Wed, 31 Dec 2014 07:16:47 +0100
Revert simplistic fix pending revisit of Mozilla integration attempt.
michael@0 | 1 | /* This Source Code Form is subject to the terms of the Mozilla Public |
michael@0 | 2 | * License, v. 2.0. If a copy of the MPL was not distributed with this |
michael@0 | 3 | * file, You can obtain one at http://mozilla.org/MPL/2.0/. */ |
michael@0 | 4 | |
michael@0 | 5 | #include "signtool.h" |
michael@0 | 6 | #include <prmem.h> |
michael@0 | 7 | #include <prio.h> |
michael@0 | 8 | #include <prenv.h> |
michael@0 | 9 | |
michael@0 | 10 | static int javascript_fn(char *relpath, char *basedir, char *reldir, |
michael@0 | 11 | char *filename, void *arg); |
michael@0 | 12 | static int extract_js (char *filename); |
michael@0 | 13 | static int copyinto (char *from, char *to); |
michael@0 | 14 | static PRStatus ensureExists (char *base, char *path); |
michael@0 | 15 | static int make_dirs(char *path, PRInt32 file_perms); |
michael@0 | 16 | |
michael@0 | 17 | static char *jartree = NULL; |
michael@0 | 18 | static int idOrdinal; |
michael@0 | 19 | static PRBool dumpParse = PR_FALSE; |
michael@0 | 20 | |
michael@0 | 21 | static char *event_handlers[] = { |
michael@0 | 22 | "onAbort", |
michael@0 | 23 | "onBlur", |
michael@0 | 24 | "onChange", |
michael@0 | 25 | "onClick", |
michael@0 | 26 | "onDblClick", |
michael@0 | 27 | "onDragDrop", |
michael@0 | 28 | "onError", |
michael@0 | 29 | "onFocus", |
michael@0 | 30 | "onKeyDown", |
michael@0 | 31 | "onKeyPress", |
michael@0 | 32 | "onKeyUp", |
michael@0 | 33 | "onLoad", |
michael@0 | 34 | "onMouseDown", |
michael@0 | 35 | "onMouseMove", |
michael@0 | 36 | "onMouseOut", |
michael@0 | 37 | "onMouseOver", |
michael@0 | 38 | "onMouseUp", |
michael@0 | 39 | "onMove", |
michael@0 | 40 | "onReset", |
michael@0 | 41 | "onResize", |
michael@0 | 42 | "onSelect", |
michael@0 | 43 | "onSubmit", |
michael@0 | 44 | "onUnload" |
michael@0 | 45 | }; |
michael@0 | 46 | |
michael@0 | 47 | |
michael@0 | 48 | static int num_handlers = 23; |
michael@0 | 49 | |
michael@0 | 50 | /* |
michael@0 | 51 | * I n l i n e J a v a S c r i p t |
michael@0 | 52 | * |
michael@0 | 53 | * Javascript signing. Instead of passing an archive to signtool, |
michael@0 | 54 | * a directory containing html files is given. Archives are created |
michael@0 | 55 | * from the archive= and src= tag attributes inside the html, |
michael@0 | 56 | * as appropriate. Then the archives are signed. |
michael@0 | 57 | * |
michael@0 | 58 | */ |
michael@0 | 59 | int |
michael@0 | 60 | InlineJavaScript(char *dir, PRBool recurse) |
michael@0 | 61 | { |
michael@0 | 62 | jartree = dir; |
michael@0 | 63 | if (verbosity >= 0) { |
michael@0 | 64 | PR_fprintf(outputFD, "\nGenerating inline signatures from HTML files in: %s\n", |
michael@0 | 65 | dir); |
michael@0 | 66 | } |
michael@0 | 67 | if (PR_GetEnv("SIGNTOOL_DUMP_PARSE")) { |
michael@0 | 68 | dumpParse = PR_TRUE; |
michael@0 | 69 | } |
michael@0 | 70 | |
michael@0 | 71 | return foreach(dir, "", javascript_fn, recurse, PR_FALSE /*include dirs*/, |
michael@0 | 72 | (void * )NULL); |
michael@0 | 73 | |
michael@0 | 74 | } |
michael@0 | 75 | |
michael@0 | 76 | |
michael@0 | 77 | /************************************************************************ |
michael@0 | 78 | * |
michael@0 | 79 | * j a v a s c r i p t _ f n |
michael@0 | 80 | */ |
michael@0 | 81 | static int javascript_fn |
michael@0 | 82 | (char *relpath, char *basedir, char *reldir, char *filename, void *arg) |
michael@0 | 83 | { |
michael@0 | 84 | char fullname [FNSIZE]; |
michael@0 | 85 | |
michael@0 | 86 | /* only process inline scripts from .htm, .html, and .shtml*/ |
michael@0 | 87 | |
michael@0 | 88 | if (!(PL_strcaserstr(filename, ".htm") == filename + strlen(filename) - |
michael@0 | 89 | 4) && |
michael@0 | 90 | !(PL_strcaserstr(filename, ".html") == filename + strlen(filename) - |
michael@0 | 91 | 5) && |
michael@0 | 92 | !(PL_strcaserstr(filename, ".shtml") == filename + strlen(filename) |
michael@0 | 93 | -6)) { |
michael@0 | 94 | return 0; |
michael@0 | 95 | } |
michael@0 | 96 | |
michael@0 | 97 | /* don't process scripts that signtool has already |
michael@0 | 98 | extracted (those that are inside .arc directories) */ |
michael@0 | 99 | |
michael@0 | 100 | if (PL_strcaserstr(filename, ".arc") == filename + strlen(filename) - 4) |
michael@0 | 101 | return 0; |
michael@0 | 102 | |
michael@0 | 103 | if (verbosity >= 0) { |
michael@0 | 104 | PR_fprintf(outputFD, "Processing HTML file: %s\n", relpath); |
michael@0 | 105 | } |
michael@0 | 106 | |
michael@0 | 107 | /* reset firstArchive at top of each HTML file */ |
michael@0 | 108 | |
michael@0 | 109 | /* skip directories that contain extracted scripts */ |
michael@0 | 110 | |
michael@0 | 111 | if (PL_strcaserstr(reldir, ".arc") == reldir + strlen(reldir) - 4) |
michael@0 | 112 | return 0; |
michael@0 | 113 | |
michael@0 | 114 | sprintf (fullname, "%s/%s", basedir, relpath); |
michael@0 | 115 | return extract_js (fullname); |
michael@0 | 116 | } |
michael@0 | 117 | |
michael@0 | 118 | |
michael@0 | 119 | /*=========================================================================== |
michael@0 | 120 | = |
michael@0 | 121 | = D A T A S T R U C T U R E S |
michael@0 | 122 | = |
michael@0 | 123 | */ |
michael@0 | 124 | typedef enum { |
michael@0 | 125 | TEXT_HTML_STATE = 0, |
michael@0 | 126 | SCRIPT_HTML_STATE |
michael@0 | 127 | } |
michael@0 | 128 | |
michael@0 | 129 | |
michael@0 | 130 | HTML_STATE ; |
michael@0 | 131 | |
michael@0 | 132 | typedef enum { |
michael@0 | 133 | /* we start in the start state */ |
michael@0 | 134 | START_STATE, |
michael@0 | 135 | |
michael@0 | 136 | /* We are looking for or reading in an attribute */ |
michael@0 | 137 | GET_ATT_STATE, |
michael@0 | 138 | |
michael@0 | 139 | /* We're burning ws before finding an attribute */ |
michael@0 | 140 | PRE_ATT_WS_STATE, |
michael@0 | 141 | |
michael@0 | 142 | /* We're burning ws after an attribute. Looking for an '='. */ |
michael@0 | 143 | POST_ATT_WS_STATE, |
michael@0 | 144 | |
michael@0 | 145 | /* We're burning ws after an '=', waiting for a value */ |
michael@0 | 146 | PRE_VAL_WS_STATE, |
michael@0 | 147 | |
michael@0 | 148 | /* We're reading in a value */ |
michael@0 | 149 | GET_VALUE_STATE, |
michael@0 | 150 | |
michael@0 | 151 | /* We're reading in a value that's inside quotes */ |
michael@0 | 152 | GET_QUOTED_VAL_STATE, |
michael@0 | 153 | |
michael@0 | 154 | /* We've encountered the closing '>' */ |
michael@0 | 155 | DONE_STATE, |
michael@0 | 156 | |
michael@0 | 157 | /* Error state */ |
michael@0 | 158 | ERR_STATE |
michael@0 | 159 | } |
michael@0 | 160 | |
michael@0 | 161 | |
michael@0 | 162 | TAG_STATE ; |
michael@0 | 163 | |
michael@0 | 164 | typedef struct AVPair_Str { |
michael@0 | 165 | char *attribute; |
michael@0 | 166 | char *value; |
michael@0 | 167 | unsigned int valueLine; /* the line that the value ends on */ |
michael@0 | 168 | struct AVPair_Str *next; |
michael@0 | 169 | } AVPair; |
michael@0 | 170 | |
michael@0 | 171 | typedef enum { |
michael@0 | 172 | APPLET_TAG, |
michael@0 | 173 | SCRIPT_TAG, |
michael@0 | 174 | LINK_TAG, |
michael@0 | 175 | STYLE_TAG, |
michael@0 | 176 | COMMENT_TAG, |
michael@0 | 177 | OTHER_TAG |
michael@0 | 178 | } |
michael@0 | 179 | |
michael@0 | 180 | |
michael@0 | 181 | TAG_TYPE ; |
michael@0 | 182 | |
michael@0 | 183 | typedef struct { |
michael@0 | 184 | TAG_TYPE type; |
michael@0 | 185 | AVPair * attList; |
michael@0 | 186 | AVPair * attListTail; |
michael@0 | 187 | char *text; |
michael@0 | 188 | } TagItem; |
michael@0 | 189 | |
michael@0 | 190 | typedef enum { |
michael@0 | 191 | TAG_ITEM, |
michael@0 | 192 | TEXT_ITEM |
michael@0 | 193 | } |
michael@0 | 194 | |
michael@0 | 195 | |
michael@0 | 196 | ITEM_TYPE ; |
michael@0 | 197 | |
michael@0 | 198 | typedef struct HTMLItem_Str { |
michael@0 | 199 | unsigned int startLine; |
michael@0 | 200 | unsigned int endLine; |
michael@0 | 201 | ITEM_TYPE type; |
michael@0 | 202 | union { |
michael@0 | 203 | TagItem *tag; |
michael@0 | 204 | char *text; |
michael@0 | 205 | } item; |
michael@0 | 206 | struct HTMLItem_Str *next; |
michael@0 | 207 | } HTMLItem; |
michael@0 | 208 | |
michael@0 | 209 | typedef struct { |
michael@0 | 210 | PRFileDesc *fd; |
michael@0 | 211 | PRInt32 curIndex; |
michael@0 | 212 | PRBool IsEOF; |
michael@0 | 213 | #define FILE_BUFFER_BUFSIZE 512 |
michael@0 | 214 | char buf[FILE_BUFFER_BUFSIZE]; |
michael@0 | 215 | PRInt32 startOffset; |
michael@0 | 216 | PRInt32 maxIndex; |
michael@0 | 217 | unsigned int lineNum; |
michael@0 | 218 | } FileBuffer; |
michael@0 | 219 | |
michael@0 | 220 | /*=========================================================================== |
michael@0 | 221 | = |
michael@0 | 222 | = F U N C T I O N S |
michael@0 | 223 | = |
michael@0 | 224 | */ |
michael@0 | 225 | static HTMLItem*CreateTextItem(char *text, unsigned int startline, |
michael@0 | 226 | unsigned int endline); |
michael@0 | 227 | static HTMLItem*CreateTagItem(TagItem*ti, unsigned int startline, |
michael@0 | 228 | unsigned int endline); |
michael@0 | 229 | static TagItem*ProcessTag(FileBuffer*fb, char **errStr); |
michael@0 | 230 | static void DestroyHTMLItem(HTMLItem *item); |
michael@0 | 231 | static void DestroyTagItem(TagItem*ti); |
michael@0 | 232 | static TAG_TYPE GetTagType(char *att); |
michael@0 | 233 | static FileBuffer*FB_Create(PRFileDesc*fd); |
michael@0 | 234 | static int FB_GetChar(FileBuffer *fb); |
michael@0 | 235 | static PRInt32 FB_GetPointer(FileBuffer *fb); |
michael@0 | 236 | static PRInt32 FB_GetRange(FileBuffer *fb, PRInt32 start, PRInt32 end, |
michael@0 | 237 | char **buf); |
michael@0 | 238 | static unsigned int FB_GetLineNum(FileBuffer *fb); |
michael@0 | 239 | static void FB_Destroy(FileBuffer *fb); |
michael@0 | 240 | static void PrintTagItem(PRFileDesc *fd, TagItem *ti); |
michael@0 | 241 | static void PrintHTMLStream(PRFileDesc *fd, HTMLItem *head); |
michael@0 | 242 | |
michael@0 | 243 | /************************************************************************ |
michael@0 | 244 | * |
michael@0 | 245 | * C r e a t e T e x t I t e m |
michael@0 | 246 | */ |
michael@0 | 247 | static HTMLItem* |
michael@0 | 248 | CreateTextItem(char *text, unsigned int startline, unsigned int endline) |
michael@0 | 249 | { |
michael@0 | 250 | HTMLItem * item; |
michael@0 | 251 | |
michael@0 | 252 | item = PR_Malloc(sizeof(HTMLItem)); |
michael@0 | 253 | if (!item) { |
michael@0 | 254 | return NULL; |
michael@0 | 255 | } |
michael@0 | 256 | |
michael@0 | 257 | item->type = TEXT_ITEM; |
michael@0 | 258 | item->item.text = text; |
michael@0 | 259 | item->next = NULL; |
michael@0 | 260 | item->startLine = startline; |
michael@0 | 261 | item->endLine = endline; |
michael@0 | 262 | |
michael@0 | 263 | return item; |
michael@0 | 264 | } |
michael@0 | 265 | |
michael@0 | 266 | |
michael@0 | 267 | /************************************************************************ |
michael@0 | 268 | * |
michael@0 | 269 | * C r e a t e T a g I t e m |
michael@0 | 270 | */ |
michael@0 | 271 | static HTMLItem* |
michael@0 | 272 | CreateTagItem(TagItem*ti, unsigned int startline, unsigned int endline) |
michael@0 | 273 | { |
michael@0 | 274 | HTMLItem * item; |
michael@0 | 275 | |
michael@0 | 276 | item = PR_Malloc(sizeof(HTMLItem)); |
michael@0 | 277 | if (!item) { |
michael@0 | 278 | return NULL; |
michael@0 | 279 | } |
michael@0 | 280 | |
michael@0 | 281 | item->type = TAG_ITEM; |
michael@0 | 282 | item->item.tag = ti; |
michael@0 | 283 | item->next = NULL; |
michael@0 | 284 | item->startLine = startline; |
michael@0 | 285 | item->endLine = endline; |
michael@0 | 286 | |
michael@0 | 287 | return item; |
michael@0 | 288 | } |
michael@0 | 289 | |
michael@0 | 290 | |
michael@0 | 291 | static PRBool |
michael@0 | 292 | isAttChar(int c) |
michael@0 | 293 | { |
michael@0 | 294 | return (isalnum(c) || c == '/' || c == '-'); |
michael@0 | 295 | } |
michael@0 | 296 | |
michael@0 | 297 | |
michael@0 | 298 | /************************************************************************ |
michael@0 | 299 | * |
michael@0 | 300 | * P r o c e s s T a g |
michael@0 | 301 | */ |
michael@0 | 302 | static TagItem* |
michael@0 | 303 | ProcessTag(FileBuffer*fb, char **errStr) |
michael@0 | 304 | { |
michael@0 | 305 | TAG_STATE state; |
michael@0 | 306 | PRInt32 startText, startID, curPos; |
michael@0 | 307 | PRBool firstAtt; |
michael@0 | 308 | int curchar; |
michael@0 | 309 | TagItem * ti = NULL; |
michael@0 | 310 | AVPair * curPair = NULL; |
michael@0 | 311 | char quotechar = '\0'; |
michael@0 | 312 | unsigned int linenum; |
michael@0 | 313 | unsigned int startline; |
michael@0 | 314 | |
michael@0 | 315 | state = START_STATE; |
michael@0 | 316 | |
michael@0 | 317 | startID = FB_GetPointer(fb); |
michael@0 | 318 | startText = startID; |
michael@0 | 319 | firstAtt = PR_TRUE; |
michael@0 | 320 | |
michael@0 | 321 | ti = (TagItem * ) PR_Malloc(sizeof(TagItem)); |
michael@0 | 322 | if (!ti) |
michael@0 | 323 | out_of_memory(); |
michael@0 | 324 | ti->type = OTHER_TAG; |
michael@0 | 325 | ti->attList = NULL; |
michael@0 | 326 | ti->attListTail = NULL; |
michael@0 | 327 | ti->text = NULL; |
michael@0 | 328 | |
michael@0 | 329 | startline = FB_GetLineNum(fb); |
michael@0 | 330 | |
michael@0 | 331 | while (state != DONE_STATE && state != ERR_STATE) { |
michael@0 | 332 | linenum = FB_GetLineNum(fb); |
michael@0 | 333 | curchar = FB_GetChar(fb); |
michael@0 | 334 | if (curchar == EOF) { |
michael@0 | 335 | *errStr = PR_smprintf( |
michael@0 | 336 | "line %d: Unexpected end-of-file while parsing tag starting at line %d.\n", |
michael@0 | 337 | linenum, startline); |
michael@0 | 338 | state = ERR_STATE; |
michael@0 | 339 | continue; |
michael@0 | 340 | } |
michael@0 | 341 | |
michael@0 | 342 | switch (state) { |
michael@0 | 343 | case START_STATE: |
michael@0 | 344 | if (curchar == '!') { |
michael@0 | 345 | /* |
michael@0 | 346 | * SGML tag or comment |
michael@0 | 347 | * Here's the general rule for SGML tags. Everything from |
michael@0 | 348 | * <! to > is the tag. Inside the tag, comments are |
michael@0 | 349 | * delimited with --. So we are looking for the first '>' |
michael@0 | 350 | * that is not commented out, that is, not inside a pair |
michael@0 | 351 | * of --: <!DOCTYPE --this is a comment >(psyche!) --> |
michael@0 | 352 | */ |
michael@0 | 353 | |
michael@0 | 354 | PRBool inComment = PR_FALSE; |
michael@0 | 355 | short hyphenCount = 0; /* number of consecutive hyphens */ |
michael@0 | 356 | |
michael@0 | 357 | while (1) { |
michael@0 | 358 | linenum = FB_GetLineNum(fb); |
michael@0 | 359 | curchar = FB_GetChar(fb); |
michael@0 | 360 | if (curchar == EOF) { |
michael@0 | 361 | /* Uh oh, EOF inside comment */ |
michael@0 | 362 | *errStr = PR_smprintf( |
michael@0 | 363 | "line %d: Unexpected end-of-file inside comment starting at line %d.\n", |
michael@0 | 364 | linenum, startline); |
michael@0 | 365 | state = ERR_STATE; |
michael@0 | 366 | break; |
michael@0 | 367 | } |
michael@0 | 368 | if (curchar == '-') { |
michael@0 | 369 | if (hyphenCount == 1) { |
michael@0 | 370 | /* This is a comment delimiter */ |
michael@0 | 371 | inComment = !inComment; |
michael@0 | 372 | hyphenCount = 0; |
michael@0 | 373 | } else { |
michael@0 | 374 | /* beginning of a comment delimiter? */ |
michael@0 | 375 | hyphenCount = 1; |
michael@0 | 376 | } |
michael@0 | 377 | } else if (curchar == '>') { |
michael@0 | 378 | if (!inComment) { |
michael@0 | 379 | /* This is the end of the tag */ |
michael@0 | 380 | state = DONE_STATE; |
michael@0 | 381 | break; |
michael@0 | 382 | } else { |
michael@0 | 383 | /* The > is inside a comment, so it's not |
michael@0 | 384 | * really the end of the tag */ |
michael@0 | 385 | hyphenCount = 0; |
michael@0 | 386 | } |
michael@0 | 387 | } else { |
michael@0 | 388 | hyphenCount = 0; |
michael@0 | 389 | } |
michael@0 | 390 | } |
michael@0 | 391 | ti->type = COMMENT_TAG; |
michael@0 | 392 | break; |
michael@0 | 393 | } |
michael@0 | 394 | /* fall through */ |
michael@0 | 395 | case GET_ATT_STATE: |
michael@0 | 396 | if (isspace(curchar) || curchar == '=' || curchar |
michael@0 | 397 | == '>') { |
michael@0 | 398 | /* end of the current attribute */ |
michael@0 | 399 | curPos = FB_GetPointer(fb) - 2; |
michael@0 | 400 | if (curPos >= startID) { |
michael@0 | 401 | /* We have an attribute */ |
michael@0 | 402 | curPair = (AVPair * )PR_Malloc(sizeof(AVPair)); |
michael@0 | 403 | if (!curPair) |
michael@0 | 404 | out_of_memory(); |
michael@0 | 405 | curPair->value = NULL; |
michael@0 | 406 | curPair->next = NULL; |
michael@0 | 407 | FB_GetRange(fb, startID, curPos, |
michael@0 | 408 | &curPair->attribute); |
michael@0 | 409 | |
michael@0 | 410 | /* Stick this attribute on the list */ |
michael@0 | 411 | if (ti->attListTail) { |
michael@0 | 412 | ti->attListTail->next = curPair; |
michael@0 | 413 | ti->attListTail = curPair; |
michael@0 | 414 | } else { |
michael@0 | 415 | ti->attList = ti->attListTail = |
michael@0 | 416 | curPair; |
michael@0 | 417 | } |
michael@0 | 418 | |
michael@0 | 419 | /* If this is the first attribute, find the type of tag |
michael@0 | 420 | * based on it. Also, start saving the text of the tag. */ |
michael@0 | 421 | if (firstAtt) { |
michael@0 | 422 | ti->type = GetTagType(curPair->attribute); |
michael@0 | 423 | startText = FB_GetPointer(fb) |
michael@0 | 424 | -1; |
michael@0 | 425 | firstAtt = PR_FALSE; |
michael@0 | 426 | } |
michael@0 | 427 | } else { |
michael@0 | 428 | if (curchar == '=') { |
michael@0 | 429 | /* If we don't have any attribute but we do have an |
michael@0 | 430 | * equal sign, that's an error */ |
michael@0 | 431 | *errStr = PR_smprintf("line %d: Malformed tag starting at line %d.\n", |
michael@0 | 432 | linenum, startline); |
michael@0 | 433 | state = ERR_STATE; |
michael@0 | 434 | break; |
michael@0 | 435 | } |
michael@0 | 436 | } |
michael@0 | 437 | |
michael@0 | 438 | /* Compute next state */ |
michael@0 | 439 | if (curchar == '=') { |
michael@0 | 440 | startID = FB_GetPointer(fb); |
michael@0 | 441 | state = PRE_VAL_WS_STATE; |
michael@0 | 442 | } else if (curchar == '>') { |
michael@0 | 443 | state = DONE_STATE; |
michael@0 | 444 | } else if (curPair) { |
michael@0 | 445 | state = POST_ATT_WS_STATE; |
michael@0 | 446 | } else { |
michael@0 | 447 | state = PRE_ATT_WS_STATE; |
michael@0 | 448 | } |
michael@0 | 449 | } else if (isAttChar(curchar)) { |
michael@0 | 450 | /* Just another char in the attribute. Do nothing */ |
michael@0 | 451 | state = GET_ATT_STATE; |
michael@0 | 452 | } else { |
michael@0 | 453 | /* bogus char */ |
michael@0 | 454 | *errStr = PR_smprintf("line %d: Bogus chararacter '%c' in tag.\n", |
michael@0 | 455 | linenum, curchar); |
michael@0 | 456 | state = ERR_STATE; |
michael@0 | 457 | break; |
michael@0 | 458 | } |
michael@0 | 459 | break; |
michael@0 | 460 | case PRE_ATT_WS_STATE: |
michael@0 | 461 | if (curchar == '>') { |
michael@0 | 462 | state = DONE_STATE; |
michael@0 | 463 | } else if (isspace(curchar)) { |
michael@0 | 464 | /* more whitespace, do nothing */ |
michael@0 | 465 | } else if (isAttChar(curchar)) { |
michael@0 | 466 | /* starting another attribute */ |
michael@0 | 467 | startID = FB_GetPointer(fb) - 1; |
michael@0 | 468 | state = GET_ATT_STATE; |
michael@0 | 469 | } else { |
michael@0 | 470 | /* bogus char */ |
michael@0 | 471 | *errStr = PR_smprintf("line %d: Bogus character '%c' in tag.\n", |
michael@0 | 472 | linenum, curchar); |
michael@0 | 473 | state = ERR_STATE; |
michael@0 | 474 | break; |
michael@0 | 475 | } |
michael@0 | 476 | break; |
michael@0 | 477 | case POST_ATT_WS_STATE: |
michael@0 | 478 | if (curchar == '>') { |
michael@0 | 479 | state = DONE_STATE; |
michael@0 | 480 | } else if (isspace(curchar)) { |
michael@0 | 481 | /* more whitespace, do nothing */ |
michael@0 | 482 | } else if (isAttChar(curchar)) { |
michael@0 | 483 | /* starting another attribute */ |
michael@0 | 484 | startID = FB_GetPointer(fb) - 1; |
michael@0 | 485 | state = GET_ATT_STATE; |
michael@0 | 486 | } else if (curchar == '=') { |
michael@0 | 487 | /* there was whitespace between the attribute and its equal |
michael@0 | 488 | * sign, which means there's a value coming up */ |
michael@0 | 489 | state = PRE_VAL_WS_STATE; |
michael@0 | 490 | } else { |
michael@0 | 491 | /* bogus char */ |
michael@0 | 492 | *errStr = PR_smprintf("line %d: Bogus character '%c' in tag.\n", |
michael@0 | 493 | linenum, curchar); |
michael@0 | 494 | state = ERR_STATE; |
michael@0 | 495 | break; |
michael@0 | 496 | } |
michael@0 | 497 | break; |
michael@0 | 498 | case PRE_VAL_WS_STATE: |
michael@0 | 499 | if (curchar == '>') { |
michael@0 | 500 | /* premature end-of-tag (sounds like a personal problem). */ |
michael@0 | 501 | *errStr = PR_smprintf( |
michael@0 | 502 | "line %d: End of tag while waiting for value.\n", |
michael@0 | 503 | linenum); |
michael@0 | 504 | state = ERR_STATE; |
michael@0 | 505 | break; |
michael@0 | 506 | } else if (isspace(curchar)) { |
michael@0 | 507 | /* more whitespace, do nothing */ |
michael@0 | 508 | break; |
michael@0 | 509 | } else { |
michael@0 | 510 | /* this must be some sort of value. Fall through |
michael@0 | 511 | * to GET_VALUE_STATE */ |
michael@0 | 512 | startID = FB_GetPointer(fb) - 1; |
michael@0 | 513 | state = GET_VALUE_STATE; |
michael@0 | 514 | } |
michael@0 | 515 | /* Fall through if we didn't break on '>' or whitespace */ |
michael@0 | 516 | case GET_VALUE_STATE: |
michael@0 | 517 | if (isspace(curchar) || curchar == '>') { |
michael@0 | 518 | /* end of value */ |
michael@0 | 519 | curPos = FB_GetPointer(fb) - 2; |
michael@0 | 520 | if (curPos >= startID) { |
michael@0 | 521 | /* Grab the value */ |
michael@0 | 522 | FB_GetRange(fb, startID, curPos, |
michael@0 | 523 | &curPair->value); |
michael@0 | 524 | curPair->valueLine = linenum; |
michael@0 | 525 | } else { |
michael@0 | 526 | /* empty value, leave as NULL */ |
michael@0 | 527 | } |
michael@0 | 528 | if (isspace(curchar)) { |
michael@0 | 529 | state = PRE_ATT_WS_STATE; |
michael@0 | 530 | } else { |
michael@0 | 531 | state = DONE_STATE; |
michael@0 | 532 | } |
michael@0 | 533 | } else if (curchar == '\"' || curchar == '\'') { |
michael@0 | 534 | /* quoted value. Start recording the value inside the quote*/ |
michael@0 | 535 | startID = FB_GetPointer(fb); |
michael@0 | 536 | state = GET_QUOTED_VAL_STATE; |
michael@0 | 537 | PORT_Assert(quotechar == '\0'); |
michael@0 | 538 | quotechar = curchar; /* look for matching quote type */ |
michael@0 | 539 | } else { |
michael@0 | 540 | /* just more value */ |
michael@0 | 541 | } |
michael@0 | 542 | break; |
michael@0 | 543 | case GET_QUOTED_VAL_STATE: |
michael@0 | 544 | PORT_Assert(quotechar != '\0'); |
michael@0 | 545 | if (curchar == quotechar) { |
michael@0 | 546 | /* end of quoted value */ |
michael@0 | 547 | curPos = FB_GetPointer(fb) - 2; |
michael@0 | 548 | if (curPos >= startID) { |
michael@0 | 549 | /* Grab the value */ |
michael@0 | 550 | FB_GetRange(fb, startID, curPos, |
michael@0 | 551 | &curPair->value); |
michael@0 | 552 | curPair->valueLine = linenum; |
michael@0 | 553 | } else { |
michael@0 | 554 | /* empty value, leave it as NULL */ |
michael@0 | 555 | } |
michael@0 | 556 | state = GET_ATT_STATE; |
michael@0 | 557 | quotechar = '\0'; |
michael@0 | 558 | startID = FB_GetPointer(fb); |
michael@0 | 559 | } else { |
michael@0 | 560 | /* more quoted value, continue */ |
michael@0 | 561 | } |
michael@0 | 562 | break; |
michael@0 | 563 | case DONE_STATE: |
michael@0 | 564 | case ERR_STATE: |
michael@0 | 565 | default: |
michael@0 | 566 | ; /* should never get here */ |
michael@0 | 567 | } |
michael@0 | 568 | } |
michael@0 | 569 | |
michael@0 | 570 | if (state == DONE_STATE) { |
michael@0 | 571 | /* Get the text of the tag */ |
michael@0 | 572 | curPos = FB_GetPointer(fb) - 1; |
michael@0 | 573 | FB_GetRange(fb, startText, curPos, &ti->text); |
michael@0 | 574 | |
michael@0 | 575 | /* Return the tag */ |
michael@0 | 576 | return ti; |
michael@0 | 577 | } |
michael@0 | 578 | |
michael@0 | 579 | /* Uh oh, an error. Kill the tag item*/ |
michael@0 | 580 | DestroyTagItem(ti); |
michael@0 | 581 | return NULL; |
michael@0 | 582 | } |
michael@0 | 583 | |
michael@0 | 584 | |
michael@0 | 585 | /************************************************************************ |
michael@0 | 586 | * |
michael@0 | 587 | * D e s t r o y H T M L I t e m |
michael@0 | 588 | */ |
michael@0 | 589 | static void |
michael@0 | 590 | DestroyHTMLItem(HTMLItem *item) |
michael@0 | 591 | { |
michael@0 | 592 | if (item->type == TAG_ITEM) { |
michael@0 | 593 | DestroyTagItem(item->item.tag); |
michael@0 | 594 | } else { |
michael@0 | 595 | if (item->item.text) { |
michael@0 | 596 | PR_Free(item->item.text); |
michael@0 | 597 | } |
michael@0 | 598 | } |
michael@0 | 599 | } |
michael@0 | 600 | |
michael@0 | 601 | |
michael@0 | 602 | /************************************************************************ |
michael@0 | 603 | * |
michael@0 | 604 | * D e s t r o y T a g I t e m |
michael@0 | 605 | */ |
michael@0 | 606 | static void |
michael@0 | 607 | DestroyTagItem(TagItem*ti) |
michael@0 | 608 | { |
michael@0 | 609 | AVPair * temp; |
michael@0 | 610 | |
michael@0 | 611 | if (ti->text) { |
michael@0 | 612 | PR_Free(ti->text); |
michael@0 | 613 | ti->text = NULL; |
michael@0 | 614 | } |
michael@0 | 615 | |
michael@0 | 616 | while (ti->attList) { |
michael@0 | 617 | temp = ti->attList; |
michael@0 | 618 | ti->attList = ti->attList->next; |
michael@0 | 619 | |
michael@0 | 620 | if (temp->attribute) { |
michael@0 | 621 | PR_Free(temp->attribute); |
michael@0 | 622 | temp->attribute = NULL; |
michael@0 | 623 | } |
michael@0 | 624 | if (temp->value) { |
michael@0 | 625 | PR_Free(temp->value); |
michael@0 | 626 | temp->value = NULL; |
michael@0 | 627 | } |
michael@0 | 628 | PR_Free(temp); |
michael@0 | 629 | } |
michael@0 | 630 | |
michael@0 | 631 | PR_Free(ti); |
michael@0 | 632 | } |
michael@0 | 633 | |
michael@0 | 634 | |
michael@0 | 635 | /************************************************************************ |
michael@0 | 636 | * |
michael@0 | 637 | * G e t T a g T y p e |
michael@0 | 638 | */ |
michael@0 | 639 | static TAG_TYPE |
michael@0 | 640 | GetTagType(char *att) |
michael@0 | 641 | { |
michael@0 | 642 | if (!PORT_Strcasecmp(att, "APPLET")) { |
michael@0 | 643 | return APPLET_TAG; |
michael@0 | 644 | } |
michael@0 | 645 | if (!PORT_Strcasecmp(att, "SCRIPT")) { |
michael@0 | 646 | return SCRIPT_TAG; |
michael@0 | 647 | } |
michael@0 | 648 | if (!PORT_Strcasecmp(att, "LINK")) { |
michael@0 | 649 | return LINK_TAG; |
michael@0 | 650 | } |
michael@0 | 651 | if (!PORT_Strcasecmp(att, "STYLE")) { |
michael@0 | 652 | return STYLE_TAG; |
michael@0 | 653 | } |
michael@0 | 654 | return OTHER_TAG; |
michael@0 | 655 | } |
michael@0 | 656 | |
michael@0 | 657 | |
michael@0 | 658 | /************************************************************************ |
michael@0 | 659 | * |
michael@0 | 660 | * F B _ C r e a t e |
michael@0 | 661 | */ |
michael@0 | 662 | static FileBuffer* |
michael@0 | 663 | FB_Create(PRFileDesc*fd) |
michael@0 | 664 | { |
michael@0 | 665 | FileBuffer * fb; |
michael@0 | 666 | PRInt32 amountRead; |
michael@0 | 667 | PRInt32 storedOffset; |
michael@0 | 668 | |
michael@0 | 669 | fb = (FileBuffer * ) PR_Malloc(sizeof(FileBuffer)); |
michael@0 | 670 | fb->fd = fd; |
michael@0 | 671 | storedOffset = PR_Seek(fd, 0, PR_SEEK_CUR); |
michael@0 | 672 | PR_Seek(fd, 0, PR_SEEK_SET); |
michael@0 | 673 | fb->startOffset = 0; |
michael@0 | 674 | amountRead = PR_Read(fd, fb->buf, FILE_BUFFER_BUFSIZE); |
michael@0 | 675 | if (amountRead == -1) |
michael@0 | 676 | goto loser; |
michael@0 | 677 | fb->maxIndex = amountRead - 1; |
michael@0 | 678 | fb->curIndex = 0; |
michael@0 | 679 | fb->IsEOF = (fb->curIndex > fb->maxIndex) ? PR_TRUE : PR_FALSE; |
michael@0 | 680 | fb->lineNum = 1; |
michael@0 | 681 | |
michael@0 | 682 | PR_Seek(fd, storedOffset, PR_SEEK_SET); |
michael@0 | 683 | return fb; |
michael@0 | 684 | loser: |
michael@0 | 685 | PR_Seek(fd, storedOffset, PR_SEEK_SET); |
michael@0 | 686 | PR_Free(fb); |
michael@0 | 687 | return NULL; |
michael@0 | 688 | } |
michael@0 | 689 | |
michael@0 | 690 | |
michael@0 | 691 | /************************************************************************ |
michael@0 | 692 | * |
michael@0 | 693 | * F B _ G e t C h a r |
michael@0 | 694 | */ |
michael@0 | 695 | static int |
michael@0 | 696 | FB_GetChar(FileBuffer *fb) |
michael@0 | 697 | { |
michael@0 | 698 | PRInt32 storedOffset; |
michael@0 | 699 | PRInt32 amountRead; |
michael@0 | 700 | int retval = -1; |
michael@0 | 701 | |
michael@0 | 702 | if (fb->IsEOF) { |
michael@0 | 703 | return EOF; |
michael@0 | 704 | } |
michael@0 | 705 | |
michael@0 | 706 | storedOffset = PR_Seek(fb->fd, 0, PR_SEEK_CUR); |
michael@0 | 707 | |
michael@0 | 708 | retval = (unsigned char) fb->buf[fb->curIndex++]; |
michael@0 | 709 | if (retval == '\n') |
michael@0 | 710 | fb->lineNum++; |
michael@0 | 711 | |
michael@0 | 712 | if (fb->curIndex > fb->maxIndex) { |
michael@0 | 713 | /* We're at the end of the buffer. Try to get some new data from the |
michael@0 | 714 | * file */ |
michael@0 | 715 | fb->startOffset += fb->maxIndex + 1; |
michael@0 | 716 | PR_Seek(fb->fd, fb->startOffset, PR_SEEK_SET); |
michael@0 | 717 | amountRead = PR_Read(fb->fd, fb->buf, FILE_BUFFER_BUFSIZE); |
michael@0 | 718 | if (amountRead == -1) |
michael@0 | 719 | goto loser; |
michael@0 | 720 | fb->maxIndex = amountRead - 1; |
michael@0 | 721 | fb->curIndex = 0; |
michael@0 | 722 | } |
michael@0 | 723 | |
michael@0 | 724 | fb->IsEOF = (fb->curIndex > fb->maxIndex) ? PR_TRUE : PR_FALSE; |
michael@0 | 725 | |
michael@0 | 726 | loser: |
michael@0 | 727 | PR_Seek(fb->fd, storedOffset, PR_SEEK_SET); |
michael@0 | 728 | return retval; |
michael@0 | 729 | } |
michael@0 | 730 | |
michael@0 | 731 | |
michael@0 | 732 | /************************************************************************ |
michael@0 | 733 | * |
michael@0 | 734 | * F B _ G e t L i n e N u m |
michael@0 | 735 | * |
michael@0 | 736 | */ |
michael@0 | 737 | static unsigned int |
michael@0 | 738 | FB_GetLineNum(FileBuffer *fb) |
michael@0 | 739 | { |
michael@0 | 740 | return fb->lineNum; |
michael@0 | 741 | } |
michael@0 | 742 | |
michael@0 | 743 | |
michael@0 | 744 | /************************************************************************ |
michael@0 | 745 | * |
michael@0 | 746 | * F B _ G e t P o i n t e r |
michael@0 | 747 | * |
michael@0 | 748 | */ |
michael@0 | 749 | static PRInt32 |
michael@0 | 750 | FB_GetPointer(FileBuffer *fb) |
michael@0 | 751 | { |
michael@0 | 752 | return fb->startOffset + fb->curIndex; |
michael@0 | 753 | } |
michael@0 | 754 | |
michael@0 | 755 | |
michael@0 | 756 | /************************************************************************ |
michael@0 | 757 | * |
michael@0 | 758 | * F B _ G e t R a n g e |
michael@0 | 759 | * |
michael@0 | 760 | */ |
michael@0 | 761 | static PRInt32 |
michael@0 | 762 | FB_GetRange(FileBuffer *fb, PRInt32 start, PRInt32 end, char **buf) |
michael@0 | 763 | { |
michael@0 | 764 | PRInt32 amountRead; |
michael@0 | 765 | PRInt32 storedOffset; |
michael@0 | 766 | |
michael@0 | 767 | *buf = PR_Malloc(end - start + 2); |
michael@0 | 768 | if (*buf == NULL) { |
michael@0 | 769 | return 0; |
michael@0 | 770 | } |
michael@0 | 771 | |
michael@0 | 772 | storedOffset = PR_Seek(fb->fd, 0, PR_SEEK_CUR); |
michael@0 | 773 | PR_Seek(fb->fd, start, PR_SEEK_SET); |
michael@0 | 774 | amountRead = PR_Read(fb->fd, *buf, end - start + 1); |
michael@0 | 775 | PR_Seek(fb->fd, storedOffset, PR_SEEK_SET); |
michael@0 | 776 | if (amountRead == -1) { |
michael@0 | 777 | PR_Free(*buf); |
michael@0 | 778 | *buf = NULL; |
michael@0 | 779 | return 0; |
michael@0 | 780 | } |
michael@0 | 781 | |
michael@0 | 782 | (*buf)[end-start+1] = '\0'; |
michael@0 | 783 | return amountRead; |
michael@0 | 784 | } |
michael@0 | 785 | |
michael@0 | 786 | |
michael@0 | 787 | /************************************************************************ |
michael@0 | 788 | * |
michael@0 | 789 | * F B _ D e s t r o y |
michael@0 | 790 | * |
michael@0 | 791 | */ |
michael@0 | 792 | static void |
michael@0 | 793 | FB_Destroy(FileBuffer *fb) |
michael@0 | 794 | { |
michael@0 | 795 | if (fb) { |
michael@0 | 796 | PR_Free(fb); |
michael@0 | 797 | } |
michael@0 | 798 | } |
michael@0 | 799 | |
michael@0 | 800 | |
michael@0 | 801 | /************************************************************************ |
michael@0 | 802 | * |
michael@0 | 803 | * P r i n t T a g I t e m |
michael@0 | 804 | * |
michael@0 | 805 | */ |
michael@0 | 806 | static void |
michael@0 | 807 | PrintTagItem(PRFileDesc *fd, TagItem *ti) |
michael@0 | 808 | { |
michael@0 | 809 | AVPair * pair; |
michael@0 | 810 | |
michael@0 | 811 | PR_fprintf(fd, "TAG:\n----\nType: "); |
michael@0 | 812 | switch (ti->type) { |
michael@0 | 813 | case APPLET_TAG: |
michael@0 | 814 | PR_fprintf(fd, "applet\n"); |
michael@0 | 815 | break; |
michael@0 | 816 | case SCRIPT_TAG: |
michael@0 | 817 | PR_fprintf(fd, "script\n"); |
michael@0 | 818 | break; |
michael@0 | 819 | case LINK_TAG: |
michael@0 | 820 | PR_fprintf(fd, "link\n"); |
michael@0 | 821 | break; |
michael@0 | 822 | case STYLE_TAG: |
michael@0 | 823 | PR_fprintf(fd, "style\n"); |
michael@0 | 824 | break; |
michael@0 | 825 | case COMMENT_TAG: |
michael@0 | 826 | PR_fprintf(fd, "comment\n"); |
michael@0 | 827 | break; |
michael@0 | 828 | case OTHER_TAG: |
michael@0 | 829 | default: |
michael@0 | 830 | PR_fprintf(fd, "other\n"); |
michael@0 | 831 | break; |
michael@0 | 832 | } |
michael@0 | 833 | |
michael@0 | 834 | PR_fprintf(fd, "Attributes:\n"); |
michael@0 | 835 | for (pair = ti->attList; pair; pair = pair->next) { |
michael@0 | 836 | PR_fprintf(fd, "\t%s=%s\n", pair->attribute, |
michael@0 | 837 | pair->value ? pair->value : ""); |
michael@0 | 838 | } |
michael@0 | 839 | PR_fprintf(fd, "Text:%s\n", ti->text ? ti->text : ""); |
michael@0 | 840 | |
michael@0 | 841 | PR_fprintf(fd, "---End of tag---\n"); |
michael@0 | 842 | } |
michael@0 | 843 | |
michael@0 | 844 | |
michael@0 | 845 | /************************************************************************ |
michael@0 | 846 | * |
michael@0 | 847 | * P r i n t H T M L S t r e a m |
michael@0 | 848 | * |
michael@0 | 849 | */ |
michael@0 | 850 | static void |
michael@0 | 851 | PrintHTMLStream(PRFileDesc *fd, HTMLItem *head) |
michael@0 | 852 | { |
michael@0 | 853 | while (head) { |
michael@0 | 854 | if (head->type == TAG_ITEM) { |
michael@0 | 855 | PrintTagItem(fd, head->item.tag); |
michael@0 | 856 | } else { |
michael@0 | 857 | PR_fprintf(fd, "\nTEXT:\n-----\n%s\n-----\n\n", head->item.text); |
michael@0 | 858 | } |
michael@0 | 859 | head = head->next; |
michael@0 | 860 | } |
michael@0 | 861 | } |
michael@0 | 862 | |
michael@0 | 863 | |
michael@0 | 864 | /************************************************************************ |
michael@0 | 865 | * |
michael@0 | 866 | * S a v e I n l i n e S c r i p t |
michael@0 | 867 | * |
michael@0 | 868 | */ |
michael@0 | 869 | static int |
michael@0 | 870 | SaveInlineScript(char *text, char *id, char *basedir, char *archiveDir) |
michael@0 | 871 | { |
michael@0 | 872 | char *filename = NULL; |
michael@0 | 873 | PRFileDesc * fd = NULL; |
michael@0 | 874 | int retval = -1; |
michael@0 | 875 | PRInt32 writeLen; |
michael@0 | 876 | char *ilDir = NULL; |
michael@0 | 877 | |
michael@0 | 878 | if (!text || !id || !archiveDir) { |
michael@0 | 879 | return - 1; |
michael@0 | 880 | } |
michael@0 | 881 | |
michael@0 | 882 | if (dumpParse) { |
michael@0 | 883 | PR_fprintf(outputFD, "SaveInlineScript: text=%s, id=%s, \n" |
michael@0 | 884 | "basedir=%s, archiveDir=%s\n", |
michael@0 | 885 | text, id, basedir, archiveDir); |
michael@0 | 886 | } |
michael@0 | 887 | |
michael@0 | 888 | /* Make sure the archive directory is around */ |
michael@0 | 889 | if (ensureExists(basedir, archiveDir) != PR_SUCCESS) { |
michael@0 | 890 | PR_fprintf(errorFD, |
michael@0 | 891 | "ERROR: Unable to create archive directory %s.\n", archiveDir); |
michael@0 | 892 | errorCount++; |
michael@0 | 893 | return - 1; |
michael@0 | 894 | } |
michael@0 | 895 | |
michael@0 | 896 | /* Make sure the inline script directory is around */ |
michael@0 | 897 | ilDir = PR_smprintf("%s/inlineScripts", archiveDir); |
michael@0 | 898 | scriptdir = "inlineScripts"; |
michael@0 | 899 | if (ensureExists(basedir, ilDir) != PR_SUCCESS) { |
michael@0 | 900 | PR_fprintf(errorFD, |
michael@0 | 901 | "ERROR: Unable to create directory %s.\n", ilDir); |
michael@0 | 902 | errorCount++; |
michael@0 | 903 | return - 1; |
michael@0 | 904 | } |
michael@0 | 905 | |
michael@0 | 906 | filename = PR_smprintf("%s/%s/%s", basedir, ilDir, id); |
michael@0 | 907 | |
michael@0 | 908 | /* If the file already exists, give a warning, then blow it away */ |
michael@0 | 909 | if (PR_Access(filename, PR_ACCESS_EXISTS) == PR_SUCCESS) { |
michael@0 | 910 | PR_fprintf(errorFD, |
michael@0 | 911 | "warning: file \"%s\" already exists--will overwrite.\n", |
michael@0 | 912 | filename); |
michael@0 | 913 | warningCount++; |
michael@0 | 914 | if (rm_dash_r(filename)) { |
michael@0 | 915 | PR_fprintf(errorFD, "ERROR: Unable to delete %s.\n", filename); |
michael@0 | 916 | errorCount++; |
michael@0 | 917 | goto finish; |
michael@0 | 918 | } |
michael@0 | 919 | } |
michael@0 | 920 | |
michael@0 | 921 | /* Write text into file with name id */ |
michael@0 | 922 | fd = PR_Open(filename, PR_WRONLY | PR_CREATE_FILE | PR_TRUNCATE, 0777); |
michael@0 | 923 | if (!fd) { |
michael@0 | 924 | PR_fprintf(errorFD, "ERROR: Unable to create file \"%s\".\n", |
michael@0 | 925 | filename); |
michael@0 | 926 | errorCount++; |
michael@0 | 927 | goto finish; |
michael@0 | 928 | } |
michael@0 | 929 | writeLen = strlen(text); |
michael@0 | 930 | if ( PR_Write(fd, text, writeLen) != writeLen) { |
michael@0 | 931 | PR_fprintf(errorFD, "ERROR: Unable to write to file \"%s\".\n", |
michael@0 | 932 | filename); |
michael@0 | 933 | errorCount++; |
michael@0 | 934 | goto finish; |
michael@0 | 935 | } |
michael@0 | 936 | |
michael@0 | 937 | retval = 0; |
michael@0 | 938 | finish: |
michael@0 | 939 | if (filename) { |
michael@0 | 940 | PR_smprintf_free(filename); |
michael@0 | 941 | } |
michael@0 | 942 | if (ilDir) { |
michael@0 | 943 | PR_smprintf_free(ilDir); |
michael@0 | 944 | } |
michael@0 | 945 | if (fd) { |
michael@0 | 946 | PR_Close(fd); |
michael@0 | 947 | } |
michael@0 | 948 | return retval; |
michael@0 | 949 | } |
michael@0 | 950 | |
michael@0 | 951 | |
michael@0 | 952 | /************************************************************************ |
michael@0 | 953 | * |
michael@0 | 954 | * S a v e U n n a m a b l e S c r i p t |
michael@0 | 955 | * |
michael@0 | 956 | */ |
michael@0 | 957 | static int |
michael@0 | 958 | SaveUnnamableScript(char *text, char *basedir, char *archiveDir, |
michael@0 | 959 | char *HTMLfilename) |
michael@0 | 960 | { |
michael@0 | 961 | char *id = NULL; |
michael@0 | 962 | char *ext = NULL; |
michael@0 | 963 | char *start = NULL; |
michael@0 | 964 | int retval = -1; |
michael@0 | 965 | |
michael@0 | 966 | if (!text || !archiveDir || !HTMLfilename) { |
michael@0 | 967 | return - 1; |
michael@0 | 968 | } |
michael@0 | 969 | |
michael@0 | 970 | if (dumpParse) { |
michael@0 | 971 | PR_fprintf(outputFD, "SaveUnnamableScript: text=%s, basedir=%s,\n" |
michael@0 | 972 | "archiveDir=%s, filename=%s\n", text, basedir, archiveDir, |
michael@0 | 973 | HTMLfilename); |
michael@0 | 974 | } |
michael@0 | 975 | |
michael@0 | 976 | /* Construct the filename */ |
michael@0 | 977 | ext = PL_strrchr(HTMLfilename, '.'); |
michael@0 | 978 | if (ext) { |
michael@0 | 979 | *ext = '\0'; |
michael@0 | 980 | } |
michael@0 | 981 | for (start = HTMLfilename; strpbrk(start, "/\\"); |
michael@0 | 982 | start = strpbrk(start, "/\\") + 1) |
michael@0 | 983 | /* do nothing */; |
michael@0 | 984 | if (*start == '\0') |
michael@0 | 985 | start = HTMLfilename; |
michael@0 | 986 | id = PR_smprintf("_%s%d", start, idOrdinal++); |
michael@0 | 987 | if (ext) { |
michael@0 | 988 | *ext = '.'; |
michael@0 | 989 | } |
michael@0 | 990 | |
michael@0 | 991 | /* Now call SaveInlineScript to do the work */ |
michael@0 | 992 | retval = SaveInlineScript(text, id, basedir, archiveDir); |
michael@0 | 993 | |
michael@0 | 994 | PR_Free(id); |
michael@0 | 995 | |
michael@0 | 996 | return retval; |
michael@0 | 997 | } |
michael@0 | 998 | |
michael@0 | 999 | |
michael@0 | 1000 | /************************************************************************ |
michael@0 | 1001 | * |
michael@0 | 1002 | * S a v e S o u r c e |
michael@0 | 1003 | * |
michael@0 | 1004 | */ |
michael@0 | 1005 | static int |
michael@0 | 1006 | SaveSource(char *src, char *codebase, char *basedir, char *archiveDir) |
michael@0 | 1007 | { |
michael@0 | 1008 | char *from = NULL, *to = NULL; |
michael@0 | 1009 | int retval = -1; |
michael@0 | 1010 | char *arcDir = NULL; |
michael@0 | 1011 | |
michael@0 | 1012 | if (!src || !archiveDir) { |
michael@0 | 1013 | return - 1; |
michael@0 | 1014 | } |
michael@0 | 1015 | |
michael@0 | 1016 | if (dumpParse) { |
michael@0 | 1017 | PR_fprintf(outputFD, "SaveSource: src=%s, codebase=%s, basedir=%s,\n" |
michael@0 | 1018 | "archiveDir=%s\n", src, codebase, basedir, archiveDir); |
michael@0 | 1019 | } |
michael@0 | 1020 | |
michael@0 | 1021 | if (codebase) { |
michael@0 | 1022 | arcDir = PR_smprintf("%s/%s/%s/", basedir, codebase, archiveDir); |
michael@0 | 1023 | } else { |
michael@0 | 1024 | arcDir = PR_smprintf("%s/%s/", basedir, archiveDir); |
michael@0 | 1025 | } |
michael@0 | 1026 | |
michael@0 | 1027 | if (codebase) { |
michael@0 | 1028 | from = PR_smprintf("%s/%s/%s", basedir, codebase, src); |
michael@0 | 1029 | to = PR_smprintf("%s%s", arcDir, src); |
michael@0 | 1030 | } else { |
michael@0 | 1031 | from = PR_smprintf("%s/%s", basedir, src); |
michael@0 | 1032 | to = PR_smprintf("%s%s", arcDir, src); |
michael@0 | 1033 | } |
michael@0 | 1034 | |
michael@0 | 1035 | if (make_dirs(to, 0777)) { |
michael@0 | 1036 | PR_fprintf(errorFD, |
michael@0 | 1037 | "ERROR: Unable to create archive directory %s.\n", archiveDir); |
michael@0 | 1038 | errorCount++; |
michael@0 | 1039 | goto finish; |
michael@0 | 1040 | } |
michael@0 | 1041 | |
michael@0 | 1042 | retval = copyinto(from, to); |
michael@0 | 1043 | finish: |
michael@0 | 1044 | if (from) |
michael@0 | 1045 | PR_Free(from); |
michael@0 | 1046 | if (to) |
michael@0 | 1047 | PR_Free(to); |
michael@0 | 1048 | if (arcDir) |
michael@0 | 1049 | PR_Free(arcDir); |
michael@0 | 1050 | return retval; |
michael@0 | 1051 | } |
michael@0 | 1052 | |
michael@0 | 1053 | |
michael@0 | 1054 | /************************************************************************ |
michael@0 | 1055 | * |
michael@0 | 1056 | * T a g T y p e T o S t r i n g |
michael@0 | 1057 | * |
michael@0 | 1058 | */ |
michael@0 | 1059 | char * |
michael@0 | 1060 | TagTypeToString(TAG_TYPE type) |
michael@0 | 1061 | { |
michael@0 | 1062 | switch (type) { |
michael@0 | 1063 | case APPLET_TAG: |
michael@0 | 1064 | return "APPLET"; |
michael@0 | 1065 | case SCRIPT_TAG: |
michael@0 | 1066 | return "SCRIPT"; |
michael@0 | 1067 | case LINK_TAG: |
michael@0 | 1068 | return "LINK"; |
michael@0 | 1069 | case STYLE_TAG: |
michael@0 | 1070 | return "STYLE"; |
michael@0 | 1071 | default: |
michael@0 | 1072 | break; |
michael@0 | 1073 | } |
michael@0 | 1074 | return "unknown"; |
michael@0 | 1075 | } |
michael@0 | 1076 | |
michael@0 | 1077 | |
michael@0 | 1078 | /************************************************************************ |
michael@0 | 1079 | * |
michael@0 | 1080 | * e x t r a c t _ j s |
michael@0 | 1081 | * |
michael@0 | 1082 | */ |
michael@0 | 1083 | static int |
michael@0 | 1084 | extract_js(char *filename) |
michael@0 | 1085 | { |
michael@0 | 1086 | PRFileDesc * fd = NULL; |
michael@0 | 1087 | FileBuffer * fb = NULL; |
michael@0 | 1088 | HTMLItem * head = NULL; |
michael@0 | 1089 | HTMLItem * tail = NULL; |
michael@0 | 1090 | HTMLItem * curitem = NULL; |
michael@0 | 1091 | HTMLItem * styleList = NULL; |
michael@0 | 1092 | HTMLItem * styleListTail = NULL; |
michael@0 | 1093 | HTMLItem * entityList = NULL; |
michael@0 | 1094 | HTMLItem * entityListTail = NULL; |
michael@0 | 1095 | TagItem * tagp = NULL; |
michael@0 | 1096 | char *text = NULL; |
michael@0 | 1097 | char *tagerr = NULL; |
michael@0 | 1098 | char *archiveDir = NULL; |
michael@0 | 1099 | char *firstArchiveDir = NULL; |
michael@0 | 1100 | char *basedir = NULL; |
michael@0 | 1101 | PRInt32 textStart; |
michael@0 | 1102 | PRInt32 curOffset; |
michael@0 | 1103 | HTML_STATE state; |
michael@0 | 1104 | int curchar; |
michael@0 | 1105 | int retval = -1; |
michael@0 | 1106 | unsigned int linenum, startLine; |
michael@0 | 1107 | |
michael@0 | 1108 | /* Initialize the implicit ID counter for each file */ |
michael@0 | 1109 | idOrdinal = 0; |
michael@0 | 1110 | |
michael@0 | 1111 | /* |
michael@0 | 1112 | * First, parse the HTML into a stream of tags and text. |
michael@0 | 1113 | */ |
michael@0 | 1114 | |
michael@0 | 1115 | fd = PR_Open(filename, PR_RDONLY, 0); |
michael@0 | 1116 | if (!fd) { |
michael@0 | 1117 | PR_fprintf(errorFD, "Unable to open %s for reading.\n", filename); |
michael@0 | 1118 | errorCount++; |
michael@0 | 1119 | return - 1; |
michael@0 | 1120 | } |
michael@0 | 1121 | |
michael@0 | 1122 | /* Construct base directory of filename. */ |
michael@0 | 1123 | { |
michael@0 | 1124 | char *cp; |
michael@0 | 1125 | |
michael@0 | 1126 | basedir = PL_strdup(filename); |
michael@0 | 1127 | |
michael@0 | 1128 | /* Remove trailing slashes */ |
michael@0 | 1129 | while ( (cp = PL_strprbrk(basedir, "/\\")) == |
michael@0 | 1130 | (basedir + strlen(basedir) - 1)) { |
michael@0 | 1131 | *cp = '\0'; |
michael@0 | 1132 | } |
michael@0 | 1133 | |
michael@0 | 1134 | /* Now remove everything from the last slash (which will be followed |
michael@0 | 1135 | * by a filename) to the end */ |
michael@0 | 1136 | cp = PL_strprbrk(basedir, "/\\"); |
michael@0 | 1137 | if (cp) { |
michael@0 | 1138 | *cp = '\0'; |
michael@0 | 1139 | } |
michael@0 | 1140 | } |
michael@0 | 1141 | |
michael@0 | 1142 | state = TEXT_HTML_STATE; |
michael@0 | 1143 | |
michael@0 | 1144 | fb = FB_Create(fd); |
michael@0 | 1145 | |
michael@0 | 1146 | textStart = 0; |
michael@0 | 1147 | startLine = 0; |
michael@0 | 1148 | while (linenum = FB_GetLineNum(fb), (curchar = FB_GetChar(fb)) != |
michael@0 | 1149 | EOF) { |
michael@0 | 1150 | switch (state) { |
michael@0 | 1151 | case TEXT_HTML_STATE: |
michael@0 | 1152 | if (curchar == '<') { |
michael@0 | 1153 | /* |
michael@0 | 1154 | * Found a tag |
michael@0 | 1155 | */ |
michael@0 | 1156 | /* Save the text so far to a new text item */ |
michael@0 | 1157 | curOffset = FB_GetPointer(fb) - 2; |
michael@0 | 1158 | if (curOffset >= textStart) { |
michael@0 | 1159 | if (FB_GetRange(fb, textStart, curOffset, |
michael@0 | 1160 | &text) != |
michael@0 | 1161 | curOffset - textStart + 1) { |
michael@0 | 1162 | PR_fprintf(errorFD, |
michael@0 | 1163 | "Unable to read from %s.\n", |
michael@0 | 1164 | filename); |
michael@0 | 1165 | errorCount++; |
michael@0 | 1166 | goto loser; |
michael@0 | 1167 | } |
michael@0 | 1168 | /* little fudge here. If the first character on a line |
michael@0 | 1169 | * is '<', meaning a new tag, the preceding text item |
michael@0 | 1170 | * actually ends on the previous line. In this case |
michael@0 | 1171 | * we will be saying that the text segment ends on the |
michael@0 | 1172 | * next line. I don't think this matters for text items. */ |
michael@0 | 1173 | curitem = CreateTextItem(text, startLine, |
michael@0 | 1174 | linenum); |
michael@0 | 1175 | text = NULL; |
michael@0 | 1176 | if (tail == NULL) { |
michael@0 | 1177 | head = tail = curitem; |
michael@0 | 1178 | } else { |
michael@0 | 1179 | tail->next = curitem; |
michael@0 | 1180 | tail = curitem; |
michael@0 | 1181 | } |
michael@0 | 1182 | } |
michael@0 | 1183 | |
michael@0 | 1184 | /* Process the tag */ |
michael@0 | 1185 | tagp = ProcessTag(fb, &tagerr); |
michael@0 | 1186 | if (!tagp) { |
michael@0 | 1187 | if (tagerr) { |
michael@0 | 1188 | PR_fprintf(errorFD, "Error in file %s: %s\n", |
michael@0 | 1189 | filename, tagerr); |
michael@0 | 1190 | errorCount++; |
michael@0 | 1191 | } else { |
michael@0 | 1192 | PR_fprintf(errorFD, |
michael@0 | 1193 | "Error in file %s, in tag starting at line %d\n", |
michael@0 | 1194 | filename, linenum); |
michael@0 | 1195 | errorCount++; |
michael@0 | 1196 | } |
michael@0 | 1197 | goto loser; |
michael@0 | 1198 | } |
michael@0 | 1199 | /* Add the tag to the list */ |
michael@0 | 1200 | curitem = CreateTagItem(tagp, linenum, FB_GetLineNum(fb)); |
michael@0 | 1201 | if (tail == NULL) { |
michael@0 | 1202 | head = tail = curitem; |
michael@0 | 1203 | } else { |
michael@0 | 1204 | tail->next = curitem; |
michael@0 | 1205 | tail = curitem; |
michael@0 | 1206 | } |
michael@0 | 1207 | |
michael@0 | 1208 | /* What's the next state */ |
michael@0 | 1209 | if (tagp->type == SCRIPT_TAG) { |
michael@0 | 1210 | state = SCRIPT_HTML_STATE; |
michael@0 | 1211 | } |
michael@0 | 1212 | |
michael@0 | 1213 | /* Start recording text from the new offset */ |
michael@0 | 1214 | textStart = FB_GetPointer(fb); |
michael@0 | 1215 | startLine = FB_GetLineNum(fb); |
michael@0 | 1216 | } else { |
michael@0 | 1217 | /* regular character. Next! */ |
michael@0 | 1218 | } |
michael@0 | 1219 | break; |
michael@0 | 1220 | case SCRIPT_HTML_STATE: |
michael@0 | 1221 | if (curchar == '<') { |
michael@0 | 1222 | char *cp; |
michael@0 | 1223 | /* |
michael@0 | 1224 | * If this is a </script> tag, then we're at the end of the |
michael@0 | 1225 | * script. Otherwise, ignore |
michael@0 | 1226 | */ |
michael@0 | 1227 | curOffset = FB_GetPointer(fb) - 1; |
michael@0 | 1228 | cp = NULL; |
michael@0 | 1229 | if (FB_GetRange(fb, curOffset, curOffset + 8, &cp) != 9) { |
michael@0 | 1230 | if (cp) { |
michael@0 | 1231 | PR_Free(cp); |
michael@0 | 1232 | cp = NULL; |
michael@0 | 1233 | } |
michael@0 | 1234 | } else { |
michael@0 | 1235 | /* compare the strings */ |
michael@0 | 1236 | if ( !PORT_Strncasecmp(cp, "</script>", 9) ) { |
michael@0 | 1237 | /* This is the end of the script. Record the text. */ |
michael@0 | 1238 | curOffset--; |
michael@0 | 1239 | if (curOffset >= textStart) { |
michael@0 | 1240 | if (FB_GetRange(fb, textStart, curOffset, &text) != |
michael@0 | 1241 | curOffset - textStart + 1) { |
michael@0 | 1242 | PR_fprintf(errorFD, "Unable to read from %s.\n", |
michael@0 | 1243 | filename); |
michael@0 | 1244 | errorCount++; |
michael@0 | 1245 | goto loser; |
michael@0 | 1246 | } |
michael@0 | 1247 | curitem = CreateTextItem(text, startLine, linenum); |
michael@0 | 1248 | text = NULL; |
michael@0 | 1249 | if (tail == NULL) { |
michael@0 | 1250 | head = tail = curitem; |
michael@0 | 1251 | } else { |
michael@0 | 1252 | tail->next = curitem; |
michael@0 | 1253 | tail = curitem; |
michael@0 | 1254 | } |
michael@0 | 1255 | } |
michael@0 | 1256 | |
michael@0 | 1257 | /* Now parse the /script tag and put it on the list */ |
michael@0 | 1258 | tagp = ProcessTag(fb, &tagerr); |
michael@0 | 1259 | if (!tagp) { |
michael@0 | 1260 | if (tagerr) { |
michael@0 | 1261 | PR_fprintf(errorFD, "Error in file %s: %s\n", |
michael@0 | 1262 | filename, tagerr); |
michael@0 | 1263 | } else { |
michael@0 | 1264 | PR_fprintf(errorFD, |
michael@0 | 1265 | "Error in file %s, in tag starting at" |
michael@0 | 1266 | " line %d\n", filename, linenum); |
michael@0 | 1267 | } |
michael@0 | 1268 | errorCount++; |
michael@0 | 1269 | goto loser; |
michael@0 | 1270 | } |
michael@0 | 1271 | curitem = CreateTagItem(tagp, linenum, |
michael@0 | 1272 | FB_GetLineNum(fb)); |
michael@0 | 1273 | if (tail == NULL) { |
michael@0 | 1274 | head = tail = curitem; |
michael@0 | 1275 | } else { |
michael@0 | 1276 | tail->next = curitem; |
michael@0 | 1277 | tail = curitem; |
michael@0 | 1278 | } |
michael@0 | 1279 | |
michael@0 | 1280 | /* go back to text state */ |
michael@0 | 1281 | state = TEXT_HTML_STATE; |
michael@0 | 1282 | |
michael@0 | 1283 | textStart = FB_GetPointer(fb); |
michael@0 | 1284 | startLine = FB_GetLineNum(fb); |
michael@0 | 1285 | } |
michael@0 | 1286 | } |
michael@0 | 1287 | } |
michael@0 | 1288 | break; |
michael@0 | 1289 | } |
michael@0 | 1290 | } |
michael@0 | 1291 | |
michael@0 | 1292 | /* End of the file. Wrap up any remaining text */ |
michael@0 | 1293 | if (state == SCRIPT_HTML_STATE) { |
michael@0 | 1294 | if (tail && tail->type == TAG_ITEM) { |
michael@0 | 1295 | PR_fprintf(errorFD, "ERROR: <SCRIPT> tag at %s:%d is not followed " |
michael@0 | 1296 | "by a </SCRIPT> tag.\n", filename, tail->startLine); |
michael@0 | 1297 | } else { |
michael@0 | 1298 | PR_fprintf(errorFD, "ERROR: <SCRIPT> tag in file %s is not followed" |
michael@0 | 1299 | " by a </SCRIPT tag.\n", filename); |
michael@0 | 1300 | } |
michael@0 | 1301 | errorCount++; |
michael@0 | 1302 | goto loser; |
michael@0 | 1303 | } |
michael@0 | 1304 | curOffset = FB_GetPointer(fb) - 1; |
michael@0 | 1305 | if (curOffset >= textStart) { |
michael@0 | 1306 | text = NULL; |
michael@0 | 1307 | if ( FB_GetRange(fb, textStart, curOffset, &text) != |
michael@0 | 1308 | curOffset - textStart + 1) { |
michael@0 | 1309 | PR_fprintf(errorFD, "Unable to read from %s.\n", filename); |
michael@0 | 1310 | errorCount++; |
michael@0 | 1311 | goto loser; |
michael@0 | 1312 | } |
michael@0 | 1313 | curitem = CreateTextItem(text, startLine, linenum); |
michael@0 | 1314 | text = NULL; |
michael@0 | 1315 | if (tail == NULL) { |
michael@0 | 1316 | head = tail = curitem; |
michael@0 | 1317 | } else { |
michael@0 | 1318 | tail->next = curitem; |
michael@0 | 1319 | tail = curitem; |
michael@0 | 1320 | } |
michael@0 | 1321 | } |
michael@0 | 1322 | |
michael@0 | 1323 | if (dumpParse) { |
michael@0 | 1324 | PrintHTMLStream(outputFD, head); |
michael@0 | 1325 | } |
michael@0 | 1326 | |
michael@0 | 1327 | /* |
michael@0 | 1328 | * Now we have a stream of tags and text. Go through and deal with each. |
michael@0 | 1329 | */ |
michael@0 | 1330 | for (curitem = head; curitem; curitem = curitem->next) { |
michael@0 | 1331 | TagItem * tagp = NULL; |
michael@0 | 1332 | AVPair * pairp = NULL; |
michael@0 | 1333 | char *src = NULL, *id = NULL, *codebase = NULL; |
michael@0 | 1334 | PRBool hasEventHandler = PR_FALSE; |
michael@0 | 1335 | int i; |
michael@0 | 1336 | |
michael@0 | 1337 | /* Reset archive directory for each tag */ |
michael@0 | 1338 | if (archiveDir) { |
michael@0 | 1339 | PR_Free(archiveDir); |
michael@0 | 1340 | archiveDir = NULL; |
michael@0 | 1341 | } |
michael@0 | 1342 | |
michael@0 | 1343 | /* We only analyze tags */ |
michael@0 | 1344 | if (curitem->type != TAG_ITEM) { |
michael@0 | 1345 | continue; |
michael@0 | 1346 | } |
michael@0 | 1347 | |
michael@0 | 1348 | tagp = curitem->item.tag; |
michael@0 | 1349 | |
michael@0 | 1350 | /* go through the attributes to get information */ |
michael@0 | 1351 | for (pairp = tagp->attList; pairp; pairp = pairp->next) { |
michael@0 | 1352 | |
michael@0 | 1353 | /* ARCHIVE= */ |
michael@0 | 1354 | if ( !PL_strcasecmp(pairp->attribute, "archive")) { |
michael@0 | 1355 | if (archiveDir) { |
michael@0 | 1356 | /* Duplicate attribute. Print warning */ |
michael@0 | 1357 | PR_fprintf(errorFD, |
michael@0 | 1358 | "warning: \"%s\" attribute overwrites previous attribute" |
michael@0 | 1359 | " in tag starting at %s:%d.\n", |
michael@0 | 1360 | pairp->attribute, filename, curitem->startLine); |
michael@0 | 1361 | warningCount++; |
michael@0 | 1362 | PR_Free(archiveDir); |
michael@0 | 1363 | } |
michael@0 | 1364 | archiveDir = PL_strdup(pairp->value); |
michael@0 | 1365 | |
michael@0 | 1366 | /* Substiture ".arc" for ".jar" */ |
michael@0 | 1367 | if ( (PL_strlen(archiveDir) < 4) || |
michael@0 | 1368 | PL_strcasecmp((archiveDir + strlen(archiveDir) -4), |
michael@0 | 1369 | ".jar")) { |
michael@0 | 1370 | PR_fprintf(errorFD, |
michael@0 | 1371 | "warning: ARCHIVE attribute should end in \".jar\" in tag" |
michael@0 | 1372 | " starting on %s:%d.\n", filename, curitem->startLine); |
michael@0 | 1373 | warningCount++; |
michael@0 | 1374 | PR_Free(archiveDir); |
michael@0 | 1375 | archiveDir = PR_smprintf("%s.arc", archiveDir); |
michael@0 | 1376 | } else { |
michael@0 | 1377 | PL_strcpy(archiveDir + strlen(archiveDir) -4, ".arc"); |
michael@0 | 1378 | } |
michael@0 | 1379 | |
michael@0 | 1380 | /* Record the first archive. This will be used later if |
michael@0 | 1381 | * the archive is not specified */ |
michael@0 | 1382 | if (firstArchiveDir == NULL) { |
michael@0 | 1383 | firstArchiveDir = PL_strdup(archiveDir); |
michael@0 | 1384 | } |
michael@0 | 1385 | } |
michael@0 | 1386 | /* CODEBASE= */ |
michael@0 | 1387 | else if ( !PL_strcasecmp(pairp->attribute, "codebase")) { |
michael@0 | 1388 | if (codebase) { |
michael@0 | 1389 | /* Duplicate attribute. Print warning */ |
michael@0 | 1390 | PR_fprintf(errorFD, |
michael@0 | 1391 | "warning: \"%s\" attribute overwrites previous attribute" |
michael@0 | 1392 | " in tag staring at %s:%d.\n", |
michael@0 | 1393 | pairp->attribute, filename, curitem->startLine); |
michael@0 | 1394 | warningCount++; |
michael@0 | 1395 | } |
michael@0 | 1396 | codebase = pairp->value; |
michael@0 | 1397 | } |
michael@0 | 1398 | /* SRC= and HREF= */ |
michael@0 | 1399 | else if ( !PORT_Strcasecmp(pairp->attribute, "src") || |
michael@0 | 1400 | !PORT_Strcasecmp(pairp->attribute, "href") ) { |
michael@0 | 1401 | if (src) { |
michael@0 | 1402 | /* Duplicate attribute. Print warning */ |
michael@0 | 1403 | PR_fprintf(errorFD, |
michael@0 | 1404 | "warning: \"%s\" attribute overwrites previous attribute" |
michael@0 | 1405 | " in tag staring at %s:%d.\n", |
michael@0 | 1406 | pairp->attribute, filename, curitem->startLine); |
michael@0 | 1407 | warningCount++; |
michael@0 | 1408 | } |
michael@0 | 1409 | src = pairp->value; |
michael@0 | 1410 | } |
michael@0 | 1411 | /* CODE= */ |
michael@0 | 1412 | else if (!PORT_Strcasecmp(pairp->attribute, "code") ) { |
michael@0 | 1413 | /*!!!XXX Change PORT to PL all over this code !!! */ |
michael@0 | 1414 | if (src) { |
michael@0 | 1415 | /* Duplicate attribute. Print warning */ |
michael@0 | 1416 | PR_fprintf(errorFD, |
michael@0 | 1417 | "warning: \"%s\" attribute overwrites previous attribute" |
michael@0 | 1418 | " ,in tag staring at %s:%d.\n", |
michael@0 | 1419 | pairp->attribute, filename, curitem->startLine); |
michael@0 | 1420 | warningCount++; |
michael@0 | 1421 | } |
michael@0 | 1422 | src = pairp->value; |
michael@0 | 1423 | |
michael@0 | 1424 | /* Append a .class if one is not already present */ |
michael@0 | 1425 | if ( (PL_strlen(src) < 6) || |
michael@0 | 1426 | PL_strcasecmp( (src + PL_strlen(src) - 6), ".class") ) { |
michael@0 | 1427 | src = PR_smprintf("%s.class", src); |
michael@0 | 1428 | /* Put this string back into the data structure so it |
michael@0 | 1429 | * will be deallocated properly */ |
michael@0 | 1430 | PR_Free(pairp->value); |
michael@0 | 1431 | pairp->value = src; |
michael@0 | 1432 | } |
michael@0 | 1433 | } |
michael@0 | 1434 | /* ID= */ |
michael@0 | 1435 | else if (!PL_strcasecmp(pairp->attribute, "id") ) { |
michael@0 | 1436 | if (id) { |
michael@0 | 1437 | /* Duplicate attribute. Print warning */ |
michael@0 | 1438 | PR_fprintf(errorFD, |
michael@0 | 1439 | "warning: \"%s\" attribute overwrites previous attribute" |
michael@0 | 1440 | " in tag staring at %s:%d.\n", |
michael@0 | 1441 | pairp->attribute, filename, curitem->startLine); |
michael@0 | 1442 | warningCount++; |
michael@0 | 1443 | } |
michael@0 | 1444 | id = pairp->value; |
michael@0 | 1445 | } |
michael@0 | 1446 | |
michael@0 | 1447 | /* STYLE= */ |
michael@0 | 1448 | /* style= attributes, along with JS entities, are stored into |
michael@0 | 1449 | * files with dynamically generated names. The filenames are |
michael@0 | 1450 | * based on the order in which the text is found in the file. |
michael@0 | 1451 | * All JS entities on all lines up to and including the line |
michael@0 | 1452 | * containing the end of the tag that has this style= attribute |
michael@0 | 1453 | * will be processed before this style=attribute. So we need |
michael@0 | 1454 | * to record the line that this _tag_ (not the attribute) ends on. |
michael@0 | 1455 | */ |
michael@0 | 1456 | else if (!PL_strcasecmp(pairp->attribute, "style") && pairp->value) |
michael@0 | 1457 | { |
michael@0 | 1458 | HTMLItem * styleItem; |
michael@0 | 1459 | /* Put this item on the style list */ |
michael@0 | 1460 | styleItem = CreateTextItem(PL_strdup(pairp->value), |
michael@0 | 1461 | curitem->startLine, curitem->endLine); |
michael@0 | 1462 | if (styleListTail == NULL) { |
michael@0 | 1463 | styleList = styleListTail = styleItem; |
michael@0 | 1464 | } else { |
michael@0 | 1465 | styleListTail->next = styleItem; |
michael@0 | 1466 | styleListTail = styleItem; |
michael@0 | 1467 | } |
michael@0 | 1468 | } |
michael@0 | 1469 | /* Event handlers */ |
michael@0 | 1470 | else { |
michael@0 | 1471 | for (i = 0; i < num_handlers; i++) { |
michael@0 | 1472 | if (!PL_strcasecmp(event_handlers[i], pairp->attribute)) { |
michael@0 | 1473 | hasEventHandler = PR_TRUE; |
michael@0 | 1474 | break; |
michael@0 | 1475 | } |
michael@0 | 1476 | } |
michael@0 | 1477 | } |
michael@0 | 1478 | |
michael@0 | 1479 | |
michael@0 | 1480 | /* JS Entity */ |
michael@0 | 1481 | { |
michael@0 | 1482 | char *entityStart, *entityEnd; |
michael@0 | 1483 | HTMLItem * entityItem; |
michael@0 | 1484 | |
michael@0 | 1485 | /* go through each JavaScript entity ( &{...}; ) and store it |
michael@0 | 1486 | * in the entityList. The important thing is to record what |
michael@0 | 1487 | * line number it's on, so we can get it in the right order |
michael@0 | 1488 | * in relation to style= attributes. |
michael@0 | 1489 | * Apparently, these can't flow across lines, so the start and |
michael@0 | 1490 | * end line will be the same. That helps matters. |
michael@0 | 1491 | */ |
michael@0 | 1492 | entityEnd = pairp->value; |
michael@0 | 1493 | while ( entityEnd && |
michael@0 | 1494 | (entityStart = PL_strstr(entityEnd, "&{")) /*}*/ != NULL) { |
michael@0 | 1495 | entityStart += 2; /* point at beginning of actual entity */ |
michael@0 | 1496 | entityEnd = PL_strchr(entityStart, '}'); |
michael@0 | 1497 | if (entityEnd) { |
michael@0 | 1498 | /* Put this item on the entity list */ |
michael@0 | 1499 | *entityEnd = '\0'; |
michael@0 | 1500 | entityItem = CreateTextItem(PL_strdup(entityStart), |
michael@0 | 1501 | pairp->valueLine, pairp->valueLine); |
michael@0 | 1502 | *entityEnd = /* { */ '}'; |
michael@0 | 1503 | if (entityListTail) { |
michael@0 | 1504 | entityListTail->next = entityItem; |
michael@0 | 1505 | entityListTail = entityItem; |
michael@0 | 1506 | } else { |
michael@0 | 1507 | entityList = entityListTail = entityItem; |
michael@0 | 1508 | } |
michael@0 | 1509 | } |
michael@0 | 1510 | } |
michael@0 | 1511 | } |
michael@0 | 1512 | } |
michael@0 | 1513 | |
michael@0 | 1514 | /* If no archive was supplied, we use the first one of the file */ |
michael@0 | 1515 | if (!archiveDir && firstArchiveDir) { |
michael@0 | 1516 | archiveDir = PL_strdup(firstArchiveDir); |
michael@0 | 1517 | } |
michael@0 | 1518 | |
michael@0 | 1519 | /* If we have an event handler, we need to archive this tag */ |
michael@0 | 1520 | if (hasEventHandler) { |
michael@0 | 1521 | if (!id) { |
michael@0 | 1522 | PR_fprintf(errorFD, |
michael@0 | 1523 | "warning: tag starting at %s:%d has event handler but" |
michael@0 | 1524 | " no ID attribute. The tag will not be signed.\n", |
michael@0 | 1525 | filename, curitem->startLine); |
michael@0 | 1526 | warningCount++; |
michael@0 | 1527 | } else if (!archiveDir) { |
michael@0 | 1528 | PR_fprintf(errorFD, |
michael@0 | 1529 | "warning: tag starting at %s:%d has event handler but" |
michael@0 | 1530 | " no ARCHIVE attribute. The tag will not be signed.\n", |
michael@0 | 1531 | filename, curitem->startLine); |
michael@0 | 1532 | warningCount++; |
michael@0 | 1533 | } else { |
michael@0 | 1534 | if (SaveInlineScript(tagp->text, id, basedir, archiveDir)) { |
michael@0 | 1535 | goto loser; |
michael@0 | 1536 | } |
michael@0 | 1537 | } |
michael@0 | 1538 | } |
michael@0 | 1539 | |
michael@0 | 1540 | switch (tagp->type) { |
michael@0 | 1541 | case APPLET_TAG: |
michael@0 | 1542 | if (!src) { |
michael@0 | 1543 | PR_fprintf(errorFD, |
michael@0 | 1544 | "error: APPLET tag starting on %s:%d has no CODE " |
michael@0 | 1545 | "attribute.\n", filename, curitem->startLine); |
michael@0 | 1546 | errorCount++; |
michael@0 | 1547 | goto loser; |
michael@0 | 1548 | } else if (!archiveDir) { |
michael@0 | 1549 | PR_fprintf(errorFD, |
michael@0 | 1550 | "error: APPLET tag starting on %s:%d has no ARCHIVE " |
michael@0 | 1551 | "attribute.\n", filename, curitem->startLine); |
michael@0 | 1552 | errorCount++; |
michael@0 | 1553 | goto loser; |
michael@0 | 1554 | } else { |
michael@0 | 1555 | if (SaveSource(src, codebase, basedir, archiveDir)) { |
michael@0 | 1556 | goto loser; |
michael@0 | 1557 | } |
michael@0 | 1558 | } |
michael@0 | 1559 | break; |
michael@0 | 1560 | case SCRIPT_TAG: |
michael@0 | 1561 | case LINK_TAG: |
michael@0 | 1562 | case STYLE_TAG: |
michael@0 | 1563 | if (!archiveDir) { |
michael@0 | 1564 | PR_fprintf(errorFD, |
michael@0 | 1565 | "error: %s tag starting on %s:%d has no ARCHIVE " |
michael@0 | 1566 | "attribute.\n", TagTypeToString(tagp->type), |
michael@0 | 1567 | filename, curitem->startLine); |
michael@0 | 1568 | errorCount++; |
michael@0 | 1569 | goto loser; |
michael@0 | 1570 | } else if (src) { |
michael@0 | 1571 | if (SaveSource(src, codebase, basedir, archiveDir)) { |
michael@0 | 1572 | goto loser; |
michael@0 | 1573 | } |
michael@0 | 1574 | } else if (id) { |
michael@0 | 1575 | /* Save the next text item */ |
michael@0 | 1576 | if (!curitem->next || (curitem->next->type != |
michael@0 | 1577 | TEXT_ITEM)) { |
michael@0 | 1578 | PR_fprintf(errorFD, |
michael@0 | 1579 | "warning: %s tag starting on %s:%d is not followed" |
michael@0 | 1580 | " by script text.\n", TagTypeToString(tagp->type), |
michael@0 | 1581 | filename, curitem->startLine); |
michael@0 | 1582 | warningCount++; |
michael@0 | 1583 | /* just create empty file */ |
michael@0 | 1584 | if (SaveInlineScript("", id, basedir, archiveDir)) { |
michael@0 | 1585 | goto loser; |
michael@0 | 1586 | } |
michael@0 | 1587 | } else { |
michael@0 | 1588 | curitem = curitem->next; |
michael@0 | 1589 | if (SaveInlineScript(curitem->item.text, |
michael@0 | 1590 | id, basedir, |
michael@0 | 1591 | archiveDir)) { |
michael@0 | 1592 | goto loser; |
michael@0 | 1593 | } |
michael@0 | 1594 | } |
michael@0 | 1595 | } else { |
michael@0 | 1596 | /* No src or id tag--warning */ |
michael@0 | 1597 | PR_fprintf(errorFD, |
michael@0 | 1598 | "warning: %s tag starting on %s:%d has no SRC or" |
michael@0 | 1599 | " ID attributes. Will not sign.\n", |
michael@0 | 1600 | TagTypeToString(tagp->type), filename, curitem->startLine); |
michael@0 | 1601 | warningCount++; |
michael@0 | 1602 | } |
michael@0 | 1603 | break; |
michael@0 | 1604 | default: |
michael@0 | 1605 | /* do nothing for other tags */ |
michael@0 | 1606 | break; |
michael@0 | 1607 | } |
michael@0 | 1608 | |
michael@0 | 1609 | } |
michael@0 | 1610 | |
michael@0 | 1611 | /* Now deal with all the unnamable scripts */ |
michael@0 | 1612 | if (firstArchiveDir) { |
michael@0 | 1613 | HTMLItem * style, *entity; |
michael@0 | 1614 | |
michael@0 | 1615 | /* Go through the lists of JS entities and style attributes. Do them |
michael@0 | 1616 | * in chronological order within a list. Pick the list with the lower |
michael@0 | 1617 | * endLine. In case of a tie, entities come first. |
michael@0 | 1618 | */ |
michael@0 | 1619 | style = styleList; |
michael@0 | 1620 | entity = entityList; |
michael@0 | 1621 | while (style || entity) { |
michael@0 | 1622 | if (!entity || (style && (style->endLine < entity->endLine))) { |
michael@0 | 1623 | /* Process style */ |
michael@0 | 1624 | SaveUnnamableScript(style->item.text, basedir, firstArchiveDir, |
michael@0 | 1625 | filename); |
michael@0 | 1626 | style = style->next; |
michael@0 | 1627 | } else { |
michael@0 | 1628 | /* Process entity */ |
michael@0 | 1629 | SaveUnnamableScript(entity->item.text, basedir, firstArchiveDir, |
michael@0 | 1630 | filename); |
michael@0 | 1631 | entity = entity->next; |
michael@0 | 1632 | } |
michael@0 | 1633 | } |
michael@0 | 1634 | } |
michael@0 | 1635 | |
michael@0 | 1636 | |
michael@0 | 1637 | retval = 0; |
michael@0 | 1638 | loser: |
michael@0 | 1639 | /* Blow away the stream */ |
michael@0 | 1640 | while (head) { |
michael@0 | 1641 | curitem = head; |
michael@0 | 1642 | head = head->next; |
michael@0 | 1643 | DestroyHTMLItem(curitem); |
michael@0 | 1644 | } |
michael@0 | 1645 | while (styleList) { |
michael@0 | 1646 | curitem = styleList; |
michael@0 | 1647 | styleList = styleList->next; |
michael@0 | 1648 | DestroyHTMLItem(curitem); |
michael@0 | 1649 | } |
michael@0 | 1650 | while (entityList) { |
michael@0 | 1651 | curitem = entityList; |
michael@0 | 1652 | entityList = entityList->next; |
michael@0 | 1653 | DestroyHTMLItem(curitem); |
michael@0 | 1654 | } |
michael@0 | 1655 | if (text) { |
michael@0 | 1656 | PR_Free(text); |
michael@0 | 1657 | text = NULL; |
michael@0 | 1658 | } |
michael@0 | 1659 | if (fb) { |
michael@0 | 1660 | FB_Destroy(fb); |
michael@0 | 1661 | fb = NULL; |
michael@0 | 1662 | } |
michael@0 | 1663 | if (fd) { |
michael@0 | 1664 | PR_Close(fd); |
michael@0 | 1665 | } |
michael@0 | 1666 | if (tagerr) { |
michael@0 | 1667 | PR_smprintf_free(tagerr); |
michael@0 | 1668 | tagerr = NULL; |
michael@0 | 1669 | } |
michael@0 | 1670 | if (archiveDir) { |
michael@0 | 1671 | PR_Free(archiveDir); |
michael@0 | 1672 | archiveDir = NULL; |
michael@0 | 1673 | } |
michael@0 | 1674 | if (firstArchiveDir) { |
michael@0 | 1675 | PR_Free(firstArchiveDir); |
michael@0 | 1676 | firstArchiveDir = NULL; |
michael@0 | 1677 | } |
michael@0 | 1678 | return retval; |
michael@0 | 1679 | } |
michael@0 | 1680 | |
michael@0 | 1681 | |
michael@0 | 1682 | /********************************************************************** |
michael@0 | 1683 | * |
michael@0 | 1684 | * e n s u r e E x i s t s |
michael@0 | 1685 | * |
michael@0 | 1686 | * Check for existence of indicated directory. If it doesn't exist, |
michael@0 | 1687 | * it will be created. |
michael@0 | 1688 | * Returns PR_SUCCESS if the directory is present, PR_FAILURE otherwise. |
michael@0 | 1689 | */ |
michael@0 | 1690 | static PRStatus |
michael@0 | 1691 | ensureExists (char *base, char *path) |
michael@0 | 1692 | { |
michael@0 | 1693 | char fn [FNSIZE]; |
michael@0 | 1694 | PRDir * dir; |
michael@0 | 1695 | sprintf (fn, "%s/%s", base, path); |
michael@0 | 1696 | |
michael@0 | 1697 | /*PR_fprintf(outputFD, "Trying to open directory %s.\n", fn);*/ |
michael@0 | 1698 | |
michael@0 | 1699 | if ( (dir = PR_OpenDir(fn)) ) { |
michael@0 | 1700 | PR_CloseDir(dir); |
michael@0 | 1701 | return PR_SUCCESS; |
michael@0 | 1702 | } |
michael@0 | 1703 | return PR_MkDir(fn, 0777); |
michael@0 | 1704 | } |
michael@0 | 1705 | |
michael@0 | 1706 | |
michael@0 | 1707 | /*************************************************************************** |
michael@0 | 1708 | * |
michael@0 | 1709 | * m a k e _ d i r s |
michael@0 | 1710 | * |
michael@0 | 1711 | * Ensure that the directory portion of the path exists. This may require |
michael@0 | 1712 | * making the directory, and its parent, and its parent's parent, etc. |
michael@0 | 1713 | */ |
michael@0 | 1714 | static int |
michael@0 | 1715 | make_dirs(char *path, int file_perms) |
michael@0 | 1716 | { |
michael@0 | 1717 | char *Path; |
michael@0 | 1718 | char *start; |
michael@0 | 1719 | char *sep; |
michael@0 | 1720 | int ret = 0; |
michael@0 | 1721 | PRFileInfo info; |
michael@0 | 1722 | |
michael@0 | 1723 | if (!path) { |
michael@0 | 1724 | return 0; |
michael@0 | 1725 | } |
michael@0 | 1726 | |
michael@0 | 1727 | Path = PL_strdup(path); |
michael@0 | 1728 | start = strpbrk(Path, "/\\"); |
michael@0 | 1729 | if (!start) { |
michael@0 | 1730 | return 0; |
michael@0 | 1731 | } |
michael@0 | 1732 | start++; /* start right after first slash */ |
michael@0 | 1733 | |
michael@0 | 1734 | /* Each time through the loop add one more directory. */ |
michael@0 | 1735 | while ( (sep = strpbrk(start, "/\\")) ) { |
michael@0 | 1736 | *sep = '\0'; |
michael@0 | 1737 | |
michael@0 | 1738 | if ( PR_GetFileInfo(Path, &info) != PR_SUCCESS) { |
michael@0 | 1739 | /* No such dir, we have to create it */ |
michael@0 | 1740 | if ( PR_MkDir(Path, file_perms) != PR_SUCCESS) { |
michael@0 | 1741 | PR_fprintf(errorFD, "ERROR: Unable to create directory %s.\n", |
michael@0 | 1742 | Path); |
michael@0 | 1743 | errorCount++; |
michael@0 | 1744 | ret = -1; |
michael@0 | 1745 | goto loser; |
michael@0 | 1746 | } |
michael@0 | 1747 | } else { |
michael@0 | 1748 | /* something exists by this name, make sure it's a directory */ |
michael@0 | 1749 | if ( info.type != PR_FILE_DIRECTORY ) { |
michael@0 | 1750 | PR_fprintf(errorFD, "ERROR: Unable to create directory %s.\n", |
michael@0 | 1751 | Path); |
michael@0 | 1752 | errorCount++; |
michael@0 | 1753 | ret = -1; |
michael@0 | 1754 | goto loser; |
michael@0 | 1755 | } |
michael@0 | 1756 | } |
michael@0 | 1757 | |
michael@0 | 1758 | start = sep + 1; /* start after the next slash */ |
michael@0 | 1759 | *sep = '/'; |
michael@0 | 1760 | } |
michael@0 | 1761 | |
michael@0 | 1762 | loser: |
michael@0 | 1763 | PR_Free(Path); |
michael@0 | 1764 | return ret; |
michael@0 | 1765 | } |
michael@0 | 1766 | |
michael@0 | 1767 | |
michael@0 | 1768 | /* |
michael@0 | 1769 | * c o p y i n t o |
michael@0 | 1770 | * |
michael@0 | 1771 | * Function to copy file "from" to path "to". |
michael@0 | 1772 | * |
michael@0 | 1773 | */ |
michael@0 | 1774 | static int |
michael@0 | 1775 | copyinto (char *from, char *to) |
michael@0 | 1776 | { |
michael@0 | 1777 | PRInt32 num; |
michael@0 | 1778 | char buf [BUFSIZ]; |
michael@0 | 1779 | PRFileDesc * infp = NULL, *outfp = NULL; |
michael@0 | 1780 | int retval = -1; |
michael@0 | 1781 | |
michael@0 | 1782 | if ((infp = PR_Open(from, PR_RDONLY, 0777)) == NULL) { |
michael@0 | 1783 | PR_fprintf(errorFD, "ERROR: Unable to open \"%s\" for reading.\n", |
michael@0 | 1784 | from); |
michael@0 | 1785 | errorCount++; |
michael@0 | 1786 | goto finish; |
michael@0 | 1787 | } |
michael@0 | 1788 | |
michael@0 | 1789 | /* If to already exists, print a warning before deleting it */ |
michael@0 | 1790 | if (PR_Access(to, PR_ACCESS_EXISTS) == PR_SUCCESS) { |
michael@0 | 1791 | PR_fprintf(errorFD, "warning: %s already exists--will overwrite\n", to); |
michael@0 | 1792 | warningCount++; |
michael@0 | 1793 | if (rm_dash_r(to)) { |
michael@0 | 1794 | PR_fprintf(errorFD, |
michael@0 | 1795 | "ERROR: Unable to remove %s.\n", to); |
michael@0 | 1796 | errorCount++; |
michael@0 | 1797 | goto finish; |
michael@0 | 1798 | } |
michael@0 | 1799 | } |
michael@0 | 1800 | |
michael@0 | 1801 | if ((outfp = PR_Open(to, PR_WRONLY | PR_CREATE_FILE | PR_TRUNCATE, 0777)) |
michael@0 | 1802 | == NULL) { |
michael@0 | 1803 | char *errBuf = NULL; |
michael@0 | 1804 | |
michael@0 | 1805 | errBuf = PR_Malloc(PR_GetErrorTextLength() + 1); |
michael@0 | 1806 | PR_fprintf(errorFD, "ERROR: Unable to open \"%s\" for writing.\n", to); |
michael@0 | 1807 | if (PR_GetErrorText(errBuf)) { |
michael@0 | 1808 | PR_fprintf(errorFD, "Cause: %s\n", errBuf); |
michael@0 | 1809 | } |
michael@0 | 1810 | if (errBuf) { |
michael@0 | 1811 | PR_Free(errBuf); |
michael@0 | 1812 | } |
michael@0 | 1813 | errorCount++; |
michael@0 | 1814 | goto finish; |
michael@0 | 1815 | } |
michael@0 | 1816 | |
michael@0 | 1817 | while ( (num = PR_Read(infp, buf, BUFSIZ)) > 0) { |
michael@0 | 1818 | if (PR_Write(outfp, buf, num) != num) { |
michael@0 | 1819 | PR_fprintf(errorFD, "ERROR: Error writing to %s.\n", to); |
michael@0 | 1820 | errorCount++; |
michael@0 | 1821 | goto finish; |
michael@0 | 1822 | } |
michael@0 | 1823 | } |
michael@0 | 1824 | |
michael@0 | 1825 | retval = 0; |
michael@0 | 1826 | finish: |
michael@0 | 1827 | if (infp) |
michael@0 | 1828 | PR_Close(infp); |
michael@0 | 1829 | if (outfp) |
michael@0 | 1830 | PR_Close(outfp); |
michael@0 | 1831 | |
michael@0 | 1832 | return retval; |
michael@0 | 1833 | } |
michael@0 | 1834 | |
michael@0 | 1835 |