|
1 /* This Source Code Form is subject to the terms of the Mozilla Public |
|
2 * License, v. 2.0. If a copy of the MPL was not distributed with this |
|
3 * file, You can obtain one at http://mozilla.org/MPL/2.0/. */ |
|
4 |
|
5 #include "signtool.h" |
|
6 #include <prmem.h> |
|
7 #include <prio.h> |
|
8 #include <prenv.h> |
|
9 |
|
10 static int javascript_fn(char *relpath, char *basedir, char *reldir, |
|
11 char *filename, void *arg); |
|
12 static int extract_js (char *filename); |
|
13 static int copyinto (char *from, char *to); |
|
14 static PRStatus ensureExists (char *base, char *path); |
|
15 static int make_dirs(char *path, PRInt32 file_perms); |
|
16 |
|
17 static char *jartree = NULL; |
|
18 static int idOrdinal; |
|
19 static PRBool dumpParse = PR_FALSE; |
|
20 |
|
21 static char *event_handlers[] = { |
|
22 "onAbort", |
|
23 "onBlur", |
|
24 "onChange", |
|
25 "onClick", |
|
26 "onDblClick", |
|
27 "onDragDrop", |
|
28 "onError", |
|
29 "onFocus", |
|
30 "onKeyDown", |
|
31 "onKeyPress", |
|
32 "onKeyUp", |
|
33 "onLoad", |
|
34 "onMouseDown", |
|
35 "onMouseMove", |
|
36 "onMouseOut", |
|
37 "onMouseOver", |
|
38 "onMouseUp", |
|
39 "onMove", |
|
40 "onReset", |
|
41 "onResize", |
|
42 "onSelect", |
|
43 "onSubmit", |
|
44 "onUnload" |
|
45 }; |
|
46 |
|
47 |
|
48 static int num_handlers = 23; |
|
49 |
|
50 /* |
|
51 * I n l i n e J a v a S c r i p t |
|
52 * |
|
53 * Javascript signing. Instead of passing an archive to signtool, |
|
54 * a directory containing html files is given. Archives are created |
|
55 * from the archive= and src= tag attributes inside the html, |
|
56 * as appropriate. Then the archives are signed. |
|
57 * |
|
58 */ |
|
59 int |
|
60 InlineJavaScript(char *dir, PRBool recurse) |
|
61 { |
|
62 jartree = dir; |
|
63 if (verbosity >= 0) { |
|
64 PR_fprintf(outputFD, "\nGenerating inline signatures from HTML files in: %s\n", |
|
65 dir); |
|
66 } |
|
67 if (PR_GetEnv("SIGNTOOL_DUMP_PARSE")) { |
|
68 dumpParse = PR_TRUE; |
|
69 } |
|
70 |
|
71 return foreach(dir, "", javascript_fn, recurse, PR_FALSE /*include dirs*/, |
|
72 (void * )NULL); |
|
73 |
|
74 } |
|
75 |
|
76 |
|
77 /************************************************************************ |
|
78 * |
|
79 * j a v a s c r i p t _ f n |
|
80 */ |
|
81 static int javascript_fn |
|
82 (char *relpath, char *basedir, char *reldir, char *filename, void *arg) |
|
83 { |
|
84 char fullname [FNSIZE]; |
|
85 |
|
86 /* only process inline scripts from .htm, .html, and .shtml*/ |
|
87 |
|
88 if (!(PL_strcaserstr(filename, ".htm") == filename + strlen(filename) - |
|
89 4) && |
|
90 !(PL_strcaserstr(filename, ".html") == filename + strlen(filename) - |
|
91 5) && |
|
92 !(PL_strcaserstr(filename, ".shtml") == filename + strlen(filename) |
|
93 -6)) { |
|
94 return 0; |
|
95 } |
|
96 |
|
97 /* don't process scripts that signtool has already |
|
98 extracted (those that are inside .arc directories) */ |
|
99 |
|
100 if (PL_strcaserstr(filename, ".arc") == filename + strlen(filename) - 4) |
|
101 return 0; |
|
102 |
|
103 if (verbosity >= 0) { |
|
104 PR_fprintf(outputFD, "Processing HTML file: %s\n", relpath); |
|
105 } |
|
106 |
|
107 /* reset firstArchive at top of each HTML file */ |
|
108 |
|
109 /* skip directories that contain extracted scripts */ |
|
110 |
|
111 if (PL_strcaserstr(reldir, ".arc") == reldir + strlen(reldir) - 4) |
|
112 return 0; |
|
113 |
|
114 sprintf (fullname, "%s/%s", basedir, relpath); |
|
115 return extract_js (fullname); |
|
116 } |
|
117 |
|
118 |
|
119 /*=========================================================================== |
|
120 = |
|
121 = D A T A S T R U C T U R E S |
|
122 = |
|
123 */ |
|
124 typedef enum { |
|
125 TEXT_HTML_STATE = 0, |
|
126 SCRIPT_HTML_STATE |
|
127 } |
|
128 |
|
129 |
|
130 HTML_STATE ; |
|
131 |
|
132 typedef enum { |
|
133 /* we start in the start state */ |
|
134 START_STATE, |
|
135 |
|
136 /* We are looking for or reading in an attribute */ |
|
137 GET_ATT_STATE, |
|
138 |
|
139 /* We're burning ws before finding an attribute */ |
|
140 PRE_ATT_WS_STATE, |
|
141 |
|
142 /* We're burning ws after an attribute. Looking for an '='. */ |
|
143 POST_ATT_WS_STATE, |
|
144 |
|
145 /* We're burning ws after an '=', waiting for a value */ |
|
146 PRE_VAL_WS_STATE, |
|
147 |
|
148 /* We're reading in a value */ |
|
149 GET_VALUE_STATE, |
|
150 |
|
151 /* We're reading in a value that's inside quotes */ |
|
152 GET_QUOTED_VAL_STATE, |
|
153 |
|
154 /* We've encountered the closing '>' */ |
|
155 DONE_STATE, |
|
156 |
|
157 /* Error state */ |
|
158 ERR_STATE |
|
159 } |
|
160 |
|
161 |
|
162 TAG_STATE ; |
|
163 |
|
164 typedef struct AVPair_Str { |
|
165 char *attribute; |
|
166 char *value; |
|
167 unsigned int valueLine; /* the line that the value ends on */ |
|
168 struct AVPair_Str *next; |
|
169 } AVPair; |
|
170 |
|
171 typedef enum { |
|
172 APPLET_TAG, |
|
173 SCRIPT_TAG, |
|
174 LINK_TAG, |
|
175 STYLE_TAG, |
|
176 COMMENT_TAG, |
|
177 OTHER_TAG |
|
178 } |
|
179 |
|
180 |
|
181 TAG_TYPE ; |
|
182 |
|
183 typedef struct { |
|
184 TAG_TYPE type; |
|
185 AVPair * attList; |
|
186 AVPair * attListTail; |
|
187 char *text; |
|
188 } TagItem; |
|
189 |
|
190 typedef enum { |
|
191 TAG_ITEM, |
|
192 TEXT_ITEM |
|
193 } |
|
194 |
|
195 |
|
196 ITEM_TYPE ; |
|
197 |
|
198 typedef struct HTMLItem_Str { |
|
199 unsigned int startLine; |
|
200 unsigned int endLine; |
|
201 ITEM_TYPE type; |
|
202 union { |
|
203 TagItem *tag; |
|
204 char *text; |
|
205 } item; |
|
206 struct HTMLItem_Str *next; |
|
207 } HTMLItem; |
|
208 |
|
209 typedef struct { |
|
210 PRFileDesc *fd; |
|
211 PRInt32 curIndex; |
|
212 PRBool IsEOF; |
|
213 #define FILE_BUFFER_BUFSIZE 512 |
|
214 char buf[FILE_BUFFER_BUFSIZE]; |
|
215 PRInt32 startOffset; |
|
216 PRInt32 maxIndex; |
|
217 unsigned int lineNum; |
|
218 } FileBuffer; |
|
219 |
|
220 /*=========================================================================== |
|
221 = |
|
222 = F U N C T I O N S |
|
223 = |
|
224 */ |
|
225 static HTMLItem*CreateTextItem(char *text, unsigned int startline, |
|
226 unsigned int endline); |
|
227 static HTMLItem*CreateTagItem(TagItem*ti, unsigned int startline, |
|
228 unsigned int endline); |
|
229 static TagItem*ProcessTag(FileBuffer*fb, char **errStr); |
|
230 static void DestroyHTMLItem(HTMLItem *item); |
|
231 static void DestroyTagItem(TagItem*ti); |
|
232 static TAG_TYPE GetTagType(char *att); |
|
233 static FileBuffer*FB_Create(PRFileDesc*fd); |
|
234 static int FB_GetChar(FileBuffer *fb); |
|
235 static PRInt32 FB_GetPointer(FileBuffer *fb); |
|
236 static PRInt32 FB_GetRange(FileBuffer *fb, PRInt32 start, PRInt32 end, |
|
237 char **buf); |
|
238 static unsigned int FB_GetLineNum(FileBuffer *fb); |
|
239 static void FB_Destroy(FileBuffer *fb); |
|
240 static void PrintTagItem(PRFileDesc *fd, TagItem *ti); |
|
241 static void PrintHTMLStream(PRFileDesc *fd, HTMLItem *head); |
|
242 |
|
243 /************************************************************************ |
|
244 * |
|
245 * C r e a t e T e x t I t e m |
|
246 */ |
|
247 static HTMLItem* |
|
248 CreateTextItem(char *text, unsigned int startline, unsigned int endline) |
|
249 { |
|
250 HTMLItem * item; |
|
251 |
|
252 item = PR_Malloc(sizeof(HTMLItem)); |
|
253 if (!item) { |
|
254 return NULL; |
|
255 } |
|
256 |
|
257 item->type = TEXT_ITEM; |
|
258 item->item.text = text; |
|
259 item->next = NULL; |
|
260 item->startLine = startline; |
|
261 item->endLine = endline; |
|
262 |
|
263 return item; |
|
264 } |
|
265 |
|
266 |
|
267 /************************************************************************ |
|
268 * |
|
269 * C r e a t e T a g I t e m |
|
270 */ |
|
271 static HTMLItem* |
|
272 CreateTagItem(TagItem*ti, unsigned int startline, unsigned int endline) |
|
273 { |
|
274 HTMLItem * item; |
|
275 |
|
276 item = PR_Malloc(sizeof(HTMLItem)); |
|
277 if (!item) { |
|
278 return NULL; |
|
279 } |
|
280 |
|
281 item->type = TAG_ITEM; |
|
282 item->item.tag = ti; |
|
283 item->next = NULL; |
|
284 item->startLine = startline; |
|
285 item->endLine = endline; |
|
286 |
|
287 return item; |
|
288 } |
|
289 |
|
290 |
|
291 static PRBool |
|
292 isAttChar(int c) |
|
293 { |
|
294 return (isalnum(c) || c == '/' || c == '-'); |
|
295 } |
|
296 |
|
297 |
|
298 /************************************************************************ |
|
299 * |
|
300 * P r o c e s s T a g |
|
301 */ |
|
302 static TagItem* |
|
303 ProcessTag(FileBuffer*fb, char **errStr) |
|
304 { |
|
305 TAG_STATE state; |
|
306 PRInt32 startText, startID, curPos; |
|
307 PRBool firstAtt; |
|
308 int curchar; |
|
309 TagItem * ti = NULL; |
|
310 AVPair * curPair = NULL; |
|
311 char quotechar = '\0'; |
|
312 unsigned int linenum; |
|
313 unsigned int startline; |
|
314 |
|
315 state = START_STATE; |
|
316 |
|
317 startID = FB_GetPointer(fb); |
|
318 startText = startID; |
|
319 firstAtt = PR_TRUE; |
|
320 |
|
321 ti = (TagItem * ) PR_Malloc(sizeof(TagItem)); |
|
322 if (!ti) |
|
323 out_of_memory(); |
|
324 ti->type = OTHER_TAG; |
|
325 ti->attList = NULL; |
|
326 ti->attListTail = NULL; |
|
327 ti->text = NULL; |
|
328 |
|
329 startline = FB_GetLineNum(fb); |
|
330 |
|
331 while (state != DONE_STATE && state != ERR_STATE) { |
|
332 linenum = FB_GetLineNum(fb); |
|
333 curchar = FB_GetChar(fb); |
|
334 if (curchar == EOF) { |
|
335 *errStr = PR_smprintf( |
|
336 "line %d: Unexpected end-of-file while parsing tag starting at line %d.\n", |
|
337 linenum, startline); |
|
338 state = ERR_STATE; |
|
339 continue; |
|
340 } |
|
341 |
|
342 switch (state) { |
|
343 case START_STATE: |
|
344 if (curchar == '!') { |
|
345 /* |
|
346 * SGML tag or comment |
|
347 * Here's the general rule for SGML tags. Everything from |
|
348 * <! to > is the tag. Inside the tag, comments are |
|
349 * delimited with --. So we are looking for the first '>' |
|
350 * that is not commented out, that is, not inside a pair |
|
351 * of --: <!DOCTYPE --this is a comment >(psyche!) --> |
|
352 */ |
|
353 |
|
354 PRBool inComment = PR_FALSE; |
|
355 short hyphenCount = 0; /* number of consecutive hyphens */ |
|
356 |
|
357 while (1) { |
|
358 linenum = FB_GetLineNum(fb); |
|
359 curchar = FB_GetChar(fb); |
|
360 if (curchar == EOF) { |
|
361 /* Uh oh, EOF inside comment */ |
|
362 *errStr = PR_smprintf( |
|
363 "line %d: Unexpected end-of-file inside comment starting at line %d.\n", |
|
364 linenum, startline); |
|
365 state = ERR_STATE; |
|
366 break; |
|
367 } |
|
368 if (curchar == '-') { |
|
369 if (hyphenCount == 1) { |
|
370 /* This is a comment delimiter */ |
|
371 inComment = !inComment; |
|
372 hyphenCount = 0; |
|
373 } else { |
|
374 /* beginning of a comment delimiter? */ |
|
375 hyphenCount = 1; |
|
376 } |
|
377 } else if (curchar == '>') { |
|
378 if (!inComment) { |
|
379 /* This is the end of the tag */ |
|
380 state = DONE_STATE; |
|
381 break; |
|
382 } else { |
|
383 /* The > is inside a comment, so it's not |
|
384 * really the end of the tag */ |
|
385 hyphenCount = 0; |
|
386 } |
|
387 } else { |
|
388 hyphenCount = 0; |
|
389 } |
|
390 } |
|
391 ti->type = COMMENT_TAG; |
|
392 break; |
|
393 } |
|
394 /* fall through */ |
|
395 case GET_ATT_STATE: |
|
396 if (isspace(curchar) || curchar == '=' || curchar |
|
397 == '>') { |
|
398 /* end of the current attribute */ |
|
399 curPos = FB_GetPointer(fb) - 2; |
|
400 if (curPos >= startID) { |
|
401 /* We have an attribute */ |
|
402 curPair = (AVPair * )PR_Malloc(sizeof(AVPair)); |
|
403 if (!curPair) |
|
404 out_of_memory(); |
|
405 curPair->value = NULL; |
|
406 curPair->next = NULL; |
|
407 FB_GetRange(fb, startID, curPos, |
|
408 &curPair->attribute); |
|
409 |
|
410 /* Stick this attribute on the list */ |
|
411 if (ti->attListTail) { |
|
412 ti->attListTail->next = curPair; |
|
413 ti->attListTail = curPair; |
|
414 } else { |
|
415 ti->attList = ti->attListTail = |
|
416 curPair; |
|
417 } |
|
418 |
|
419 /* If this is the first attribute, find the type of tag |
|
420 * based on it. Also, start saving the text of the tag. */ |
|
421 if (firstAtt) { |
|
422 ti->type = GetTagType(curPair->attribute); |
|
423 startText = FB_GetPointer(fb) |
|
424 -1; |
|
425 firstAtt = PR_FALSE; |
|
426 } |
|
427 } else { |
|
428 if (curchar == '=') { |
|
429 /* If we don't have any attribute but we do have an |
|
430 * equal sign, that's an error */ |
|
431 *errStr = PR_smprintf("line %d: Malformed tag starting at line %d.\n", |
|
432 linenum, startline); |
|
433 state = ERR_STATE; |
|
434 break; |
|
435 } |
|
436 } |
|
437 |
|
438 /* Compute next state */ |
|
439 if (curchar == '=') { |
|
440 startID = FB_GetPointer(fb); |
|
441 state = PRE_VAL_WS_STATE; |
|
442 } else if (curchar == '>') { |
|
443 state = DONE_STATE; |
|
444 } else if (curPair) { |
|
445 state = POST_ATT_WS_STATE; |
|
446 } else { |
|
447 state = PRE_ATT_WS_STATE; |
|
448 } |
|
449 } else if (isAttChar(curchar)) { |
|
450 /* Just another char in the attribute. Do nothing */ |
|
451 state = GET_ATT_STATE; |
|
452 } else { |
|
453 /* bogus char */ |
|
454 *errStr = PR_smprintf("line %d: Bogus chararacter '%c' in tag.\n", |
|
455 linenum, curchar); |
|
456 state = ERR_STATE; |
|
457 break; |
|
458 } |
|
459 break; |
|
460 case PRE_ATT_WS_STATE: |
|
461 if (curchar == '>') { |
|
462 state = DONE_STATE; |
|
463 } else if (isspace(curchar)) { |
|
464 /* more whitespace, do nothing */ |
|
465 } else if (isAttChar(curchar)) { |
|
466 /* starting another attribute */ |
|
467 startID = FB_GetPointer(fb) - 1; |
|
468 state = GET_ATT_STATE; |
|
469 } else { |
|
470 /* bogus char */ |
|
471 *errStr = PR_smprintf("line %d: Bogus character '%c' in tag.\n", |
|
472 linenum, curchar); |
|
473 state = ERR_STATE; |
|
474 break; |
|
475 } |
|
476 break; |
|
477 case POST_ATT_WS_STATE: |
|
478 if (curchar == '>') { |
|
479 state = DONE_STATE; |
|
480 } else if (isspace(curchar)) { |
|
481 /* more whitespace, do nothing */ |
|
482 } else if (isAttChar(curchar)) { |
|
483 /* starting another attribute */ |
|
484 startID = FB_GetPointer(fb) - 1; |
|
485 state = GET_ATT_STATE; |
|
486 } else if (curchar == '=') { |
|
487 /* there was whitespace between the attribute and its equal |
|
488 * sign, which means there's a value coming up */ |
|
489 state = PRE_VAL_WS_STATE; |
|
490 } else { |
|
491 /* bogus char */ |
|
492 *errStr = PR_smprintf("line %d: Bogus character '%c' in tag.\n", |
|
493 linenum, curchar); |
|
494 state = ERR_STATE; |
|
495 break; |
|
496 } |
|
497 break; |
|
498 case PRE_VAL_WS_STATE: |
|
499 if (curchar == '>') { |
|
500 /* premature end-of-tag (sounds like a personal problem). */ |
|
501 *errStr = PR_smprintf( |
|
502 "line %d: End of tag while waiting for value.\n", |
|
503 linenum); |
|
504 state = ERR_STATE; |
|
505 break; |
|
506 } else if (isspace(curchar)) { |
|
507 /* more whitespace, do nothing */ |
|
508 break; |
|
509 } else { |
|
510 /* this must be some sort of value. Fall through |
|
511 * to GET_VALUE_STATE */ |
|
512 startID = FB_GetPointer(fb) - 1; |
|
513 state = GET_VALUE_STATE; |
|
514 } |
|
515 /* Fall through if we didn't break on '>' or whitespace */ |
|
516 case GET_VALUE_STATE: |
|
517 if (isspace(curchar) || curchar == '>') { |
|
518 /* end of value */ |
|
519 curPos = FB_GetPointer(fb) - 2; |
|
520 if (curPos >= startID) { |
|
521 /* Grab the value */ |
|
522 FB_GetRange(fb, startID, curPos, |
|
523 &curPair->value); |
|
524 curPair->valueLine = linenum; |
|
525 } else { |
|
526 /* empty value, leave as NULL */ |
|
527 } |
|
528 if (isspace(curchar)) { |
|
529 state = PRE_ATT_WS_STATE; |
|
530 } else { |
|
531 state = DONE_STATE; |
|
532 } |
|
533 } else if (curchar == '\"' || curchar == '\'') { |
|
534 /* quoted value. Start recording the value inside the quote*/ |
|
535 startID = FB_GetPointer(fb); |
|
536 state = GET_QUOTED_VAL_STATE; |
|
537 PORT_Assert(quotechar == '\0'); |
|
538 quotechar = curchar; /* look for matching quote type */ |
|
539 } else { |
|
540 /* just more value */ |
|
541 } |
|
542 break; |
|
543 case GET_QUOTED_VAL_STATE: |
|
544 PORT_Assert(quotechar != '\0'); |
|
545 if (curchar == quotechar) { |
|
546 /* end of quoted value */ |
|
547 curPos = FB_GetPointer(fb) - 2; |
|
548 if (curPos >= startID) { |
|
549 /* Grab the value */ |
|
550 FB_GetRange(fb, startID, curPos, |
|
551 &curPair->value); |
|
552 curPair->valueLine = linenum; |
|
553 } else { |
|
554 /* empty value, leave it as NULL */ |
|
555 } |
|
556 state = GET_ATT_STATE; |
|
557 quotechar = '\0'; |
|
558 startID = FB_GetPointer(fb); |
|
559 } else { |
|
560 /* more quoted value, continue */ |
|
561 } |
|
562 break; |
|
563 case DONE_STATE: |
|
564 case ERR_STATE: |
|
565 default: |
|
566 ; /* should never get here */ |
|
567 } |
|
568 } |
|
569 |
|
570 if (state == DONE_STATE) { |
|
571 /* Get the text of the tag */ |
|
572 curPos = FB_GetPointer(fb) - 1; |
|
573 FB_GetRange(fb, startText, curPos, &ti->text); |
|
574 |
|
575 /* Return the tag */ |
|
576 return ti; |
|
577 } |
|
578 |
|
579 /* Uh oh, an error. Kill the tag item*/ |
|
580 DestroyTagItem(ti); |
|
581 return NULL; |
|
582 } |
|
583 |
|
584 |
|
585 /************************************************************************ |
|
586 * |
|
587 * D e s t r o y H T M L I t e m |
|
588 */ |
|
589 static void |
|
590 DestroyHTMLItem(HTMLItem *item) |
|
591 { |
|
592 if (item->type == TAG_ITEM) { |
|
593 DestroyTagItem(item->item.tag); |
|
594 } else { |
|
595 if (item->item.text) { |
|
596 PR_Free(item->item.text); |
|
597 } |
|
598 } |
|
599 } |
|
600 |
|
601 |
|
602 /************************************************************************ |
|
603 * |
|
604 * D e s t r o y T a g I t e m |
|
605 */ |
|
606 static void |
|
607 DestroyTagItem(TagItem*ti) |
|
608 { |
|
609 AVPair * temp; |
|
610 |
|
611 if (ti->text) { |
|
612 PR_Free(ti->text); |
|
613 ti->text = NULL; |
|
614 } |
|
615 |
|
616 while (ti->attList) { |
|
617 temp = ti->attList; |
|
618 ti->attList = ti->attList->next; |
|
619 |
|
620 if (temp->attribute) { |
|
621 PR_Free(temp->attribute); |
|
622 temp->attribute = NULL; |
|
623 } |
|
624 if (temp->value) { |
|
625 PR_Free(temp->value); |
|
626 temp->value = NULL; |
|
627 } |
|
628 PR_Free(temp); |
|
629 } |
|
630 |
|
631 PR_Free(ti); |
|
632 } |
|
633 |
|
634 |
|
635 /************************************************************************ |
|
636 * |
|
637 * G e t T a g T y p e |
|
638 */ |
|
639 static TAG_TYPE |
|
640 GetTagType(char *att) |
|
641 { |
|
642 if (!PORT_Strcasecmp(att, "APPLET")) { |
|
643 return APPLET_TAG; |
|
644 } |
|
645 if (!PORT_Strcasecmp(att, "SCRIPT")) { |
|
646 return SCRIPT_TAG; |
|
647 } |
|
648 if (!PORT_Strcasecmp(att, "LINK")) { |
|
649 return LINK_TAG; |
|
650 } |
|
651 if (!PORT_Strcasecmp(att, "STYLE")) { |
|
652 return STYLE_TAG; |
|
653 } |
|
654 return OTHER_TAG; |
|
655 } |
|
656 |
|
657 |
|
658 /************************************************************************ |
|
659 * |
|
660 * F B _ C r e a t e |
|
661 */ |
|
662 static FileBuffer* |
|
663 FB_Create(PRFileDesc*fd) |
|
664 { |
|
665 FileBuffer * fb; |
|
666 PRInt32 amountRead; |
|
667 PRInt32 storedOffset; |
|
668 |
|
669 fb = (FileBuffer * ) PR_Malloc(sizeof(FileBuffer)); |
|
670 fb->fd = fd; |
|
671 storedOffset = PR_Seek(fd, 0, PR_SEEK_CUR); |
|
672 PR_Seek(fd, 0, PR_SEEK_SET); |
|
673 fb->startOffset = 0; |
|
674 amountRead = PR_Read(fd, fb->buf, FILE_BUFFER_BUFSIZE); |
|
675 if (amountRead == -1) |
|
676 goto loser; |
|
677 fb->maxIndex = amountRead - 1; |
|
678 fb->curIndex = 0; |
|
679 fb->IsEOF = (fb->curIndex > fb->maxIndex) ? PR_TRUE : PR_FALSE; |
|
680 fb->lineNum = 1; |
|
681 |
|
682 PR_Seek(fd, storedOffset, PR_SEEK_SET); |
|
683 return fb; |
|
684 loser: |
|
685 PR_Seek(fd, storedOffset, PR_SEEK_SET); |
|
686 PR_Free(fb); |
|
687 return NULL; |
|
688 } |
|
689 |
|
690 |
|
691 /************************************************************************ |
|
692 * |
|
693 * F B _ G e t C h a r |
|
694 */ |
|
695 static int |
|
696 FB_GetChar(FileBuffer *fb) |
|
697 { |
|
698 PRInt32 storedOffset; |
|
699 PRInt32 amountRead; |
|
700 int retval = -1; |
|
701 |
|
702 if (fb->IsEOF) { |
|
703 return EOF; |
|
704 } |
|
705 |
|
706 storedOffset = PR_Seek(fb->fd, 0, PR_SEEK_CUR); |
|
707 |
|
708 retval = (unsigned char) fb->buf[fb->curIndex++]; |
|
709 if (retval == '\n') |
|
710 fb->lineNum++; |
|
711 |
|
712 if (fb->curIndex > fb->maxIndex) { |
|
713 /* We're at the end of the buffer. Try to get some new data from the |
|
714 * file */ |
|
715 fb->startOffset += fb->maxIndex + 1; |
|
716 PR_Seek(fb->fd, fb->startOffset, PR_SEEK_SET); |
|
717 amountRead = PR_Read(fb->fd, fb->buf, FILE_BUFFER_BUFSIZE); |
|
718 if (amountRead == -1) |
|
719 goto loser; |
|
720 fb->maxIndex = amountRead - 1; |
|
721 fb->curIndex = 0; |
|
722 } |
|
723 |
|
724 fb->IsEOF = (fb->curIndex > fb->maxIndex) ? PR_TRUE : PR_FALSE; |
|
725 |
|
726 loser: |
|
727 PR_Seek(fb->fd, storedOffset, PR_SEEK_SET); |
|
728 return retval; |
|
729 } |
|
730 |
|
731 |
|
732 /************************************************************************ |
|
733 * |
|
734 * F B _ G e t L i n e N u m |
|
735 * |
|
736 */ |
|
737 static unsigned int |
|
738 FB_GetLineNum(FileBuffer *fb) |
|
739 { |
|
740 return fb->lineNum; |
|
741 } |
|
742 |
|
743 |
|
744 /************************************************************************ |
|
745 * |
|
746 * F B _ G e t P o i n t e r |
|
747 * |
|
748 */ |
|
749 static PRInt32 |
|
750 FB_GetPointer(FileBuffer *fb) |
|
751 { |
|
752 return fb->startOffset + fb->curIndex; |
|
753 } |
|
754 |
|
755 |
|
756 /************************************************************************ |
|
757 * |
|
758 * F B _ G e t R a n g e |
|
759 * |
|
760 */ |
|
761 static PRInt32 |
|
762 FB_GetRange(FileBuffer *fb, PRInt32 start, PRInt32 end, char **buf) |
|
763 { |
|
764 PRInt32 amountRead; |
|
765 PRInt32 storedOffset; |
|
766 |
|
767 *buf = PR_Malloc(end - start + 2); |
|
768 if (*buf == NULL) { |
|
769 return 0; |
|
770 } |
|
771 |
|
772 storedOffset = PR_Seek(fb->fd, 0, PR_SEEK_CUR); |
|
773 PR_Seek(fb->fd, start, PR_SEEK_SET); |
|
774 amountRead = PR_Read(fb->fd, *buf, end - start + 1); |
|
775 PR_Seek(fb->fd, storedOffset, PR_SEEK_SET); |
|
776 if (amountRead == -1) { |
|
777 PR_Free(*buf); |
|
778 *buf = NULL; |
|
779 return 0; |
|
780 } |
|
781 |
|
782 (*buf)[end-start+1] = '\0'; |
|
783 return amountRead; |
|
784 } |
|
785 |
|
786 |
|
787 /************************************************************************ |
|
788 * |
|
789 * F B _ D e s t r o y |
|
790 * |
|
791 */ |
|
792 static void |
|
793 FB_Destroy(FileBuffer *fb) |
|
794 { |
|
795 if (fb) { |
|
796 PR_Free(fb); |
|
797 } |
|
798 } |
|
799 |
|
800 |
|
801 /************************************************************************ |
|
802 * |
|
803 * P r i n t T a g I t e m |
|
804 * |
|
805 */ |
|
806 static void |
|
807 PrintTagItem(PRFileDesc *fd, TagItem *ti) |
|
808 { |
|
809 AVPair * pair; |
|
810 |
|
811 PR_fprintf(fd, "TAG:\n----\nType: "); |
|
812 switch (ti->type) { |
|
813 case APPLET_TAG: |
|
814 PR_fprintf(fd, "applet\n"); |
|
815 break; |
|
816 case SCRIPT_TAG: |
|
817 PR_fprintf(fd, "script\n"); |
|
818 break; |
|
819 case LINK_TAG: |
|
820 PR_fprintf(fd, "link\n"); |
|
821 break; |
|
822 case STYLE_TAG: |
|
823 PR_fprintf(fd, "style\n"); |
|
824 break; |
|
825 case COMMENT_TAG: |
|
826 PR_fprintf(fd, "comment\n"); |
|
827 break; |
|
828 case OTHER_TAG: |
|
829 default: |
|
830 PR_fprintf(fd, "other\n"); |
|
831 break; |
|
832 } |
|
833 |
|
834 PR_fprintf(fd, "Attributes:\n"); |
|
835 for (pair = ti->attList; pair; pair = pair->next) { |
|
836 PR_fprintf(fd, "\t%s=%s\n", pair->attribute, |
|
837 pair->value ? pair->value : ""); |
|
838 } |
|
839 PR_fprintf(fd, "Text:%s\n", ti->text ? ti->text : ""); |
|
840 |
|
841 PR_fprintf(fd, "---End of tag---\n"); |
|
842 } |
|
843 |
|
844 |
|
845 /************************************************************************ |
|
846 * |
|
847 * P r i n t H T M L S t r e a m |
|
848 * |
|
849 */ |
|
850 static void |
|
851 PrintHTMLStream(PRFileDesc *fd, HTMLItem *head) |
|
852 { |
|
853 while (head) { |
|
854 if (head->type == TAG_ITEM) { |
|
855 PrintTagItem(fd, head->item.tag); |
|
856 } else { |
|
857 PR_fprintf(fd, "\nTEXT:\n-----\n%s\n-----\n\n", head->item.text); |
|
858 } |
|
859 head = head->next; |
|
860 } |
|
861 } |
|
862 |
|
863 |
|
864 /************************************************************************ |
|
865 * |
|
866 * S a v e I n l i n e S c r i p t |
|
867 * |
|
868 */ |
|
869 static int |
|
870 SaveInlineScript(char *text, char *id, char *basedir, char *archiveDir) |
|
871 { |
|
872 char *filename = NULL; |
|
873 PRFileDesc * fd = NULL; |
|
874 int retval = -1; |
|
875 PRInt32 writeLen; |
|
876 char *ilDir = NULL; |
|
877 |
|
878 if (!text || !id || !archiveDir) { |
|
879 return - 1; |
|
880 } |
|
881 |
|
882 if (dumpParse) { |
|
883 PR_fprintf(outputFD, "SaveInlineScript: text=%s, id=%s, \n" |
|
884 "basedir=%s, archiveDir=%s\n", |
|
885 text, id, basedir, archiveDir); |
|
886 } |
|
887 |
|
888 /* Make sure the archive directory is around */ |
|
889 if (ensureExists(basedir, archiveDir) != PR_SUCCESS) { |
|
890 PR_fprintf(errorFD, |
|
891 "ERROR: Unable to create archive directory %s.\n", archiveDir); |
|
892 errorCount++; |
|
893 return - 1; |
|
894 } |
|
895 |
|
896 /* Make sure the inline script directory is around */ |
|
897 ilDir = PR_smprintf("%s/inlineScripts", archiveDir); |
|
898 scriptdir = "inlineScripts"; |
|
899 if (ensureExists(basedir, ilDir) != PR_SUCCESS) { |
|
900 PR_fprintf(errorFD, |
|
901 "ERROR: Unable to create directory %s.\n", ilDir); |
|
902 errorCount++; |
|
903 return - 1; |
|
904 } |
|
905 |
|
906 filename = PR_smprintf("%s/%s/%s", basedir, ilDir, id); |
|
907 |
|
908 /* If the file already exists, give a warning, then blow it away */ |
|
909 if (PR_Access(filename, PR_ACCESS_EXISTS) == PR_SUCCESS) { |
|
910 PR_fprintf(errorFD, |
|
911 "warning: file \"%s\" already exists--will overwrite.\n", |
|
912 filename); |
|
913 warningCount++; |
|
914 if (rm_dash_r(filename)) { |
|
915 PR_fprintf(errorFD, "ERROR: Unable to delete %s.\n", filename); |
|
916 errorCount++; |
|
917 goto finish; |
|
918 } |
|
919 } |
|
920 |
|
921 /* Write text into file with name id */ |
|
922 fd = PR_Open(filename, PR_WRONLY | PR_CREATE_FILE | PR_TRUNCATE, 0777); |
|
923 if (!fd) { |
|
924 PR_fprintf(errorFD, "ERROR: Unable to create file \"%s\".\n", |
|
925 filename); |
|
926 errorCount++; |
|
927 goto finish; |
|
928 } |
|
929 writeLen = strlen(text); |
|
930 if ( PR_Write(fd, text, writeLen) != writeLen) { |
|
931 PR_fprintf(errorFD, "ERROR: Unable to write to file \"%s\".\n", |
|
932 filename); |
|
933 errorCount++; |
|
934 goto finish; |
|
935 } |
|
936 |
|
937 retval = 0; |
|
938 finish: |
|
939 if (filename) { |
|
940 PR_smprintf_free(filename); |
|
941 } |
|
942 if (ilDir) { |
|
943 PR_smprintf_free(ilDir); |
|
944 } |
|
945 if (fd) { |
|
946 PR_Close(fd); |
|
947 } |
|
948 return retval; |
|
949 } |
|
950 |
|
951 |
|
952 /************************************************************************ |
|
953 * |
|
954 * S a v e U n n a m a b l e S c r i p t |
|
955 * |
|
956 */ |
|
957 static int |
|
958 SaveUnnamableScript(char *text, char *basedir, char *archiveDir, |
|
959 char *HTMLfilename) |
|
960 { |
|
961 char *id = NULL; |
|
962 char *ext = NULL; |
|
963 char *start = NULL; |
|
964 int retval = -1; |
|
965 |
|
966 if (!text || !archiveDir || !HTMLfilename) { |
|
967 return - 1; |
|
968 } |
|
969 |
|
970 if (dumpParse) { |
|
971 PR_fprintf(outputFD, "SaveUnnamableScript: text=%s, basedir=%s,\n" |
|
972 "archiveDir=%s, filename=%s\n", text, basedir, archiveDir, |
|
973 HTMLfilename); |
|
974 } |
|
975 |
|
976 /* Construct the filename */ |
|
977 ext = PL_strrchr(HTMLfilename, '.'); |
|
978 if (ext) { |
|
979 *ext = '\0'; |
|
980 } |
|
981 for (start = HTMLfilename; strpbrk(start, "/\\"); |
|
982 start = strpbrk(start, "/\\") + 1) |
|
983 /* do nothing */; |
|
984 if (*start == '\0') |
|
985 start = HTMLfilename; |
|
986 id = PR_smprintf("_%s%d", start, idOrdinal++); |
|
987 if (ext) { |
|
988 *ext = '.'; |
|
989 } |
|
990 |
|
991 /* Now call SaveInlineScript to do the work */ |
|
992 retval = SaveInlineScript(text, id, basedir, archiveDir); |
|
993 |
|
994 PR_Free(id); |
|
995 |
|
996 return retval; |
|
997 } |
|
998 |
|
999 |
|
1000 /************************************************************************ |
|
1001 * |
|
1002 * S a v e S o u r c e |
|
1003 * |
|
1004 */ |
|
1005 static int |
|
1006 SaveSource(char *src, char *codebase, char *basedir, char *archiveDir) |
|
1007 { |
|
1008 char *from = NULL, *to = NULL; |
|
1009 int retval = -1; |
|
1010 char *arcDir = NULL; |
|
1011 |
|
1012 if (!src || !archiveDir) { |
|
1013 return - 1; |
|
1014 } |
|
1015 |
|
1016 if (dumpParse) { |
|
1017 PR_fprintf(outputFD, "SaveSource: src=%s, codebase=%s, basedir=%s,\n" |
|
1018 "archiveDir=%s\n", src, codebase, basedir, archiveDir); |
|
1019 } |
|
1020 |
|
1021 if (codebase) { |
|
1022 arcDir = PR_smprintf("%s/%s/%s/", basedir, codebase, archiveDir); |
|
1023 } else { |
|
1024 arcDir = PR_smprintf("%s/%s/", basedir, archiveDir); |
|
1025 } |
|
1026 |
|
1027 if (codebase) { |
|
1028 from = PR_smprintf("%s/%s/%s", basedir, codebase, src); |
|
1029 to = PR_smprintf("%s%s", arcDir, src); |
|
1030 } else { |
|
1031 from = PR_smprintf("%s/%s", basedir, src); |
|
1032 to = PR_smprintf("%s%s", arcDir, src); |
|
1033 } |
|
1034 |
|
1035 if (make_dirs(to, 0777)) { |
|
1036 PR_fprintf(errorFD, |
|
1037 "ERROR: Unable to create archive directory %s.\n", archiveDir); |
|
1038 errorCount++; |
|
1039 goto finish; |
|
1040 } |
|
1041 |
|
1042 retval = copyinto(from, to); |
|
1043 finish: |
|
1044 if (from) |
|
1045 PR_Free(from); |
|
1046 if (to) |
|
1047 PR_Free(to); |
|
1048 if (arcDir) |
|
1049 PR_Free(arcDir); |
|
1050 return retval; |
|
1051 } |
|
1052 |
|
1053 |
|
1054 /************************************************************************ |
|
1055 * |
|
1056 * T a g T y p e T o S t r i n g |
|
1057 * |
|
1058 */ |
|
1059 char * |
|
1060 TagTypeToString(TAG_TYPE type) |
|
1061 { |
|
1062 switch (type) { |
|
1063 case APPLET_TAG: |
|
1064 return "APPLET"; |
|
1065 case SCRIPT_TAG: |
|
1066 return "SCRIPT"; |
|
1067 case LINK_TAG: |
|
1068 return "LINK"; |
|
1069 case STYLE_TAG: |
|
1070 return "STYLE"; |
|
1071 default: |
|
1072 break; |
|
1073 } |
|
1074 return "unknown"; |
|
1075 } |
|
1076 |
|
1077 |
|
1078 /************************************************************************ |
|
1079 * |
|
1080 * e x t r a c t _ j s |
|
1081 * |
|
1082 */ |
|
1083 static int |
|
1084 extract_js(char *filename) |
|
1085 { |
|
1086 PRFileDesc * fd = NULL; |
|
1087 FileBuffer * fb = NULL; |
|
1088 HTMLItem * head = NULL; |
|
1089 HTMLItem * tail = NULL; |
|
1090 HTMLItem * curitem = NULL; |
|
1091 HTMLItem * styleList = NULL; |
|
1092 HTMLItem * styleListTail = NULL; |
|
1093 HTMLItem * entityList = NULL; |
|
1094 HTMLItem * entityListTail = NULL; |
|
1095 TagItem * tagp = NULL; |
|
1096 char *text = NULL; |
|
1097 char *tagerr = NULL; |
|
1098 char *archiveDir = NULL; |
|
1099 char *firstArchiveDir = NULL; |
|
1100 char *basedir = NULL; |
|
1101 PRInt32 textStart; |
|
1102 PRInt32 curOffset; |
|
1103 HTML_STATE state; |
|
1104 int curchar; |
|
1105 int retval = -1; |
|
1106 unsigned int linenum, startLine; |
|
1107 |
|
1108 /* Initialize the implicit ID counter for each file */ |
|
1109 idOrdinal = 0; |
|
1110 |
|
1111 /* |
|
1112 * First, parse the HTML into a stream of tags and text. |
|
1113 */ |
|
1114 |
|
1115 fd = PR_Open(filename, PR_RDONLY, 0); |
|
1116 if (!fd) { |
|
1117 PR_fprintf(errorFD, "Unable to open %s for reading.\n", filename); |
|
1118 errorCount++; |
|
1119 return - 1; |
|
1120 } |
|
1121 |
|
1122 /* Construct base directory of filename. */ |
|
1123 { |
|
1124 char *cp; |
|
1125 |
|
1126 basedir = PL_strdup(filename); |
|
1127 |
|
1128 /* Remove trailing slashes */ |
|
1129 while ( (cp = PL_strprbrk(basedir, "/\\")) == |
|
1130 (basedir + strlen(basedir) - 1)) { |
|
1131 *cp = '\0'; |
|
1132 } |
|
1133 |
|
1134 /* Now remove everything from the last slash (which will be followed |
|
1135 * by a filename) to the end */ |
|
1136 cp = PL_strprbrk(basedir, "/\\"); |
|
1137 if (cp) { |
|
1138 *cp = '\0'; |
|
1139 } |
|
1140 } |
|
1141 |
|
1142 state = TEXT_HTML_STATE; |
|
1143 |
|
1144 fb = FB_Create(fd); |
|
1145 |
|
1146 textStart = 0; |
|
1147 startLine = 0; |
|
1148 while (linenum = FB_GetLineNum(fb), (curchar = FB_GetChar(fb)) != |
|
1149 EOF) { |
|
1150 switch (state) { |
|
1151 case TEXT_HTML_STATE: |
|
1152 if (curchar == '<') { |
|
1153 /* |
|
1154 * Found a tag |
|
1155 */ |
|
1156 /* Save the text so far to a new text item */ |
|
1157 curOffset = FB_GetPointer(fb) - 2; |
|
1158 if (curOffset >= textStart) { |
|
1159 if (FB_GetRange(fb, textStart, curOffset, |
|
1160 &text) != |
|
1161 curOffset - textStart + 1) { |
|
1162 PR_fprintf(errorFD, |
|
1163 "Unable to read from %s.\n", |
|
1164 filename); |
|
1165 errorCount++; |
|
1166 goto loser; |
|
1167 } |
|
1168 /* little fudge here. If the first character on a line |
|
1169 * is '<', meaning a new tag, the preceding text item |
|
1170 * actually ends on the previous line. In this case |
|
1171 * we will be saying that the text segment ends on the |
|
1172 * next line. I don't think this matters for text items. */ |
|
1173 curitem = CreateTextItem(text, startLine, |
|
1174 linenum); |
|
1175 text = NULL; |
|
1176 if (tail == NULL) { |
|
1177 head = tail = curitem; |
|
1178 } else { |
|
1179 tail->next = curitem; |
|
1180 tail = curitem; |
|
1181 } |
|
1182 } |
|
1183 |
|
1184 /* Process the tag */ |
|
1185 tagp = ProcessTag(fb, &tagerr); |
|
1186 if (!tagp) { |
|
1187 if (tagerr) { |
|
1188 PR_fprintf(errorFD, "Error in file %s: %s\n", |
|
1189 filename, tagerr); |
|
1190 errorCount++; |
|
1191 } else { |
|
1192 PR_fprintf(errorFD, |
|
1193 "Error in file %s, in tag starting at line %d\n", |
|
1194 filename, linenum); |
|
1195 errorCount++; |
|
1196 } |
|
1197 goto loser; |
|
1198 } |
|
1199 /* Add the tag to the list */ |
|
1200 curitem = CreateTagItem(tagp, linenum, FB_GetLineNum(fb)); |
|
1201 if (tail == NULL) { |
|
1202 head = tail = curitem; |
|
1203 } else { |
|
1204 tail->next = curitem; |
|
1205 tail = curitem; |
|
1206 } |
|
1207 |
|
1208 /* What's the next state */ |
|
1209 if (tagp->type == SCRIPT_TAG) { |
|
1210 state = SCRIPT_HTML_STATE; |
|
1211 } |
|
1212 |
|
1213 /* Start recording text from the new offset */ |
|
1214 textStart = FB_GetPointer(fb); |
|
1215 startLine = FB_GetLineNum(fb); |
|
1216 } else { |
|
1217 /* regular character. Next! */ |
|
1218 } |
|
1219 break; |
|
1220 case SCRIPT_HTML_STATE: |
|
1221 if (curchar == '<') { |
|
1222 char *cp; |
|
1223 /* |
|
1224 * If this is a </script> tag, then we're at the end of the |
|
1225 * script. Otherwise, ignore |
|
1226 */ |
|
1227 curOffset = FB_GetPointer(fb) - 1; |
|
1228 cp = NULL; |
|
1229 if (FB_GetRange(fb, curOffset, curOffset + 8, &cp) != 9) { |
|
1230 if (cp) { |
|
1231 PR_Free(cp); |
|
1232 cp = NULL; |
|
1233 } |
|
1234 } else { |
|
1235 /* compare the strings */ |
|
1236 if ( !PORT_Strncasecmp(cp, "</script>", 9) ) { |
|
1237 /* This is the end of the script. Record the text. */ |
|
1238 curOffset--; |
|
1239 if (curOffset >= textStart) { |
|
1240 if (FB_GetRange(fb, textStart, curOffset, &text) != |
|
1241 curOffset - textStart + 1) { |
|
1242 PR_fprintf(errorFD, "Unable to read from %s.\n", |
|
1243 filename); |
|
1244 errorCount++; |
|
1245 goto loser; |
|
1246 } |
|
1247 curitem = CreateTextItem(text, startLine, linenum); |
|
1248 text = NULL; |
|
1249 if (tail == NULL) { |
|
1250 head = tail = curitem; |
|
1251 } else { |
|
1252 tail->next = curitem; |
|
1253 tail = curitem; |
|
1254 } |
|
1255 } |
|
1256 |
|
1257 /* Now parse the /script tag and put it on the list */ |
|
1258 tagp = ProcessTag(fb, &tagerr); |
|
1259 if (!tagp) { |
|
1260 if (tagerr) { |
|
1261 PR_fprintf(errorFD, "Error in file %s: %s\n", |
|
1262 filename, tagerr); |
|
1263 } else { |
|
1264 PR_fprintf(errorFD, |
|
1265 "Error in file %s, in tag starting at" |
|
1266 " line %d\n", filename, linenum); |
|
1267 } |
|
1268 errorCount++; |
|
1269 goto loser; |
|
1270 } |
|
1271 curitem = CreateTagItem(tagp, linenum, |
|
1272 FB_GetLineNum(fb)); |
|
1273 if (tail == NULL) { |
|
1274 head = tail = curitem; |
|
1275 } else { |
|
1276 tail->next = curitem; |
|
1277 tail = curitem; |
|
1278 } |
|
1279 |
|
1280 /* go back to text state */ |
|
1281 state = TEXT_HTML_STATE; |
|
1282 |
|
1283 textStart = FB_GetPointer(fb); |
|
1284 startLine = FB_GetLineNum(fb); |
|
1285 } |
|
1286 } |
|
1287 } |
|
1288 break; |
|
1289 } |
|
1290 } |
|
1291 |
|
1292 /* End of the file. Wrap up any remaining text */ |
|
1293 if (state == SCRIPT_HTML_STATE) { |
|
1294 if (tail && tail->type == TAG_ITEM) { |
|
1295 PR_fprintf(errorFD, "ERROR: <SCRIPT> tag at %s:%d is not followed " |
|
1296 "by a </SCRIPT> tag.\n", filename, tail->startLine); |
|
1297 } else { |
|
1298 PR_fprintf(errorFD, "ERROR: <SCRIPT> tag in file %s is not followed" |
|
1299 " by a </SCRIPT tag.\n", filename); |
|
1300 } |
|
1301 errorCount++; |
|
1302 goto loser; |
|
1303 } |
|
1304 curOffset = FB_GetPointer(fb) - 1; |
|
1305 if (curOffset >= textStart) { |
|
1306 text = NULL; |
|
1307 if ( FB_GetRange(fb, textStart, curOffset, &text) != |
|
1308 curOffset - textStart + 1) { |
|
1309 PR_fprintf(errorFD, "Unable to read from %s.\n", filename); |
|
1310 errorCount++; |
|
1311 goto loser; |
|
1312 } |
|
1313 curitem = CreateTextItem(text, startLine, linenum); |
|
1314 text = NULL; |
|
1315 if (tail == NULL) { |
|
1316 head = tail = curitem; |
|
1317 } else { |
|
1318 tail->next = curitem; |
|
1319 tail = curitem; |
|
1320 } |
|
1321 } |
|
1322 |
|
1323 if (dumpParse) { |
|
1324 PrintHTMLStream(outputFD, head); |
|
1325 } |
|
1326 |
|
1327 /* |
|
1328 * Now we have a stream of tags and text. Go through and deal with each. |
|
1329 */ |
|
1330 for (curitem = head; curitem; curitem = curitem->next) { |
|
1331 TagItem * tagp = NULL; |
|
1332 AVPair * pairp = NULL; |
|
1333 char *src = NULL, *id = NULL, *codebase = NULL; |
|
1334 PRBool hasEventHandler = PR_FALSE; |
|
1335 int i; |
|
1336 |
|
1337 /* Reset archive directory for each tag */ |
|
1338 if (archiveDir) { |
|
1339 PR_Free(archiveDir); |
|
1340 archiveDir = NULL; |
|
1341 } |
|
1342 |
|
1343 /* We only analyze tags */ |
|
1344 if (curitem->type != TAG_ITEM) { |
|
1345 continue; |
|
1346 } |
|
1347 |
|
1348 tagp = curitem->item.tag; |
|
1349 |
|
1350 /* go through the attributes to get information */ |
|
1351 for (pairp = tagp->attList; pairp; pairp = pairp->next) { |
|
1352 |
|
1353 /* ARCHIVE= */ |
|
1354 if ( !PL_strcasecmp(pairp->attribute, "archive")) { |
|
1355 if (archiveDir) { |
|
1356 /* Duplicate attribute. Print warning */ |
|
1357 PR_fprintf(errorFD, |
|
1358 "warning: \"%s\" attribute overwrites previous attribute" |
|
1359 " in tag starting at %s:%d.\n", |
|
1360 pairp->attribute, filename, curitem->startLine); |
|
1361 warningCount++; |
|
1362 PR_Free(archiveDir); |
|
1363 } |
|
1364 archiveDir = PL_strdup(pairp->value); |
|
1365 |
|
1366 /* Substiture ".arc" for ".jar" */ |
|
1367 if ( (PL_strlen(archiveDir) < 4) || |
|
1368 PL_strcasecmp((archiveDir + strlen(archiveDir) -4), |
|
1369 ".jar")) { |
|
1370 PR_fprintf(errorFD, |
|
1371 "warning: ARCHIVE attribute should end in \".jar\" in tag" |
|
1372 " starting on %s:%d.\n", filename, curitem->startLine); |
|
1373 warningCount++; |
|
1374 PR_Free(archiveDir); |
|
1375 archiveDir = PR_smprintf("%s.arc", archiveDir); |
|
1376 } else { |
|
1377 PL_strcpy(archiveDir + strlen(archiveDir) -4, ".arc"); |
|
1378 } |
|
1379 |
|
1380 /* Record the first archive. This will be used later if |
|
1381 * the archive is not specified */ |
|
1382 if (firstArchiveDir == NULL) { |
|
1383 firstArchiveDir = PL_strdup(archiveDir); |
|
1384 } |
|
1385 } |
|
1386 /* CODEBASE= */ |
|
1387 else if ( !PL_strcasecmp(pairp->attribute, "codebase")) { |
|
1388 if (codebase) { |
|
1389 /* Duplicate attribute. Print warning */ |
|
1390 PR_fprintf(errorFD, |
|
1391 "warning: \"%s\" attribute overwrites previous attribute" |
|
1392 " in tag staring at %s:%d.\n", |
|
1393 pairp->attribute, filename, curitem->startLine); |
|
1394 warningCount++; |
|
1395 } |
|
1396 codebase = pairp->value; |
|
1397 } |
|
1398 /* SRC= and HREF= */ |
|
1399 else if ( !PORT_Strcasecmp(pairp->attribute, "src") || |
|
1400 !PORT_Strcasecmp(pairp->attribute, "href") ) { |
|
1401 if (src) { |
|
1402 /* Duplicate attribute. Print warning */ |
|
1403 PR_fprintf(errorFD, |
|
1404 "warning: \"%s\" attribute overwrites previous attribute" |
|
1405 " in tag staring at %s:%d.\n", |
|
1406 pairp->attribute, filename, curitem->startLine); |
|
1407 warningCount++; |
|
1408 } |
|
1409 src = pairp->value; |
|
1410 } |
|
1411 /* CODE= */ |
|
1412 else if (!PORT_Strcasecmp(pairp->attribute, "code") ) { |
|
1413 /*!!!XXX Change PORT to PL all over this code !!! */ |
|
1414 if (src) { |
|
1415 /* Duplicate attribute. Print warning */ |
|
1416 PR_fprintf(errorFD, |
|
1417 "warning: \"%s\" attribute overwrites previous attribute" |
|
1418 " ,in tag staring at %s:%d.\n", |
|
1419 pairp->attribute, filename, curitem->startLine); |
|
1420 warningCount++; |
|
1421 } |
|
1422 src = pairp->value; |
|
1423 |
|
1424 /* Append a .class if one is not already present */ |
|
1425 if ( (PL_strlen(src) < 6) || |
|
1426 PL_strcasecmp( (src + PL_strlen(src) - 6), ".class") ) { |
|
1427 src = PR_smprintf("%s.class", src); |
|
1428 /* Put this string back into the data structure so it |
|
1429 * will be deallocated properly */ |
|
1430 PR_Free(pairp->value); |
|
1431 pairp->value = src; |
|
1432 } |
|
1433 } |
|
1434 /* ID= */ |
|
1435 else if (!PL_strcasecmp(pairp->attribute, "id") ) { |
|
1436 if (id) { |
|
1437 /* Duplicate attribute. Print warning */ |
|
1438 PR_fprintf(errorFD, |
|
1439 "warning: \"%s\" attribute overwrites previous attribute" |
|
1440 " in tag staring at %s:%d.\n", |
|
1441 pairp->attribute, filename, curitem->startLine); |
|
1442 warningCount++; |
|
1443 } |
|
1444 id = pairp->value; |
|
1445 } |
|
1446 |
|
1447 /* STYLE= */ |
|
1448 /* style= attributes, along with JS entities, are stored into |
|
1449 * files with dynamically generated names. The filenames are |
|
1450 * based on the order in which the text is found in the file. |
|
1451 * All JS entities on all lines up to and including the line |
|
1452 * containing the end of the tag that has this style= attribute |
|
1453 * will be processed before this style=attribute. So we need |
|
1454 * to record the line that this _tag_ (not the attribute) ends on. |
|
1455 */ |
|
1456 else if (!PL_strcasecmp(pairp->attribute, "style") && pairp->value) |
|
1457 { |
|
1458 HTMLItem * styleItem; |
|
1459 /* Put this item on the style list */ |
|
1460 styleItem = CreateTextItem(PL_strdup(pairp->value), |
|
1461 curitem->startLine, curitem->endLine); |
|
1462 if (styleListTail == NULL) { |
|
1463 styleList = styleListTail = styleItem; |
|
1464 } else { |
|
1465 styleListTail->next = styleItem; |
|
1466 styleListTail = styleItem; |
|
1467 } |
|
1468 } |
|
1469 /* Event handlers */ |
|
1470 else { |
|
1471 for (i = 0; i < num_handlers; i++) { |
|
1472 if (!PL_strcasecmp(event_handlers[i], pairp->attribute)) { |
|
1473 hasEventHandler = PR_TRUE; |
|
1474 break; |
|
1475 } |
|
1476 } |
|
1477 } |
|
1478 |
|
1479 |
|
1480 /* JS Entity */ |
|
1481 { |
|
1482 char *entityStart, *entityEnd; |
|
1483 HTMLItem * entityItem; |
|
1484 |
|
1485 /* go through each JavaScript entity ( &{...}; ) and store it |
|
1486 * in the entityList. The important thing is to record what |
|
1487 * line number it's on, so we can get it in the right order |
|
1488 * in relation to style= attributes. |
|
1489 * Apparently, these can't flow across lines, so the start and |
|
1490 * end line will be the same. That helps matters. |
|
1491 */ |
|
1492 entityEnd = pairp->value; |
|
1493 while ( entityEnd && |
|
1494 (entityStart = PL_strstr(entityEnd, "&{")) /*}*/ != NULL) { |
|
1495 entityStart += 2; /* point at beginning of actual entity */ |
|
1496 entityEnd = PL_strchr(entityStart, '}'); |
|
1497 if (entityEnd) { |
|
1498 /* Put this item on the entity list */ |
|
1499 *entityEnd = '\0'; |
|
1500 entityItem = CreateTextItem(PL_strdup(entityStart), |
|
1501 pairp->valueLine, pairp->valueLine); |
|
1502 *entityEnd = /* { */ '}'; |
|
1503 if (entityListTail) { |
|
1504 entityListTail->next = entityItem; |
|
1505 entityListTail = entityItem; |
|
1506 } else { |
|
1507 entityList = entityListTail = entityItem; |
|
1508 } |
|
1509 } |
|
1510 } |
|
1511 } |
|
1512 } |
|
1513 |
|
1514 /* If no archive was supplied, we use the first one of the file */ |
|
1515 if (!archiveDir && firstArchiveDir) { |
|
1516 archiveDir = PL_strdup(firstArchiveDir); |
|
1517 } |
|
1518 |
|
1519 /* If we have an event handler, we need to archive this tag */ |
|
1520 if (hasEventHandler) { |
|
1521 if (!id) { |
|
1522 PR_fprintf(errorFD, |
|
1523 "warning: tag starting at %s:%d has event handler but" |
|
1524 " no ID attribute. The tag will not be signed.\n", |
|
1525 filename, curitem->startLine); |
|
1526 warningCount++; |
|
1527 } else if (!archiveDir) { |
|
1528 PR_fprintf(errorFD, |
|
1529 "warning: tag starting at %s:%d has event handler but" |
|
1530 " no ARCHIVE attribute. The tag will not be signed.\n", |
|
1531 filename, curitem->startLine); |
|
1532 warningCount++; |
|
1533 } else { |
|
1534 if (SaveInlineScript(tagp->text, id, basedir, archiveDir)) { |
|
1535 goto loser; |
|
1536 } |
|
1537 } |
|
1538 } |
|
1539 |
|
1540 switch (tagp->type) { |
|
1541 case APPLET_TAG: |
|
1542 if (!src) { |
|
1543 PR_fprintf(errorFD, |
|
1544 "error: APPLET tag starting on %s:%d has no CODE " |
|
1545 "attribute.\n", filename, curitem->startLine); |
|
1546 errorCount++; |
|
1547 goto loser; |
|
1548 } else if (!archiveDir) { |
|
1549 PR_fprintf(errorFD, |
|
1550 "error: APPLET tag starting on %s:%d has no ARCHIVE " |
|
1551 "attribute.\n", filename, curitem->startLine); |
|
1552 errorCount++; |
|
1553 goto loser; |
|
1554 } else { |
|
1555 if (SaveSource(src, codebase, basedir, archiveDir)) { |
|
1556 goto loser; |
|
1557 } |
|
1558 } |
|
1559 break; |
|
1560 case SCRIPT_TAG: |
|
1561 case LINK_TAG: |
|
1562 case STYLE_TAG: |
|
1563 if (!archiveDir) { |
|
1564 PR_fprintf(errorFD, |
|
1565 "error: %s tag starting on %s:%d has no ARCHIVE " |
|
1566 "attribute.\n", TagTypeToString(tagp->type), |
|
1567 filename, curitem->startLine); |
|
1568 errorCount++; |
|
1569 goto loser; |
|
1570 } else if (src) { |
|
1571 if (SaveSource(src, codebase, basedir, archiveDir)) { |
|
1572 goto loser; |
|
1573 } |
|
1574 } else if (id) { |
|
1575 /* Save the next text item */ |
|
1576 if (!curitem->next || (curitem->next->type != |
|
1577 TEXT_ITEM)) { |
|
1578 PR_fprintf(errorFD, |
|
1579 "warning: %s tag starting on %s:%d is not followed" |
|
1580 " by script text.\n", TagTypeToString(tagp->type), |
|
1581 filename, curitem->startLine); |
|
1582 warningCount++; |
|
1583 /* just create empty file */ |
|
1584 if (SaveInlineScript("", id, basedir, archiveDir)) { |
|
1585 goto loser; |
|
1586 } |
|
1587 } else { |
|
1588 curitem = curitem->next; |
|
1589 if (SaveInlineScript(curitem->item.text, |
|
1590 id, basedir, |
|
1591 archiveDir)) { |
|
1592 goto loser; |
|
1593 } |
|
1594 } |
|
1595 } else { |
|
1596 /* No src or id tag--warning */ |
|
1597 PR_fprintf(errorFD, |
|
1598 "warning: %s tag starting on %s:%d has no SRC or" |
|
1599 " ID attributes. Will not sign.\n", |
|
1600 TagTypeToString(tagp->type), filename, curitem->startLine); |
|
1601 warningCount++; |
|
1602 } |
|
1603 break; |
|
1604 default: |
|
1605 /* do nothing for other tags */ |
|
1606 break; |
|
1607 } |
|
1608 |
|
1609 } |
|
1610 |
|
1611 /* Now deal with all the unnamable scripts */ |
|
1612 if (firstArchiveDir) { |
|
1613 HTMLItem * style, *entity; |
|
1614 |
|
1615 /* Go through the lists of JS entities and style attributes. Do them |
|
1616 * in chronological order within a list. Pick the list with the lower |
|
1617 * endLine. In case of a tie, entities come first. |
|
1618 */ |
|
1619 style = styleList; |
|
1620 entity = entityList; |
|
1621 while (style || entity) { |
|
1622 if (!entity || (style && (style->endLine < entity->endLine))) { |
|
1623 /* Process style */ |
|
1624 SaveUnnamableScript(style->item.text, basedir, firstArchiveDir, |
|
1625 filename); |
|
1626 style = style->next; |
|
1627 } else { |
|
1628 /* Process entity */ |
|
1629 SaveUnnamableScript(entity->item.text, basedir, firstArchiveDir, |
|
1630 filename); |
|
1631 entity = entity->next; |
|
1632 } |
|
1633 } |
|
1634 } |
|
1635 |
|
1636 |
|
1637 retval = 0; |
|
1638 loser: |
|
1639 /* Blow away the stream */ |
|
1640 while (head) { |
|
1641 curitem = head; |
|
1642 head = head->next; |
|
1643 DestroyHTMLItem(curitem); |
|
1644 } |
|
1645 while (styleList) { |
|
1646 curitem = styleList; |
|
1647 styleList = styleList->next; |
|
1648 DestroyHTMLItem(curitem); |
|
1649 } |
|
1650 while (entityList) { |
|
1651 curitem = entityList; |
|
1652 entityList = entityList->next; |
|
1653 DestroyHTMLItem(curitem); |
|
1654 } |
|
1655 if (text) { |
|
1656 PR_Free(text); |
|
1657 text = NULL; |
|
1658 } |
|
1659 if (fb) { |
|
1660 FB_Destroy(fb); |
|
1661 fb = NULL; |
|
1662 } |
|
1663 if (fd) { |
|
1664 PR_Close(fd); |
|
1665 } |
|
1666 if (tagerr) { |
|
1667 PR_smprintf_free(tagerr); |
|
1668 tagerr = NULL; |
|
1669 } |
|
1670 if (archiveDir) { |
|
1671 PR_Free(archiveDir); |
|
1672 archiveDir = NULL; |
|
1673 } |
|
1674 if (firstArchiveDir) { |
|
1675 PR_Free(firstArchiveDir); |
|
1676 firstArchiveDir = NULL; |
|
1677 } |
|
1678 return retval; |
|
1679 } |
|
1680 |
|
1681 |
|
1682 /********************************************************************** |
|
1683 * |
|
1684 * e n s u r e E x i s t s |
|
1685 * |
|
1686 * Check for existence of indicated directory. If it doesn't exist, |
|
1687 * it will be created. |
|
1688 * Returns PR_SUCCESS if the directory is present, PR_FAILURE otherwise. |
|
1689 */ |
|
1690 static PRStatus |
|
1691 ensureExists (char *base, char *path) |
|
1692 { |
|
1693 char fn [FNSIZE]; |
|
1694 PRDir * dir; |
|
1695 sprintf (fn, "%s/%s", base, path); |
|
1696 |
|
1697 /*PR_fprintf(outputFD, "Trying to open directory %s.\n", fn);*/ |
|
1698 |
|
1699 if ( (dir = PR_OpenDir(fn)) ) { |
|
1700 PR_CloseDir(dir); |
|
1701 return PR_SUCCESS; |
|
1702 } |
|
1703 return PR_MkDir(fn, 0777); |
|
1704 } |
|
1705 |
|
1706 |
|
1707 /*************************************************************************** |
|
1708 * |
|
1709 * m a k e _ d i r s |
|
1710 * |
|
1711 * Ensure that the directory portion of the path exists. This may require |
|
1712 * making the directory, and its parent, and its parent's parent, etc. |
|
1713 */ |
|
1714 static int |
|
1715 make_dirs(char *path, int file_perms) |
|
1716 { |
|
1717 char *Path; |
|
1718 char *start; |
|
1719 char *sep; |
|
1720 int ret = 0; |
|
1721 PRFileInfo info; |
|
1722 |
|
1723 if (!path) { |
|
1724 return 0; |
|
1725 } |
|
1726 |
|
1727 Path = PL_strdup(path); |
|
1728 start = strpbrk(Path, "/\\"); |
|
1729 if (!start) { |
|
1730 return 0; |
|
1731 } |
|
1732 start++; /* start right after first slash */ |
|
1733 |
|
1734 /* Each time through the loop add one more directory. */ |
|
1735 while ( (sep = strpbrk(start, "/\\")) ) { |
|
1736 *sep = '\0'; |
|
1737 |
|
1738 if ( PR_GetFileInfo(Path, &info) != PR_SUCCESS) { |
|
1739 /* No such dir, we have to create it */ |
|
1740 if ( PR_MkDir(Path, file_perms) != PR_SUCCESS) { |
|
1741 PR_fprintf(errorFD, "ERROR: Unable to create directory %s.\n", |
|
1742 Path); |
|
1743 errorCount++; |
|
1744 ret = -1; |
|
1745 goto loser; |
|
1746 } |
|
1747 } else { |
|
1748 /* something exists by this name, make sure it's a directory */ |
|
1749 if ( info.type != PR_FILE_DIRECTORY ) { |
|
1750 PR_fprintf(errorFD, "ERROR: Unable to create directory %s.\n", |
|
1751 Path); |
|
1752 errorCount++; |
|
1753 ret = -1; |
|
1754 goto loser; |
|
1755 } |
|
1756 } |
|
1757 |
|
1758 start = sep + 1; /* start after the next slash */ |
|
1759 *sep = '/'; |
|
1760 } |
|
1761 |
|
1762 loser: |
|
1763 PR_Free(Path); |
|
1764 return ret; |
|
1765 } |
|
1766 |
|
1767 |
|
1768 /* |
|
1769 * c o p y i n t o |
|
1770 * |
|
1771 * Function to copy file "from" to path "to". |
|
1772 * |
|
1773 */ |
|
1774 static int |
|
1775 copyinto (char *from, char *to) |
|
1776 { |
|
1777 PRInt32 num; |
|
1778 char buf [BUFSIZ]; |
|
1779 PRFileDesc * infp = NULL, *outfp = NULL; |
|
1780 int retval = -1; |
|
1781 |
|
1782 if ((infp = PR_Open(from, PR_RDONLY, 0777)) == NULL) { |
|
1783 PR_fprintf(errorFD, "ERROR: Unable to open \"%s\" for reading.\n", |
|
1784 from); |
|
1785 errorCount++; |
|
1786 goto finish; |
|
1787 } |
|
1788 |
|
1789 /* If to already exists, print a warning before deleting it */ |
|
1790 if (PR_Access(to, PR_ACCESS_EXISTS) == PR_SUCCESS) { |
|
1791 PR_fprintf(errorFD, "warning: %s already exists--will overwrite\n", to); |
|
1792 warningCount++; |
|
1793 if (rm_dash_r(to)) { |
|
1794 PR_fprintf(errorFD, |
|
1795 "ERROR: Unable to remove %s.\n", to); |
|
1796 errorCount++; |
|
1797 goto finish; |
|
1798 } |
|
1799 } |
|
1800 |
|
1801 if ((outfp = PR_Open(to, PR_WRONLY | PR_CREATE_FILE | PR_TRUNCATE, 0777)) |
|
1802 == NULL) { |
|
1803 char *errBuf = NULL; |
|
1804 |
|
1805 errBuf = PR_Malloc(PR_GetErrorTextLength() + 1); |
|
1806 PR_fprintf(errorFD, "ERROR: Unable to open \"%s\" for writing.\n", to); |
|
1807 if (PR_GetErrorText(errBuf)) { |
|
1808 PR_fprintf(errorFD, "Cause: %s\n", errBuf); |
|
1809 } |
|
1810 if (errBuf) { |
|
1811 PR_Free(errBuf); |
|
1812 } |
|
1813 errorCount++; |
|
1814 goto finish; |
|
1815 } |
|
1816 |
|
1817 while ( (num = PR_Read(infp, buf, BUFSIZ)) > 0) { |
|
1818 if (PR_Write(outfp, buf, num) != num) { |
|
1819 PR_fprintf(errorFD, "ERROR: Error writing to %s.\n", to); |
|
1820 errorCount++; |
|
1821 goto finish; |
|
1822 } |
|
1823 } |
|
1824 |
|
1825 retval = 0; |
|
1826 finish: |
|
1827 if (infp) |
|
1828 PR_Close(infp); |
|
1829 if (outfp) |
|
1830 PR_Close(outfp); |
|
1831 |
|
1832 return retval; |
|
1833 } |
|
1834 |
|
1835 |