1 | // ==================================================================== // |
---|
2 | // // |
---|
3 | // File : arb_help2xml.cxx // |
---|
4 | // Purpose : Converts old ARB help format to XML // |
---|
5 | // // |
---|
6 | // Coded by Ralf Westram (coder@reallysoft.de) in October 2001 // |
---|
7 | // Copyright Department of Microbiology (Technical University Munich) // |
---|
8 | // // |
---|
9 | // Visit our web site at: http://www.arb-home.de/ // |
---|
10 | // // |
---|
11 | // ==================================================================== // |
---|
12 | |
---|
13 | #include <xml.hxx> |
---|
14 | |
---|
15 | #include <list> |
---|
16 | #include <set> |
---|
17 | #include <iostream> |
---|
18 | #include <fstream> |
---|
19 | |
---|
20 | #include <cstdlib> |
---|
21 | #include <cstdarg> |
---|
22 | #include <cstring> |
---|
23 | #include <climits> |
---|
24 | |
---|
25 | #include <sys/stat.h> |
---|
26 | |
---|
27 | using namespace std; |
---|
28 | |
---|
29 | #define h2x_assert(bed) arb_assert(bed) |
---|
30 | |
---|
31 | #if defined(DEBUG) |
---|
32 | #define WARN_MISSING_HELP |
---|
33 | #endif // DEBUG |
---|
34 | |
---|
35 | |
---|
36 | // #define DUMP_DATA // use this to see internal data (class Helpfile) |
---|
37 | #define MAX_LINE_LENGTH 200 // maximum length of lines in input stream |
---|
38 | #define TABSIZE 8 |
---|
39 | |
---|
40 | static const char *knownSections[] = { "OCCURRENCE", "DESCRIPTION", "NOTES", "EXAMPLES", "WARNINGS", "BUGS", |
---|
41 | "QUESTION", "ANSWER", "SECTION", |
---|
42 | 0 }; |
---|
43 | |
---|
44 | enum SectionType { |
---|
45 | SEC_OCCURRENCE, |
---|
46 | SEC_DESCRIPTION, |
---|
47 | SEC_NOTES, |
---|
48 | SEC_EXAMPLES, |
---|
49 | SEC_WARNINGS, |
---|
50 | SEC_BUGS, |
---|
51 | SEC_QUESTION, |
---|
52 | SEC_ANSWER, |
---|
53 | SEC_SECTION, |
---|
54 | |
---|
55 | SEC_NONE, |
---|
56 | |
---|
57 | SEC_FAKE, |
---|
58 | }; |
---|
59 | |
---|
60 | |
---|
61 | STATIC_ATTRIBUTED(__ATTR__VFORMAT(1), string vstrf(const char *format, va_list argPtr)) { |
---|
62 | static size_t buf_size = 256; |
---|
63 | static char *buffer = new char[buf_size]; |
---|
64 | |
---|
65 | size_t length; |
---|
66 | while (1) { |
---|
67 | if (!buffer) { |
---|
68 | h2x_assert(buffer); // to stop when debugging |
---|
69 | throw string("out of memory"); |
---|
70 | } |
---|
71 | |
---|
72 | length = vsnprintf(buffer, buf_size, format, argPtr); |
---|
73 | if (length < buf_size) break; // string fits into current buffer |
---|
74 | |
---|
75 | // otherwise resize buffer : |
---|
76 | buf_size += buf_size/2; |
---|
77 | delete [] buffer; |
---|
78 | buffer = new char[buf_size]; |
---|
79 | } |
---|
80 | |
---|
81 | return string(buffer, length); |
---|
82 | } |
---|
83 | |
---|
84 | STATIC_ATTRIBUTED(__ATTR__FORMAT(1), string strf(const char *format, ...)) { |
---|
85 | va_list argPtr; |
---|
86 | va_start(argPtr, format); |
---|
87 | string result = vstrf(format, argPtr); |
---|
88 | va_end(argPtr); |
---|
89 | |
---|
90 | return result; |
---|
91 | } |
---|
92 | |
---|
93 | // -------------------------------------- |
---|
94 | // warnings |
---|
95 | // -------------------------------------- |
---|
96 | |
---|
97 | class LineAttachedMessage { |
---|
98 | string message; |
---|
99 | size_t lineno; |
---|
100 | |
---|
101 | public: |
---|
102 | LineAttachedMessage(const string& message_, size_t lineno_) |
---|
103 | : message(message_) |
---|
104 | , lineno(lineno_) |
---|
105 | {} |
---|
106 | |
---|
107 | const string& Message() const { return message; } |
---|
108 | size_t Lineno() const { return lineno; } |
---|
109 | }; |
---|
110 | |
---|
111 | |
---|
112 | static list<LineAttachedMessage> warnings; |
---|
113 | |
---|
114 | inline LineAttachedMessage make_warning(const string& warning, size_t lineno) { |
---|
115 | return LineAttachedMessage(string("Warning: ")+warning, lineno); |
---|
116 | } |
---|
117 | inline void add_warning(const string& warning, size_t lineno) { |
---|
118 | warnings.push_back(make_warning(warning, lineno)); |
---|
119 | } |
---|
120 | inline void preadd_warning(const string& warning, size_t lineno) { |
---|
121 | LineAttachedMessage line_message = make_warning(warning, lineno); |
---|
122 | if (warnings.size() < 2) { |
---|
123 | warnings.push_front(line_message); |
---|
124 | } |
---|
125 | else { |
---|
126 | LineAttachedMessage prev_message = warnings.back(); |
---|
127 | warnings.pop_back(); |
---|
128 | warnings.push_back(line_message); |
---|
129 | warnings.push_back(prev_message); |
---|
130 | } |
---|
131 | } |
---|
132 | |
---|
133 | // ---------------------- |
---|
134 | // class Reader |
---|
135 | // ---------------------- |
---|
136 | |
---|
137 | class Reader { |
---|
138 | private: |
---|
139 | istream& in; |
---|
140 | char lineBuffer[MAX_LINE_LENGTH]; |
---|
141 | char lineBuffer2[MAX_LINE_LENGTH]; |
---|
142 | bool readAgain; |
---|
143 | bool eof; |
---|
144 | int lineNo; |
---|
145 | |
---|
146 | void getline() { |
---|
147 | if (!eof) { |
---|
148 | if (in.eof()) eof = true; |
---|
149 | else { |
---|
150 | h2x_assert(in.good()); |
---|
151 | |
---|
152 | in.getline(lineBuffer, MAX_LINE_LENGTH); |
---|
153 | lineNo++; |
---|
154 | |
---|
155 | if (in.eof()) eof = true; |
---|
156 | else if (in.fail()) throw "line too long"; |
---|
157 | |
---|
158 | if (strchr(lineBuffer, '\t')) { |
---|
159 | int o2 = 0; |
---|
160 | |
---|
161 | for (int o = 0; lineBuffer[o]; ++o) { |
---|
162 | if (lineBuffer[o] == '\t') { |
---|
163 | int spaces = TABSIZE - (o2 % TABSIZE); |
---|
164 | while (spaces--) lineBuffer2[o2++] = ' '; |
---|
165 | } |
---|
166 | else { |
---|
167 | lineBuffer2[o2++] = lineBuffer[o]; |
---|
168 | } |
---|
169 | } |
---|
170 | lineBuffer2[o2] = 0; |
---|
171 | strcpy(lineBuffer, lineBuffer2); |
---|
172 | } |
---|
173 | |
---|
174 | char *eol = strchr(lineBuffer, 0)-1; |
---|
175 | while (eol >= lineBuffer && eol[0] == ' ') eol--; |
---|
176 | if (eol > lineBuffer) { |
---|
177 | // now eol points to last character |
---|
178 | if (eol[0] == '-' && isalnum(eol[-1])) { |
---|
179 | add_warning("manual hyphenation detected", lineNo); |
---|
180 | } |
---|
181 | } |
---|
182 | } |
---|
183 | } |
---|
184 | } |
---|
185 | |
---|
186 | public: |
---|
187 | Reader(istream& in_) : in(in_), readAgain(true), eof(false), lineNo(0) { getline(); } |
---|
188 | virtual ~Reader() {} |
---|
189 | |
---|
190 | const char *getNext() { |
---|
191 | if (readAgain) readAgain = false; |
---|
192 | else getline(); |
---|
193 | return eof ? 0 : lineBuffer; |
---|
194 | } |
---|
195 | |
---|
196 | void back() { |
---|
197 | h2x_assert(!readAgain); |
---|
198 | readAgain = true; |
---|
199 | } |
---|
200 | |
---|
201 | int getLineNo() const { return lineNo; } |
---|
202 | }; |
---|
203 | |
---|
204 | |
---|
205 | typedef list<string> Strings; |
---|
206 | |
---|
207 | class Section { |
---|
208 | SectionType type; |
---|
209 | Strings content; |
---|
210 | size_t start_lineno; |
---|
211 | |
---|
212 | public: |
---|
213 | Section(SectionType type_, size_t start_lineno_) : type(type_), start_lineno(start_lineno_) {} |
---|
214 | |
---|
215 | const Strings& Content() const { return content; } |
---|
216 | Strings& Content() { return content; } |
---|
217 | |
---|
218 | SectionType get_type() const { return type; } |
---|
219 | |
---|
220 | size_t StartLineno() const { return start_lineno; } |
---|
221 | void set_StartLineno(size_t start_lineno_) { start_lineno = start_lineno_; } |
---|
222 | }; |
---|
223 | |
---|
224 | |
---|
225 | #if defined(WARN_MISSING_HELP) |
---|
226 | void check_TODO(const char *line, const Reader& reader) { |
---|
227 | if (strstr(line, "@@@") != NULL || strstr(line, "TODO") != NULL) { |
---|
228 | string warn = strf("TODO: %s", line); |
---|
229 | add_warning(warn.c_str(), reader.getLineNo()); |
---|
230 | } |
---|
231 | } |
---|
232 | #endif // WARN_MISSING_HELP |
---|
233 | |
---|
234 | // --------------------------- |
---|
235 | // class NamedSection |
---|
236 | // --------------------------- |
---|
237 | class NamedSection { |
---|
238 | private: |
---|
239 | string name; |
---|
240 | Section section; |
---|
241 | |
---|
242 | public: |
---|
243 | NamedSection(const string& name_, const Section& section_) |
---|
244 | : name(name_) |
---|
245 | , section(section_) |
---|
246 | {} |
---|
247 | virtual ~NamedSection() {} |
---|
248 | |
---|
249 | const Section& getSection() const { return section; } |
---|
250 | const string& getName() const { return name; } |
---|
251 | }; |
---|
252 | |
---|
253 | typedef list<NamedSection> NamedSections; |
---|
254 | |
---|
255 | class Link { |
---|
256 | string target; |
---|
257 | size_t source_lineno; |
---|
258 | |
---|
259 | public: |
---|
260 | Link(const string& target_, size_t source_lineno_) |
---|
261 | : target(target_) |
---|
262 | , source_lineno(source_lineno_) |
---|
263 | {} |
---|
264 | |
---|
265 | const string& Target() const { return target; } |
---|
266 | size_t SourceLineno() const { return source_lineno; } |
---|
267 | }; |
---|
268 | |
---|
269 | typedef list<Link> Links; |
---|
270 | |
---|
271 | // ----------------------- |
---|
272 | // class Helpfile |
---|
273 | // ----------------------- |
---|
274 | class Helpfile { |
---|
275 | private: |
---|
276 | Links uplinks; |
---|
277 | Links references; |
---|
278 | Links auto_references; |
---|
279 | Section title; |
---|
280 | NamedSections sections; |
---|
281 | string inputfile; |
---|
282 | |
---|
283 | public: |
---|
284 | Helpfile() : title(SEC_FAKE, -1U) {} |
---|
285 | virtual ~Helpfile() {} |
---|
286 | |
---|
287 | void readHelp(istream& in, const string& filename); |
---|
288 | void writeXML(FILE *out, const string& page_name); |
---|
289 | void extractInternalLinks(); |
---|
290 | |
---|
291 | const Section& get_title() const { return title; } |
---|
292 | }; |
---|
293 | |
---|
294 | inline bool isWhite(char c) { return c == ' '; } |
---|
295 | |
---|
296 | #if defined(DUMP_DATA) |
---|
297 | static void display(const Strings& strings, const string& title, FILE *out) { |
---|
298 | fprintf(out, " %s:\n", title.c_str()); |
---|
299 | for (Strings::const_iterator s = strings.begin(); s != strings.end(); ++s) { |
---|
300 | fprintf(out, " '%s'\n", s->c_str()); |
---|
301 | } |
---|
302 | } |
---|
303 | static void display(const Sections& sections, const string& title, FILE *out) { |
---|
304 | fprintf(out, "%s:\n", title.c_str()); |
---|
305 | for (Sections::const_iterator s = sections.begin(); s != sections.end(); ++s) { |
---|
306 | display(s->second, s->first, out); |
---|
307 | } |
---|
308 | } |
---|
309 | #endif // DUMP_DATA |
---|
310 | |
---|
311 | inline bool isEmptyOrComment(const char *s) { |
---|
312 | if (s[0] == '#') return true; |
---|
313 | for (int off = 0; ; ++off) { |
---|
314 | if (s[off] == 0) return true; |
---|
315 | if (!isWhite(s[off])) break; |
---|
316 | } |
---|
317 | |
---|
318 | return false; |
---|
319 | } |
---|
320 | |
---|
321 | inline const char *extractKeyword(const char *line, string& keyword) { |
---|
322 | // returns NULL if no keyword was found |
---|
323 | // otherwise returns position behind keyword and sets value of 'keyword' |
---|
324 | |
---|
325 | const char *space = strchr(line, ' '); |
---|
326 | if (space && space>line) { |
---|
327 | keyword = string(line, 0, space-line); |
---|
328 | return space; |
---|
329 | } |
---|
330 | else if (!space) { // test for keyword w/o content behind |
---|
331 | if (line[0]) { // not empty |
---|
332 | keyword = line; |
---|
333 | return strchr(line, 0); |
---|
334 | } |
---|
335 | } |
---|
336 | return 0; |
---|
337 | } |
---|
338 | |
---|
339 | inline const char *eatWhite(const char *line) { |
---|
340 | // skips whitespace |
---|
341 | while (isWhite(*line)) ++line; |
---|
342 | return line; |
---|
343 | } |
---|
344 | |
---|
345 | inline void pushParagraph(Section& sec, string& paragraph) { |
---|
346 | if (paragraph.length()) { |
---|
347 | sec.Content().push_back(paragraph); |
---|
348 | paragraph = ""; |
---|
349 | } |
---|
350 | } |
---|
351 | |
---|
352 | inline const char *firstChar(const char *s) { |
---|
353 | while (isWhite(s[0])) ++s; |
---|
354 | return s; |
---|
355 | } |
---|
356 | |
---|
357 | static void parseSection(Section& sec, const char *line, int indentation, Reader& reader) { |
---|
358 | string paragraph = line; |
---|
359 | int lines_in_paragraph = 1; |
---|
360 | |
---|
361 | if (sec.StartLineno() == -1U) { |
---|
362 | sec.set_StartLineno(reader.getLineNo()); |
---|
363 | } |
---|
364 | |
---|
365 | while (1) { |
---|
366 | line = reader.getNext(); |
---|
367 | if (!line) break; |
---|
368 | if (isEmptyOrComment(line)) { |
---|
369 | pushParagraph(sec, paragraph); lines_in_paragraph = 0; |
---|
370 | } |
---|
371 | else { |
---|
372 | string keyword; |
---|
373 | const char *rest = extractKeyword(line, keyword); |
---|
374 | |
---|
375 | if (rest) { // a new keyword |
---|
376 | reader.back(); |
---|
377 | break; |
---|
378 | } |
---|
379 | |
---|
380 | #if defined(WARN_MISSING_HELP) |
---|
381 | check_TODO(line, reader); |
---|
382 | #endif // WARN_MISSING_HELP |
---|
383 | |
---|
384 | string Line = line; |
---|
385 | |
---|
386 | if (sec.get_type() == SEC_OCCURRENCE) { |
---|
387 | pushParagraph(sec, paragraph); lines_in_paragraph = 0; |
---|
388 | } |
---|
389 | else { |
---|
390 | const char *first = firstChar(line); |
---|
391 | if (first[0] == '-') { |
---|
392 | pushParagraph(sec, paragraph); lines_in_paragraph = 0; |
---|
393 | Line[first-line] = ' '; |
---|
394 | } |
---|
395 | } |
---|
396 | |
---|
397 | if (paragraph.length()) { |
---|
398 | paragraph = paragraph+"\n"+Line; |
---|
399 | } |
---|
400 | else { |
---|
401 | paragraph = string("\n")+Line; |
---|
402 | } |
---|
403 | lines_in_paragraph++; |
---|
404 | } |
---|
405 | } |
---|
406 | |
---|
407 | pushParagraph(sec, paragraph); |
---|
408 | |
---|
409 | if (sec.Content().size()>0 && indentation>0) { |
---|
410 | string spaces; |
---|
411 | spaces.reserve(indentation); |
---|
412 | spaces.append(indentation, ' '); |
---|
413 | |
---|
414 | Strings::iterator p = sec.Content().begin(); |
---|
415 | |
---|
416 | *p = string("\n")+spaces+*p; |
---|
417 | } |
---|
418 | } |
---|
419 | inline void check_duplicates(const string& link, const char * /* where */, const Links& existing, bool add_warnings) { |
---|
420 | for (Links::const_iterator ex = existing.begin(); ex != existing.end(); ++ex) { |
---|
421 | if (ex->Target() == link) { |
---|
422 | if (add_warnings) add_warning(strf("First Link to '%s' was found here.", ex->Target().c_str()), ex->SourceLineno()); |
---|
423 | throw strf("Link to '%s' duplicated here.", link.c_str()); |
---|
424 | } |
---|
425 | } |
---|
426 | } |
---|
427 | inline void check_duplicates(const string& link, const Links& uplinks, const Links& references, bool add_warnings) { |
---|
428 | check_duplicates(link, "UP", uplinks, add_warnings); |
---|
429 | check_duplicates(link, "SUB", references, add_warnings); |
---|
430 | } |
---|
431 | |
---|
432 | void Helpfile::readHelp(istream& in, const string& filename) { |
---|
433 | Reader read(in); |
---|
434 | |
---|
435 | inputfile = filename; // remember file read (for comment) |
---|
436 | |
---|
437 | const char *line; |
---|
438 | const char *name_only = strrchr(filename.c_str(), '/'); |
---|
439 | |
---|
440 | h2x_assert(name_only); |
---|
441 | ++name_only; |
---|
442 | |
---|
443 | try { |
---|
444 | while (1) { |
---|
445 | line = read.getNext(); |
---|
446 | if (!line) break; |
---|
447 | |
---|
448 | if (isEmptyOrComment(line)) { |
---|
449 | continue; |
---|
450 | } |
---|
451 | |
---|
452 | #if defined(WARN_MISSING_HELP) |
---|
453 | check_TODO(line, read); |
---|
454 | #endif // WARN_MISSING_HELP |
---|
455 | |
---|
456 | string keyword; |
---|
457 | const char *rest = extractKeyword(line, keyword); |
---|
458 | |
---|
459 | if (rest) { // found a keyword |
---|
460 | if (keyword == "UP") { |
---|
461 | rest = eatWhite(rest); |
---|
462 | if (strlen(rest)) { |
---|
463 | check_duplicates(rest, uplinks, references, true); |
---|
464 | if (strcmp(name_only, rest) == 0) throw "UP link to self"; |
---|
465 | |
---|
466 | uplinks.push_back(Link(rest, read.getLineNo())); |
---|
467 | } |
---|
468 | } |
---|
469 | else if (keyword == "SUB") { |
---|
470 | rest = eatWhite(rest); |
---|
471 | if (strlen(rest)) { |
---|
472 | check_duplicates(rest, uplinks, references, true); |
---|
473 | if (strcmp(name_only, rest) == 0) throw "SUB link to self"; |
---|
474 | |
---|
475 | references.push_back(Link(rest, read.getLineNo())); |
---|
476 | } |
---|
477 | } |
---|
478 | else if (keyword == "TITLE") { |
---|
479 | rest = eatWhite(rest); |
---|
480 | // parseSection(title, rest, rest-line, read); |
---|
481 | parseSection(title, rest, 0, read); |
---|
482 | |
---|
483 | if (title.Content().empty()) throw "empty TITLE not allowed"; |
---|
484 | |
---|
485 | const char *t = title.Content().front().c_str(); |
---|
486 | |
---|
487 | if (strstr(t, "Standard help file form") != 0) { |
---|
488 | throw strf("Illegal title for help file: '%s'", t); |
---|
489 | } |
---|
490 | } |
---|
491 | else { |
---|
492 | if (keyword == "NOTE") keyword = "NOTES"; |
---|
493 | if (keyword == "EXAMPLE") keyword = "EXAMPLES"; |
---|
494 | if (keyword == "WARNING") keyword = "WARNINGS"; |
---|
495 | |
---|
496 | SectionType stype = SEC_NONE; |
---|
497 | int idx; |
---|
498 | for (idx = 0; knownSections[idx]; ++idx) { |
---|
499 | if (knownSections[idx] == keyword) { |
---|
500 | stype = SectionType(idx); |
---|
501 | break; |
---|
502 | } |
---|
503 | } |
---|
504 | |
---|
505 | if (knownSections[idx]) { |
---|
506 | if (stype == SEC_SECTION) { |
---|
507 | string section_name = eatWhite(rest); |
---|
508 | Section sec(stype, read.getLineNo()); |
---|
509 | |
---|
510 | parseSection(sec, "", 0, read); |
---|
511 | sections.push_back(NamedSection(section_name, sec)); |
---|
512 | } |
---|
513 | else { |
---|
514 | Section sec(stype, read.getLineNo()); |
---|
515 | |
---|
516 | rest = eatWhite(rest); |
---|
517 | parseSection(sec, rest, rest-line, read); |
---|
518 | sections.push_back(NamedSection(keyword, sec)); |
---|
519 | } |
---|
520 | } |
---|
521 | else { |
---|
522 | throw strf("unknown keyword '%s'", keyword.c_str()); |
---|
523 | } |
---|
524 | } |
---|
525 | } |
---|
526 | else { |
---|
527 | throw strf("Unhandled line"); |
---|
528 | } |
---|
529 | } |
---|
530 | } |
---|
531 | |
---|
532 | catch (string& err) { throw LineAttachedMessage(err, read.getLineNo()); } |
---|
533 | catch (const char *err) { throw LineAttachedMessage(err, read.getLineNo()); } |
---|
534 | } |
---|
535 | |
---|
536 | static bool shouldReflow(const string& s, int& foundIndentation) { |
---|
537 | // foundIndentation is only valid if shouldReflow() returns true |
---|
538 | enum { START, CHAR, SPACE, MULTIPLE, DOT, DOTSPACE } state = START; |
---|
539 | bool equal_indent = true; |
---|
540 | int lastIndent = -1; |
---|
541 | int thisIndent = 0; |
---|
542 | |
---|
543 | for (string::const_iterator c = s.begin(); c != s.end(); ++c, ++thisIndent) { |
---|
544 | if (*c == '\n') { |
---|
545 | state = START; |
---|
546 | thisIndent = 0; |
---|
547 | } |
---|
548 | else if (isWhite(*c)) { |
---|
549 | if (state == DOT || state == DOTSPACE) state = DOTSPACE; // multiple spaces after DOT are allowed |
---|
550 | else if (state == SPACE) state = MULTIPLE; // now seen multiple spaces |
---|
551 | else if (state == CHAR) state = SPACE; // now seen 1 space |
---|
552 | } |
---|
553 | else { |
---|
554 | if (state == MULTIPLE) return false; // character after multiple spaces |
---|
555 | if (state == START) { |
---|
556 | if (lastIndent == -1) lastIndent = thisIndent; |
---|
557 | else if (lastIndent != thisIndent) equal_indent = false; |
---|
558 | } |
---|
559 | if (*c == '.' || *c == ',') state = DOT; |
---|
560 | else state = CHAR; |
---|
561 | } |
---|
562 | } |
---|
563 | |
---|
564 | if (equal_indent) { |
---|
565 | foundIndentation = lastIndent-1; |
---|
566 | h2x_assert(lastIndent >= 0); |
---|
567 | } |
---|
568 | return equal_indent; |
---|
569 | } |
---|
570 | |
---|
571 | static bool startsWithNumber(string& s, long long &number, bool do_erase = true) { |
---|
572 | // tests if first line starts with 'number.' |
---|
573 | // if true then the number is removed |
---|
574 | |
---|
575 | size_t off = s.find_first_not_of(" \n"); |
---|
576 | if (off == string::npos) return false; |
---|
577 | if (!isdigit(s[off])) return false; |
---|
578 | |
---|
579 | size_t num_start = off; |
---|
580 | number = 0; |
---|
581 | |
---|
582 | for (; isdigit(s[off]); ++off) { |
---|
583 | number = number*10 + (s[off]-'0'); |
---|
584 | } |
---|
585 | |
---|
586 | if (s[off] != '.') return false; |
---|
587 | if (s[off+1] != ' ') return false; |
---|
588 | |
---|
589 | if (do_erase) { |
---|
590 | // remove 'number.' from string : |
---|
591 | for (size_t erase = num_start; erase <= off; ++erase) { |
---|
592 | s[erase] = ' '; |
---|
593 | } |
---|
594 | } |
---|
595 | |
---|
596 | return true; |
---|
597 | } |
---|
598 | |
---|
599 | static int get_first_indentation(const string& s) { |
---|
600 | // returns the indentation of the first line containing other than spaces |
---|
601 | size_t text_start = s.find_first_not_of(" \n"); |
---|
602 | size_t prev_lineend = s.find_last_of('\n', text_start); |
---|
603 | |
---|
604 | if (prev_lineend == string::npos) return text_start; |
---|
605 | return text_start-prev_lineend-1; |
---|
606 | } |
---|
607 | |
---|
608 | static string correctSpaces(const string& text, int change) { |
---|
609 | h2x_assert(text.find('\n') == string::npos); |
---|
610 | |
---|
611 | if (!change) return text; |
---|
612 | |
---|
613 | size_t first = text.find_first_not_of(' '); |
---|
614 | if (first == string::npos) return ""; // empty line |
---|
615 | |
---|
616 | if (change<0) { |
---|
617 | int remove = -change; |
---|
618 | h2x_assert(remove <= int(first)); |
---|
619 | return text.substr(remove); |
---|
620 | } |
---|
621 | |
---|
622 | h2x_assert(change>0); // add spaces |
---|
623 | return string(change, ' ')+text; |
---|
624 | } |
---|
625 | |
---|
626 | static string correctIndentation(const string& text, int change) { |
---|
627 | // removes 'remove' spaces from evry line |
---|
628 | |
---|
629 | size_t this_lineend = text.find('\n'); |
---|
630 | string result; |
---|
631 | |
---|
632 | if (this_lineend == string::npos) { |
---|
633 | result = correctSpaces(text, change); |
---|
634 | } |
---|
635 | else { |
---|
636 | result = correctSpaces(text.substr(0, this_lineend), change); |
---|
637 | |
---|
638 | while (this_lineend != string::npos) { |
---|
639 | size_t next_lineend = text.find('\n', this_lineend+1); |
---|
640 | if (next_lineend == string::npos) { // last line |
---|
641 | result = result+"\n"+correctSpaces(text.substr(this_lineend+1), change); |
---|
642 | } |
---|
643 | else { |
---|
644 | result = result+"\n"+correctSpaces(text.substr(this_lineend+1, next_lineend-this_lineend-1), change); |
---|
645 | } |
---|
646 | this_lineend = next_lineend; |
---|
647 | } |
---|
648 | } |
---|
649 | return result; |
---|
650 | } |
---|
651 | |
---|
652 | inline size_t countSpaces(const string& text) { |
---|
653 | size_t first = text.find_first_not_of(' '); |
---|
654 | if (first == string::npos) return INT_MAX; // empty line |
---|
655 | return first; |
---|
656 | } |
---|
657 | |
---|
658 | static size_t scanMinIndentation(const string& text) { |
---|
659 | size_t this_lineend = text.find('\n'); |
---|
660 | size_t min_indent = INT_MAX; |
---|
661 | |
---|
662 | if (this_lineend == string::npos) { |
---|
663 | min_indent = countSpaces(text); |
---|
664 | } |
---|
665 | else { |
---|
666 | while (this_lineend != string::npos) { |
---|
667 | size_t next_lineend = text.find('\n', this_lineend+1); |
---|
668 | if (next_lineend == string::npos) { |
---|
669 | min_indent = min(min_indent, countSpaces(text.substr(this_lineend+1))); |
---|
670 | } |
---|
671 | else { |
---|
672 | min_indent = min(min_indent, countSpaces(text.substr(this_lineend+1, next_lineend-this_lineend-1))); |
---|
673 | } |
---|
674 | this_lineend = next_lineend; |
---|
675 | } |
---|
676 | } |
---|
677 | |
---|
678 | if (min_indent == INT_MAX) min_indent = 0; // only empty lines |
---|
679 | return min_indent; |
---|
680 | } |
---|
681 | |
---|
682 | // ---------------------------- |
---|
683 | // class ParagraphTree |
---|
684 | // ---------------------------- |
---|
685 | class ParagraphTree : virtual Noncopyable { |
---|
686 | private: |
---|
687 | ParagraphTree *brother; // has same indentation as this |
---|
688 | ParagraphTree *son; // indentation + 1 |
---|
689 | |
---|
690 | bool is_enumerated; // 1., 2., usw. |
---|
691 | long long enumeration; // the value of the enumeration (undefined if !is_enumerated) |
---|
692 | |
---|
693 | bool reflow; // should the paragraph be reflown ? (true if indentation is equal for all lines of text) |
---|
694 | int indentation; // the real indentation of the black (after enumeration was removed) |
---|
695 | |
---|
696 | string text; // text of the Section (containing linefeeds) |
---|
697 | size_t lineNo; // line number where Paragraph starts |
---|
698 | |
---|
699 | ParagraphTree(Strings::const_iterator begin, const Strings::const_iterator end, size_t beginLineNo) { |
---|
700 | h2x_assert(begin != end); |
---|
701 | |
---|
702 | text = *begin; |
---|
703 | son = 0; |
---|
704 | |
---|
705 | enumeration = 0; |
---|
706 | is_enumerated = startsWithNumber(text, enumeration); |
---|
707 | |
---|
708 | lineNo = beginLineNo; |
---|
709 | |
---|
710 | if (is_enumerated) { |
---|
711 | size_t text_start = text.find_first_not_of(" \n"); |
---|
712 | size_t next_linestart = text.find('\n', text_start); |
---|
713 | |
---|
714 | if (next_linestart != string::npos) { |
---|
715 | // more than one line -> set indent of 1st line to indent of 2nd line : |
---|
716 | ++next_linestart; // point behind \n |
---|
717 | size_t ind = get_first_indentation(text.substr(next_linestart)); |
---|
718 | text = string("\n")+string(ind, ' ') + text.substr(text_start); |
---|
719 | } |
---|
720 | } |
---|
721 | |
---|
722 | indentation = 0; |
---|
723 | reflow = shouldReflow(text, indentation); |
---|
724 | |
---|
725 | if (!reflow) { |
---|
726 | size_t reststart = text.find('\n', 1); |
---|
727 | if (reststart != string::npos) { |
---|
728 | int rest_indent = -1; |
---|
729 | string rest = text.substr(reststart); |
---|
730 | bool rest_reflow = shouldReflow(rest, rest_indent); |
---|
731 | |
---|
732 | if (rest_reflow) { |
---|
733 | int first_indent = countSpaces(text.substr(1)); |
---|
734 | size_t last = text.find_last_not_of(' ', reststart-1); |
---|
735 | bool is_header = last != string::npos && text[last] == ':'; |
---|
736 | |
---|
737 | if (!is_header && rest_indent == (first_indent+8)) { |
---|
738 | #if defined(DEBUG) |
---|
739 | size_t textstart = text.find_first_not_of(" \n"); |
---|
740 | h2x_assert(textstart != string::npos); |
---|
741 | #endif // DEBUG |
---|
742 | |
---|
743 | text = text.substr(0, reststart)+correctIndentation(rest, -8); |
---|
744 | reflow = shouldReflow(text, indentation); |
---|
745 | } |
---|
746 | } |
---|
747 | } |
---|
748 | } |
---|
749 | |
---|
750 | if (!reflow) { |
---|
751 | indentation = scanMinIndentation(text); |
---|
752 | } |
---|
753 | |
---|
754 | text = correctIndentation(text, -indentation); |
---|
755 | brother = buildParagraphTree(++begin, end, beginLineNo); |
---|
756 | } |
---|
757 | |
---|
758 | public: |
---|
759 | virtual ~ParagraphTree() { |
---|
760 | delete brother; |
---|
761 | delete son; |
---|
762 | } |
---|
763 | |
---|
764 | size_t countTextNodes() { |
---|
765 | size_t nodes = 1; // this |
---|
766 | if (son) nodes += son->countTextNodes(); |
---|
767 | if (brother) nodes += brother->countTextNodes(); |
---|
768 | return nodes; |
---|
769 | } |
---|
770 | |
---|
771 | void dump(ostream& out) { |
---|
772 | out << "text='" << text << "'\n"; |
---|
773 | out << "is_enumerated='" << is_enumerated << "'"; |
---|
774 | out << "enumeration='" << enumeration << "'"; |
---|
775 | out << "reflow='" << reflow << "'"; |
---|
776 | out << "indentation='" << indentation << "'\n"; |
---|
777 | |
---|
778 | if (brother) { cout << "\nbrother:\n"; brother->dump(out); cout << "\n"; } |
---|
779 | if (son) { cout << "\nson:\n"; son->dump(out); cout << "\n"; } |
---|
780 | } |
---|
781 | |
---|
782 | static ParagraphTree* buildParagraphTree(Strings::const_iterator begin, const Strings::const_iterator end, size_t beginLineNo) { |
---|
783 | if (begin == end) return 0; |
---|
784 | return new ParagraphTree(begin, end, beginLineNo); |
---|
785 | } |
---|
786 | static ParagraphTree* buildParagraphTree(const NamedSection& N) { |
---|
787 | ParagraphTree *ptree = 0; |
---|
788 | const Strings& S = N.getSection().Content(); |
---|
789 | if (S.empty()) throw string("Tried to build an empty ParagraphTree (Section=")+N.getName()+")"; |
---|
790 | else ptree = buildParagraphTree(S.begin(), S.end(), N.getSection().StartLineno()); |
---|
791 | return ptree; |
---|
792 | } |
---|
793 | |
---|
794 | bool contains(ParagraphTree *that) { |
---|
795 | return |
---|
796 | this == that || |
---|
797 | (son && son->contains(that)) || |
---|
798 | (brother && brother->contains(that)); |
---|
799 | } |
---|
800 | |
---|
801 | ParagraphTree *predeccessor(ParagraphTree *before_this) { |
---|
802 | if (brother == before_this) return this; |
---|
803 | if (!brother) return 0; |
---|
804 | return brother->predeccessor(before_this); |
---|
805 | } |
---|
806 | |
---|
807 | void append(ParagraphTree *new_brother) { |
---|
808 | if (!brother) brother = new_brother; |
---|
809 | else brother->append(new_brother); |
---|
810 | } |
---|
811 | |
---|
812 | ParagraphTree* removeTill(ParagraphTree *after) { |
---|
813 | ParagraphTree *removed = this; |
---|
814 | ParagraphTree *after_pred = this; |
---|
815 | |
---|
816 | while (1) { |
---|
817 | h2x_assert(after_pred); |
---|
818 | h2x_assert(after_pred->brother); // removeTill called with non-existing 'after' |
---|
819 | |
---|
820 | if (after_pred->brother == after) { // found after |
---|
821 | after_pred->brother = 0; // unlink |
---|
822 | break; |
---|
823 | } |
---|
824 | after_pred = after_pred->brother; |
---|
825 | } |
---|
826 | |
---|
827 | return removed; |
---|
828 | } |
---|
829 | |
---|
830 | ParagraphTree *firstEnumerated() { |
---|
831 | if (is_enumerated) return this; |
---|
832 | if (brother) return brother->firstEnumerated(); |
---|
833 | return 0; |
---|
834 | } |
---|
835 | ParagraphTree *nextEnumerated() { |
---|
836 | h2x_assert(is_enumerated); |
---|
837 | if (brother) { |
---|
838 | ParagraphTree *next = brother->firstEnumerated(); |
---|
839 | if (next && next->enumeration == (enumeration+1)) { |
---|
840 | return next; |
---|
841 | } |
---|
842 | } |
---|
843 | return 0; |
---|
844 | } |
---|
845 | |
---|
846 | ParagraphTree* firstWithSameOrLowerIndent(int wanted_indentation) { |
---|
847 | if (indentation <= wanted_indentation) return this; |
---|
848 | if (!brother) return 0; |
---|
849 | return brother->firstWithSameOrLowerIndent(wanted_indentation); |
---|
850 | } |
---|
851 | |
---|
852 | ParagraphTree* format_indentations(); |
---|
853 | ParagraphTree* format_enums(); |
---|
854 | private: |
---|
855 | static ParagraphTree* buildNewParagraph(const string& Text, size_t beginLineNo) { |
---|
856 | Strings S; |
---|
857 | S.push_back(Text); |
---|
858 | return new ParagraphTree(S.begin(), S.end(), beginLineNo); |
---|
859 | } |
---|
860 | ParagraphTree *extractEmbeddedEnum(int lookfor) { |
---|
861 | size_t this_lineend = text.find('\n'); |
---|
862 | |
---|
863 | while (this_lineend != string::npos) { |
---|
864 | long long number; |
---|
865 | string embedded = text.substr(this_lineend); |
---|
866 | if (startsWithNumber(embedded, number, false)) { |
---|
867 | if (number == lookfor) { |
---|
868 | text.erase(this_lineend); |
---|
869 | return buildNewParagraph(embedded, lineNo); |
---|
870 | } |
---|
871 | break; |
---|
872 | } |
---|
873 | this_lineend = text.find('\n', this_lineend+1); |
---|
874 | } |
---|
875 | |
---|
876 | return 0; |
---|
877 | } |
---|
878 | public: |
---|
879 | static size_t embeddedCounter; |
---|
880 | |
---|
881 | void xml_write(bool ignore_enumerated = false, bool write_as_entry = false); |
---|
882 | }; |
---|
883 | |
---|
884 | size_t ParagraphTree::embeddedCounter = 0; |
---|
885 | |
---|
886 | ParagraphTree* ParagraphTree::format_enums() { |
---|
887 | // reformats tree such that all enums are brothers |
---|
888 | ParagraphTree *enum_this = firstEnumerated(); |
---|
889 | |
---|
890 | if (enum_this) { // we have enumeration |
---|
891 | ParagraphTree *before_enum = predeccessor(enum_this); |
---|
892 | |
---|
893 | if (before_enum) { |
---|
894 | h2x_assert(before_enum->son == 0); |
---|
895 | before_enum->son = before_enum->brother; |
---|
896 | before_enum->brother = 0; |
---|
897 | } |
---|
898 | |
---|
899 | for (ParagraphTree *enum_next = enum_this->nextEnumerated(); |
---|
900 | enum_next; |
---|
901 | enum_this = enum_next, enum_next = enum_this->nextEnumerated()) |
---|
902 | { |
---|
903 | if (enum_next != enum_this->brother) { |
---|
904 | h2x_assert(enum_this->son == 0); |
---|
905 | enum_this->son = enum_this->brother->removeTill(enum_next); |
---|
906 | enum_this->brother = enum_next; |
---|
907 | } |
---|
908 | } |
---|
909 | |
---|
910 | // enum_this is the last enumeration |
---|
911 | h2x_assert(!enum_this->son); |
---|
912 | |
---|
913 | if (enum_this->brother) { // there are more sections behind enum |
---|
914 | ParagraphTree *after_enum = enum_this->firstWithSameOrLowerIndent(enum_this->indentation-1); |
---|
915 | |
---|
916 | if (after_enum) { // indent should go back after enum |
---|
917 | h2x_assert(!enum_this->son); |
---|
918 | |
---|
919 | if (after_enum != enum_this->brother) { |
---|
920 | enum_this->son = enum_this->brother->removeTill(after_enum); |
---|
921 | } |
---|
922 | enum_this->brother = 0; |
---|
923 | |
---|
924 | h2x_assert(before_enum); |
---|
925 | h2x_assert(before_enum->brother == 0); |
---|
926 | before_enum->brother = after_enum->format_enums(); |
---|
927 | } |
---|
928 | else { // nothing after enum -> take all as children |
---|
929 | h2x_assert(enum_this->son == 0); |
---|
930 | enum_this->son = enum_this->brother; |
---|
931 | enum_this->brother = 0; |
---|
932 | } |
---|
933 | if (enum_this->son) enum_this->son = enum_this->son->format_enums(); |
---|
934 | } |
---|
935 | else { |
---|
936 | if (before_enum) { |
---|
937 | if (before_enum->son) before_enum->son->append(before_enum->brother); |
---|
938 | before_enum->brother = 0; |
---|
939 | } |
---|
940 | } |
---|
941 | |
---|
942 | if (enum_this->enumeration == 1) { // oops - only '1.' search for enum inside block |
---|
943 | ParagraphTree *lookin = enum_this; |
---|
944 | int lookfor; |
---|
945 | |
---|
946 | for (lookfor = 2; ; ++lookfor) { |
---|
947 | ParagraphTree *next_enum = lookin->extractEmbeddedEnum(lookfor); |
---|
948 | if (!next_enum) break; |
---|
949 | |
---|
950 | embeddedCounter++; |
---|
951 | next_enum->brother = lookin->brother; |
---|
952 | lookin->brother = next_enum; |
---|
953 | lookin = next_enum; |
---|
954 | } |
---|
955 | } |
---|
956 | |
---|
957 | return this; |
---|
958 | } |
---|
959 | |
---|
960 | // no enumerations found |
---|
961 | return this; |
---|
962 | } |
---|
963 | |
---|
964 | ParagraphTree* ParagraphTree::format_indentations() { |
---|
965 | if (brother) { |
---|
966 | if (is_enumerated) { |
---|
967 | if (brother) brother = brother->format_indentations(); |
---|
968 | if (son) son = son->format_indentations(); |
---|
969 | } |
---|
970 | else { |
---|
971 | ParagraphTree *same_indent = brother->firstWithSameOrLowerIndent(indentation); |
---|
972 | if (same_indent) { |
---|
973 | if (same_indent == brother) { |
---|
974 | brother = brother->format_indentations(); |
---|
975 | if (son) son = son->format_indentations(); |
---|
976 | } |
---|
977 | else { |
---|
978 | ParagraphTree *children = brother->removeTill(same_indent); |
---|
979 | brother = same_indent->format_indentations(); |
---|
980 | h2x_assert(!son); |
---|
981 | son = children->format_indentations(); |
---|
982 | } |
---|
983 | } |
---|
984 | else { // none with same indent |
---|
985 | h2x_assert(!son); |
---|
986 | son = brother->format_indentations(); |
---|
987 | brother = 0; |
---|
988 | } |
---|
989 | } |
---|
990 | } |
---|
991 | else { // no brother |
---|
992 | if (son) son = son->format_indentations(); |
---|
993 | } |
---|
994 | return this; |
---|
995 | } |
---|
996 | |
---|
997 | // ----------------- |
---|
998 | // LinkType |
---|
999 | |
---|
1000 | enum LinkType { |
---|
1001 | LT_UNKNOWN = 0, |
---|
1002 | LT_HTTP = 1, |
---|
1003 | LT_FTP = 2, |
---|
1004 | LT_FILE = 4, |
---|
1005 | LT_EMAIL = 8, |
---|
1006 | LT_HLP = 16, |
---|
1007 | LT_PS = 32, |
---|
1008 | LT_PDF = 64 |
---|
1009 | }; |
---|
1010 | |
---|
1011 | static const char *link_id[] = { |
---|
1012 | "unknown", |
---|
1013 | "www", |
---|
1014 | "www", |
---|
1015 | "www", |
---|
1016 | "email", |
---|
1017 | "hlp", |
---|
1018 | "ps", |
---|
1019 | "pdf", |
---|
1020 | }; |
---|
1021 | |
---|
1022 | static string LinkType2id(LinkType type) { |
---|
1023 | int idx = 0; |
---|
1024 | while (type >= 1) { |
---|
1025 | idx++; |
---|
1026 | type = LinkType(type>>1); |
---|
1027 | } |
---|
1028 | return link_id[idx]; |
---|
1029 | } |
---|
1030 | |
---|
1031 | inline const char *getExtension(const string& name) { |
---|
1032 | size_t last_dot = name.find_last_of('.'); |
---|
1033 | if (last_dot == string::npos) { |
---|
1034 | return NULL; |
---|
1035 | } |
---|
1036 | return name.c_str()+last_dot+1; |
---|
1037 | } |
---|
1038 | |
---|
1039 | static LinkType detectLinkType(const string& link_target) { |
---|
1040 | LinkType type = LT_UNKNOWN; |
---|
1041 | const char *ext = getExtension(link_target); |
---|
1042 | |
---|
1043 | if (ext && strcasecmp(ext, "hlp") == 0) type = LT_HLP; |
---|
1044 | else if (link_target.find("http://") == 0) type = LT_HTTP; |
---|
1045 | else if (link_target.find("ftp://") == 0) type = LT_FTP; |
---|
1046 | else if (link_target.find("file://") == 0) type = LT_FILE; |
---|
1047 | else if (link_target.find('@') != string::npos) type = LT_EMAIL; |
---|
1048 | else if (ext && strcasecmp(ext, "ps") == 0) type = LT_PS; |
---|
1049 | else if (ext && strcasecmp(ext, "pdf") == 0) type = LT_PDF; |
---|
1050 | |
---|
1051 | return type; |
---|
1052 | } |
---|
1053 | |
---|
1054 | // -------------------------------------------------------------------------------- |
---|
1055 | |
---|
1056 | |
---|
1057 | |
---|
1058 | static string locate_helpfile(const string& helpname) { |
---|
1059 | // search for 'helpname' in various helpfile locations |
---|
1060 | |
---|
1061 | #define PATHS 2 |
---|
1062 | static string path[PATHS] = { "oldhelp/", "genhelp/" }; |
---|
1063 | struct stat st; |
---|
1064 | |
---|
1065 | for (size_t p = 0; p<PATHS; p++) { |
---|
1066 | string fullname = path[p]+helpname; |
---|
1067 | if (stat(fullname.c_str(), &st) == 0) { |
---|
1068 | return fullname; |
---|
1069 | } |
---|
1070 | } |
---|
1071 | return ""; |
---|
1072 | #undef PATHS |
---|
1073 | } |
---|
1074 | |
---|
1075 | static string locate_document(const string& docname) { |
---|
1076 | // search for 'docname' or 'docname.gz' in various helpfile locations |
---|
1077 | |
---|
1078 | string located = locate_helpfile(docname); |
---|
1079 | if (located.empty()) { |
---|
1080 | located = locate_helpfile(docname+".gz"); |
---|
1081 | } |
---|
1082 | return located; |
---|
1083 | } |
---|
1084 | |
---|
1085 | static void add_link_attributes(XML_Tag& link, LinkType type, const string& dest, size_t source_line) { |
---|
1086 | if (type == LT_UNKNOWN) { |
---|
1087 | string msg = string("Invalid link (dest='")+dest+"')"; |
---|
1088 | throw LineAttachedMessage(msg, source_line); |
---|
1089 | } |
---|
1090 | |
---|
1091 | link.add_attribute("dest", dest); |
---|
1092 | link.add_attribute("type", LinkType2id(type)); |
---|
1093 | link.add_attribute("source_line", source_line); |
---|
1094 | |
---|
1095 | if (type&(LT_HLP|LT_PDF|LT_PS)) { // other links (www, email) cannot be checked for existence here |
---|
1096 | string fullhelp = ((type<_HLP) ? locate_helpfile : locate_document)(dest); |
---|
1097 | if (fullhelp.empty()) { |
---|
1098 | link.add_attribute("missing", "1"); |
---|
1099 | string warning = strf("Dead link to '%s'", dest.c_str()); |
---|
1100 | h2x_assert(source_line<1000); // illegal line number ? |
---|
1101 | add_warning(warning, source_line); |
---|
1102 | } |
---|
1103 | } |
---|
1104 | } |
---|
1105 | |
---|
1106 | static void print_XML_Text_expanding_links(const string& text, size_t lineNo) { |
---|
1107 | size_t found = text.find("LINK{", 0); |
---|
1108 | if (found != string::npos) { |
---|
1109 | size_t inside_link = found+5; |
---|
1110 | size_t close = text.find('}', inside_link); |
---|
1111 | |
---|
1112 | if (close != string::npos) { |
---|
1113 | string link_target = text.substr(inside_link, close-inside_link); |
---|
1114 | LinkType type = detectLinkType(link_target); |
---|
1115 | string dest = link_target; |
---|
1116 | |
---|
1117 | // cppcheck-suppress unusedScopedObject |
---|
1118 | XML_Text(text.substr(0, found)); |
---|
1119 | |
---|
1120 | { |
---|
1121 | XML_Tag link("LINK"); |
---|
1122 | link.set_on_extra_line(false); |
---|
1123 | add_link_attributes(link, type, dest, lineNo); |
---|
1124 | } |
---|
1125 | |
---|
1126 | return print_XML_Text_expanding_links(text.substr(close+1), lineNo); |
---|
1127 | } |
---|
1128 | } |
---|
1129 | |
---|
1130 | XML_Text t(text); |
---|
1131 | } |
---|
1132 | |
---|
1133 | void ParagraphTree::xml_write(bool ignore_enumerated, bool write_as_entry) { |
---|
1134 | if (is_enumerated && !ignore_enumerated) { |
---|
1135 | XML_Tag e("ENUM"); |
---|
1136 | xml_write(true); |
---|
1137 | } |
---|
1138 | else { |
---|
1139 | { |
---|
1140 | XML_Tag para(is_enumerated || write_as_entry ? "ENTRY" : "P"); |
---|
1141 | { |
---|
1142 | XML_Tag textblock("T"); |
---|
1143 | textblock.add_attribute("reflow", reflow ? "1" : "0"); |
---|
1144 | { |
---|
1145 | string usedText; |
---|
1146 | if (reflow) { |
---|
1147 | usedText = correctIndentation(text, (textblock.Indent()+1) * the_XML_Document->indentation_per_level); |
---|
1148 | } |
---|
1149 | else { |
---|
1150 | usedText = text; |
---|
1151 | } |
---|
1152 | print_XML_Text_expanding_links(usedText, lineNo); |
---|
1153 | } |
---|
1154 | } |
---|
1155 | if (son) { |
---|
1156 | if (!son->is_enumerated && son->brother) { |
---|
1157 | XML_Tag sontag("LIST"); |
---|
1158 | son->xml_write(false, true); |
---|
1159 | } |
---|
1160 | else { |
---|
1161 | son->xml_write(false); |
---|
1162 | } |
---|
1163 | } |
---|
1164 | } |
---|
1165 | if (brother) brother->xml_write(ignore_enumerated, write_as_entry); |
---|
1166 | } |
---|
1167 | } |
---|
1168 | |
---|
1169 | static void create_top_links(const Links& links, const char *tag) { |
---|
1170 | for (Links::const_iterator s = links.begin(); s != links.end(); ++s) { |
---|
1171 | XML_Tag link(tag); |
---|
1172 | add_link_attributes(link, detectLinkType(s->Target()), s->Target(), s->SourceLineno()); |
---|
1173 | } |
---|
1174 | } |
---|
1175 | |
---|
1176 | void Helpfile::writeXML(FILE *out, const string& page_name) { |
---|
1177 | #if defined(DUMP_DATA) |
---|
1178 | display(uplinks, "Uplinks", stdout); |
---|
1179 | display(references, "References", stdout); |
---|
1180 | display(auto_references, "Auto-References", stdout); |
---|
1181 | display(title, "Title", stdout); |
---|
1182 | display(sections, "Sections", stdout); |
---|
1183 | #endif // DUMP_DATA |
---|
1184 | |
---|
1185 | XML_Document xml("PAGE", "arb_help.dtd", out); |
---|
1186 | |
---|
1187 | xml.skip_empty_tags = true; |
---|
1188 | xml.indentation_per_level = 2; |
---|
1189 | |
---|
1190 | xml.getRoot().add_attribute("name", page_name); |
---|
1191 | #if defined(DEBUG) |
---|
1192 | xml.getRoot().add_attribute("edit_warning", "devel"); // inserts a edit warning into development version |
---|
1193 | #else |
---|
1194 | xml.getRoot().add_attribute("edit_warning", "release"); // inserts a different edit warning into release version |
---|
1195 | #endif // DEBUG |
---|
1196 | |
---|
1197 | { |
---|
1198 | // cppcheck-suppress unusedScopedObject |
---|
1199 | XML_Comment(string("automatically generated from ../")+inputfile+' '); |
---|
1200 | } |
---|
1201 | |
---|
1202 | create_top_links(uplinks, "UP"); |
---|
1203 | create_top_links(references, "SUB"); |
---|
1204 | create_top_links(auto_references, "SUB"); |
---|
1205 | |
---|
1206 | { |
---|
1207 | XML_Tag title_tag("TITLE"); |
---|
1208 | Strings& T = title.Content(); |
---|
1209 | for (Strings::const_iterator s = T.begin(); s != T.end(); ++s) { |
---|
1210 | if (s != T.begin()) { XML_Text text("\n"); } |
---|
1211 | XML_Text text(*s); |
---|
1212 | } |
---|
1213 | } |
---|
1214 | |
---|
1215 | for (NamedSections::const_iterator named_sec = sections.begin(); named_sec != sections.end(); ++named_sec) { |
---|
1216 | XML_Tag section_tag("SECTION"); |
---|
1217 | section_tag.add_attribute("name", named_sec->getName()); |
---|
1218 | |
---|
1219 | ParagraphTree *ptree = ParagraphTree::buildParagraphTree(*named_sec); |
---|
1220 | |
---|
1221 | ParagraphTree::embeddedCounter = 0; |
---|
1222 | |
---|
1223 | #if defined(DEBUG) |
---|
1224 | size_t textnodes = ptree->countTextNodes(); |
---|
1225 | #endif // DEBUG |
---|
1226 | |
---|
1227 | ptree = ptree->format_enums(); |
---|
1228 | |
---|
1229 | #if defined(DEBUG) |
---|
1230 | size_t textnodes2 = ptree->countTextNodes(); |
---|
1231 | h2x_assert(textnodes2 == (textnodes+ParagraphTree::embeddedCounter)); // if this occurs format_enums has an error |
---|
1232 | #endif // DEBUG |
---|
1233 | |
---|
1234 | ptree = ptree->format_indentations(); |
---|
1235 | |
---|
1236 | #if defined(DEBUG) |
---|
1237 | size_t textnodes3 = ptree->countTextNodes(); |
---|
1238 | h2x_assert(textnodes3 == textnodes2); // if this occurs format_indentations has an error |
---|
1239 | #endif // DEBUG |
---|
1240 | |
---|
1241 | ptree->xml_write(); |
---|
1242 | |
---|
1243 | delete ptree; |
---|
1244 | } |
---|
1245 | } |
---|
1246 | |
---|
1247 | void Helpfile::extractInternalLinks() { |
---|
1248 | for (NamedSections::const_iterator named_sec = sections.begin(); named_sec != sections.end(); ++named_sec) { |
---|
1249 | const Section& sec = named_sec->getSection(); |
---|
1250 | try { |
---|
1251 | const Strings& s = sec.Content(); |
---|
1252 | |
---|
1253 | for (Strings::const_iterator li = s.begin(); li != s.end(); ++li) { |
---|
1254 | const string& line = *li; |
---|
1255 | size_t start = 0; |
---|
1256 | |
---|
1257 | while (1) { |
---|
1258 | size_t found = line.find("LINK{", start); |
---|
1259 | if (found == string::npos) break; |
---|
1260 | found += 5; |
---|
1261 | size_t close = line.find('}', found); |
---|
1262 | if (close == string::npos) break; |
---|
1263 | |
---|
1264 | string link_target = line.substr(found, close-found); |
---|
1265 | |
---|
1266 | if (link_target.find("http://") == string::npos && |
---|
1267 | link_target.find("ftp://") == string::npos && |
---|
1268 | link_target.find("file://") == string::npos && |
---|
1269 | link_target.find('@') == string::npos) |
---|
1270 | { |
---|
1271 | try { |
---|
1272 | check_duplicates(link_target, "SUB", references, true); // check only sublinks here |
---|
1273 | check_duplicates(link_target, "UP", uplinks, false); // check only sublinks here |
---|
1274 | check_duplicates(link_target, "AUTO-SUB", auto_references, false); // check only sublinks here |
---|
1275 | auto_references.push_back(Link(link_target, sec.StartLineno())); |
---|
1276 | } |
---|
1277 | catch (string& err) { |
---|
1278 | ; // silently ignore inlined |
---|
1279 | } |
---|
1280 | } |
---|
1281 | start = close+1; |
---|
1282 | } |
---|
1283 | } |
---|
1284 | } |
---|
1285 | catch (string& err) { |
---|
1286 | throw LineAttachedMessage(string("'"+err+"' while scanning LINK{} in SECTION '"+named_sec->getName()+'\''), |
---|
1287 | sec.StartLineno()); |
---|
1288 | } |
---|
1289 | } |
---|
1290 | } |
---|
1291 | |
---|
1292 | static void show_err(const string& err, size_t lineno, const string& helpfile) { |
---|
1293 | if (err.find(helpfile+':') != string::npos) { |
---|
1294 | cerr << err; |
---|
1295 | } |
---|
1296 | else if (lineno == -1U) { |
---|
1297 | cerr << helpfile << ":1: [in unknown line] " << err; |
---|
1298 | } |
---|
1299 | else { |
---|
1300 | cerr << helpfile << ":" << lineno << ": " << err; |
---|
1301 | } |
---|
1302 | cerr << '\n'; |
---|
1303 | } |
---|
1304 | static void show_err(const LineAttachedMessage& line_err, const string& helpfile) { |
---|
1305 | show_err(line_err.Message(), line_err.Lineno(), helpfile); |
---|
1306 | } |
---|
1307 | |
---|
1308 | static void show_warnings(const string& helpfile) { |
---|
1309 | for (list<LineAttachedMessage>::const_iterator wi = warnings.begin(); wi != warnings.end(); ++wi) { |
---|
1310 | show_err(*wi, helpfile); |
---|
1311 | } |
---|
1312 | } |
---|
1313 | |
---|
1314 | static void show_warnings_and_error(const LineAttachedMessage& error, const string& helpfile) { |
---|
1315 | show_warnings(helpfile); |
---|
1316 | show_err(error, helpfile); |
---|
1317 | } |
---|
1318 | |
---|
1319 | static void show_warnings_and_error(const string& error, const string& helpfile) { |
---|
1320 | show_warnings_and_error(LineAttachedMessage(error, -1U), helpfile); |
---|
1321 | } |
---|
1322 | |
---|
1323 | int ARB_main(int argc, const char *argv[]) { |
---|
1324 | Helpfile help; |
---|
1325 | string arb_help; |
---|
1326 | |
---|
1327 | try { |
---|
1328 | if (argc != 3) { |
---|
1329 | cerr << "Usage: arb_help2xml <ARB helpfile> <XML output>\n"; |
---|
1330 | return EXIT_FAILURE; |
---|
1331 | } |
---|
1332 | |
---|
1333 | arb_help = argv[1]; |
---|
1334 | string xml_output = argv[2]; |
---|
1335 | |
---|
1336 | { |
---|
1337 | ifstream in(arb_help.c_str()); |
---|
1338 | help.readHelp(in, arb_help); |
---|
1339 | } |
---|
1340 | |
---|
1341 | help.extractInternalLinks(); |
---|
1342 | |
---|
1343 | { |
---|
1344 | FILE *out = std::fopen(xml_output.c_str(), "wt"); |
---|
1345 | if (!out) throw string("Can't open '")+xml_output+'\''; |
---|
1346 | |
---|
1347 | try { |
---|
1348 | // arb_help contains 'oldhelp/name.hlp' |
---|
1349 | size_t slash = arb_help.find('/'); |
---|
1350 | size_t dot = arb_help.find_last_of('.'); |
---|
1351 | |
---|
1352 | if (slash == string::npos || dot == string::npos) { |
---|
1353 | throw string("parameter <ARB helpfile> has to be in format 'oldhelp/name.hlp' (not '"+arb_help+"')"); |
---|
1354 | } |
---|
1355 | |
---|
1356 | string page_name(arb_help, slash+1, dot-slash-1); |
---|
1357 | help.writeXML(out, page_name); |
---|
1358 | fclose(out); |
---|
1359 | } |
---|
1360 | catch (...) { |
---|
1361 | fclose(out); |
---|
1362 | remove(xml_output.c_str()); |
---|
1363 | throw; |
---|
1364 | } |
---|
1365 | } |
---|
1366 | |
---|
1367 | show_warnings(arb_help); |
---|
1368 | |
---|
1369 | return EXIT_SUCCESS; |
---|
1370 | } |
---|
1371 | catch (LineAttachedMessage& err) { show_warnings_and_error(err, arb_help); } |
---|
1372 | catch (string& err) { show_warnings_and_error(err, arb_help); } |
---|
1373 | catch (const char * err) { show_warnings_and_error(err, arb_help); } |
---|
1374 | catch (...) { show_warnings_and_error("unknown exception in arb_help2xml", arb_help); } |
---|
1375 | |
---|
1376 | return EXIT_FAILURE; |
---|
1377 | } |
---|
1378 | |
---|
1379 | |
---|
1380 | // -------------------------------------------------------------------------------- |
---|
1381 | |
---|
1382 | #ifdef UNIT_TESTS |
---|
1383 | #include <test_unit.h> |
---|
1384 | |
---|
1385 | void TEST_hlp2xml_conversion() { |
---|
1386 | // oldhelp/ad_align.hlp |
---|
1387 | string arb_help = "../../HELP_SOURCE/oldhelp/ad_align.hlp"; |
---|
1388 | ifstream in(arb_help.c_str()); |
---|
1389 | |
---|
1390 | Helpfile help; |
---|
1391 | help.readHelp(in, arb_help); |
---|
1392 | |
---|
1393 | Section title = help.get_title(); |
---|
1394 | const Strings& strings = title.Content(); |
---|
1395 | |
---|
1396 | TEST_ASSERT_EQUAL(strings.front().c_str(), "Alignment Administration"); |
---|
1397 | TEST_ASSERT(strings.size() == 1); |
---|
1398 | } |
---|
1399 | |
---|
1400 | #endif // UNIT_TESTS |
---|