1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/security/nss/cmd/signtool/javascript.c Wed Dec 31 06:09:35 2014 +0100
1.3 @@ -0,0 +1,1835 @@
1.4 +/* This Source Code Form is subject to the terms of the Mozilla Public
1.5 + * License, v. 2.0. If a copy of the MPL was not distributed with this
1.6 + * file, You can obtain one at http://mozilla.org/MPL/2.0/. */
1.7 +
1.8 +#include "signtool.h"
1.9 +#include <prmem.h>
1.10 +#include <prio.h>
1.11 +#include <prenv.h>
1.12 +
1.13 +static int javascript_fn(char *relpath, char *basedir, char *reldir,
1.14 +char *filename, void *arg);
1.15 +static int extract_js (char *filename);
1.16 +static int copyinto (char *from, char *to);
1.17 +static PRStatus ensureExists (char *base, char *path);
1.18 +static int make_dirs(char *path, PRInt32 file_perms);
1.19 +
1.20 +static char *jartree = NULL;
1.21 +static int idOrdinal;
1.22 +static PRBool dumpParse = PR_FALSE;
1.23 +
1.24 +static char *event_handlers[] = {
1.25 + "onAbort",
1.26 + "onBlur",
1.27 + "onChange",
1.28 + "onClick",
1.29 + "onDblClick",
1.30 + "onDragDrop",
1.31 + "onError",
1.32 + "onFocus",
1.33 + "onKeyDown",
1.34 + "onKeyPress",
1.35 + "onKeyUp",
1.36 + "onLoad",
1.37 + "onMouseDown",
1.38 + "onMouseMove",
1.39 + "onMouseOut",
1.40 + "onMouseOver",
1.41 + "onMouseUp",
1.42 + "onMove",
1.43 + "onReset",
1.44 + "onResize",
1.45 + "onSelect",
1.46 + "onSubmit",
1.47 + "onUnload"
1.48 +};
1.49 +
1.50 +
1.51 +static int num_handlers = 23;
1.52 +
1.53 +/*
1.54 + * I n l i n e J a v a S c r i p t
1.55 + *
1.56 + * Javascript signing. Instead of passing an archive to signtool,
1.57 + * a directory containing html files is given. Archives are created
1.58 + * from the archive= and src= tag attributes inside the html,
1.59 + * as appropriate. Then the archives are signed.
1.60 + *
1.61 + */
1.62 +int
1.63 +InlineJavaScript(char *dir, PRBool recurse)
1.64 +{
1.65 + jartree = dir;
1.66 + if (verbosity >= 0) {
1.67 + PR_fprintf(outputFD, "\nGenerating inline signatures from HTML files in: %s\n",
1.68 + dir);
1.69 + }
1.70 + if (PR_GetEnv("SIGNTOOL_DUMP_PARSE")) {
1.71 + dumpParse = PR_TRUE;
1.72 + }
1.73 +
1.74 + return foreach(dir, "", javascript_fn, recurse, PR_FALSE /*include dirs*/,
1.75 + (void * )NULL);
1.76 +
1.77 +}
1.78 +
1.79 +
1.80 +/************************************************************************
1.81 + *
1.82 + * j a v a s c r i p t _ f n
1.83 + */
1.84 +static int javascript_fn
1.85 +(char *relpath, char *basedir, char *reldir, char *filename, void *arg)
1.86 +{
1.87 + char fullname [FNSIZE];
1.88 +
1.89 + /* only process inline scripts from .htm, .html, and .shtml*/
1.90 +
1.91 + if (!(PL_strcaserstr(filename, ".htm") == filename + strlen(filename) -
1.92 + 4) &&
1.93 + !(PL_strcaserstr(filename, ".html") == filename + strlen(filename) -
1.94 + 5) &&
1.95 + !(PL_strcaserstr(filename, ".shtml") == filename + strlen(filename)
1.96 + -6)) {
1.97 + return 0;
1.98 + }
1.99 +
1.100 + /* don't process scripts that signtool has already
1.101 + extracted (those that are inside .arc directories) */
1.102 +
1.103 + if (PL_strcaserstr(filename, ".arc") == filename + strlen(filename) - 4)
1.104 + return 0;
1.105 +
1.106 + if (verbosity >= 0) {
1.107 + PR_fprintf(outputFD, "Processing HTML file: %s\n", relpath);
1.108 + }
1.109 +
1.110 + /* reset firstArchive at top of each HTML file */
1.111 +
1.112 + /* skip directories that contain extracted scripts */
1.113 +
1.114 + if (PL_strcaserstr(reldir, ".arc") == reldir + strlen(reldir) - 4)
1.115 + return 0;
1.116 +
1.117 + sprintf (fullname, "%s/%s", basedir, relpath);
1.118 + return extract_js (fullname);
1.119 +}
1.120 +
1.121 +
1.122 +/*===========================================================================
1.123 + =
1.124 + = D A T A S T R U C T U R E S
1.125 + =
1.126 +*/
1.127 +typedef enum {
1.128 + TEXT_HTML_STATE = 0,
1.129 + SCRIPT_HTML_STATE
1.130 +}
1.131 +
1.132 +
1.133 +HTML_STATE ;
1.134 +
1.135 +typedef enum {
1.136 + /* we start in the start state */
1.137 + START_STATE,
1.138 +
1.139 + /* We are looking for or reading in an attribute */
1.140 + GET_ATT_STATE,
1.141 +
1.142 + /* We're burning ws before finding an attribute */
1.143 + PRE_ATT_WS_STATE,
1.144 +
1.145 + /* We're burning ws after an attribute. Looking for an '='. */
1.146 + POST_ATT_WS_STATE,
1.147 +
1.148 + /* We're burning ws after an '=', waiting for a value */
1.149 + PRE_VAL_WS_STATE,
1.150 +
1.151 + /* We're reading in a value */
1.152 + GET_VALUE_STATE,
1.153 +
1.154 + /* We're reading in a value that's inside quotes */
1.155 + GET_QUOTED_VAL_STATE,
1.156 +
1.157 + /* We've encountered the closing '>' */
1.158 + DONE_STATE,
1.159 +
1.160 + /* Error state */
1.161 + ERR_STATE
1.162 +}
1.163 +
1.164 +
1.165 +TAG_STATE ;
1.166 +
1.167 +typedef struct AVPair_Str {
1.168 + char *attribute;
1.169 + char *value;
1.170 + unsigned int valueLine; /* the line that the value ends on */
1.171 + struct AVPair_Str *next;
1.172 +} AVPair;
1.173 +
1.174 +typedef enum {
1.175 + APPLET_TAG,
1.176 + SCRIPT_TAG,
1.177 + LINK_TAG,
1.178 + STYLE_TAG,
1.179 + COMMENT_TAG,
1.180 + OTHER_TAG
1.181 +}
1.182 +
1.183 +
1.184 +TAG_TYPE ;
1.185 +
1.186 +typedef struct {
1.187 + TAG_TYPE type;
1.188 + AVPair * attList;
1.189 + AVPair * attListTail;
1.190 + char *text;
1.191 +} TagItem;
1.192 +
1.193 +typedef enum {
1.194 + TAG_ITEM,
1.195 + TEXT_ITEM
1.196 +}
1.197 +
1.198 +
1.199 +ITEM_TYPE ;
1.200 +
1.201 +typedef struct HTMLItem_Str {
1.202 + unsigned int startLine;
1.203 + unsigned int endLine;
1.204 + ITEM_TYPE type;
1.205 + union {
1.206 + TagItem *tag;
1.207 + char *text;
1.208 + } item;
1.209 + struct HTMLItem_Str *next;
1.210 +} HTMLItem;
1.211 +
1.212 +typedef struct {
1.213 + PRFileDesc *fd;
1.214 + PRInt32 curIndex;
1.215 + PRBool IsEOF;
1.216 +#define FILE_BUFFER_BUFSIZE 512
1.217 + char buf[FILE_BUFFER_BUFSIZE];
1.218 + PRInt32 startOffset;
1.219 + PRInt32 maxIndex;
1.220 + unsigned int lineNum;
1.221 +} FileBuffer;
1.222 +
1.223 +/*===========================================================================
1.224 + =
1.225 + = F U N C T I O N S
1.226 + =
1.227 +*/
1.228 +static HTMLItem*CreateTextItem(char *text, unsigned int startline,
1.229 +unsigned int endline);
1.230 +static HTMLItem*CreateTagItem(TagItem*ti, unsigned int startline,
1.231 +unsigned int endline);
1.232 +static TagItem*ProcessTag(FileBuffer*fb, char **errStr);
1.233 +static void DestroyHTMLItem(HTMLItem *item);
1.234 +static void DestroyTagItem(TagItem*ti);
1.235 +static TAG_TYPE GetTagType(char *att);
1.236 +static FileBuffer*FB_Create(PRFileDesc*fd);
1.237 +static int FB_GetChar(FileBuffer *fb);
1.238 +static PRInt32 FB_GetPointer(FileBuffer *fb);
1.239 +static PRInt32 FB_GetRange(FileBuffer *fb, PRInt32 start, PRInt32 end,
1.240 +char **buf);
1.241 +static unsigned int FB_GetLineNum(FileBuffer *fb);
1.242 +static void FB_Destroy(FileBuffer *fb);
1.243 +static void PrintTagItem(PRFileDesc *fd, TagItem *ti);
1.244 +static void PrintHTMLStream(PRFileDesc *fd, HTMLItem *head);
1.245 +
1.246 +/************************************************************************
1.247 + *
1.248 + * C r e a t e T e x t I t e m
1.249 + */
1.250 +static HTMLItem*
1.251 +CreateTextItem(char *text, unsigned int startline, unsigned int endline)
1.252 +{
1.253 + HTMLItem * item;
1.254 +
1.255 + item = PR_Malloc(sizeof(HTMLItem));
1.256 + if (!item) {
1.257 + return NULL;
1.258 + }
1.259 +
1.260 + item->type = TEXT_ITEM;
1.261 + item->item.text = text;
1.262 + item->next = NULL;
1.263 + item->startLine = startline;
1.264 + item->endLine = endline;
1.265 +
1.266 + return item;
1.267 +}
1.268 +
1.269 +
1.270 +/************************************************************************
1.271 + *
1.272 + * C r e a t e T a g I t e m
1.273 + */
1.274 +static HTMLItem*
1.275 +CreateTagItem(TagItem*ti, unsigned int startline, unsigned int endline)
1.276 +{
1.277 + HTMLItem * item;
1.278 +
1.279 + item = PR_Malloc(sizeof(HTMLItem));
1.280 + if (!item) {
1.281 + return NULL;
1.282 + }
1.283 +
1.284 + item->type = TAG_ITEM;
1.285 + item->item.tag = ti;
1.286 + item->next = NULL;
1.287 + item->startLine = startline;
1.288 + item->endLine = endline;
1.289 +
1.290 + return item;
1.291 +}
1.292 +
1.293 +
1.294 +static PRBool
1.295 +isAttChar(int c)
1.296 +{
1.297 + return (isalnum(c) || c == '/' || c == '-');
1.298 +}
1.299 +
1.300 +
1.301 +/************************************************************************
1.302 + *
1.303 + * P r o c e s s T a g
1.304 + */
1.305 +static TagItem*
1.306 +ProcessTag(FileBuffer*fb, char **errStr)
1.307 +{
1.308 + TAG_STATE state;
1.309 + PRInt32 startText, startID, curPos;
1.310 + PRBool firstAtt;
1.311 + int curchar;
1.312 + TagItem * ti = NULL;
1.313 + AVPair * curPair = NULL;
1.314 + char quotechar = '\0';
1.315 + unsigned int linenum;
1.316 + unsigned int startline;
1.317 +
1.318 + state = START_STATE;
1.319 +
1.320 + startID = FB_GetPointer(fb);
1.321 + startText = startID;
1.322 + firstAtt = PR_TRUE;
1.323 +
1.324 + ti = (TagItem * ) PR_Malloc(sizeof(TagItem));
1.325 + if (!ti)
1.326 + out_of_memory();
1.327 + ti->type = OTHER_TAG;
1.328 + ti->attList = NULL;
1.329 + ti->attListTail = NULL;
1.330 + ti->text = NULL;
1.331 +
1.332 + startline = FB_GetLineNum(fb);
1.333 +
1.334 + while (state != DONE_STATE && state != ERR_STATE) {
1.335 + linenum = FB_GetLineNum(fb);
1.336 + curchar = FB_GetChar(fb);
1.337 + if (curchar == EOF) {
1.338 + *errStr = PR_smprintf(
1.339 + "line %d: Unexpected end-of-file while parsing tag starting at line %d.\n",
1.340 + linenum, startline);
1.341 + state = ERR_STATE;
1.342 + continue;
1.343 + }
1.344 +
1.345 + switch (state) {
1.346 + case START_STATE:
1.347 + if (curchar == '!') {
1.348 + /*
1.349 + * SGML tag or comment
1.350 + * Here's the general rule for SGML tags. Everything from
1.351 + * <! to > is the tag. Inside the tag, comments are
1.352 + * delimited with --. So we are looking for the first '>'
1.353 + * that is not commented out, that is, not inside a pair
1.354 + * of --: <!DOCTYPE --this is a comment >(psyche!) -->
1.355 + */
1.356 +
1.357 + PRBool inComment = PR_FALSE;
1.358 + short hyphenCount = 0; /* number of consecutive hyphens */
1.359 +
1.360 + while (1) {
1.361 + linenum = FB_GetLineNum(fb);
1.362 + curchar = FB_GetChar(fb);
1.363 + if (curchar == EOF) {
1.364 + /* Uh oh, EOF inside comment */
1.365 + *errStr = PR_smprintf(
1.366 + "line %d: Unexpected end-of-file inside comment starting at line %d.\n",
1.367 + linenum, startline);
1.368 + state = ERR_STATE;
1.369 + break;
1.370 + }
1.371 + if (curchar == '-') {
1.372 + if (hyphenCount == 1) {
1.373 + /* This is a comment delimiter */
1.374 + inComment = !inComment;
1.375 + hyphenCount = 0;
1.376 + } else {
1.377 + /* beginning of a comment delimiter? */
1.378 + hyphenCount = 1;
1.379 + }
1.380 + } else if (curchar == '>') {
1.381 + if (!inComment) {
1.382 + /* This is the end of the tag */
1.383 + state = DONE_STATE;
1.384 + break;
1.385 + } else {
1.386 + /* The > is inside a comment, so it's not
1.387 + * really the end of the tag */
1.388 + hyphenCount = 0;
1.389 + }
1.390 + } else {
1.391 + hyphenCount = 0;
1.392 + }
1.393 + }
1.394 + ti->type = COMMENT_TAG;
1.395 + break;
1.396 + }
1.397 + /* fall through */
1.398 + case GET_ATT_STATE:
1.399 + if (isspace(curchar) || curchar == '=' || curchar
1.400 + == '>') {
1.401 + /* end of the current attribute */
1.402 + curPos = FB_GetPointer(fb) - 2;
1.403 + if (curPos >= startID) {
1.404 + /* We have an attribute */
1.405 + curPair = (AVPair * )PR_Malloc(sizeof(AVPair));
1.406 + if (!curPair)
1.407 + out_of_memory();
1.408 + curPair->value = NULL;
1.409 + curPair->next = NULL;
1.410 + FB_GetRange(fb, startID, curPos,
1.411 + &curPair->attribute);
1.412 +
1.413 + /* Stick this attribute on the list */
1.414 + if (ti->attListTail) {
1.415 + ti->attListTail->next = curPair;
1.416 + ti->attListTail = curPair;
1.417 + } else {
1.418 + ti->attList = ti->attListTail =
1.419 + curPair;
1.420 + }
1.421 +
1.422 + /* If this is the first attribute, find the type of tag
1.423 + * based on it. Also, start saving the text of the tag. */
1.424 + if (firstAtt) {
1.425 + ti->type = GetTagType(curPair->attribute);
1.426 + startText = FB_GetPointer(fb)
1.427 + -1;
1.428 + firstAtt = PR_FALSE;
1.429 + }
1.430 + } else {
1.431 + if (curchar == '=') {
1.432 + /* If we don't have any attribute but we do have an
1.433 + * equal sign, that's an error */
1.434 + *errStr = PR_smprintf("line %d: Malformed tag starting at line %d.\n",
1.435 + linenum, startline);
1.436 + state = ERR_STATE;
1.437 + break;
1.438 + }
1.439 + }
1.440 +
1.441 + /* Compute next state */
1.442 + if (curchar == '=') {
1.443 + startID = FB_GetPointer(fb);
1.444 + state = PRE_VAL_WS_STATE;
1.445 + } else if (curchar == '>') {
1.446 + state = DONE_STATE;
1.447 + } else if (curPair) {
1.448 + state = POST_ATT_WS_STATE;
1.449 + } else {
1.450 + state = PRE_ATT_WS_STATE;
1.451 + }
1.452 + } else if (isAttChar(curchar)) {
1.453 + /* Just another char in the attribute. Do nothing */
1.454 + state = GET_ATT_STATE;
1.455 + } else {
1.456 + /* bogus char */
1.457 + *errStr = PR_smprintf("line %d: Bogus chararacter '%c' in tag.\n",
1.458 + linenum, curchar);
1.459 + state = ERR_STATE;
1.460 + break;
1.461 + }
1.462 + break;
1.463 + case PRE_ATT_WS_STATE:
1.464 + if (curchar == '>') {
1.465 + state = DONE_STATE;
1.466 + } else if (isspace(curchar)) {
1.467 + /* more whitespace, do nothing */
1.468 + } else if (isAttChar(curchar)) {
1.469 + /* starting another attribute */
1.470 + startID = FB_GetPointer(fb) - 1;
1.471 + state = GET_ATT_STATE;
1.472 + } else {
1.473 + /* bogus char */
1.474 + *errStr = PR_smprintf("line %d: Bogus character '%c' in tag.\n",
1.475 + linenum, curchar);
1.476 + state = ERR_STATE;
1.477 + break;
1.478 + }
1.479 + break;
1.480 + case POST_ATT_WS_STATE:
1.481 + if (curchar == '>') {
1.482 + state = DONE_STATE;
1.483 + } else if (isspace(curchar)) {
1.484 + /* more whitespace, do nothing */
1.485 + } else if (isAttChar(curchar)) {
1.486 + /* starting another attribute */
1.487 + startID = FB_GetPointer(fb) - 1;
1.488 + state = GET_ATT_STATE;
1.489 + } else if (curchar == '=') {
1.490 + /* there was whitespace between the attribute and its equal
1.491 + * sign, which means there's a value coming up */
1.492 + state = PRE_VAL_WS_STATE;
1.493 + } else {
1.494 + /* bogus char */
1.495 + *errStr = PR_smprintf("line %d: Bogus character '%c' in tag.\n",
1.496 + linenum, curchar);
1.497 + state = ERR_STATE;
1.498 + break;
1.499 + }
1.500 + break;
1.501 + case PRE_VAL_WS_STATE:
1.502 + if (curchar == '>') {
1.503 + /* premature end-of-tag (sounds like a personal problem). */
1.504 + *errStr = PR_smprintf(
1.505 + "line %d: End of tag while waiting for value.\n",
1.506 + linenum);
1.507 + state = ERR_STATE;
1.508 + break;
1.509 + } else if (isspace(curchar)) {
1.510 + /* more whitespace, do nothing */
1.511 + break;
1.512 + } else {
1.513 + /* this must be some sort of value. Fall through
1.514 + * to GET_VALUE_STATE */
1.515 + startID = FB_GetPointer(fb) - 1;
1.516 + state = GET_VALUE_STATE;
1.517 + }
1.518 + /* Fall through if we didn't break on '>' or whitespace */
1.519 + case GET_VALUE_STATE:
1.520 + if (isspace(curchar) || curchar == '>') {
1.521 + /* end of value */
1.522 + curPos = FB_GetPointer(fb) - 2;
1.523 + if (curPos >= startID) {
1.524 + /* Grab the value */
1.525 + FB_GetRange(fb, startID, curPos,
1.526 + &curPair->value);
1.527 + curPair->valueLine = linenum;
1.528 + } else {
1.529 + /* empty value, leave as NULL */
1.530 + }
1.531 + if (isspace(curchar)) {
1.532 + state = PRE_ATT_WS_STATE;
1.533 + } else {
1.534 + state = DONE_STATE;
1.535 + }
1.536 + } else if (curchar == '\"' || curchar == '\'') {
1.537 + /* quoted value. Start recording the value inside the quote*/
1.538 + startID = FB_GetPointer(fb);
1.539 + state = GET_QUOTED_VAL_STATE;
1.540 + PORT_Assert(quotechar == '\0');
1.541 + quotechar = curchar; /* look for matching quote type */
1.542 + } else {
1.543 + /* just more value */
1.544 + }
1.545 + break;
1.546 + case GET_QUOTED_VAL_STATE:
1.547 + PORT_Assert(quotechar != '\0');
1.548 + if (curchar == quotechar) {
1.549 + /* end of quoted value */
1.550 + curPos = FB_GetPointer(fb) - 2;
1.551 + if (curPos >= startID) {
1.552 + /* Grab the value */
1.553 + FB_GetRange(fb, startID, curPos,
1.554 + &curPair->value);
1.555 + curPair->valueLine = linenum;
1.556 + } else {
1.557 + /* empty value, leave it as NULL */
1.558 + }
1.559 + state = GET_ATT_STATE;
1.560 + quotechar = '\0';
1.561 + startID = FB_GetPointer(fb);
1.562 + } else {
1.563 + /* more quoted value, continue */
1.564 + }
1.565 + break;
1.566 + case DONE_STATE:
1.567 + case ERR_STATE:
1.568 + default:
1.569 + ; /* should never get here */
1.570 + }
1.571 + }
1.572 +
1.573 + if (state == DONE_STATE) {
1.574 + /* Get the text of the tag */
1.575 + curPos = FB_GetPointer(fb) - 1;
1.576 + FB_GetRange(fb, startText, curPos, &ti->text);
1.577 +
1.578 + /* Return the tag */
1.579 + return ti;
1.580 + }
1.581 +
1.582 + /* Uh oh, an error. Kill the tag item*/
1.583 + DestroyTagItem(ti);
1.584 + return NULL;
1.585 +}
1.586 +
1.587 +
1.588 +/************************************************************************
1.589 + *
1.590 + * D e s t r o y H T M L I t e m
1.591 + */
1.592 +static void
1.593 +DestroyHTMLItem(HTMLItem *item)
1.594 +{
1.595 + if (item->type == TAG_ITEM) {
1.596 + DestroyTagItem(item->item.tag);
1.597 + } else {
1.598 + if (item->item.text) {
1.599 + PR_Free(item->item.text);
1.600 + }
1.601 + }
1.602 +}
1.603 +
1.604 +
1.605 +/************************************************************************
1.606 + *
1.607 + * D e s t r o y T a g I t e m
1.608 + */
1.609 +static void
1.610 +DestroyTagItem(TagItem*ti)
1.611 +{
1.612 + AVPair * temp;
1.613 +
1.614 + if (ti->text) {
1.615 + PR_Free(ti->text);
1.616 + ti->text = NULL;
1.617 + }
1.618 +
1.619 + while (ti->attList) {
1.620 + temp = ti->attList;
1.621 + ti->attList = ti->attList->next;
1.622 +
1.623 + if (temp->attribute) {
1.624 + PR_Free(temp->attribute);
1.625 + temp->attribute = NULL;
1.626 + }
1.627 + if (temp->value) {
1.628 + PR_Free(temp->value);
1.629 + temp->value = NULL;
1.630 + }
1.631 + PR_Free(temp);
1.632 + }
1.633 +
1.634 + PR_Free(ti);
1.635 +}
1.636 +
1.637 +
1.638 +/************************************************************************
1.639 + *
1.640 + * G e t T a g T y p e
1.641 + */
1.642 +static TAG_TYPE
1.643 +GetTagType(char *att)
1.644 +{
1.645 + if (!PORT_Strcasecmp(att, "APPLET")) {
1.646 + return APPLET_TAG;
1.647 + }
1.648 + if (!PORT_Strcasecmp(att, "SCRIPT")) {
1.649 + return SCRIPT_TAG;
1.650 + }
1.651 + if (!PORT_Strcasecmp(att, "LINK")) {
1.652 + return LINK_TAG;
1.653 + }
1.654 + if (!PORT_Strcasecmp(att, "STYLE")) {
1.655 + return STYLE_TAG;
1.656 + }
1.657 + return OTHER_TAG;
1.658 +}
1.659 +
1.660 +
1.661 +/************************************************************************
1.662 + *
1.663 + * F B _ C r e a t e
1.664 + */
1.665 +static FileBuffer*
1.666 +FB_Create(PRFileDesc*fd)
1.667 +{
1.668 + FileBuffer * fb;
1.669 + PRInt32 amountRead;
1.670 + PRInt32 storedOffset;
1.671 +
1.672 + fb = (FileBuffer * ) PR_Malloc(sizeof(FileBuffer));
1.673 + fb->fd = fd;
1.674 + storedOffset = PR_Seek(fd, 0, PR_SEEK_CUR);
1.675 + PR_Seek(fd, 0, PR_SEEK_SET);
1.676 + fb->startOffset = 0;
1.677 + amountRead = PR_Read(fd, fb->buf, FILE_BUFFER_BUFSIZE);
1.678 + if (amountRead == -1)
1.679 + goto loser;
1.680 + fb->maxIndex = amountRead - 1;
1.681 + fb->curIndex = 0;
1.682 + fb->IsEOF = (fb->curIndex > fb->maxIndex) ? PR_TRUE : PR_FALSE;
1.683 + fb->lineNum = 1;
1.684 +
1.685 + PR_Seek(fd, storedOffset, PR_SEEK_SET);
1.686 + return fb;
1.687 +loser:
1.688 + PR_Seek(fd, storedOffset, PR_SEEK_SET);
1.689 + PR_Free(fb);
1.690 + return NULL;
1.691 +}
1.692 +
1.693 +
1.694 +/************************************************************************
1.695 + *
1.696 + * F B _ G e t C h a r
1.697 + */
1.698 +static int
1.699 +FB_GetChar(FileBuffer *fb)
1.700 +{
1.701 + PRInt32 storedOffset;
1.702 + PRInt32 amountRead;
1.703 + int retval = -1;
1.704 +
1.705 + if (fb->IsEOF) {
1.706 + return EOF;
1.707 + }
1.708 +
1.709 + storedOffset = PR_Seek(fb->fd, 0, PR_SEEK_CUR);
1.710 +
1.711 + retval = (unsigned char) fb->buf[fb->curIndex++];
1.712 + if (retval == '\n')
1.713 + fb->lineNum++;
1.714 +
1.715 + if (fb->curIndex > fb->maxIndex) {
1.716 + /* We're at the end of the buffer. Try to get some new data from the
1.717 + * file */
1.718 + fb->startOffset += fb->maxIndex + 1;
1.719 + PR_Seek(fb->fd, fb->startOffset, PR_SEEK_SET);
1.720 + amountRead = PR_Read(fb->fd, fb->buf, FILE_BUFFER_BUFSIZE);
1.721 + if (amountRead == -1)
1.722 + goto loser;
1.723 + fb->maxIndex = amountRead - 1;
1.724 + fb->curIndex = 0;
1.725 + }
1.726 +
1.727 + fb->IsEOF = (fb->curIndex > fb->maxIndex) ? PR_TRUE : PR_FALSE;
1.728 +
1.729 +loser:
1.730 + PR_Seek(fb->fd, storedOffset, PR_SEEK_SET);
1.731 + return retval;
1.732 +}
1.733 +
1.734 +
1.735 +/************************************************************************
1.736 + *
1.737 + * F B _ G e t L i n e N u m
1.738 + *
1.739 + */
1.740 +static unsigned int
1.741 +FB_GetLineNum(FileBuffer *fb)
1.742 +{
1.743 + return fb->lineNum;
1.744 +}
1.745 +
1.746 +
1.747 +/************************************************************************
1.748 + *
1.749 + * F B _ G e t P o i n t e r
1.750 + *
1.751 + */
1.752 +static PRInt32
1.753 +FB_GetPointer(FileBuffer *fb)
1.754 +{
1.755 + return fb->startOffset + fb->curIndex;
1.756 +}
1.757 +
1.758 +
1.759 +/************************************************************************
1.760 + *
1.761 + * F B _ G e t R a n g e
1.762 + *
1.763 + */
1.764 +static PRInt32
1.765 +FB_GetRange(FileBuffer *fb, PRInt32 start, PRInt32 end, char **buf)
1.766 +{
1.767 + PRInt32 amountRead;
1.768 + PRInt32 storedOffset;
1.769 +
1.770 + *buf = PR_Malloc(end - start + 2);
1.771 + if (*buf == NULL) {
1.772 + return 0;
1.773 + }
1.774 +
1.775 + storedOffset = PR_Seek(fb->fd, 0, PR_SEEK_CUR);
1.776 + PR_Seek(fb->fd, start, PR_SEEK_SET);
1.777 + amountRead = PR_Read(fb->fd, *buf, end - start + 1);
1.778 + PR_Seek(fb->fd, storedOffset, PR_SEEK_SET);
1.779 + if (amountRead == -1) {
1.780 + PR_Free(*buf);
1.781 + *buf = NULL;
1.782 + return 0;
1.783 + }
1.784 +
1.785 + (*buf)[end-start+1] = '\0';
1.786 + return amountRead;
1.787 +}
1.788 +
1.789 +
1.790 +/************************************************************************
1.791 + *
1.792 + * F B _ D e s t r o y
1.793 + *
1.794 + */
1.795 +static void
1.796 +FB_Destroy(FileBuffer *fb)
1.797 +{
1.798 + if (fb) {
1.799 + PR_Free(fb);
1.800 + }
1.801 +}
1.802 +
1.803 +
1.804 +/************************************************************************
1.805 + *
1.806 + * P r i n t T a g I t e m
1.807 + *
1.808 + */
1.809 +static void
1.810 +PrintTagItem(PRFileDesc *fd, TagItem *ti)
1.811 +{
1.812 + AVPair * pair;
1.813 +
1.814 + PR_fprintf(fd, "TAG:\n----\nType: ");
1.815 + switch (ti->type) {
1.816 + case APPLET_TAG:
1.817 + PR_fprintf(fd, "applet\n");
1.818 + break;
1.819 + case SCRIPT_TAG:
1.820 + PR_fprintf(fd, "script\n");
1.821 + break;
1.822 + case LINK_TAG:
1.823 + PR_fprintf(fd, "link\n");
1.824 + break;
1.825 + case STYLE_TAG:
1.826 + PR_fprintf(fd, "style\n");
1.827 + break;
1.828 + case COMMENT_TAG:
1.829 + PR_fprintf(fd, "comment\n");
1.830 + break;
1.831 + case OTHER_TAG:
1.832 + default:
1.833 + PR_fprintf(fd, "other\n");
1.834 + break;
1.835 + }
1.836 +
1.837 + PR_fprintf(fd, "Attributes:\n");
1.838 + for (pair = ti->attList; pair; pair = pair->next) {
1.839 + PR_fprintf(fd, "\t%s=%s\n", pair->attribute,
1.840 + pair->value ? pair->value : "");
1.841 + }
1.842 + PR_fprintf(fd, "Text:%s\n", ti->text ? ti->text : "");
1.843 +
1.844 + PR_fprintf(fd, "---End of tag---\n");
1.845 +}
1.846 +
1.847 +
1.848 +/************************************************************************
1.849 + *
1.850 + * P r i n t H T M L S t r e a m
1.851 + *
1.852 + */
1.853 +static void
1.854 +PrintHTMLStream(PRFileDesc *fd, HTMLItem *head)
1.855 +{
1.856 + while (head) {
1.857 + if (head->type == TAG_ITEM) {
1.858 + PrintTagItem(fd, head->item.tag);
1.859 + } else {
1.860 + PR_fprintf(fd, "\nTEXT:\n-----\n%s\n-----\n\n", head->item.text);
1.861 + }
1.862 + head = head->next;
1.863 + }
1.864 +}
1.865 +
1.866 +
1.867 +/************************************************************************
1.868 + *
1.869 + * S a v e I n l i n e S c r i p t
1.870 + *
1.871 + */
1.872 +static int
1.873 +SaveInlineScript(char *text, char *id, char *basedir, char *archiveDir)
1.874 +{
1.875 + char *filename = NULL;
1.876 + PRFileDesc * fd = NULL;
1.877 + int retval = -1;
1.878 + PRInt32 writeLen;
1.879 + char *ilDir = NULL;
1.880 +
1.881 + if (!text || !id || !archiveDir) {
1.882 + return - 1;
1.883 + }
1.884 +
1.885 + if (dumpParse) {
1.886 + PR_fprintf(outputFD, "SaveInlineScript: text=%s, id=%s, \n"
1.887 + "basedir=%s, archiveDir=%s\n",
1.888 + text, id, basedir, archiveDir);
1.889 + }
1.890 +
1.891 + /* Make sure the archive directory is around */
1.892 + if (ensureExists(basedir, archiveDir) != PR_SUCCESS) {
1.893 + PR_fprintf(errorFD,
1.894 + "ERROR: Unable to create archive directory %s.\n", archiveDir);
1.895 + errorCount++;
1.896 + return - 1;
1.897 + }
1.898 +
1.899 + /* Make sure the inline script directory is around */
1.900 + ilDir = PR_smprintf("%s/inlineScripts", archiveDir);
1.901 + scriptdir = "inlineScripts";
1.902 + if (ensureExists(basedir, ilDir) != PR_SUCCESS) {
1.903 + PR_fprintf(errorFD,
1.904 + "ERROR: Unable to create directory %s.\n", ilDir);
1.905 + errorCount++;
1.906 + return - 1;
1.907 + }
1.908 +
1.909 + filename = PR_smprintf("%s/%s/%s", basedir, ilDir, id);
1.910 +
1.911 + /* If the file already exists, give a warning, then blow it away */
1.912 + if (PR_Access(filename, PR_ACCESS_EXISTS) == PR_SUCCESS) {
1.913 + PR_fprintf(errorFD,
1.914 + "warning: file \"%s\" already exists--will overwrite.\n",
1.915 + filename);
1.916 + warningCount++;
1.917 + if (rm_dash_r(filename)) {
1.918 + PR_fprintf(errorFD, "ERROR: Unable to delete %s.\n", filename);
1.919 + errorCount++;
1.920 + goto finish;
1.921 + }
1.922 + }
1.923 +
1.924 + /* Write text into file with name id */
1.925 + fd = PR_Open(filename, PR_WRONLY | PR_CREATE_FILE | PR_TRUNCATE, 0777);
1.926 + if (!fd) {
1.927 + PR_fprintf(errorFD, "ERROR: Unable to create file \"%s\".\n",
1.928 + filename);
1.929 + errorCount++;
1.930 + goto finish;
1.931 + }
1.932 + writeLen = strlen(text);
1.933 + if ( PR_Write(fd, text, writeLen) != writeLen) {
1.934 + PR_fprintf(errorFD, "ERROR: Unable to write to file \"%s\".\n",
1.935 + filename);
1.936 + errorCount++;
1.937 + goto finish;
1.938 + }
1.939 +
1.940 + retval = 0;
1.941 +finish:
1.942 + if (filename) {
1.943 + PR_smprintf_free(filename);
1.944 + }
1.945 + if (ilDir) {
1.946 + PR_smprintf_free(ilDir);
1.947 + }
1.948 + if (fd) {
1.949 + PR_Close(fd);
1.950 + }
1.951 + return retval;
1.952 +}
1.953 +
1.954 +
1.955 +/************************************************************************
1.956 + *
1.957 + * S a v e U n n a m a b l e S c r i p t
1.958 + *
1.959 + */
1.960 +static int
1.961 +SaveUnnamableScript(char *text, char *basedir, char *archiveDir,
1.962 +char *HTMLfilename)
1.963 +{
1.964 + char *id = NULL;
1.965 + char *ext = NULL;
1.966 + char *start = NULL;
1.967 + int retval = -1;
1.968 +
1.969 + if (!text || !archiveDir || !HTMLfilename) {
1.970 + return - 1;
1.971 + }
1.972 +
1.973 + if (dumpParse) {
1.974 + PR_fprintf(outputFD, "SaveUnnamableScript: text=%s, basedir=%s,\n"
1.975 + "archiveDir=%s, filename=%s\n", text, basedir, archiveDir,
1.976 + HTMLfilename);
1.977 + }
1.978 +
1.979 + /* Construct the filename */
1.980 + ext = PL_strrchr(HTMLfilename, '.');
1.981 + if (ext) {
1.982 + *ext = '\0';
1.983 + }
1.984 + for (start = HTMLfilename; strpbrk(start, "/\\");
1.985 + start = strpbrk(start, "/\\") + 1)
1.986 + /* do nothing */;
1.987 + if (*start == '\0')
1.988 + start = HTMLfilename;
1.989 + id = PR_smprintf("_%s%d", start, idOrdinal++);
1.990 + if (ext) {
1.991 + *ext = '.';
1.992 + }
1.993 +
1.994 + /* Now call SaveInlineScript to do the work */
1.995 + retval = SaveInlineScript(text, id, basedir, archiveDir);
1.996 +
1.997 + PR_Free(id);
1.998 +
1.999 + return retval;
1.1000 +}
1.1001 +
1.1002 +
1.1003 +/************************************************************************
1.1004 + *
1.1005 + * S a v e S o u r c e
1.1006 + *
1.1007 + */
1.1008 +static int
1.1009 +SaveSource(char *src, char *codebase, char *basedir, char *archiveDir)
1.1010 +{
1.1011 + char *from = NULL, *to = NULL;
1.1012 + int retval = -1;
1.1013 + char *arcDir = NULL;
1.1014 +
1.1015 + if (!src || !archiveDir) {
1.1016 + return - 1;
1.1017 + }
1.1018 +
1.1019 + if (dumpParse) {
1.1020 + PR_fprintf(outputFD, "SaveSource: src=%s, codebase=%s, basedir=%s,\n"
1.1021 + "archiveDir=%s\n", src, codebase, basedir, archiveDir);
1.1022 + }
1.1023 +
1.1024 + if (codebase) {
1.1025 + arcDir = PR_smprintf("%s/%s/%s/", basedir, codebase, archiveDir);
1.1026 + } else {
1.1027 + arcDir = PR_smprintf("%s/%s/", basedir, archiveDir);
1.1028 + }
1.1029 +
1.1030 + if (codebase) {
1.1031 + from = PR_smprintf("%s/%s/%s", basedir, codebase, src);
1.1032 + to = PR_smprintf("%s%s", arcDir, src);
1.1033 + } else {
1.1034 + from = PR_smprintf("%s/%s", basedir, src);
1.1035 + to = PR_smprintf("%s%s", arcDir, src);
1.1036 + }
1.1037 +
1.1038 + if (make_dirs(to, 0777)) {
1.1039 + PR_fprintf(errorFD,
1.1040 + "ERROR: Unable to create archive directory %s.\n", archiveDir);
1.1041 + errorCount++;
1.1042 + goto finish;
1.1043 + }
1.1044 +
1.1045 + retval = copyinto(from, to);
1.1046 +finish:
1.1047 + if (from)
1.1048 + PR_Free(from);
1.1049 + if (to)
1.1050 + PR_Free(to);
1.1051 + if (arcDir)
1.1052 + PR_Free(arcDir);
1.1053 + return retval;
1.1054 +}
1.1055 +
1.1056 +
1.1057 +/************************************************************************
1.1058 + *
1.1059 + * T a g T y p e T o S t r i n g
1.1060 + *
1.1061 + */
1.1062 +char *
1.1063 +TagTypeToString(TAG_TYPE type)
1.1064 +{
1.1065 + switch (type) {
1.1066 + case APPLET_TAG:
1.1067 + return "APPLET";
1.1068 + case SCRIPT_TAG:
1.1069 + return "SCRIPT";
1.1070 + case LINK_TAG:
1.1071 + return "LINK";
1.1072 + case STYLE_TAG:
1.1073 + return "STYLE";
1.1074 + default:
1.1075 + break;
1.1076 + }
1.1077 + return "unknown";
1.1078 +}
1.1079 +
1.1080 +
1.1081 +/************************************************************************
1.1082 + *
1.1083 + * e x t r a c t _ j s
1.1084 + *
1.1085 + */
1.1086 +static int
1.1087 +extract_js(char *filename)
1.1088 +{
1.1089 + PRFileDesc * fd = NULL;
1.1090 + FileBuffer * fb = NULL;
1.1091 + HTMLItem * head = NULL;
1.1092 + HTMLItem * tail = NULL;
1.1093 + HTMLItem * curitem = NULL;
1.1094 + HTMLItem * styleList = NULL;
1.1095 + HTMLItem * styleListTail = NULL;
1.1096 + HTMLItem * entityList = NULL;
1.1097 + HTMLItem * entityListTail = NULL;
1.1098 + TagItem * tagp = NULL;
1.1099 + char *text = NULL;
1.1100 + char *tagerr = NULL;
1.1101 + char *archiveDir = NULL;
1.1102 + char *firstArchiveDir = NULL;
1.1103 + char *basedir = NULL;
1.1104 + PRInt32 textStart;
1.1105 + PRInt32 curOffset;
1.1106 + HTML_STATE state;
1.1107 + int curchar;
1.1108 + int retval = -1;
1.1109 + unsigned int linenum, startLine;
1.1110 +
1.1111 + /* Initialize the implicit ID counter for each file */
1.1112 + idOrdinal = 0;
1.1113 +
1.1114 + /*
1.1115 + * First, parse the HTML into a stream of tags and text.
1.1116 + */
1.1117 +
1.1118 + fd = PR_Open(filename, PR_RDONLY, 0);
1.1119 + if (!fd) {
1.1120 + PR_fprintf(errorFD, "Unable to open %s for reading.\n", filename);
1.1121 + errorCount++;
1.1122 + return - 1;
1.1123 + }
1.1124 +
1.1125 + /* Construct base directory of filename. */
1.1126 + {
1.1127 + char *cp;
1.1128 +
1.1129 + basedir = PL_strdup(filename);
1.1130 +
1.1131 + /* Remove trailing slashes */
1.1132 + while ( (cp = PL_strprbrk(basedir, "/\\")) ==
1.1133 + (basedir + strlen(basedir) - 1)) {
1.1134 + *cp = '\0';
1.1135 + }
1.1136 +
1.1137 + /* Now remove everything from the last slash (which will be followed
1.1138 + * by a filename) to the end */
1.1139 + cp = PL_strprbrk(basedir, "/\\");
1.1140 + if (cp) {
1.1141 + *cp = '\0';
1.1142 + }
1.1143 + }
1.1144 +
1.1145 + state = TEXT_HTML_STATE;
1.1146 +
1.1147 + fb = FB_Create(fd);
1.1148 +
1.1149 + textStart = 0;
1.1150 + startLine = 0;
1.1151 + while (linenum = FB_GetLineNum(fb), (curchar = FB_GetChar(fb)) !=
1.1152 + EOF) {
1.1153 + switch (state) {
1.1154 + case TEXT_HTML_STATE:
1.1155 + if (curchar == '<') {
1.1156 + /*
1.1157 + * Found a tag
1.1158 + */
1.1159 + /* Save the text so far to a new text item */
1.1160 + curOffset = FB_GetPointer(fb) - 2;
1.1161 + if (curOffset >= textStart) {
1.1162 + if (FB_GetRange(fb, textStart, curOffset,
1.1163 + &text) !=
1.1164 + curOffset - textStart + 1) {
1.1165 + PR_fprintf(errorFD,
1.1166 + "Unable to read from %s.\n",
1.1167 + filename);
1.1168 + errorCount++;
1.1169 + goto loser;
1.1170 + }
1.1171 + /* little fudge here. If the first character on a line
1.1172 + * is '<', meaning a new tag, the preceding text item
1.1173 + * actually ends on the previous line. In this case
1.1174 + * we will be saying that the text segment ends on the
1.1175 + * next line. I don't think this matters for text items. */
1.1176 + curitem = CreateTextItem(text, startLine,
1.1177 + linenum);
1.1178 + text = NULL;
1.1179 + if (tail == NULL) {
1.1180 + head = tail = curitem;
1.1181 + } else {
1.1182 + tail->next = curitem;
1.1183 + tail = curitem;
1.1184 + }
1.1185 + }
1.1186 +
1.1187 + /* Process the tag */
1.1188 + tagp = ProcessTag(fb, &tagerr);
1.1189 + if (!tagp) {
1.1190 + if (tagerr) {
1.1191 + PR_fprintf(errorFD, "Error in file %s: %s\n",
1.1192 + filename, tagerr);
1.1193 + errorCount++;
1.1194 + } else {
1.1195 + PR_fprintf(errorFD,
1.1196 + "Error in file %s, in tag starting at line %d\n",
1.1197 + filename, linenum);
1.1198 + errorCount++;
1.1199 + }
1.1200 + goto loser;
1.1201 + }
1.1202 + /* Add the tag to the list */
1.1203 + curitem = CreateTagItem(tagp, linenum, FB_GetLineNum(fb));
1.1204 + if (tail == NULL) {
1.1205 + head = tail = curitem;
1.1206 + } else {
1.1207 + tail->next = curitem;
1.1208 + tail = curitem;
1.1209 + }
1.1210 +
1.1211 + /* What's the next state */
1.1212 + if (tagp->type == SCRIPT_TAG) {
1.1213 + state = SCRIPT_HTML_STATE;
1.1214 + }
1.1215 +
1.1216 + /* Start recording text from the new offset */
1.1217 + textStart = FB_GetPointer(fb);
1.1218 + startLine = FB_GetLineNum(fb);
1.1219 + } else {
1.1220 + /* regular character. Next! */
1.1221 + }
1.1222 + break;
1.1223 + case SCRIPT_HTML_STATE:
1.1224 + if (curchar == '<') {
1.1225 + char *cp;
1.1226 + /*
1.1227 + * If this is a </script> tag, then we're at the end of the
1.1228 + * script. Otherwise, ignore
1.1229 + */
1.1230 + curOffset = FB_GetPointer(fb) - 1;
1.1231 + cp = NULL;
1.1232 + if (FB_GetRange(fb, curOffset, curOffset + 8, &cp) != 9) {
1.1233 + if (cp) {
1.1234 + PR_Free(cp);
1.1235 + cp = NULL;
1.1236 + }
1.1237 + } else {
1.1238 + /* compare the strings */
1.1239 + if ( !PORT_Strncasecmp(cp, "</script>", 9) ) {
1.1240 + /* This is the end of the script. Record the text. */
1.1241 + curOffset--;
1.1242 + if (curOffset >= textStart) {
1.1243 + if (FB_GetRange(fb, textStart, curOffset, &text) !=
1.1244 + curOffset - textStart + 1) {
1.1245 + PR_fprintf(errorFD, "Unable to read from %s.\n",
1.1246 + filename);
1.1247 + errorCount++;
1.1248 + goto loser;
1.1249 + }
1.1250 + curitem = CreateTextItem(text, startLine, linenum);
1.1251 + text = NULL;
1.1252 + if (tail == NULL) {
1.1253 + head = tail = curitem;
1.1254 + } else {
1.1255 + tail->next = curitem;
1.1256 + tail = curitem;
1.1257 + }
1.1258 + }
1.1259 +
1.1260 + /* Now parse the /script tag and put it on the list */
1.1261 + tagp = ProcessTag(fb, &tagerr);
1.1262 + if (!tagp) {
1.1263 + if (tagerr) {
1.1264 + PR_fprintf(errorFD, "Error in file %s: %s\n",
1.1265 + filename, tagerr);
1.1266 + } else {
1.1267 + PR_fprintf(errorFD,
1.1268 + "Error in file %s, in tag starting at"
1.1269 + " line %d\n", filename, linenum);
1.1270 + }
1.1271 + errorCount++;
1.1272 + goto loser;
1.1273 + }
1.1274 + curitem = CreateTagItem(tagp, linenum,
1.1275 + FB_GetLineNum(fb));
1.1276 + if (tail == NULL) {
1.1277 + head = tail = curitem;
1.1278 + } else {
1.1279 + tail->next = curitem;
1.1280 + tail = curitem;
1.1281 + }
1.1282 +
1.1283 + /* go back to text state */
1.1284 + state = TEXT_HTML_STATE;
1.1285 +
1.1286 + textStart = FB_GetPointer(fb);
1.1287 + startLine = FB_GetLineNum(fb);
1.1288 + }
1.1289 + }
1.1290 + }
1.1291 + break;
1.1292 + }
1.1293 + }
1.1294 +
1.1295 + /* End of the file. Wrap up any remaining text */
1.1296 + if (state == SCRIPT_HTML_STATE) {
1.1297 + if (tail && tail->type == TAG_ITEM) {
1.1298 + PR_fprintf(errorFD, "ERROR: <SCRIPT> tag at %s:%d is not followed "
1.1299 + "by a </SCRIPT> tag.\n", filename, tail->startLine);
1.1300 + } else {
1.1301 + PR_fprintf(errorFD, "ERROR: <SCRIPT> tag in file %s is not followed"
1.1302 + " by a </SCRIPT tag.\n", filename);
1.1303 + }
1.1304 + errorCount++;
1.1305 + goto loser;
1.1306 + }
1.1307 + curOffset = FB_GetPointer(fb) - 1;
1.1308 + if (curOffset >= textStart) {
1.1309 + text = NULL;
1.1310 + if ( FB_GetRange(fb, textStart, curOffset, &text) !=
1.1311 + curOffset - textStart + 1) {
1.1312 + PR_fprintf(errorFD, "Unable to read from %s.\n", filename);
1.1313 + errorCount++;
1.1314 + goto loser;
1.1315 + }
1.1316 + curitem = CreateTextItem(text, startLine, linenum);
1.1317 + text = NULL;
1.1318 + if (tail == NULL) {
1.1319 + head = tail = curitem;
1.1320 + } else {
1.1321 + tail->next = curitem;
1.1322 + tail = curitem;
1.1323 + }
1.1324 + }
1.1325 +
1.1326 + if (dumpParse) {
1.1327 + PrintHTMLStream(outputFD, head);
1.1328 + }
1.1329 +
1.1330 + /*
1.1331 + * Now we have a stream of tags and text. Go through and deal with each.
1.1332 + */
1.1333 + for (curitem = head; curitem; curitem = curitem->next) {
1.1334 + TagItem * tagp = NULL;
1.1335 + AVPair * pairp = NULL;
1.1336 + char *src = NULL, *id = NULL, *codebase = NULL;
1.1337 + PRBool hasEventHandler = PR_FALSE;
1.1338 + int i;
1.1339 +
1.1340 + /* Reset archive directory for each tag */
1.1341 + if (archiveDir) {
1.1342 + PR_Free(archiveDir);
1.1343 + archiveDir = NULL;
1.1344 + }
1.1345 +
1.1346 + /* We only analyze tags */
1.1347 + if (curitem->type != TAG_ITEM) {
1.1348 + continue;
1.1349 + }
1.1350 +
1.1351 + tagp = curitem->item.tag;
1.1352 +
1.1353 + /* go through the attributes to get information */
1.1354 + for (pairp = tagp->attList; pairp; pairp = pairp->next) {
1.1355 +
1.1356 + /* ARCHIVE= */
1.1357 + if ( !PL_strcasecmp(pairp->attribute, "archive")) {
1.1358 + if (archiveDir) {
1.1359 + /* Duplicate attribute. Print warning */
1.1360 + PR_fprintf(errorFD,
1.1361 + "warning: \"%s\" attribute overwrites previous attribute"
1.1362 + " in tag starting at %s:%d.\n",
1.1363 + pairp->attribute, filename, curitem->startLine);
1.1364 + warningCount++;
1.1365 + PR_Free(archiveDir);
1.1366 + }
1.1367 + archiveDir = PL_strdup(pairp->value);
1.1368 +
1.1369 + /* Substiture ".arc" for ".jar" */
1.1370 + if ( (PL_strlen(archiveDir) < 4) ||
1.1371 + PL_strcasecmp((archiveDir + strlen(archiveDir) -4),
1.1372 + ".jar")) {
1.1373 + PR_fprintf(errorFD,
1.1374 + "warning: ARCHIVE attribute should end in \".jar\" in tag"
1.1375 + " starting on %s:%d.\n", filename, curitem->startLine);
1.1376 + warningCount++;
1.1377 + PR_Free(archiveDir);
1.1378 + archiveDir = PR_smprintf("%s.arc", archiveDir);
1.1379 + } else {
1.1380 + PL_strcpy(archiveDir + strlen(archiveDir) -4, ".arc");
1.1381 + }
1.1382 +
1.1383 + /* Record the first archive. This will be used later if
1.1384 + * the archive is not specified */
1.1385 + if (firstArchiveDir == NULL) {
1.1386 + firstArchiveDir = PL_strdup(archiveDir);
1.1387 + }
1.1388 + }
1.1389 + /* CODEBASE= */
1.1390 + else if ( !PL_strcasecmp(pairp->attribute, "codebase")) {
1.1391 + if (codebase) {
1.1392 + /* Duplicate attribute. Print warning */
1.1393 + PR_fprintf(errorFD,
1.1394 + "warning: \"%s\" attribute overwrites previous attribute"
1.1395 + " in tag staring at %s:%d.\n",
1.1396 + pairp->attribute, filename, curitem->startLine);
1.1397 + warningCount++;
1.1398 + }
1.1399 + codebase = pairp->value;
1.1400 + }
1.1401 + /* SRC= and HREF= */
1.1402 + else if ( !PORT_Strcasecmp(pairp->attribute, "src") ||
1.1403 + !PORT_Strcasecmp(pairp->attribute, "href") ) {
1.1404 + if (src) {
1.1405 + /* Duplicate attribute. Print warning */
1.1406 + PR_fprintf(errorFD,
1.1407 + "warning: \"%s\" attribute overwrites previous attribute"
1.1408 + " in tag staring at %s:%d.\n",
1.1409 + pairp->attribute, filename, curitem->startLine);
1.1410 + warningCount++;
1.1411 + }
1.1412 + src = pairp->value;
1.1413 + }
1.1414 + /* CODE= */
1.1415 + else if (!PORT_Strcasecmp(pairp->attribute, "code") ) {
1.1416 + /*!!!XXX Change PORT to PL all over this code !!! */
1.1417 + if (src) {
1.1418 + /* Duplicate attribute. Print warning */
1.1419 + PR_fprintf(errorFD,
1.1420 + "warning: \"%s\" attribute overwrites previous attribute"
1.1421 + " ,in tag staring at %s:%d.\n",
1.1422 + pairp->attribute, filename, curitem->startLine);
1.1423 + warningCount++;
1.1424 + }
1.1425 + src = pairp->value;
1.1426 +
1.1427 + /* Append a .class if one is not already present */
1.1428 + if ( (PL_strlen(src) < 6) ||
1.1429 + PL_strcasecmp( (src + PL_strlen(src) - 6), ".class") ) {
1.1430 + src = PR_smprintf("%s.class", src);
1.1431 + /* Put this string back into the data structure so it
1.1432 + * will be deallocated properly */
1.1433 + PR_Free(pairp->value);
1.1434 + pairp->value = src;
1.1435 + }
1.1436 + }
1.1437 + /* ID= */
1.1438 + else if (!PL_strcasecmp(pairp->attribute, "id") ) {
1.1439 + if (id) {
1.1440 + /* Duplicate attribute. Print warning */
1.1441 + PR_fprintf(errorFD,
1.1442 + "warning: \"%s\" attribute overwrites previous attribute"
1.1443 + " in tag staring at %s:%d.\n",
1.1444 + pairp->attribute, filename, curitem->startLine);
1.1445 + warningCount++;
1.1446 + }
1.1447 + id = pairp->value;
1.1448 + }
1.1449 +
1.1450 + /* STYLE= */
1.1451 + /* style= attributes, along with JS entities, are stored into
1.1452 + * files with dynamically generated names. The filenames are
1.1453 + * based on the order in which the text is found in the file.
1.1454 + * All JS entities on all lines up to and including the line
1.1455 + * containing the end of the tag that has this style= attribute
1.1456 + * will be processed before this style=attribute. So we need
1.1457 + * to record the line that this _tag_ (not the attribute) ends on.
1.1458 + */
1.1459 + else if (!PL_strcasecmp(pairp->attribute, "style") && pairp->value)
1.1460 + {
1.1461 + HTMLItem * styleItem;
1.1462 + /* Put this item on the style list */
1.1463 + styleItem = CreateTextItem(PL_strdup(pairp->value),
1.1464 + curitem->startLine, curitem->endLine);
1.1465 + if (styleListTail == NULL) {
1.1466 + styleList = styleListTail = styleItem;
1.1467 + } else {
1.1468 + styleListTail->next = styleItem;
1.1469 + styleListTail = styleItem;
1.1470 + }
1.1471 + }
1.1472 + /* Event handlers */
1.1473 + else {
1.1474 + for (i = 0; i < num_handlers; i++) {
1.1475 + if (!PL_strcasecmp(event_handlers[i], pairp->attribute)) {
1.1476 + hasEventHandler = PR_TRUE;
1.1477 + break;
1.1478 + }
1.1479 + }
1.1480 + }
1.1481 +
1.1482 +
1.1483 + /* JS Entity */
1.1484 + {
1.1485 + char *entityStart, *entityEnd;
1.1486 + HTMLItem * entityItem;
1.1487 +
1.1488 + /* go through each JavaScript entity ( &{...}; ) and store it
1.1489 + * in the entityList. The important thing is to record what
1.1490 + * line number it's on, so we can get it in the right order
1.1491 + * in relation to style= attributes.
1.1492 + * Apparently, these can't flow across lines, so the start and
1.1493 + * end line will be the same. That helps matters.
1.1494 + */
1.1495 + entityEnd = pairp->value;
1.1496 + while ( entityEnd &&
1.1497 + (entityStart = PL_strstr(entityEnd, "&{")) /*}*/ != NULL) {
1.1498 + entityStart += 2; /* point at beginning of actual entity */
1.1499 + entityEnd = PL_strchr(entityStart, '}');
1.1500 + if (entityEnd) {
1.1501 + /* Put this item on the entity list */
1.1502 + *entityEnd = '\0';
1.1503 + entityItem = CreateTextItem(PL_strdup(entityStart),
1.1504 + pairp->valueLine, pairp->valueLine);
1.1505 + *entityEnd = /* { */ '}';
1.1506 + if (entityListTail) {
1.1507 + entityListTail->next = entityItem;
1.1508 + entityListTail = entityItem;
1.1509 + } else {
1.1510 + entityList = entityListTail = entityItem;
1.1511 + }
1.1512 + }
1.1513 + }
1.1514 + }
1.1515 + }
1.1516 +
1.1517 + /* If no archive was supplied, we use the first one of the file */
1.1518 + if (!archiveDir && firstArchiveDir) {
1.1519 + archiveDir = PL_strdup(firstArchiveDir);
1.1520 + }
1.1521 +
1.1522 + /* If we have an event handler, we need to archive this tag */
1.1523 + if (hasEventHandler) {
1.1524 + if (!id) {
1.1525 + PR_fprintf(errorFD,
1.1526 + "warning: tag starting at %s:%d has event handler but"
1.1527 + " no ID attribute. The tag will not be signed.\n",
1.1528 + filename, curitem->startLine);
1.1529 + warningCount++;
1.1530 + } else if (!archiveDir) {
1.1531 + PR_fprintf(errorFD,
1.1532 + "warning: tag starting at %s:%d has event handler but"
1.1533 + " no ARCHIVE attribute. The tag will not be signed.\n",
1.1534 + filename, curitem->startLine);
1.1535 + warningCount++;
1.1536 + } else {
1.1537 + if (SaveInlineScript(tagp->text, id, basedir, archiveDir)) {
1.1538 + goto loser;
1.1539 + }
1.1540 + }
1.1541 + }
1.1542 +
1.1543 + switch (tagp->type) {
1.1544 + case APPLET_TAG:
1.1545 + if (!src) {
1.1546 + PR_fprintf(errorFD,
1.1547 + "error: APPLET tag starting on %s:%d has no CODE "
1.1548 + "attribute.\n", filename, curitem->startLine);
1.1549 + errorCount++;
1.1550 + goto loser;
1.1551 + } else if (!archiveDir) {
1.1552 + PR_fprintf(errorFD,
1.1553 + "error: APPLET tag starting on %s:%d has no ARCHIVE "
1.1554 + "attribute.\n", filename, curitem->startLine);
1.1555 + errorCount++;
1.1556 + goto loser;
1.1557 + } else {
1.1558 + if (SaveSource(src, codebase, basedir, archiveDir)) {
1.1559 + goto loser;
1.1560 + }
1.1561 + }
1.1562 + break;
1.1563 + case SCRIPT_TAG:
1.1564 + case LINK_TAG:
1.1565 + case STYLE_TAG:
1.1566 + if (!archiveDir) {
1.1567 + PR_fprintf(errorFD,
1.1568 + "error: %s tag starting on %s:%d has no ARCHIVE "
1.1569 + "attribute.\n", TagTypeToString(tagp->type),
1.1570 + filename, curitem->startLine);
1.1571 + errorCount++;
1.1572 + goto loser;
1.1573 + } else if (src) {
1.1574 + if (SaveSource(src, codebase, basedir, archiveDir)) {
1.1575 + goto loser;
1.1576 + }
1.1577 + } else if (id) {
1.1578 + /* Save the next text item */
1.1579 + if (!curitem->next || (curitem->next->type !=
1.1580 + TEXT_ITEM)) {
1.1581 + PR_fprintf(errorFD,
1.1582 + "warning: %s tag starting on %s:%d is not followed"
1.1583 + " by script text.\n", TagTypeToString(tagp->type),
1.1584 + filename, curitem->startLine);
1.1585 + warningCount++;
1.1586 + /* just create empty file */
1.1587 + if (SaveInlineScript("", id, basedir, archiveDir)) {
1.1588 + goto loser;
1.1589 + }
1.1590 + } else {
1.1591 + curitem = curitem->next;
1.1592 + if (SaveInlineScript(curitem->item.text,
1.1593 + id, basedir,
1.1594 + archiveDir)) {
1.1595 + goto loser;
1.1596 + }
1.1597 + }
1.1598 + } else {
1.1599 + /* No src or id tag--warning */
1.1600 + PR_fprintf(errorFD,
1.1601 + "warning: %s tag starting on %s:%d has no SRC or"
1.1602 + " ID attributes. Will not sign.\n",
1.1603 + TagTypeToString(tagp->type), filename, curitem->startLine);
1.1604 + warningCount++;
1.1605 + }
1.1606 + break;
1.1607 + default:
1.1608 + /* do nothing for other tags */
1.1609 + break;
1.1610 + }
1.1611 +
1.1612 + }
1.1613 +
1.1614 + /* Now deal with all the unnamable scripts */
1.1615 + if (firstArchiveDir) {
1.1616 + HTMLItem * style, *entity;
1.1617 +
1.1618 + /* Go through the lists of JS entities and style attributes. Do them
1.1619 + * in chronological order within a list. Pick the list with the lower
1.1620 + * endLine. In case of a tie, entities come first.
1.1621 + */
1.1622 + style = styleList;
1.1623 + entity = entityList;
1.1624 + while (style || entity) {
1.1625 + if (!entity || (style && (style->endLine < entity->endLine))) {
1.1626 + /* Process style */
1.1627 + SaveUnnamableScript(style->item.text, basedir, firstArchiveDir,
1.1628 + filename);
1.1629 + style = style->next;
1.1630 + } else {
1.1631 + /* Process entity */
1.1632 + SaveUnnamableScript(entity->item.text, basedir, firstArchiveDir,
1.1633 + filename);
1.1634 + entity = entity->next;
1.1635 + }
1.1636 + }
1.1637 + }
1.1638 +
1.1639 +
1.1640 + retval = 0;
1.1641 +loser:
1.1642 + /* Blow away the stream */
1.1643 + while (head) {
1.1644 + curitem = head;
1.1645 + head = head->next;
1.1646 + DestroyHTMLItem(curitem);
1.1647 + }
1.1648 + while (styleList) {
1.1649 + curitem = styleList;
1.1650 + styleList = styleList->next;
1.1651 + DestroyHTMLItem(curitem);
1.1652 + }
1.1653 + while (entityList) {
1.1654 + curitem = entityList;
1.1655 + entityList = entityList->next;
1.1656 + DestroyHTMLItem(curitem);
1.1657 + }
1.1658 + if (text) {
1.1659 + PR_Free(text);
1.1660 + text = NULL;
1.1661 + }
1.1662 + if (fb) {
1.1663 + FB_Destroy(fb);
1.1664 + fb = NULL;
1.1665 + }
1.1666 + if (fd) {
1.1667 + PR_Close(fd);
1.1668 + }
1.1669 + if (tagerr) {
1.1670 + PR_smprintf_free(tagerr);
1.1671 + tagerr = NULL;
1.1672 + }
1.1673 + if (archiveDir) {
1.1674 + PR_Free(archiveDir);
1.1675 + archiveDir = NULL;
1.1676 + }
1.1677 + if (firstArchiveDir) {
1.1678 + PR_Free(firstArchiveDir);
1.1679 + firstArchiveDir = NULL;
1.1680 + }
1.1681 + return retval;
1.1682 +}
1.1683 +
1.1684 +
1.1685 +/**********************************************************************
1.1686 + *
1.1687 + * e n s u r e E x i s t s
1.1688 + *
1.1689 + * Check for existence of indicated directory. If it doesn't exist,
1.1690 + * it will be created.
1.1691 + * Returns PR_SUCCESS if the directory is present, PR_FAILURE otherwise.
1.1692 + */
1.1693 +static PRStatus
1.1694 +ensureExists (char *base, char *path)
1.1695 +{
1.1696 + char fn [FNSIZE];
1.1697 + PRDir * dir;
1.1698 + sprintf (fn, "%s/%s", base, path);
1.1699 +
1.1700 + /*PR_fprintf(outputFD, "Trying to open directory %s.\n", fn);*/
1.1701 +
1.1702 + if ( (dir = PR_OpenDir(fn)) ) {
1.1703 + PR_CloseDir(dir);
1.1704 + return PR_SUCCESS;
1.1705 + }
1.1706 + return PR_MkDir(fn, 0777);
1.1707 +}
1.1708 +
1.1709 +
1.1710 +/***************************************************************************
1.1711 + *
1.1712 + * m a k e _ d i r s
1.1713 + *
1.1714 + * Ensure that the directory portion of the path exists. This may require
1.1715 + * making the directory, and its parent, and its parent's parent, etc.
1.1716 + */
1.1717 +static int
1.1718 +make_dirs(char *path, int file_perms)
1.1719 +{
1.1720 + char *Path;
1.1721 + char *start;
1.1722 + char *sep;
1.1723 + int ret = 0;
1.1724 + PRFileInfo info;
1.1725 +
1.1726 + if (!path) {
1.1727 + return 0;
1.1728 + }
1.1729 +
1.1730 + Path = PL_strdup(path);
1.1731 + start = strpbrk(Path, "/\\");
1.1732 + if (!start) {
1.1733 + return 0;
1.1734 + }
1.1735 + start++; /* start right after first slash */
1.1736 +
1.1737 + /* Each time through the loop add one more directory. */
1.1738 + while ( (sep = strpbrk(start, "/\\")) ) {
1.1739 + *sep = '\0';
1.1740 +
1.1741 + if ( PR_GetFileInfo(Path, &info) != PR_SUCCESS) {
1.1742 + /* No such dir, we have to create it */
1.1743 + if ( PR_MkDir(Path, file_perms) != PR_SUCCESS) {
1.1744 + PR_fprintf(errorFD, "ERROR: Unable to create directory %s.\n",
1.1745 + Path);
1.1746 + errorCount++;
1.1747 + ret = -1;
1.1748 + goto loser;
1.1749 + }
1.1750 + } else {
1.1751 + /* something exists by this name, make sure it's a directory */
1.1752 + if ( info.type != PR_FILE_DIRECTORY ) {
1.1753 + PR_fprintf(errorFD, "ERROR: Unable to create directory %s.\n",
1.1754 + Path);
1.1755 + errorCount++;
1.1756 + ret = -1;
1.1757 + goto loser;
1.1758 + }
1.1759 + }
1.1760 +
1.1761 + start = sep + 1; /* start after the next slash */
1.1762 + *sep = '/';
1.1763 + }
1.1764 +
1.1765 +loser:
1.1766 + PR_Free(Path);
1.1767 + return ret;
1.1768 +}
1.1769 +
1.1770 +
1.1771 +/*
1.1772 + * c o p y i n t o
1.1773 + *
1.1774 + * Function to copy file "from" to path "to".
1.1775 + *
1.1776 + */
1.1777 +static int
1.1778 +copyinto (char *from, char *to)
1.1779 +{
1.1780 + PRInt32 num;
1.1781 + char buf [BUFSIZ];
1.1782 + PRFileDesc * infp = NULL, *outfp = NULL;
1.1783 + int retval = -1;
1.1784 +
1.1785 + if ((infp = PR_Open(from, PR_RDONLY, 0777)) == NULL) {
1.1786 + PR_fprintf(errorFD, "ERROR: Unable to open \"%s\" for reading.\n",
1.1787 + from);
1.1788 + errorCount++;
1.1789 + goto finish;
1.1790 + }
1.1791 +
1.1792 + /* If to already exists, print a warning before deleting it */
1.1793 + if (PR_Access(to, PR_ACCESS_EXISTS) == PR_SUCCESS) {
1.1794 + PR_fprintf(errorFD, "warning: %s already exists--will overwrite\n", to);
1.1795 + warningCount++;
1.1796 + if (rm_dash_r(to)) {
1.1797 + PR_fprintf(errorFD,
1.1798 + "ERROR: Unable to remove %s.\n", to);
1.1799 + errorCount++;
1.1800 + goto finish;
1.1801 + }
1.1802 + }
1.1803 +
1.1804 + if ((outfp = PR_Open(to, PR_WRONLY | PR_CREATE_FILE | PR_TRUNCATE, 0777))
1.1805 + == NULL) {
1.1806 + char *errBuf = NULL;
1.1807 +
1.1808 + errBuf = PR_Malloc(PR_GetErrorTextLength() + 1);
1.1809 + PR_fprintf(errorFD, "ERROR: Unable to open \"%s\" for writing.\n", to);
1.1810 + if (PR_GetErrorText(errBuf)) {
1.1811 + PR_fprintf(errorFD, "Cause: %s\n", errBuf);
1.1812 + }
1.1813 + if (errBuf) {
1.1814 + PR_Free(errBuf);
1.1815 + }
1.1816 + errorCount++;
1.1817 + goto finish;
1.1818 + }
1.1819 +
1.1820 + while ( (num = PR_Read(infp, buf, BUFSIZ)) > 0) {
1.1821 + if (PR_Write(outfp, buf, num) != num) {
1.1822 + PR_fprintf(errorFD, "ERROR: Error writing to %s.\n", to);
1.1823 + errorCount++;
1.1824 + goto finish;
1.1825 + }
1.1826 + }
1.1827 +
1.1828 + retval = 0;
1.1829 +finish:
1.1830 + if (infp)
1.1831 + PR_Close(infp);
1.1832 + if (outfp)
1.1833 + PR_Close(outfp);
1.1834 +
1.1835 + return retval;
1.1836 +}
1.1837 +
1.1838 +
