d7482997 |
1 | /* |
2 | * xhtml backend for Halibut |
3 | * (initial implementation by James Aylett) |
4 | * |
5 | * Still to do: |
6 | * |
7 | * +++ doesn't handle non-breaking hyphens. Not sure how to yet. |
8 | * +++ entity names (from a file -- ideally supply normal SGML files) |
9 | * +++ configuration directive to file split where the current layout |
10 | * code wouldn't. Needs changes to _ponder_layout() and _do_paras(), |
11 | * perhaps others. |
12 | * |
13 | * Limitations: |
14 | * |
15 | * +++ biblio/index references target the nearest section marker, rather |
16 | * than having a dedicated target themselves. In large bibliographies |
17 | * this will cause problems. (The solution is to fake up a response |
18 | * from xhtml_find_section(), probably linking it into the sections |
19 | * chain just in case we need it again, and to make freeing it up |
20 | * easier.) docsrc.pl used to work as we do, however, and SGT agrees that |
21 | * this is acceptable for now. |
22 | * +++ can't cope with leaf-level == 0. It's all to do with the |
23 | * top-level file not being normal, probably not even having a valid |
24 | * section level, and stuff like that. I question whether this is an |
25 | * issue, frankly; small manuals that fit on one page should probably |
26 | * not be written in halibut at all. |
27 | */ |
28 | |
29 | #include <stdio.h> |
30 | #include <stdlib.h> |
677e18a2 |
31 | #include <string.h> |
d7482997 |
32 | #include <assert.h> |
33 | #include "halibut.h" |
34 | |
50d6b4bd |
35 | /* |
36 | * FILENAME_TEMPLATE (overridable in config of course) allows you |
37 | * to choose the general form for your HTML file names. It is |
38 | * slightly printf-styled (% followed by a single character is a |
39 | * formatting directive, %% is a literal %). Formatting directives |
40 | * are: |
41 | * |
ba9c1487 |
42 | * - %n is the section type-plus-number, minus whitespace (`Chapter1.2'). |
50d6b4bd |
43 | * - %b is the section number on its own (`1.2'). |
44 | * - %k is the section's _internal_ keyword. |
45 | * - %N is the section's visible title in the output, again minus |
46 | * whitespace. |
47 | * |
48 | * %n, %b and %k will all default to %N if the section is |
49 | * unnumbered (`Bibliography' is often a good example). |
6d6d850c |
50 | * |
51 | * FRAGMENT_TEMPLATE is the same, but defines the <a name="foo"> |
52 | * markers used to cross-reference to particular subsections of a |
53 | * file. |
50d6b4bd |
54 | */ |
55 | |
56 | #define FILENAME_SINGLE "Manual.html" |
57 | #define FILENAME_CONTENTS "Contents.html" |
58 | #define FILENAME_INDEX "IndexPage.html" |
59 | #define FILENAME_TEMPLATE "%n.html" |
6d6d850c |
60 | #define FRAGMENT_TEMPLATE "%b" |
50d6b4bd |
61 | |
d7482997 |
62 | struct xhtmlsection_Struct { |
63 | struct xhtmlsection_Struct *next; /* next sibling (NULL if split across files) */ |
64 | struct xhtmlsection_Struct *child; /* NULL if split across files */ |
65 | struct xhtmlsection_Struct *parent; /* NULL if split across files */ |
66 | struct xhtmlsection_Struct *chain; /* single structure independent of weird trees */ |
67 | paragraph *para; |
68 | struct xhtmlfile_Struct *file; /* which file is this a part of? */ |
69 | char *fragment; /* fragment id within the file */ |
70 | int level; |
71 | }; |
72 | |
73 | struct xhtmlfile_Struct { |
74 | struct xhtmlfile_Struct *next; |
75 | struct xhtmlfile_Struct *child; |
76 | struct xhtmlfile_Struct *parent; |
77 | char *filename; |
78 | struct xhtmlsection_Struct *sections; /* sections within this file (only one for non-leaf) */ |
79 | int is_leaf; /* is this file a leaf file, ie does it not have any children? */ |
80 | }; |
81 | |
82 | typedef struct xhtmlsection_Struct xhtmlsection; |
83 | typedef struct xhtmlfile_Struct xhtmlfile; |
84 | typedef struct xhtmlindex_Struct xhtmlindex; |
85 | |
86 | struct xhtmlindex_Struct { |
87 | int nsection; |
88 | int size; |
89 | xhtmlsection **sections; |
90 | }; |
91 | |
92 | typedef struct { |
5d9cc07b |
93 | int just_numbers; |
94 | wchar_t *number_suffix; |
95 | } xhtmlheadfmt; |
96 | |
97 | typedef struct { |
d7482997 |
98 | int contents_depth[6]; |
99 | int leaf_contains_contents; |
100 | int leaf_level; |
101 | int leaf_smallest_contents; |
102 | int include_version_id; |
103 | wchar_t *author, *description; |
104 | wchar_t *head_end, *body, *body_start, *body_end, *address_start, *address_end, *nav_attrs; |
105 | int suppress_address; |
5d9cc07b |
106 | xhtmlheadfmt fchapter, *fsect; |
107 | int nfsect; |
50d6b4bd |
108 | char *contents_filename, *index_filename; |
6d6d850c |
109 | char *single_filename, *template_filename, *template_fragment; |
d7482997 |
110 | } xhtmlconfig; |
111 | |
112 | /*static void xhtml_level(paragraph *, int); |
113 | static void xhtml_level_0(paragraph *); |
114 | static void xhtml_docontents(FILE *, paragraph *, int); |
115 | static void xhtml_dosections(FILE *, paragraph *, int); |
116 | static void xhtml_dobody(FILE *, paragraph *, int);*/ |
117 | |
118 | static void xhtml_doheader(FILE *, word *); |
119 | static void xhtml_dofooter(FILE *); |
120 | static void xhtml_versionid(FILE *, word *, int); |
121 | |
122 | static void xhtml_utostr(wchar_t *, char **); |
123 | static int xhtml_para_level(paragraph *); |
124 | static int xhtml_reservedchar(int); |
125 | |
4b3c5afb |
126 | static int xhtml_convert(wchar_t *, int, char **, int); |
ce9921d6 |
127 | static void xhtml_rdaddwc(rdstringc *, word *, word *, int); |
128 | static void xhtml_para(FILE *, word *, int); |
d7482997 |
129 | static void xhtml_codepara(FILE *, word *); |
ce9921d6 |
130 | static void xhtml_heading(FILE *, paragraph *, int); |
d7482997 |
131 | |
132 | /* File-global variables are much easier than passing these things |
133 | * all over the place. Evil, but easier. We can replace this with a single |
134 | * structure at some point. |
135 | */ |
136 | static xhtmlconfig conf; |
137 | static keywordlist *keywords; |
138 | static indexdata *idx; |
139 | static xhtmlfile *topfile; |
140 | static xhtmlsection *topsection; |
141 | static paragraph *sourceparas; |
142 | static xhtmlfile *lastfile; |
143 | static xhtmlfile *xhtml_last_file = NULL; |
c8c7926b |
144 | static int last_level=-1, start_level; |
d7482997 |
145 | static xhtmlsection *currentsection; |
146 | |
147 | static xhtmlconfig xhtml_configure(paragraph *source) |
148 | { |
149 | xhtmlconfig ret; |
150 | |
151 | /* |
152 | * Defaults. |
153 | */ |
154 | ret.contents_depth[0] = 2; |
155 | ret.contents_depth[1] = 3; |
156 | ret.contents_depth[2] = 4; |
157 | ret.contents_depth[3] = 5; |
158 | ret.contents_depth[4] = 6; |
159 | ret.contents_depth[5] = 7; |
160 | ret.leaf_level = 2; |
161 | ret.leaf_smallest_contents = 4; |
162 | ret.leaf_contains_contents = FALSE; |
163 | ret.include_version_id = TRUE; |
164 | ret.author = NULL; |
165 | ret.description = NULL; |
166 | ret.head_end = NULL; |
167 | ret.body = NULL; |
168 | ret.body_start = NULL; |
169 | ret.body_end = NULL; |
170 | ret.address_start = NULL; |
171 | ret.address_end = NULL; |
172 | ret.nav_attrs = NULL; |
173 | ret.suppress_address = FALSE; |
174 | |
5d9cc07b |
175 | ret.fchapter.just_numbers = FALSE; |
e5e6bf9d |
176 | ret.fchapter.number_suffix = L": "; |
5d9cc07b |
177 | ret.nfsect = 2; |
178 | ret.fsect = mknewa(xhtmlheadfmt, ret.nfsect); |
179 | ret.fsect[0].just_numbers = FALSE; |
e5e6bf9d |
180 | ret.fsect[0].number_suffix = L": "; |
5d9cc07b |
181 | ret.fsect[1].just_numbers = TRUE; |
e5e6bf9d |
182 | ret.fsect[1].number_suffix = L" "; |
50d6b4bd |
183 | ret.contents_filename = strdup(FILENAME_CONTENTS); |
184 | ret.single_filename = strdup(FILENAME_SINGLE); |
185 | ret.index_filename = strdup(FILENAME_INDEX); |
186 | ret.template_filename = strdup(FILENAME_TEMPLATE); |
6d6d850c |
187 | ret.template_fragment = strdup(FRAGMENT_TEMPLATE); |
5d9cc07b |
188 | |
d7482997 |
189 | for (; source; source = source->next) |
190 | { |
191 | if (source->type == para_Config) |
192 | { |
50d6b4bd |
193 | if (!ustricmp(source->keyword, L"xhtml-contents-filename")) { |
194 | sfree(ret.contents_filename); |
e4ea58f8 |
195 | ret.contents_filename = dupstr(adv(source->origkeyword)); |
50d6b4bd |
196 | } else if (!ustricmp(source->keyword, L"xhtml-single-filename")) { |
197 | sfree(ret.single_filename); |
e4ea58f8 |
198 | ret.single_filename = dupstr(adv(source->origkeyword)); |
50d6b4bd |
199 | } else if (!ustricmp(source->keyword, L"xhtml-index-filename")) { |
200 | sfree(ret.index_filename); |
e4ea58f8 |
201 | ret.index_filename = dupstr(adv(source->origkeyword)); |
50d6b4bd |
202 | } else if (!ustricmp(source->keyword, L"xhtml-template-filename")) { |
203 | sfree(ret.template_filename); |
e4ea58f8 |
204 | ret.template_filename = dupstr(adv(source->origkeyword)); |
6d6d850c |
205 | } else if (!ustricmp(source->keyword, L"xhtml-template-fragment")) { |
206 | sfree(ret.template_fragment); |
e4ea58f8 |
207 | ret.template_fragment = utoa_dup(uadv(source->keyword), CS_ASCII); |
50d6b4bd |
208 | } else if (!ustricmp(source->keyword, L"xhtml-contents-depth-0")) { |
d7482997 |
209 | ret.contents_depth[0] = utoi(uadv(source->keyword)); |
210 | } else if (!ustricmp(source->keyword, L"xhtml-contents-depth-1")) { |
211 | ret.contents_depth[1] = utoi(uadv(source->keyword)); |
212 | } else if (!ustricmp(source->keyword, L"xhtml-contents-depth-2")) { |
213 | ret.contents_depth[2] = utoi(uadv(source->keyword)); |
214 | } else if (!ustricmp(source->keyword, L"xhtml-contents-depth-3")) { |
215 | ret.contents_depth[3] = utoi(uadv(source->keyword)); |
216 | } else if (!ustricmp(source->keyword, L"xhtml-contents-depth-4")) { |
217 | ret.contents_depth[4] = utoi(uadv(source->keyword)); |
218 | } else if (!ustricmp(source->keyword, L"xhtml-contents-depth-5")) { |
219 | ret.contents_depth[5] = utoi(uadv(source->keyword)); |
220 | } else if (!ustricmp(source->keyword, L"xhtml-leaf-level")) { |
221 | ret.leaf_level = utoi(uadv(source->keyword)); |
d7482997 |
222 | } else if (!ustricmp(source->keyword, L"xhtml-leaf-smallest-contents")) { |
223 | ret.leaf_smallest_contents = utoi(uadv(source->keyword)); |
224 | } else if (!ustricmp(source->keyword, L"xhtml-versionid")) { |
225 | ret.include_version_id = utob(uadv(source->keyword)); |
226 | } else if (!ustricmp(source->keyword, L"xhtml-leaf-contains-contents")) { |
227 | ret.leaf_contains_contents = utob(uadv(source->keyword)); |
228 | } else if (!ustricmp(source->keyword, L"xhtml-suppress-address")) { |
229 | ret.suppress_address = utob(uadv(source->keyword)); |
230 | } else if (!ustricmp(source->keyword, L"xhtml-author")) { |
231 | ret.author = uadv(source->keyword); |
232 | } else if (!ustricmp(source->keyword, L"xhtml-description")) { |
233 | ret.description = uadv(source->keyword); |
234 | } else if (!ustricmp(source->keyword, L"xhtml-head-end")) { |
235 | ret.head_end = uadv(source->keyword); |
236 | } else if (!ustricmp(source->keyword, L"xhtml-body-start")) { |
237 | ret.body_start = uadv(source->keyword); |
238 | } else if (!ustricmp(source->keyword, L"xhtml-body-tag")) { |
239 | ret.body = uadv(source->keyword); |
240 | } else if (!ustricmp(source->keyword, L"xhtml-body-end")) { |
241 | ret.body_end = uadv(source->keyword); |
242 | } else if (!ustricmp(source->keyword, L"xhtml-address-start")) { |
243 | ret.address_start = uadv(source->keyword); |
244 | } else if (!ustricmp(source->keyword, L"xhtml-address-end")) { |
245 | ret.address_end = uadv(source->keyword); |
246 | } else if (!ustricmp(source->keyword, L"xhtml-navigation-attributes")) { |
247 | ret.nav_attrs = uadv(source->keyword); |
5d9cc07b |
248 | } else if (!ustricmp(source->keyword, L"xhtml-chapter-numeric")) { |
249 | ret.fchapter.just_numbers = utob(uadv(source->keyword)); |
250 | } else if (!ustricmp(source->keyword, L"xhtml-chapter-suffix")) { |
e5e6bf9d |
251 | ret.fchapter.number_suffix = uadv(source->keyword); |
5d9cc07b |
252 | } else if (!ustricmp(source->keyword, L"xhtml-section-numeric")) { |
253 | wchar_t *p = uadv(source->keyword); |
254 | int n = 0; |
255 | if (uisdigit(*p)) { |
256 | n = utoi(p); |
257 | p = uadv(p); |
258 | } |
259 | if (n >= ret.nfsect) { |
260 | int i; |
261 | ret.fsect = resize(ret.fsect, n+1); |
262 | for (i = ret.nfsect; i <= n; i++) |
263 | ret.fsect[i] = ret.fsect[ret.nfsect-1]; |
264 | ret.nfsect = n+1; |
265 | } |
266 | ret.fsect[n].just_numbers = utob(p); |
267 | } else if (!ustricmp(source->keyword, L"xhtml-section-suffix")) { |
268 | wchar_t *p = uadv(source->keyword); |
269 | int n = 0; |
270 | if (uisdigit(*p)) { |
271 | n = utoi(p); |
272 | p = uadv(p); |
273 | } |
274 | if (n >= ret.nfsect) { |
275 | int i; |
276 | ret.fsect = resize(ret.fsect, n+1); |
277 | for (i = ret.nfsect; i <= n; i++) |
278 | ret.fsect[i] = ret.fsect[ret.nfsect-1]; |
279 | ret.nfsect = n+1; |
280 | } |
e5e6bf9d |
281 | ret.fsect[n].number_suffix = p; |
d7482997 |
282 | } |
283 | } |
284 | } |
285 | |
286 | /* printf(" !!! leaf_level = %i\n", ret.leaf_level); |
287 | printf(" !!! contentdepth-0 = %i\n", ret.contents_depth[0]); |
288 | printf(" !!! contentdepth-1 = %i\n", ret.contents_depth[1]); |
289 | printf(" !!! contentdepth-2 = %i\n", ret.contents_depth[2]); |
290 | printf(" !!! contentdepth-3 = %i\n", ret.contents_depth[3]); |
291 | printf(" !!! contentdepth-4 = %i\n", ret.contents_depth[4]); |
292 | printf(" !!! contentdepth-5 = %i\n", ret.contents_depth[5]); |
293 | printf(" !!! leaf_contains_contents = %i\n", ret.leaf_contains_contents);*/ |
294 | return ret; |
295 | } |
296 | |
ba9c1487 |
297 | paragraph *xhtml_config_filename(char *filename) |
298 | { |
299 | /* |
300 | * If the user passes in a single filename as a parameter to |
301 | * the `--html' command-line option, then we should assume it |
302 | * to imply _two_ config directives: |
303 | * \cfg{xhtml-single-filename}{whatever} and |
304 | * \cfg{xhtml-leaf-level}{0}; the rationale being that the user |
305 | * wants their output _in that file_. |
306 | */ |
e4ea58f8 |
307 | paragraph *p, *q; |
ba9c1487 |
308 | |
e4ea58f8 |
309 | p = cmdline_cfg_simple("xhtml-single-filename", filename, NULL); |
310 | q = cmdline_cfg_simple("xhtml-leaf-level", "0", NULL); |
311 | p->next = q; |
312 | return p; |
ba9c1487 |
313 | } |
314 | |
d7482997 |
315 | static xhtmlsection *xhtml_new_section(xhtmlsection *last) |
316 | { |
317 | xhtmlsection *ret = mknew(xhtmlsection); |
318 | ret->next=NULL; |
319 | ret->child=NULL; |
320 | ret->parent=NULL; |
321 | ret->chain=last; |
322 | ret->para=NULL; |
323 | ret->file=NULL; |
324 | ret->fragment=NULL; |
325 | ret->level=-1; /* marker: end of chain */ |
326 | return ret; |
327 | } |
328 | |
329 | /* Returns NULL or the section that marks that paragraph */ |
330 | static xhtmlsection *xhtml_find_section(paragraph *p) |
331 | { |
332 | xhtmlsection *ret = topsection; |
333 | if (xhtml_para_level(p)==-1) { /* first, we back-track to a section paragraph */ |
334 | paragraph *p2 = sourceparas; |
335 | paragraph *p3 = NULL; |
336 | while (p2 && p2!=p) { |
337 | if (xhtml_para_level(p2)!=-1) { |
338 | p3 = p2; |
339 | } |
340 | p2=p2->next; |
341 | } |
342 | if (p3==NULL) { /* for some reason, we couldn't find a section before this paragraph ... ? */ |
343 | /* Note that this can happen, if you have a cross-reference to before the first chapter starts. |
344 | * So don't do that, then. |
345 | */ |
346 | return NULL; |
347 | } |
348 | p=p3; |
349 | } |
350 | while (ret && ret->para != p) { |
351 | /* printf(" xhtml_find_section(): checking %s for para @ %p\n", ret->fragment, p);*/ |
352 | ret=ret->chain; |
353 | } |
354 | return ret; |
355 | } |
356 | |
6d6d850c |
357 | static void xhtml_format(paragraph *p, char *template_string, rdstringc *r) |
358 | { |
359 | char *c, *t; |
360 | word *w; |
361 | wchar_t *ws; |
362 | |
363 | t = template_string; |
364 | while (*t) { |
365 | if (*t == '%' && t[1]) { |
366 | int fmt; |
367 | |
368 | t++; |
369 | fmt = *t++; |
370 | |
371 | if (fmt == '%') { |
372 | rdaddc(r, fmt); |
373 | continue; |
374 | } |
375 | |
376 | w = NULL; |
377 | ws = NULL; |
378 | |
379 | if (p->kwtext && fmt == 'n') |
380 | w = p->kwtext; |
381 | else if (p->kwtext2 && fmt == 'b') |
382 | w = p->kwtext2; |
383 | else if (p->keyword && *p->keyword && fmt == 'k') |
384 | ws = p->keyword; |
385 | else |
386 | w = p->words; |
387 | |
388 | while (w) { |
389 | switch (removeattr(w->type)) |
390 | { |
391 | case word_Normal: |
392 | /*case word_Emph: |
393 | case word_Code: |
394 | case word_WeakCode:*/ |
395 | xhtml_utostr(w->text, &c); |
396 | rdaddsc(r,c); |
397 | sfree(c); |
398 | break; |
399 | } |
400 | w = w->next; |
401 | } |
402 | if (ws) { |
403 | xhtml_utostr(ws, &c); |
404 | rdaddsc(r,c); |
405 | sfree(c); |
406 | } |
407 | } else { |
408 | rdaddc(r, *t++); |
409 | } |
410 | } |
411 | } |
412 | |
d7482997 |
413 | static xhtmlfile *xhtml_new_file(xhtmlsection *sect) |
414 | { |
415 | xhtmlfile *ret = mknew(xhtmlfile); |
416 | |
417 | ret->next=NULL; |
418 | ret->child=NULL; |
419 | ret->parent=NULL; |
420 | ret->filename=NULL; |
421 | ret->sections=sect; |
422 | ret->is_leaf=(sect!=NULL && sect->level==conf.leaf_level); |
423 | if (sect==NULL) { |
424 | if (conf.leaf_level==0) { /* currently unused */ |
50d6b4bd |
425 | ret->filename = smalloc(strlen(conf.single_filename)+1); |
426 | sprintf(ret->filename, conf.single_filename); |
d7482997 |
427 | } else { |
50d6b4bd |
428 | ret->filename = smalloc(strlen(conf.contents_filename)+1); |
429 | sprintf(ret->filename, conf.contents_filename); |
d7482997 |
430 | } |
431 | } else { |
432 | paragraph *p = sect->para; |
433 | rdstringc fname_c = { 0, 0, NULL }; |
6d6d850c |
434 | xhtml_format(p, conf.template_filename, &fname_c); |
d7482997 |
435 | ret->filename = rdtrimc(&fname_c); |
436 | } |
437 | /* printf(" ! new file '%s', is_leaf == %s\n", ret->filename, (ret->is_leaf)?("true"):("false"));*/ |
438 | return ret; |
439 | } |
440 | |
441 | /* |
442 | * Walk the tree fixing up files which are actually leaf (ie |
443 | * have no children) but aren't at leaf level, so they have the |
444 | * leaf flag set. |
445 | */ |
446 | void xhtml_fixup_layout(xhtmlfile* file) |
447 | { |
448 | if (file->child==NULL) { |
449 | file->is_leaf = TRUE; |
450 | } else { |
451 | xhtml_fixup_layout(file->child); |
452 | } |
453 | if (file->next) |
454 | xhtml_fixup_layout(file->next); |
455 | } |
456 | |
457 | /* |
458 | * Create the tree structure so we know where everything goes. |
459 | * Method: |
460 | * |
461 | * Ignoring file splitting, we have three choices with each new section: |
462 | * |
463 | * +-----------------+-----------------+ |
464 | * | | | |
465 | * X +----X----+ (1) |
466 | * | | |
5d9cc07b |
467 | * Y (2) |
d7482997 |
468 | * | |
469 | * (3) |
470 | * |
471 | * Y is the last section we added (currentsect). |
472 | * If sect is the section we want to add, then: |
473 | * |
474 | * (1) if sect->level < currentsect->level |
475 | * (2) if sect->level == currentsect->level |
476 | * (3) if sect->level > currentsect->level |
477 | * |
478 | * This requires the constraint that you never skip section numbers |
479 | * (so you can't have a.b.c.d without all of a, a.b and a.b.c existing). |
480 | * |
481 | * Note that you _can_ have 1.1.1.1 followed by 1.2 - you can change |
482 | * more than one level at a time. Lots of asserts, and probably part of |
483 | * the algorithm here, rely on this being true. (It currently isn't |
484 | * enforced by halibut, however.) |
485 | * |
486 | * File splitting makes this harder. For instance, say we added at (3) |
487 | * above and now need to add another section. We are splitting at level |
488 | * 2, ie the level of Y. Z is the last section we added: |
489 | * |
490 | * +-----------------+-----------------+ |
491 | * | | | |
492 | * X +----X----+ (1) |
493 | * | | |
494 | * +----Y----+ (1) |
495 | * | | |
496 | * Z (2) |
497 | * | |
498 | * (3) |
499 | * |
500 | * The (1) case is now split; we need to search upwards to find where |
501 | * to actually link in. The other two cases remain the same (and will |
502 | * always be like this). |
503 | * |
504 | * File splitting makes this harder, however. The decision of whether |
505 | * to split to a new file is always on the same condition, however (is |
506 | * the level of this section higher than the leaf_level configuration |
507 | * value or not). |
508 | * |
509 | * Treating the cases backwards: |
510 | * |
511 | * (3) same file if sect->level > conf.leaf_level, otherwise new file |
512 | * |
513 | * if in the same file, currentsect->child points to sect |
514 | * otherwise the linking is done through the file tree (which works |
515 | * in more or less the same way, ie currentfile->child points to |
516 | * the new file) |
517 | * |
518 | * (2) same file if sect->level > conf.leaf_level, otherwise new file |
519 | * |
520 | * if in the same file, currentsect->next points to sect |
521 | * otherwise file linking and currentfile->next points to the new |
522 | * file (we know that Z must have caused a new file to be created) |
523 | * |
524 | * (1) same file if sect->level > conf.leaf_level, otherwise new file |
525 | * |
526 | * this is actually effectively the same case as (2) here, |
527 | * except that we first have to travel up the sections to figure |
528 | * out which section this new one will be a sibling of. In doing |
529 | * so, we may disappear off the top of a file and have to go up |
530 | * to its parent in the file tree. |
531 | * |
532 | */ |
533 | static void xhtml_ponder_layout(paragraph *p) |
534 | { |
535 | xhtmlsection *lastsection; |
536 | xhtmlsection *currentsect; |
537 | xhtmlfile *currentfile; |
538 | |
539 | lastfile = NULL; |
540 | topsection = xhtml_new_section(NULL); |
541 | topfile = xhtml_new_file(NULL); |
542 | lastsection = topsection; |
543 | currentfile = topfile; |
544 | currentsect = topsection; |
545 | |
d2e74722 |
546 | if (conf.leaf_level == 0) { |
547 | topfile->is_leaf = 1; |
548 | topfile->sections = topsection; |
549 | topsection->file = topfile; |
550 | } |
551 | |
d7482997 |
552 | for (; p; p=p->next) |
553 | { |
554 | int level = xhtml_para_level(p); |
555 | if (level>0) /* actually a section */ |
556 | { |
557 | xhtmlsection *sect; |
6d6d850c |
558 | rdstringc frag_c = { 0, 0, NULL }; |
d7482997 |
559 | |
560 | sect = xhtml_new_section(lastsection); |
561 | lastsection = sect; |
562 | sect->para = p; |
6d6d850c |
563 | |
564 | xhtml_format(p, conf.template_fragment, &frag_c); |
565 | sect->fragment = rdtrimc(&frag_c); |
d7482997 |
566 | sect->level = level; |
567 | /* printf(" ! adding para @ %p as sect %s, level %i\n", sect->para, sect->fragment, level);*/ |
568 | |
569 | if (level>currentsect->level) { /* case (3) */ |
570 | if (level>conf.leaf_level) { /* same file */ |
571 | assert(currentfile->is_leaf); |
572 | currentsect->child = sect; |
573 | sect->parent=currentsect; |
574 | sect->file=currentfile; |
575 | /* printf("connected '%s' to existing file '%s' [I]\n", sect->fragment, currentfile->filename);*/ |
576 | currentsect=sect; |
577 | } else { /* new file */ |
578 | xhtmlfile *file = xhtml_new_file(sect); |
579 | assert(!currentfile->is_leaf); |
580 | currentfile->child=file; |
581 | sect->file=file; |
582 | file->parent=currentfile; |
583 | /* printf("connected '%s' to new file '%s' [I]\n", sect->fragment, file->filename);*/ |
584 | currentfile=file; |
585 | currentsect=sect; |
586 | } |
587 | } else if (level >= currentsect->file->sections->level) { |
588 | /* Case (1) or (2) *AND* still under the section that starts |
589 | * the current file. |
590 | * |
591 | * I'm not convinced that this couldn't be rolled in with the |
592 | * final else {} leg further down. It seems a lot of effort |
593 | * this way. |
594 | */ |
595 | if (level>conf.leaf_level) { /* stick within the same file */ |
596 | assert(currentfile->is_leaf); |
597 | sect->file = currentfile; |
598 | while (currentsect && currentsect->level > level && |
599 | currentsect->file==currentsect->parent->file) { |
600 | currentsect = currentsect->parent; |
601 | } |
602 | assert(currentsect); |
603 | currentsect->next = sect; |
604 | assert(currentsect->level == sect->level); |
605 | sect->parent = currentsect->parent; |
606 | currentsect = sect; |
607 | /* printf("connected '%s' to existing file '%s' [II]\n", sect->fragment, currentfile->filename);*/ |
608 | } else { /* new file */ |
609 | xhtmlfile *file = xhtml_new_file(sect); |
610 | sect->file=file; |
611 | currentfile->next=file; |
612 | file->parent=currentfile->parent; |
613 | file->is_leaf=(level==conf.leaf_level); |
614 | file->sections=sect; |
615 | /* printf("connected '%s' to new file '%s' [II]\n", sect->fragment, file->filename);*/ |
616 | currentfile=file; |
617 | currentsect=sect; |
618 | } |
619 | } else { /* Case (1) or (2) and we must move up the file tree first */ |
620 | /* this loop is now probably irrelevant - we know we can't connect |
621 | * to anything in the current file */ |
622 | while (currentsect && level<currentsect->level) { |
623 | currentsect=currentsect->parent; |
624 | if (currentsect) { |
625 | /* printf(" * up one level to '%s'\n", currentsect->fragment);*/ |
626 | } else { |
627 | /* printf(" * up one level (off top of current file)\n");*/ |
628 | } |
629 | } |
630 | if (currentsect) { |
631 | /* I'm pretty sure this can now never fire */ |
632 | assert(currentfile->is_leaf); |
633 | /* printf("connected '%s' to existing file '%s' [III]\n", sect->fragment, currentfile->filename);*/ |
634 | sect->file = currentfile; |
635 | currentsect->next=sect; |
636 | currentsect=sect; |
637 | } else { /* find a file we can attach to */ |
638 | while (currentfile && currentfile->sections && level<currentfile->sections->level) { |
639 | currentfile=currentfile->parent; |
640 | if (currentfile) { |
641 | /* printf(" * up one file level to '%s'\n", currentfile->filename);*/ |
642 | } else { |
643 | /* printf(" * up one file level (off top of tree)\n");*/ |
644 | } |
645 | } |
646 | if (currentfile) { /* new file (we had to skip up a file to |
647 | get here, so we must be dealing with a |
648 | level no lower than the configured |
649 | leaf_level */ |
650 | xhtmlfile *file = xhtml_new_file(sect); |
651 | currentfile->next=file; |
652 | sect->file=file; |
653 | file->parent=currentfile->parent; |
654 | file->is_leaf=(level==conf.leaf_level); |
655 | file->sections=sect; |
656 | /* printf("connected '%s' to new file '%s' [III]\n", sect->fragment, file->filename);*/ |
657 | currentfile=file; |
658 | currentsect=sect; |
659 | } else { |
660 | fatal(err_whatever, "Ran off the top trying to connect sibling: strange document."); |
661 | } |
662 | } |
663 | } |
664 | } |
665 | } |
666 | topsection = lastsection; /* get correct end of the chain */ |
667 | xhtml_fixup_layout(topfile); /* leaf files not at leaf level marked as such */ |
668 | } |
669 | |
670 | static void xhtml_do_index(); |
671 | static void xhtml_do_file(xhtmlfile *file); |
672 | static void xhtml_do_top_file(xhtmlfile *file, paragraph *sourceform); |
ce9921d6 |
673 | static void xhtml_do_paras(FILE *fp, paragraph *p, paragraph *end, int indexable); |
d7482997 |
674 | static int xhtml_do_contents_limit(FILE *fp, xhtmlfile *file, int limit); |
675 | static int xhtml_do_contents_section_limit(FILE *fp, xhtmlsection *section, int limit); |
676 | static int xhtml_add_contents_entry(FILE *fp, xhtmlsection *section, int limit); |
677 | static int xhtml_do_contents(FILE *fp, xhtmlfile *file); |
678 | static int xhtml_do_naked_contents(FILE *fp, xhtmlfile *file); |
679 | static void xhtml_do_sections(FILE *fp, xhtmlsection *sections); |
680 | |
681 | /* |
682 | * Do all the files in this structure. |
683 | */ |
684 | static void xhtml_do_files(xhtmlfile *file) |
685 | { |
686 | xhtml_do_file(file); |
687 | if (file->child) |
688 | xhtml_do_files(file->child); |
689 | if (file->next) |
690 | xhtml_do_files(file->next); |
691 | } |
692 | |
693 | /* |
694 | * Free up all memory used by the file tree from 'xfile' downwards |
695 | */ |
696 | static void xhtml_free_file(xhtmlfile* xfile) |
697 | { |
698 | if (xfile==NULL) { |
699 | return; |
700 | } |
701 | |
702 | if (xfile->filename) { |
703 | sfree(xfile->filename); |
704 | } |
705 | xhtml_free_file(xfile->child); |
706 | xhtml_free_file(xfile->next); |
707 | sfree(xfile); |
708 | } |
709 | |
710 | /* |
711 | * Main function. |
712 | */ |
713 | void xhtml_backend(paragraph *sourceform, keywordlist *in_keywords, |
43341922 |
714 | indexdata *in_idx, void *unused) |
d7482997 |
715 | { |
716 | /* int i;*/ |
717 | indexentry *ientry; |
718 | int ti; |
719 | xhtmlsection *xsect; |
720 | |
43341922 |
721 | IGNORE(unused); |
722 | |
d7482997 |
723 | sourceparas = sourceform; |
724 | conf = xhtml_configure(sourceform); |
725 | keywords = in_keywords; |
726 | idx = in_idx; |
727 | |
728 | /* Clear up the index entries backend data pointers */ |
729 | for (ti=0; (ientry = (indexentry *)index234(idx->entries, ti))!=NULL; ti++) { |
730 | ientry->backend_data=NULL; |
731 | } |
732 | |
733 | xhtml_ponder_layout(sourceform); |
734 | |
735 | /* old system ... (writes to *.alt, but gets some stuff wrong and is ugly) */ |
736 | /* xhtml_level_0(sourceform); |
737 | for (i=1; i<=conf.leaf_level; i++) |
738 | { |
739 | xhtml_level(sourceform, i); |
740 | }*/ |
741 | |
742 | /* new system ... (writes to *.html, but isn't fully trusted) */ |
743 | xhtml_do_top_file(topfile, sourceform); |
744 | assert(!topfile->next); /* shouldn't have a sibling at all */ |
d2e74722 |
745 | if (topfile->child) { |
746 | xhtml_do_files(topfile->child); |
747 | xhtml_do_index(); |
748 | } |
d7482997 |
749 | |
750 | /* release file, section, index data structures */ |
751 | xsect = topsection; |
752 | while (xsect) { |
753 | xhtmlsection *tmp = xsect->chain; |
754 | if (xsect->fragment) { |
755 | sfree(xsect->fragment); |
756 | } |
757 | sfree(xsect); |
758 | xsect = tmp; |
759 | } |
760 | xhtml_free_file(topfile); |
761 | for (ti = 0; (ientry=(indexentry *)index234(idx->entries, ti))!=NULL; ti++) { |
762 | if (ientry->backend_data!=NULL) { |
763 | xhtmlindex *xi = (xhtmlindex*) ientry->backend_data; |
764 | if (xi->sections!=NULL) { |
765 | sfree(xi->sections); |
766 | } |
767 | sfree(xi); |
768 | } |
769 | ientry->backend_data = NULL; |
770 | } |
e5e6bf9d |
771 | sfree(conf.fsect); |
d7482997 |
772 | } |
773 | |
774 | static int xhtml_para_level(paragraph *p) |
775 | { |
776 | switch (p->type) |
777 | { |
d9d3dd95 |
778 | case para_Title: |
779 | return 0; |
780 | break; |
d7482997 |
781 | case para_UnnumberedChapter: |
782 | case para_Chapter: |
783 | case para_Appendix: |
784 | return 1; |
785 | break; |
786 | /* case para_BiblioCited: |
787 | return 2; |
788 | break;*/ |
789 | case para_Heading: |
790 | case para_Subsect: |
791 | return p->aux+2; |
792 | break; |
793 | default: |
794 | return -1; |
795 | break; |
796 | } |
797 | } |
798 | |
d7482997 |
799 | /* Output the nav links for the current file. |
800 | * file == NULL means we're doing the index |
801 | */ |
802 | static void xhtml_donavlinks(FILE *fp, xhtmlfile *file) |
803 | { |
804 | xhtmlfile *xhtml_next_file = NULL; |
805 | fprintf(fp, "<p"); |
806 | if (conf.nav_attrs!=NULL) { |
807 | fprintf(fp, " %ls>", conf.nav_attrs); |
808 | } else { |
809 | fprintf(fp, ">"); |
810 | } |
811 | if (xhtml_last_file==NULL) { |
812 | fprintf(fp, "Previous | "); |
813 | } else { |
503e4ab0 |
814 | fprintf(fp, "<a href=\"%s\">Previous</a> | ", xhtml_last_file->filename); |
d7482997 |
815 | } |
503e4ab0 |
816 | fprintf(fp, "<a href=\"%s\">Contents</a> | ", conf.contents_filename); |
25acf71d |
817 | if (file == NULL) { |
818 | fprintf(fp, "Index | "); |
819 | } else { |
503e4ab0 |
820 | fprintf(fp, "<a href=\"%s\">Index</a> | ", conf.index_filename); |
25acf71d |
821 | } |
d7482997 |
822 | if (file != NULL) { /* otherwise we're doing nav links for the index */ |
823 | if (xhtml_next_file==NULL) |
824 | xhtml_next_file = file->child; |
825 | if (xhtml_next_file==NULL) |
826 | xhtml_next_file = file->next; |
827 | if (xhtml_next_file==NULL) |
828 | xhtml_next_file = file->parent->next; |
829 | } |
830 | if (xhtml_next_file==NULL) { |
831 | if (file==NULL) { /* index, so no next file */ |
832 | fprintf(fp, "Next "); |
833 | } else { |
503e4ab0 |
834 | fprintf(fp, "<a href=\"%s\">Next</a>", conf.index_filename); |
d7482997 |
835 | } |
836 | } else { |
503e4ab0 |
837 | fprintf(fp, "<a href=\"%s\">Next</a>", xhtml_next_file->filename); |
d7482997 |
838 | } |
839 | fprintf(fp, "</p>\n"); |
840 | } |
841 | |
842 | /* Write out the index file */ |
d2e74722 |
843 | static void xhtml_do_index_body(FILE *fp) |
d7482997 |
844 | { |
d7482997 |
845 | indexentry *y; |
846 | int ti; |
d7482997 |
847 | |
d2e74722 |
848 | if (count234(idx->entries) == 0) |
849 | return; /* don't write anything at all */ |
d7482997 |
850 | |
851 | fprintf(fp, "<dl>\n"); |
852 | /* iterate over idx->entries using the tree functions and display everything */ |
853 | for (ti = 0; (y = (indexentry *)index234(idx->entries, ti)) != NULL; ti++) { |
854 | if (y->backend_data) { |
855 | int i; |
856 | xhtmlindex *xi; |
857 | |
858 | fprintf(fp, "<dt>"); |
ce9921d6 |
859 | xhtml_para(fp, y->text, FALSE); |
d7482997 |
860 | fprintf(fp, "</dt>\n<dd>"); |
861 | |
862 | xi = (xhtmlindex*) y->backend_data; |
863 | for (i=0; i<xi->nsection; i++) { |
864 | xhtmlsection *sect = xi->sections[i]; |
865 | if (sect) { |
503e4ab0 |
866 | fprintf(fp, "<a href=\"%s#%s\">", sect->file->filename, sect->fragment); |
d7482997 |
867 | if (sect->para->kwtext) { |
ce9921d6 |
868 | xhtml_para(fp, sect->para->kwtext, FALSE); |
d7482997 |
869 | } else if (sect->para->words) { |
ce9921d6 |
870 | xhtml_para(fp, sect->para->words, FALSE); |
d7482997 |
871 | } |
872 | fprintf(fp, "</a>"); |
873 | if (i+1<xi->nsection) { |
874 | fprintf(fp, ", "); |
875 | } |
876 | } |
877 | } |
878 | fprintf(fp, "</dd>\n"); |
879 | } |
880 | } |
881 | fprintf(fp, "</dl>\n"); |
d2e74722 |
882 | } |
883 | static void xhtml_do_index() |
884 | { |
5dd44dce |
885 | word temp_word = { NULL, NULL, word_Normal, 0, 0, L"Index", |
886 | { NULL, 0, 0}, NULL }; |
50d6b4bd |
887 | FILE *fp = fopen(conf.index_filename, "w"); |
d2e74722 |
888 | |
889 | if (fp==NULL) |
50d6b4bd |
890 | fatal(err_cantopenw, conf.index_filename); |
d2e74722 |
891 | xhtml_doheader(fp, &temp_word); |
892 | xhtml_donavlinks(fp, NULL); |
893 | |
894 | xhtml_do_index_body(fp); |
d7482997 |
895 | |
896 | xhtml_donavlinks(fp, NULL); |
897 | xhtml_dofooter(fp); |
898 | fclose(fp); |
899 | } |
900 | |
901 | /* Output the given file. This includes whatever contents at beginning and end, etc. etc. */ |
902 | static void xhtml_do_file(xhtmlfile *file) |
903 | { |
904 | FILE *fp = fopen(file->filename, "w"); |
905 | if (fp==NULL) |
906 | fatal(err_cantopenw, file->filename); |
907 | |
908 | if (file->sections->para->words) { |
909 | xhtml_doheader(fp, file->sections->para->words); |
910 | } else if (file->sections->para->kwtext) { |
911 | xhtml_doheader(fp, file->sections->para->kwtext); |
912 | } else { |
913 | xhtml_doheader(fp, NULL); |
914 | } |
915 | |
916 | xhtml_donavlinks(fp, file); |
917 | |
d2e74722 |
918 | if (file->is_leaf && conf.leaf_contains_contents && |
919 | xhtml_do_contents(NULL, file)>=conf.leaf_smallest_contents) |
d7482997 |
920 | xhtml_do_contents(fp, file); |
921 | xhtml_do_sections(fp, file->sections); |
922 | if (!file->is_leaf) |
923 | xhtml_do_naked_contents(fp, file); |
924 | |
925 | xhtml_donavlinks(fp, file); |
926 | |
927 | xhtml_dofooter(fp); |
928 | fclose(fp); |
929 | |
930 | xhtml_last_file = file; |
931 | } |
932 | |
933 | /* Output the top-level file. */ |
934 | static void xhtml_do_top_file(xhtmlfile *file, paragraph *sourceform) |
935 | { |
936 | paragraph *p; |
937 | int done=FALSE; |
938 | FILE *fp = fopen(file->filename, "w"); |
939 | if (fp==NULL) |
940 | fatal(err_cantopenw, file->filename); |
941 | |
942 | /* Do the title -- only one allowed */ |
943 | for (p = sourceform; p && !done; p = p->next) |
944 | { |
945 | if (p->type == para_Title) |
946 | { |
947 | xhtml_doheader(fp, p->words); |
948 | done=TRUE; |
949 | } |
950 | } |
951 | if (!done) |
952 | xhtml_doheader(fp, NULL /* Eek! */); |
953 | |
d2e74722 |
954 | /* |
955 | * Display the title. |
956 | */ |
957 | for (p = sourceform; p; p = p->next) |
958 | { |
959 | if (p->type == para_Title) { |
ce9921d6 |
960 | xhtml_heading(fp, p, FALSE); |
d2e74722 |
961 | break; |
962 | } |
963 | } |
964 | |
9057a0a8 |
965 | /* Do the preamble */ |
d7482997 |
966 | for (p = sourceform; p; p = p->next) |
967 | { |
8902e0ed |
968 | if (p->type == para_Chapter || p->type == para_Heading || |
969 | p->type == para_Subsect || p->type == para_Appendix || |
970 | p->type == para_UnnumberedChapter) { |
971 | /* |
972 | * We've found the end of the preamble. Do every normal |
973 | * paragraph up to there. |
974 | */ |
ce9921d6 |
975 | xhtml_do_paras(fp, sourceform, p, FALSE); |
8902e0ed |
976 | break; |
d7482997 |
977 | } |
978 | } |
d7482997 |
979 | |
980 | xhtml_do_contents(fp, file); |
981 | xhtml_do_sections(fp, file->sections); |
d2e74722 |
982 | |
5d9cc07b |
983 | /* |
984 | * Put the index in the top file if we're in single-file mode |
985 | * (leaf-level 0). |
986 | */ |
987 | if (conf.leaf_level == 0 && count234(idx->entries) > 0) { |
d2e74722 |
988 | fprintf(fp, "<a name=\"index\"></a><h1>Index</h1>\n"); |
989 | xhtml_do_index_body(fp); |
990 | } |
991 | |
d7482997 |
992 | xhtml_dofooter(fp); |
993 | fclose(fp); |
994 | } |
995 | |
996 | /* Convert a Unicode string to an ASCII one. '?' is |
997 | * used for unmappable characters. |
998 | */ |
999 | static void xhtml_utostr(wchar_t *in, char **out) |
1000 | { |
1001 | int l = ustrlen(in); |
1002 | int i; |
1003 | *out = smalloc(l+1); |
1004 | for (i=0; i<l; i++) |
1005 | { |
1006 | if (in[i]>=32 && in[i]<=126) |
1007 | (*out)[i]=(char)in[i]; |
1008 | else |
1009 | (*out)[i]='?'; |
1010 | } |
1011 | (*out)[i]=0; |
1012 | } |
1013 | |
1014 | /* |
1015 | * Write contents for the given file, and subfiles, down to |
1016 | * the appropriate contents depth. Returns the number of |
1017 | * entries written. |
1018 | */ |
1019 | static int xhtml_do_contents(FILE *fp, xhtmlfile *file) |
1020 | { |
c8c7926b |
1021 | int level, limit, count = 0; |
d7482997 |
1022 | if (!file) |
1023 | return 0; |
1024 | |
1025 | level = (file->sections)?(file->sections->level):(0); |
1026 | limit = conf.contents_depth[(level>5)?(5):(level)]; |
1027 | start_level = (file->is_leaf) ? (level-1) : (level); |
1028 | last_level = start_level; |
1029 | |
1030 | count += xhtml_do_contents_section_limit(fp, file->sections, limit); |
1031 | count += xhtml_do_contents_limit(fp, file->child, limit); |
1032 | if (fp!=NULL) { |
1033 | while (last_level > start_level) { |
1034 | last_level--; |
c8c7926b |
1035 | fprintf(fp, "</li></ul>\n"); |
d7482997 |
1036 | } |
1037 | } |
1038 | return count; |
1039 | } |
1040 | |
1041 | /* As above, but doesn't do anything in the current file */ |
1042 | static int xhtml_do_naked_contents(FILE *fp, xhtmlfile *file) |
1043 | { |
1044 | int level, limit, start_level, count = 0; |
1045 | if (!file) |
1046 | return 0; |
1047 | |
1048 | level = (file->sections)?(file->sections->level):(0); |
1049 | limit = conf.contents_depth[(level>5)?(5):(level)]; |
1050 | start_level = (file->is_leaf) ? (level-1) : (level); |
1051 | last_level = start_level; |
1052 | |
1053 | count = xhtml_do_contents_limit(fp, file->child, limit); |
1054 | if (fp!=NULL) { |
1055 | while (last_level > start_level) { |
1056 | last_level--; |
c8c7926b |
1057 | fprintf(fp, "</li></ul>\n"); |
d7482997 |
1058 | } |
1059 | } |
1060 | return count; |
1061 | } |
1062 | |
1063 | /* |
1064 | * Write contents for the given file, children, and siblings, down to |
1065 | * given limit contents depth. |
1066 | */ |
1067 | static int xhtml_do_contents_limit(FILE *fp, xhtmlfile *file, int limit) |
1068 | { |
1069 | int count = 0; |
1070 | while (file) { |
1071 | count += xhtml_do_contents_section_limit(fp, file->sections, limit); |
1072 | count += xhtml_do_contents_limit(fp, file->child, limit); |
1073 | file = file->next; |
1074 | } |
1075 | return count; |
1076 | } |
1077 | |
1078 | /* |
1079 | * Write contents entries for the given section tree, down to the |
1080 | * limit contents depth. |
1081 | */ |
1082 | static int xhtml_do_contents_section_deep_limit(FILE *fp, xhtmlsection *section, int limit) |
1083 | { |
1084 | int count = 0; |
1085 | while (section) { |
1086 | if (!xhtml_add_contents_entry(fp, section, limit)) |
1087 | return 0; |
1088 | else |
1089 | count++; |
1090 | count += xhtml_do_contents_section_deep_limit(fp, section->child, limit); |
1091 | section = section->next; |
1092 | } |
1093 | return count; |
1094 | } |
1095 | |
1096 | /* |
1097 | * Write contents entries for the given section tree, down to the |
1098 | * limit contents depth. |
1099 | */ |
1100 | static int xhtml_do_contents_section_limit(FILE *fp, xhtmlsection *section, int limit) |
1101 | { |
1102 | int count = 0; |
1103 | if (!section) |
1104 | return 0; |
1105 | xhtml_add_contents_entry(fp, section, limit); |
1106 | count=1; |
1107 | count += xhtml_do_contents_section_deep_limit(fp, section->child, limit); |
1108 | /* section=section->child; |
1109 | while (section && xhtml_add_contents_entry(fp, section, limit)) { |
1110 | section = section->next; |
1111 | }*/ |
1112 | return count; |
1113 | } |
1114 | |
1115 | /* |
1116 | * Add a section entry, unless we're exceeding the limit, in which |
1117 | * case return FALSE (otherwise return TRUE). |
1118 | */ |
1119 | static int xhtml_add_contents_entry(FILE *fp, xhtmlsection *section, int limit) |
1120 | { |
1121 | if (!section || section->level > limit) |
1122 | return FALSE; |
5d9cc07b |
1123 | if (fp==NULL || section->level < 0) |
d7482997 |
1124 | return TRUE; |
c8c7926b |
1125 | if (last_level > section->level) { |
1126 | while (last_level > section->level) { |
1127 | last_level--; |
1128 | fprintf(fp, "</li></ul>\n"); |
1129 | } |
1130 | fprintf(fp, "</li>\n"); |
1131 | } else if (last_level < section->level) { |
1132 | assert(last_level == section->level - 1); |
d7482997 |
1133 | last_level++; |
1134 | fprintf(fp, "<ul>\n"); |
c8c7926b |
1135 | } else { |
1136 | fprintf(fp, "</li>\n"); |
d7482997 |
1137 | } |
1138 | fprintf(fp, "<li><a href=\"%s#%s\">", section->file->filename, section->fragment); |
1139 | if (section->para->kwtext) { |
ce9921d6 |
1140 | xhtml_para(fp, section->para->kwtext, FALSE); |
d7482997 |
1141 | if (section->para->words) { |
1142 | fprintf(fp, ": "); |
1143 | } |
1144 | } |
1145 | if (section->para->words) { |
ce9921d6 |
1146 | xhtml_para(fp, section->para->words, FALSE); |
d7482997 |
1147 | } |
c8c7926b |
1148 | fprintf(fp, "</a>\n"); |
d7482997 |
1149 | return TRUE; |
1150 | } |
1151 | |
1152 | /* |
1153 | * Write all the sections in this file. Do all paragraphs in this section, then all |
1154 | * children (recursively), then go on to the next one (tail recursively). |
1155 | */ |
1156 | static void xhtml_do_sections(FILE *fp, xhtmlsection *sections) |
1157 | { |
1158 | while (sections) { |
1159 | currentsection = sections; |
ce9921d6 |
1160 | xhtml_do_paras(fp, sections->para, NULL, TRUE); |
d7482997 |
1161 | xhtml_do_sections(fp, sections->child); |
1162 | sections = sections->next; |
1163 | } |
1164 | } |
1165 | |
1166 | /* Write this list of paragraphs. Close off all lists at the end. */ |
ce9921d6 |
1167 | static void xhtml_do_paras(FILE *fp, paragraph *p, paragraph *end, |
1168 | int indexable) |
d7482997 |
1169 | { |
7136a6c7 |
1170 | int last_type = -1, ptype, first=TRUE; |
1171 | stack lcont_stack = stk_new(); |
d7482997 |
1172 | if (!p) |
1173 | return; |
1174 | |
1175 | /* for (; p && (xhtml_para_level(p)>limit || xhtml_para_level(p)==-1 || first); p=p->next) {*/ |
8902e0ed |
1176 | for (; p && p != end && (xhtml_para_level(p)==-1 || first); p=p->next) { |
d7482997 |
1177 | first=FALSE; |
7136a6c7 |
1178 | switch (ptype = p->type) |
d7482997 |
1179 | { |
1180 | /* |
1181 | * Things we ignore because we've already processed them or |
1182 | * aren't going to touch them in this pass. |
1183 | */ |
1184 | case para_IM: |
1185 | case para_BR: |
1186 | case para_Biblio: /* only touch BiblioCited */ |
1187 | case para_VersionID: |
d7482997 |
1188 | case para_NoCite: |
1189 | case para_Title: |
1190 | break; |
1191 | |
1192 | /* |
1193 | * Chapter titles. |
1194 | */ |
1195 | case para_Chapter: |
1196 | case para_Appendix: |
1197 | case para_UnnumberedChapter: |
ce9921d6 |
1198 | xhtml_heading(fp, p, indexable); |
d7482997 |
1199 | break; |
1200 | |
1201 | case para_Heading: |
1202 | case para_Subsect: |
ce9921d6 |
1203 | xhtml_heading(fp, p, indexable); |
d7482997 |
1204 | break; |
1205 | |
1206 | case para_Rule: |
1207 | fprintf(fp, "\n<hr />\n"); |
1208 | break; |
1209 | |
1210 | case para_Normal: |
9057a0a8 |
1211 | case para_Copyright: |
d7482997 |
1212 | fprintf(fp, "\n<p>"); |
ce9921d6 |
1213 | xhtml_para(fp, p->words, indexable); |
d7482997 |
1214 | fprintf(fp, "</p>\n"); |
1215 | break; |
1216 | |
7136a6c7 |
1217 | case para_LcontPush: |
1218 | { |
1219 | int *p; |
1220 | p = mknew(int); |
1221 | *p = last_type; |
1222 | stk_push(lcont_stack, p); |
1223 | last_type = para_Normal; |
1224 | } |
1225 | break; |
1226 | case para_LcontPop: |
1227 | { |
1228 | int *p = stk_pop(lcont_stack); |
1229 | assert(p); |
1230 | ptype = last_type = *p; |
1231 | sfree(p); |
1232 | goto closeofflist; /* ick */ |
1233 | } |
1234 | break; |
2614b01d |
1235 | case para_QuotePush: |
1236 | fprintf(fp, "<blockquote>\n"); |
1237 | break; |
1238 | case para_QuotePop: |
1239 | fprintf(fp, "</blockquote>\n"); |
1240 | break; |
7136a6c7 |
1241 | |
d7482997 |
1242 | case para_Bullet: |
1243 | case para_NumberedList: |
7136a6c7 |
1244 | case para_Description: |
1245 | case para_DescribedThing: |
d7482997 |
1246 | case para_BiblioCited: |
c8c7926b |
1247 | if (last_type!=p->type && |
1248 | !(last_type==para_DescribedThing && p->type==para_Description) && |
1249 | !(last_type==para_Description && p->type==para_DescribedThing)) { |
d7482997 |
1250 | /* start up list if necessary */ |
1251 | if (p->type == para_Bullet) { |
1252 | fprintf(fp, "<ul>\n"); |
1253 | } else if (p->type == para_NumberedList) { |
1254 | fprintf(fp, "<ol>\n"); |
7136a6c7 |
1255 | } else if (p->type == para_BiblioCited || |
1256 | p->type == para_DescribedThing || |
1257 | p->type == para_Description) { |
d7482997 |
1258 | fprintf(fp, "<dl>\n"); |
1259 | } |
1260 | } |
7136a6c7 |
1261 | if (p->type == para_Bullet || p->type == para_NumberedList) { |
d7482997 |
1262 | fprintf(fp, "<li>"); |
7136a6c7 |
1263 | } else if (p->type == para_DescribedThing) { |
1264 | fprintf(fp, "<dt>"); |
1265 | } else if (p->type == para_Description) { |
1266 | fprintf(fp, "<dd>"); |
1267 | } else if (p->type == para_BiblioCited) { |
d7482997 |
1268 | fprintf(fp, "<dt>"); |
ce9921d6 |
1269 | xhtml_para(fp, p->kwtext, indexable); |
d7482997 |
1270 | fprintf(fp, "</dt>\n<dd>"); |
1271 | } |
ce9921d6 |
1272 | xhtml_para(fp, p->words, indexable); |
7136a6c7 |
1273 | { |
1274 | paragraph *p2 = p->next; |
1275 | if (p2 && xhtml_para_level(p2)==-1 && p2->type == para_LcontPush) |
1276 | break; |
1277 | } |
1278 | |
1279 | closeofflist: |
1280 | if (ptype == para_BiblioCited) { |
d7482997 |
1281 | fprintf(fp, "</dd>\n"); |
c8c7926b |
1282 | } else if (ptype == para_DescribedThing) { |
7136a6c7 |
1283 | fprintf(fp, "</dt>"); |
c8c7926b |
1284 | } else if (ptype == para_Description) { |
7136a6c7 |
1285 | fprintf(fp, "</dd>"); |
1286 | } else if (ptype == para_Bullet || ptype == para_NumberedList) { |
d7482997 |
1287 | fprintf(fp, "</li>"); |
1288 | } |
7136a6c7 |
1289 | if (ptype == para_Bullet || ptype == para_NumberedList || |
1290 | ptype == para_BiblioCited || ptype == para_Description || |
1291 | ptype == para_DescribedThing) |
d7482997 |
1292 | /* close off list if necessary */ |
1293 | { |
1294 | paragraph *p2 = p->next; |
1295 | int close_off=FALSE; |
1296 | /* if (p2 && (xhtml_para_level(p2)>limit || xhtml_para_level(p2)==-1)) {*/ |
1297 | if (p2 && xhtml_para_level(p2)==-1) { |
c8c7926b |
1298 | if (p2->type != ptype && |
1299 | !(p2->type==para_DescribedThing && ptype==para_Description) && |
1300 | !(p2->type==para_Description && ptype==para_DescribedThing) && |
1301 | p2->type != para_LcontPush) |
d7482997 |
1302 | close_off=TRUE; |
1303 | } else { |
1304 | close_off=TRUE; |
1305 | } |
1306 | if (close_off) { |
7136a6c7 |
1307 | if (ptype == para_Bullet) { |
d7482997 |
1308 | fprintf(fp, "</ul>\n"); |
7136a6c7 |
1309 | } else if (ptype == para_NumberedList) { |
d7482997 |
1310 | fprintf(fp, "</ol>\n"); |
7136a6c7 |
1311 | } else if (ptype == para_BiblioCited || |
1312 | ptype == para_Description || |
1313 | ptype == para_DescribedThing) { |
d7482997 |
1314 | fprintf(fp, "</dl>\n"); |
1315 | } |
1316 | } |
1317 | } |
1318 | break; |
1319 | |
1320 | case para_Code: |
1321 | xhtml_codepara(fp, p->words); |
1322 | break; |
1323 | } |
7136a6c7 |
1324 | last_type = ptype; |
d7482997 |
1325 | } |
7136a6c7 |
1326 | |
1327 | stk_free(lcont_stack); |
d7482997 |
1328 | } |
1329 | |
1330 | /* |
1331 | * Output a header for this XHTML file. |
1332 | */ |
1333 | static void xhtml_doheader(FILE *fp, word *title) |
1334 | { |
1335 | fprintf(fp, "<!DOCTYPE html PUBLIC \"-//W3C//DTD XHTML 1.0 Transitional//EN\"\n"); |
1336 | fprintf(fp, "\"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd\">\n"); |
503e4ab0 |
1337 | fprintf(fp, "<html xmlns=\"http://www.w3.org/1999/xhtml\">\n\n<head>\n<title>"); |
d7482997 |
1338 | if (title==NULL) |
1339 | fprintf(fp, "The thing with no name!"); |
1340 | else |
ce9921d6 |
1341 | xhtml_para(fp, title, FALSE); |
d7482997 |
1342 | fprintf(fp, "</title>\n"); |
1343 | fprintf(fp, "<meta name=\"generator\" content=\"Halibut %s xhtml-backend\" />\n", version); |
1344 | if (conf.author) |
1345 | fprintf(fp, "<meta name=\"author\" content=\"%ls\" />\n", conf.author); |
1346 | if (conf.description) |
1347 | fprintf(fp, "<meta name=\"description\" content=\"%ls\" />\n", conf.description); |
1348 | if (conf.head_end) |
1349 | fprintf(fp, "%ls\n", conf.head_end); |
1350 | fprintf(fp, "</head>\n\n"); |
1351 | if (conf.body) |
1352 | fprintf(fp, "%ls\n", conf.body); |
1353 | else |
1354 | fprintf(fp, "<body>\n"); |
1355 | if (conf.body_start) |
1356 | fprintf(fp, "%ls\n", conf.body_start); |
1357 | } |
1358 | |
1359 | /* |
1360 | * Output a footer for this XHTML file. |
1361 | */ |
1362 | static void xhtml_dofooter(FILE *fp) |
1363 | { |
1364 | fprintf(fp, "\n<hr />\n\n"); |
1365 | if (conf.body_end) |
1366 | fprintf(fp, "%ls\n", conf.body_end); |
1367 | if (!conf.suppress_address) { |
1368 | fprintf(fp,"<address>\n"); |
1369 | if (conf.address_start) |
1370 | fprintf(fp, "%ls\n", conf.address_start); |
1371 | /* Do the version ID */ |
1372 | if (conf.include_version_id) { |
1373 | paragraph *p; |
1374 | int started = 0; |
1375 | for (p = sourceparas; p; p = p->next) |
1376 | if (p->type == para_VersionID) { |
1377 | xhtml_versionid(fp, p->words, started); |
1378 | started = 1; |
1379 | } |
1380 | } |
1381 | if (conf.address_end) |
1382 | fprintf(fp, "%ls\n", conf.address_end); |
1383 | fprintf(fp, "</address>\n"); |
1384 | } |
1385 | fprintf(fp, "</body>\n\n</html>\n"); |
1386 | } |
1387 | |
1388 | /* |
1389 | * Output the versionid paragraph. Typically this is a version control |
1390 | * ID string (such as $Id...$ in RCS). |
1391 | */ |
1392 | static void xhtml_versionid(FILE *fp, word *text, int started) |
1393 | { |
1394 | rdstringc t = { 0, 0, NULL }; |
1395 | |
1396 | rdaddc(&t, '['); /* FIXME: configurability */ |
ce9921d6 |
1397 | xhtml_rdaddwc(&t, text, NULL, FALSE); |
d7482997 |
1398 | rdaddc(&t, ']'); /* FIXME: configurability */ |
1399 | |
1400 | if (started) |
c8c7926b |
1401 | fprintf(fp, "<br />\n"); |
d7482997 |
1402 | fprintf(fp, "%s\n", t.text); |
1403 | sfree(t.text); |
1404 | } |
1405 | |
1406 | /* Is this an XHTML reserved character? */ |
1407 | static int xhtml_reservedchar(int c) |
1408 | { |
1409 | if (c=='&' || c=='<' || c=='>' || c=='"') |
1410 | return TRUE; |
1411 | else |
1412 | return FALSE; |
1413 | } |
1414 | |
1415 | /* |
1416 | * Convert a wide string into valid XHTML: Anything outside ASCII will |
1417 | * be fixed up as an entity. Currently we don't worry about constraining the |
1418 | * encoded character set, which we should probably do at some point (we can |
1419 | * still fix up and return FALSE - see the last comment here). We also don't |
1420 | * currently |
1421 | * |
1422 | * Because this is only used for words, spaces are HARD spaces (any other |
1423 | * spaces will be word_Whitespace not word_Normal). So they become |
1424 | * Unless hard_spaces is FALSE, of course (code paragraphs break the above |
1425 | * rule). |
1426 | * |
1427 | * If `result' is non-NULL, mallocs the resulting string and stores a pointer to |
1428 | * it in `*result'. If `result' is NULL, merely checks whether all |
1429 | * characters in the string are feasible. |
1430 | * |
1431 | * Return is nonzero if all characters are OK. If not all |
1432 | * characters are OK but `result' is non-NULL, a result _will_ |
1433 | * still be generated! |
1434 | */ |
4b3c5afb |
1435 | static int xhtml_convert(wchar_t *s, int maxlen, char **result, |
1436 | int hard_spaces) { |
d7482997 |
1437 | int doing = (result != 0); |
1438 | int ok = TRUE; |
1439 | char *p = NULL; |
1440 | int plen = 0, psize = 0; |
1441 | |
4b3c5afb |
1442 | if (maxlen <= 0) |
1443 | maxlen = -1; |
1444 | |
1445 | for (; *s && maxlen != 0; s++, maxlen--) { |
d7482997 |
1446 | wchar_t c = *s; |
1447 | |
1448 | #define ensure_size(i) if (i>=psize) { psize = i+256; p = resize(p, psize); } |
1449 | |
1450 | if (((c == 32 && !hard_spaces) || (c > 32 && c <= 126 && !xhtml_reservedchar(c)))) { |
1451 | /* Char is OK. */ |
1452 | if (doing) |
1453 | { |
1454 | ensure_size(plen); |
1455 | p[plen++] = (char)c; |
1456 | } |
1457 | } else { |
1458 | /* Char needs fixing up. */ |
1459 | /* ok = FALSE; -- currently we never return FALSE; we |
1460 | * might want to when considering a character set for the |
1461 | * encoded document. |
1462 | */ |
1463 | if (doing) |
1464 | { |
1465 | if (c==32) { /* a space in a word is a hard space */ |
1466 | ensure_size(plen+6); /* includes space for the NUL, which is subsequently stomped on */ |
1467 | sprintf(p+plen, " "); |
1468 | plen+=6; |
1469 | } else { |
1470 | /* FIXME: entity names! */ |
1471 | ensure_size(plen+8); /* includes space for the NUL, which is subsequently stomped on */ |
1472 | plen+=sprintf(p+plen, "&#%04i;", (int)c); |
1473 | } |
1474 | } |
1475 | } |
1476 | } |
1477 | if (doing) { |
1478 | p = resize(p, plen+1); |
1479 | p[plen] = '\0'; |
1480 | *result = p; |
1481 | } |
1482 | return ok; |
1483 | } |
1484 | |
1485 | /* |
1486 | * This formats the given words as XHTML. |
ce9921d6 |
1487 | * |
1488 | * `indexable', if FALSE, prohibits adding any index references. |
1489 | * You might use this, for example, if an index reference occurred |
1490 | * in a section title, to prevent phony index references when the |
1491 | * section title is processed in strange places such as contents |
1492 | * sections. |
d7482997 |
1493 | */ |
ce9921d6 |
1494 | static void xhtml_rdaddwc(rdstringc *rs, word *text, word *end, int indexable) { |
d7482997 |
1495 | char *c; |
1496 | keyword *kwl; |
1497 | xhtmlsection *sect; |
1498 | indextag *itag; |
1499 | int ti; |
1500 | |
1501 | for (; text && text != end; text = text->next) { |
1502 | switch (text->type) { |
1503 | case word_HyperLink: |
1504 | xhtml_utostr(text->text, &c); |
1505 | rdaddsc(rs, "<a href=\""); |
1506 | rdaddsc(rs, c); |
1507 | rdaddsc(rs, "\">"); |
1508 | sfree(c); |
1509 | break; |
1510 | |
1511 | case word_UpperXref: |
1512 | case word_LowerXref: |
1513 | kwl = kw_lookup(keywords, text->text); |
1514 | if (kwl) { |
1515 | sect=xhtml_find_section(kwl->para); |
1516 | if (sect) { |
1517 | rdaddsc(rs, "<a href=\""); |
1518 | rdaddsc(rs, sect->file->filename); |
1519 | rdaddc(rs, '#'); |
1520 | rdaddsc(rs, sect->fragment); |
1521 | rdaddsc(rs, "\">"); |
1522 | } else { |
1523 | rdaddsc(rs, "<a href=\"Apologies.html\"><!-- probably a bibliography cross reference -->"); |
1524 | error(err_whatever, "Couldn't locate cross-reference! (Probably a bibliography entry.)"); |
1525 | } |
1526 | } else { |
1527 | rdaddsc(rs, "<a href=\"Apologies.html\"><!-- unknown cross-reference -->"); |
1528 | error(err_whatever, "Couldn't locate cross-reference! (Wasn't in source file.)"); |
1529 | } |
1530 | break; |
1531 | |
1532 | case word_IndexRef: /* in theory we could make an index target here */ |
1533 | /* rdaddsc(rs, "<a name=\"idx-"); |
1534 | xhtml_utostr(text->text, &c); |
1535 | rdaddsc(rs, c); |
1536 | sfree(c); |
1537 | rdaddsc(rs, "\"></a>");*/ |
1538 | /* what we _do_ need to do is to fix up the backend data |
1539 | * for any indexentry this points to. |
1540 | */ |
ce9921d6 |
1541 | if (!indexable) |
1542 | break; |
1543 | |
d7482997 |
1544 | for (ti=0; (itag = (indextag *)index234(idx->tags, ti))!=NULL; ti++) { |
1545 | /* FIXME: really ustricmp() and not ustrcmp()? */ |
1546 | if (ustricmp(itag->name, text->text)==0) { |
1547 | break; |
1548 | } |
1549 | } |
1550 | if (itag!=NULL) { |
1551 | if (itag->refs!=NULL) { |
1552 | int i; |
1553 | for (i=0; i<itag->nrefs; i++) { |
1554 | xhtmlindex *idx_ref; |
1555 | indexentry *ientry; |
1556 | |
1557 | ientry = itag->refs[i]; |
1558 | if (ientry->backend_data==NULL) { |
1559 | idx_ref = (xhtmlindex*) smalloc(sizeof(xhtmlindex)); |
1560 | if (idx_ref==NULL) |
1561 | fatal(err_nomemory); |
1562 | idx_ref->nsection = 0; |
1563 | idx_ref->size = 4; |
1564 | idx_ref->sections = (xhtmlsection**) smalloc(idx_ref->size * sizeof(xhtmlsection*)); |
1565 | if (idx_ref->sections==NULL) |
1566 | fatal(err_nomemory); |
1567 | ientry->backend_data = idx_ref; |
1568 | } else { |
1569 | idx_ref = ientry->backend_data; |
1570 | if (idx_ref->nsection+1 > idx_ref->size) { |
1571 | int new_size = idx_ref->size * 2; |
1572 | idx_ref->sections = srealloc(idx_ref->sections, new_size * sizeof(xhtmlsection)); |
1573 | if (idx_ref->sections==NULL) { |
1574 | fatal(err_nomemory); |
1575 | } |
1576 | idx_ref->size = new_size; |
1577 | } |
1578 | } |
1579 | idx_ref->sections[idx_ref->nsection++] = currentsection; |
1580 | #if 0 |
1581 | #endif |
1582 | } |
1583 | } else { |
1584 | fatal(err_whatever, "Index tag had no entries!"); |
1585 | } |
1586 | } else { |
1587 | fprintf(stderr, "Looking for index entry '%ls'\n", text->text); |
1588 | fatal(err_whatever, "Couldn't locate index entry! (Wasn't in index.)"); |
1589 | } |
1590 | break; |
1591 | |
1592 | case word_HyperEnd: |
1593 | case word_XrefEnd: |
1594 | rdaddsc(rs, "</a>"); |
1595 | break; |
1596 | |
1597 | case word_Normal: |
1598 | case word_Emph: |
1599 | case word_Code: |
1600 | case word_WeakCode: |
1601 | case word_WhiteSpace: |
1602 | case word_EmphSpace: |
1603 | case word_CodeSpace: |
1604 | case word_WkCodeSpace: |
1605 | case word_Quote: |
1606 | case word_EmphQuote: |
1607 | case word_CodeQuote: |
1608 | case word_WkCodeQuote: |
1609 | assert(text->type != word_CodeQuote && |
1610 | text->type != word_WkCodeQuote); |
1611 | if (towordstyle(text->type) == word_Emph && |
1612 | (attraux(text->aux) == attr_First || |
1613 | attraux(text->aux) == attr_Only)) |
1614 | rdaddsc(rs, "<em>"); |
1615 | else if ((towordstyle(text->type) == word_Code || towordstyle(text->type) == word_WeakCode) && |
1616 | (attraux(text->aux) == attr_First || |
1617 | attraux(text->aux) == attr_Only)) |
1618 | rdaddsc(rs, "<code>"); |
1619 | |
1620 | if (removeattr(text->type) == word_Normal) { |
12efc259 |
1621 | if (xhtml_convert(text->text, 0, &c, TRUE) || !text->alt) |
1622 | /* spaces in the word are hard */ |
d7482997 |
1623 | rdaddsc(rs, c); |
1624 | else |
ce9921d6 |
1625 | xhtml_rdaddwc(rs, text->alt, NULL, indexable); |
d7482997 |
1626 | sfree(c); |
1627 | } else if (removeattr(text->type) == word_WhiteSpace) { |
1628 | rdaddc(rs, ' '); |
1629 | } else if (removeattr(text->type) == word_Quote) { |
1630 | rdaddsc(rs, """); |
1631 | } |
1632 | |
1633 | if (towordstyle(text->type) == word_Emph && |
1634 | (attraux(text->aux) == attr_Last || |
1635 | attraux(text->aux) == attr_Only)) |
1636 | rdaddsc(rs, "</em>"); |
1637 | else if ((towordstyle(text->type) == word_Code || towordstyle(text->type) == word_WeakCode) && |
1638 | (attraux(text->aux) == attr_Last || |
1639 | attraux(text->aux) == attr_Only)) |
1640 | rdaddsc(rs, "</code>"); |
1641 | break; |
1642 | } |
1643 | } |
1644 | } |
1645 | |
1646 | /* Output a heading, formatted as XHTML. |
1647 | */ |
ce9921d6 |
1648 | static void xhtml_heading(FILE *fp, paragraph *p, int indexable) |
d7482997 |
1649 | { |
1650 | rdstringc t = { 0, 0, NULL }; |
1651 | word *tprefix = p->kwtext; |
1652 | word *nprefix = p->kwtext2; |
1653 | word *text = p->words; |
1654 | int level = xhtml_para_level(p); |
1655 | xhtmlsection *sect = xhtml_find_section(p); |
5d9cc07b |
1656 | xhtmlheadfmt *fmt; |
d7482997 |
1657 | char *fragment; |
1658 | if (sect) { |
1659 | fragment = sect->fragment; |
1660 | } else { |
d2e74722 |
1661 | if (p->type == para_Title) |
1662 | fragment = "title"; |
1663 | else { |
1664 | fragment = ""; /* FIXME: what else can we do? */ |
1665 | error(err_whatever, "Couldn't locate heading cross-reference!"); |
1666 | } |
d7482997 |
1667 | } |
1668 | |
5d9cc07b |
1669 | if (p->type == para_Title) |
1670 | fmt = NULL; |
1671 | else if (level == 1) |
1672 | fmt = &conf.fchapter; |
1673 | else if (level-1 < conf.nfsect) |
1674 | fmt = &conf.fsect[level-1]; |
1675 | else |
1676 | fmt = &conf.fsect[conf.nfsect-1]; |
1677 | |
1678 | if (fmt && fmt->just_numbers && nprefix) { |
ce9921d6 |
1679 | xhtml_rdaddwc(&t, nprefix, NULL, indexable); |
5d9cc07b |
1680 | if (fmt) { |
1681 | char *c; |
4b3c5afb |
1682 | if (xhtml_convert(fmt->number_suffix, 0, &c, FALSE)) { |
5d9cc07b |
1683 | rdaddsc(&t, c); |
1684 | sfree(c); |
1685 | } |
1686 | } |
1687 | } else if (fmt && !fmt->just_numbers && tprefix) { |
ce9921d6 |
1688 | xhtml_rdaddwc(&t, tprefix, NULL, indexable); |
5d9cc07b |
1689 | if (fmt) { |
1690 | char *c; |
4b3c5afb |
1691 | if (xhtml_convert(fmt->number_suffix, 0, &c, FALSE)) { |
5d9cc07b |
1692 | rdaddsc(&t, c); |
1693 | sfree(c); |
1694 | } |
1695 | } |
d7482997 |
1696 | } |
ce9921d6 |
1697 | xhtml_rdaddwc(&t, text, NULL, indexable); |
d9d3dd95 |
1698 | /* |
1699 | * If we're outputting in single-file mode, we need to lower |
1700 | * the level of each heading by one, because the overall |
1701 | * document title will be sitting right at the top as an <h1> |
1702 | * and so chapters and sections should start at <h2>. |
1703 | * |
1704 | * Even if not, the document title will come back from |
1705 | * xhtml_para_level() as level zero, so we must increment that |
1706 | * no matter what leaf_level is set to. |
1707 | */ |
1708 | if (conf.leaf_level == 0 || level == 0) |
1709 | level++; |
d7482997 |
1710 | fprintf(fp, "<a name=\"%s\"></a><h%i>%s</h%i>\n", fragment, level, t.text, level); |
1711 | sfree(t.text); |
1712 | } |
1713 | |
1714 | /* Output a paragraph. Styles are handled by xhtml_rdaddwc(). |
1715 | * This looks pretty simple; I may have missed something ... |
1716 | */ |
ce9921d6 |
1717 | static void xhtml_para(FILE *fp, word *text, int indexable) |
d7482997 |
1718 | { |
1719 | rdstringc out = { 0, 0, NULL }; |
ce9921d6 |
1720 | xhtml_rdaddwc(&out, text, NULL, indexable); |
d7482997 |
1721 | fprintf(fp, "%s", out.text); |
1722 | sfree(out.text); |
1723 | } |
1724 | |
1725 | /* Output a code paragraph. I'm treating this as preformatted, which |
1726 | * may not be entirely correct. See xhtml_para() for my worries about |
1727 | * this being overly-simple; however I think that most of the complexity |
1728 | * of the text backend came entirely out of word wrapping anyway. |
1729 | */ |
1730 | static void xhtml_codepara(FILE *fp, word *text) |
1731 | { |
1732 | fprintf(fp, "<pre>"); |
1733 | for (; text; text = text->next) if (text->type == word_WeakCode) { |
4b3c5afb |
1734 | word *here, *next; |
d7482997 |
1735 | char *c; |
4b3c5afb |
1736 | |
1737 | /* |
1738 | * See if this WeakCode is followed by an Emph to indicate |
1739 | * emphasis. |
1740 | */ |
1741 | here = text; |
1742 | if (text->next && text->next->type == word_Emph) { |
1743 | next = text = text->next; |
1744 | } else |
1745 | next = NULL; |
1746 | |
1747 | if (next) { |
1748 | wchar_t *t, *e; |
1749 | int n; |
1750 | |
1751 | t = here->text; |
1752 | e = next->text; |
1753 | |
1754 | while (*e) { |
1755 | int ec = *e; |
1756 | |
1757 | for (n = 0; t[n] && e[n] && e[n] == ec; n++); |
1758 | xhtml_convert(t, n, &c, FALSE); |
1759 | fprintf(fp, "%s%s%s", |
1760 | (ec == 'i' ? "<em>" : ec == 'b' ? "<b>" : ""), |
1761 | c, |
1762 | (ec == 'i' ? "</em>" : ec == 'b' ? "</b>" : "")); |
1763 | sfree(c); |
1764 | |
1765 | t += n; |
1766 | e += n; |
1767 | } |
1768 | |
1769 | xhtml_convert(t, 0, &c, FALSE); |
1770 | fprintf(fp, "%s\n", c); |
1771 | sfree(c); |
1772 | } else { |
1773 | xhtml_convert(here->text, 0, &c, FALSE); |
1774 | fprintf(fp, "%s\n", c); |
1775 | sfree(c); |
1776 | } |
d7482997 |
1777 | } |
1778 | fprintf(fp, "</pre>\n"); |
1779 | } |