic && node->parent && node->parent->content != node)
2020
+ {
2021
+ TY_(PFlushLine)( doc, indent );
2022
+ }
2023
+ PPrintTag( doc, mode, indent, node );
2024
+
2025
+ indent = 0;
2026
+ TY_(PFlushLine)( doc, indent );
2027
+
2028
+ for ( content = node->content; content; content = content->next )
2029
+ {
2030
+ TY_(PPrintTree)( doc, (mode | PREFORMATTED | NOWRAP),
2031
+ indent, content );
2032
+ }
2033
+ PCondFlushLine( doc, indent );
2034
+ indent = indprev;
2035
+ PPrintEndTag( doc, mode, indent, node );
2036
+
2037
+ if ( cfgAutoBool(doc, TidyIndentContent) == TidyNoState
2038
+ && node->next != NULL )
2039
+ TY_(PFlushLine)( doc, indent );
2040
+ }
2041
+ else if ( nodeIsSTYLE(node) || nodeIsSCRIPT(node) )
2042
+ {
2043
+ PPrintScriptStyle( doc, (mode | PREFORMATTED | NOWRAP | CDATA),
2044
+ indent, node );
2045
+ }
2046
+ else if ( TY_(nodeCMIsInline)(node) )
2047
+ {
2048
+ if ( cfgBool(doc, TidyMakeClean) )
2049
+ {
2050
+ /* replace <nobr>...</nobr> by &nbsp; or &#160; etc. */
2051
+ if ( nodeIsNOBR(node) )
2052
+ {
2053
+ for ( content = node->content;
2054
+ content != NULL;
2055
+ content = content->next)
2056
+ TY_(PPrintTree)( doc, mode|NOWRAP, indent, content );
2057
+ return;
2058
+ }
2059
+ }
2060
+
2061
+ /* otherwise a normal inline element */
2062
+ PPrintTag( doc, mode, indent, node );
2063
+
2064
+ /* indent content for SELECT, TEXTAREA, MAP, OBJECT and APPLET */
2065
+ if ( ShouldIndent(doc, node) )
2066
+ {
2067
+ indent += spaces;
2068
+ PCondFlushLine( doc, indent );
2069
+
2070
+ for ( content = node->content;
2071
+ content != NULL;
2072
+ content = content->next )
2073
+ TY_(PPrintTree)( doc, mode, indent, content );
2074
+
2075
+ indent -= spaces;
2076
+ PCondFlushLine( doc, indent );
2077
+ /* PCondFlushLine( doc, indent ); */
2078
+ }
2079
+ else
2080
+ {
2081
+ for ( content = node->content;
2082
+ content != NULL;
2083
+ content = content->next )
2084
+ TY_(PPrintTree)( doc, mode, indent, content );
2085
+ }
2086
+ PPrintEndTag( doc, mode, indent, node );
2087
+ }
2088
+ else /* other tags */
2089
+ {
2090
+ Bool indcont = ( cfgAutoBool(doc, TidyIndentContent) != TidyNoState );
2091
+ Bool indsmart = ( cfgAutoBool(doc, TidyIndentContent) == TidyAutoState );
2092
+ Bool hideend = cfgBool( doc, TidyHideEndTags );
2093
+ Bool classic = cfgBool( doc, TidyVertSpace );
2094
+ uint contentIndent = indent;
2095
+
2096
+ /* insert extra newline for classic formatting */
2097
+ if (classic && node->parent && node->parent->content != node && !nodeIsHTML(node))
2098
+ {
2099
+ TY_(PFlushLine)( doc, indent );
2100
+ }
2101
+
2102
+ if ( ShouldIndent(doc, node) )
2103
+ contentIndent += spaces;
2104
+
2105
+ PCondFlushLine( doc, indent );
2106
+ if ( indsmart && node->prev != NULL )
2107
+ TY_(PFlushLine)( doc, indent );
2108
+
2109
+ /* do not omit elements with attributes */
2110
+ if ( !hideend || !TY_(nodeHasCM)(node, CM_OMITST) ||
2111
+ node->attributes != NULL )
2112
+ {
2113
+ PPrintTag( doc, mode, indent, node );
2114
+
2115
+ if ( ShouldIndent(doc, node) )
2116
+ {
2117
+ /* fix for bug 530791, don't wrap after */
2118
+ /* <li> if first child is text node */
2119
+ if (!(nodeIsLI(node) && TY_(nodeIsText)(node->content)))
2120
+ PCondFlushLine( doc, contentIndent );
2121
+ }
2122
+ else if ( TY_(nodeHasCM)(node, CM_HTML) || nodeIsNOFRAMES(node) ||
2123
+ (TY_(nodeHasCM)(node, CM_HEAD) && !nodeIsTITLE(node)) )
2124
+ TY_(PFlushLine)( doc, contentIndent );
2125
+ }
2126
+
2127
+ last = NULL;
2128
+ for ( content = node->content; content; content = content->next )
2129
+ {
2130
+ /* kludge for naked text before block level tag */
2131
+ if ( last && !indcont && TY_(nodeIsText)(last) &&
2132
+ content->tag && !TY_(nodeHasCM)(content, CM_INLINE) )
2133
+ {
2134
+ /* TY_(PFlushLine)(fout, indent); */
2135
+ TY_(PFlushLine)( doc, contentIndent );
2136
+ }
2137
+
2138
+ TY_(PPrintTree)( doc, mode, contentIndent, content );
2139
+ last = content;
2140
+ }
2141
+
2142
+ /* don't flush line for td and th */
2143
+ if ( ShouldIndent(doc, node) ||
2144
+ ( !hideend &&
2145
+ ( TY_(nodeHasCM)(node, CM_HTML) ||
2146
+ nodeIsNOFRAMES(node) ||
2147
+ (TY_(nodeHasCM)(node, CM_HEAD) && !nodeIsTITLE(node))
2148
+ )
2149
+ )
2150
+ )
2151
+ {
2152
+ PCondFlushLine( doc, indent );
2153
+ if ( !hideend || !TY_(nodeHasCM)(node, CM_OPT) )
2154
+ {
2155
+ PPrintEndTag( doc, mode, indent, node );
2156
+ /* TY_(PFlushLine)( doc, indent ); */
2157
+ }
2158
+ }
2159
+ else
2160
+ {
2161
+ if ( !hideend || !TY_(nodeHasCM)(node, CM_OPT) )
2162
+ {
2163
+ /* newline before endtag for classic formatting */
2164
+ if ( classic && !HasMixedContent(node) )
2165
+ TY_(PFlushLine)( doc, indent );
2166
+ PPrintEndTag( doc, mode, indent, node );
2167
+ }
2168
+ }
2169
+
2170
+ if (!indcont && !hideend && !nodeIsHTML(node) && !classic)
2171
+ TY_(PFlushLine)( doc, indent );
2172
+ else if (classic && node->next != NULL && TY_(nodeHasCM)(node, CM_LIST|CM_DEFLIST|CM_TABLE|CM_BLOCK/*|CM_HEADING*/))
2173
+ TY_(PFlushLine)( doc, indent );
2174
+ }
2175
+ }
2176
+ }
2177
+
2178
+ void TY_(PPrintXMLTree)( TidyDocImpl* doc, uint mode, uint indent, Node *node )
2179
+ {
2180
+ Bool xhtmlOut = cfgBool( doc, TidyXhtmlOut );
2181
+ if (node == NULL)
2182
+ return;
2183
+
2184
+ if ( node->type == TextNode)
2185
+ {
2186
+ PPrintText( doc, mode, indent, node );
2187
+ }
2188
+ else if ( node->type == CommentTag )
2189
+ {
2190
+ PCondFlushLine( doc, indent );
2191
+ PPrintComment( doc, indent, node);
2192
+ /* PCondFlushLine( doc, 0 ); */
2193
+ }
2194
+ else if ( node->type == RootNode )
2195
+ {
2196
+ Node *content;
2197
+ for ( content = node->content;
2198
+ content != NULL;
2199
+ content = content->next )
2200
+ TY_(PPrintXMLTree)( doc, mode, indent, content );
2201
+ }
2202
+ else if ( node->type == DocTypeTag )
2203
+ PPrintDocType( doc, indent, node );
2204
+ else if ( node->type == ProcInsTag )
2205
+ PPrintPI( doc, indent, node );
2206
+ else if ( node->type == XmlDecl )
2207
+ PPrintXmlDecl( doc, indent, node );
2208
+ else if ( node->type == CDATATag )
2209
+ PPrintCDATA( doc, indent, node );
2210
+ else if ( node->type == SectionTag )
2211
+ PPrintSection( doc, indent, node );
2212
+ else if ( node->type == AspTag )
2213
+ PPrintAsp( doc, indent, node );
2214
+ else if ( node->type == JsteTag)
2215
+ PPrintJste( doc, indent, node );
2216
+ else if ( node->type == PhpTag)
2217
+ PPrintPhp( doc, indent, node );
2218
+ else if ( TY_(nodeHasCM)(node, CM_EMPTY) ||
2219
+ (node->type == StartEndTag && !xhtmlOut) )
2220
+ {
2221
+ PCondFlushLine( doc, indent );
2222
+ PPrintTag( doc, mode, indent, node );
2223
+ /* TY_(PFlushLine)( doc, indent ); */
2224
+ }
2225
+ else /* some kind of container element */
2226
+ {
2227
+ uint spaces = cfg( doc, TidyIndentSpaces );
2228
+ Node *content;
2229
+ Bool mixed = no;
2230
+ uint cindent;
2231
+
2232
+ for ( content = node->content; content; content = content->next )
2233
+ {
2234
+ if ( TY_(nodeIsText)(content) )
2235
+ {
2236
+ mixed = yes;
2237
+ break;
2238
+ }
2239
+ }
2240
+
2241
+ PCondFlushLine( doc, indent );
2242
+
2243
+ if ( TY_(XMLPreserveWhiteSpace)(doc, node) )
2244
+ {
2245
+ indent = 0;
2246
+ mixed = no;
2247
+ cindent = 0;
2248
+ }
2249
+ else if (mixed)
2250
+ cindent = indent;
2251
+ else
2252
+ cindent = indent + spaces;
2253
+
2254
+ PPrintTag( doc, mode, indent, node );
2255
+ if ( !mixed && node->content )
2256
+ TY_(PFlushLine)( doc, cindent );
2257
+
2258
+ for ( content = node->content; content; content = content->next )
2259
+ TY_(PPrintXMLTree)( doc, mode, cindent, content );
2260
+
2261
+ if ( !mixed && node->content )
2262
+ PCondFlushLine( doc, indent );
2263
+
2264
+ PPrintEndTag( doc, mode, indent, node );
2265
+ /* PCondFlushLine( doc, indent ); */
2266
+ }
2267
+ }
2268
+
2269
+ /*
2270
+ * local variables:
2271
+ * mode: c
2272
+ * indent-tabs-mode: nil
2273
+ * c-basic-offset: 4
2274
+ * eval: (c-set-offset 'substatement-open 0)
2275
+ * end:
2276
+ */