1 /*
2 * Copyright (c) 2015, 2018, Oracle and/or its affiliates. All rights reserved.
3 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
4 *
5 * This code is free software; you can redistribute it and/or modify it
6 * under the terms of the GNU General Public License version 2 only, as
7 * published by the Free Software Foundation. Oracle designates this
8 * particular file as subject to the "Classpath" exception as provided
9 * by Oracle in the LICENSE file that accompanied this code.
10 *
11 * This code is distributed in the hope that it will be useful, but WITHOUT
12 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
13 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
14 * version 2 for more details (a copy is included in the LICENSE file that
15 * accompanied this code).
16 *
17 * You should have received a copy of the GNU General Public License version
18 * 2 along with this work; if not, write to the Free Software Foundation,
19 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
20 *
21 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
22 * or visit www.oracle.com if you need additional information or have any
23 * questions.
24 */
25
26 #include <stdlib.h>
27 #include <stdio.h>
28 #include <assert.h>
29 #include <sys/stat.h>
30 #include <ctype.h>
31
32 #ifdef DEBUG_ARGFILE
33 #ifndef NO_JNI
34 #define NO_JNI
35 #endif
36 #define JLI_ReportMessage(...) printf(__VA_ARGS__)
37 #define JDK_JAVA_OPTIONS "JDK_JAVA_OPTIONS"
38 int IsWhiteSpaceOption(const char* name) { return 1; }
39 #else
40 #include "java.h"
41 #include "jni.h"
42 #endif
43
44 #include "jli_util.h"
45 #include "emessages.h"
46
47 #define MAX_ARGF_SIZE 0x7fffffffL
48
49 static char* clone_substring(const char *begin, size_t len) {
50 char *rv = (char *) JLI_MemAlloc(len + 1);
51 memcpy(rv, begin, len);
52 rv[len] = '\0';
53 return rv;
54 }
55
56 enum STATE {
57 FIND_NEXT,
58 IN_COMMENT,
59 IN_QUOTE,
60 IN_ESCAPE,
61 SKIP_LEAD_WS,
62 IN_TOKEN
63 };
64
65 typedef struct {
66 enum STATE state;
67 const char* cptr;
68 const char* eob;
69 char quote_char;
70 JLI_List parts;
71 } __ctx_args;
72
73 #define NOT_FOUND -1
74 static int firstAppArgIndex = NOT_FOUND;
75
76 static jboolean expectingNoDashArg = JNI_FALSE;
77 // Initialize to 1, as the first argument is the app name and not preprocessed
78 static size_t argsCount = 1;
79 static jboolean stopExpansion = JNI_FALSE;
80 static jboolean relaunch = JNI_FALSE;
81
82 /*
83 * Prototypes for internal functions.
84 */
85 static jboolean expand(JLI_List args, const char *str, const char *var_name);
86
87 JNIEXPORT void JNICALL
88 JLI_InitArgProcessing(jboolean hasJavaArgs, jboolean disableArgFile) {
89 // No expansion for relaunch
90 if (argsCount != 1) {
91 relaunch = JNI_TRUE;
92 stopExpansion = JNI_TRUE;
93 argsCount = 1;
94 } else {
95 stopExpansion = disableArgFile;
96 }
97
98 expectingNoDashArg = JNI_FALSE;
99
100 // for tools, this value remains 0 all the time.
101 firstAppArgIndex = hasJavaArgs ? 0: NOT_FOUND;
102 }
103
104 JNIEXPORT int JNICALL
105 JLI_GetAppArgIndex() {
106 // Will be 0 for tools
107 return firstAppArgIndex;
108 }
109
110 static void checkArg(const char *arg) {
111 size_t idx = 0;
112 argsCount++;
113
114 // All arguments arrive here must be a launcher argument,
115 // ie. by now, all argfile expansions must have been performed.
116 if (*arg == '-') {
117 expectingNoDashArg = JNI_FALSE;
118 if (IsWhiteSpaceOption(arg)) {
119 // expect an argument
120 expectingNoDashArg = JNI_TRUE;
121
122 if (JLI_StrCmp(arg, "-jar") == 0 ||
123 JLI_StrCmp(arg, "--module") == 0 ||
124 JLI_StrCmp(arg, "-m") == 0) {
125 // This is tricky, we do expect NoDashArg
126 // But that is considered main class to stop expansion
127 expectingNoDashArg = JNI_FALSE;
128 // We can not just update the idx here because if -jar @file
129 // still need expansion of @file to get the argument for -jar
130 }
131 } else if (JLI_StrCmp(arg, "--disable-@files") == 0) {
132 stopExpansion = JNI_TRUE;
133 }
134 } else {
135 if (!expectingNoDashArg) {
136 // this is main class, argsCount is index to next arg
137 idx = argsCount;
138 }
139 expectingNoDashArg = JNI_FALSE;
140 }
141 // only update on java mode and not yet found main class
142 if (firstAppArgIndex == NOT_FOUND && idx != 0) {
143 firstAppArgIndex = (int) idx;
144 }
145 }
146
147 /*
148 [\n\r] +------------+ +------------+ [\n\r]
149 +---------+ IN_COMMENT +<------+ | IN_ESCAPE +---------+
150 | +------------+ | +------------+ |
151 | [#] ^ |[#] ^ | |
152 | +----------+ | [\\]| |[^\n\r] |
153 v | | | v |
154 +------------+ [^ \t\n\r\f] +------------+['"]> +------------+ |
155 | FIND_NEXT +-------------->+ IN_TOKEN +-----------+ IN_QUOTE + |
156 +------------+ +------------+ <[quote]+------------+ |
157 | ^ | | ^ ^ |
158 | | [ \t\n\r\f]| [\n\r]| | |[^ \t\n\r\f]v
159 | +--------------------------+-----------------------+ | +--------------+
160 | ['"] | | SKIP_LEAD_WS |
161 +---------------------------------------------------------+ +--------------+
162 */
163 static char* nextToken(__ctx_args *pctx) {
164 const char* nextc = pctx->cptr;
165 const char* const eob = pctx->eob;
166 const char* anchor = nextc;
167 char *token;
168
169 for (; nextc < eob; nextc++) {
170 register char ch = *nextc;
171
172 // Skip white space characters
173 if (pctx->state == FIND_NEXT || pctx->state == SKIP_LEAD_WS) {
174 while (ch == ' ' || ch == '\n' || ch == '\r' || ch == '\t' || ch == '\f') {
175 nextc++;
176 if (nextc >= eob) {
177 return NULL;
178 }
179 ch = *nextc;
180 }
181 pctx->state = (pctx->state == FIND_NEXT) ? IN_TOKEN : IN_QUOTE;
182 anchor = nextc;
183 // Deal with escape sequences
184 } else if (pctx->state == IN_ESCAPE) {
185 // concatenation directive
186 if (ch == '\n' || ch == '\r') {
187 pctx->state = SKIP_LEAD_WS;
188 } else {
189 // escaped character
190 char* escaped = (char*) JLI_MemAlloc(2 * sizeof(char));
191 escaped[1] = '\0';
192 switch (ch) {
193 case 'n':
194 escaped[0] = '\n';
195 break;
196 case 'r':
197 escaped[0] = '\r';
198 break;
199 case 't':
200 escaped[0] = '\t';
201 break;
202 case 'f':
203 escaped[0] = '\f';
204 break;
205 default:
206 escaped[0] = ch;
207 break;
208 }
209 JLI_List_add(pctx->parts, escaped);
210 pctx->state = IN_QUOTE;
211 }
212 // anchor to next character
213 anchor = nextc + 1;
214 continue;
215 // ignore comment to EOL
216 } else if (pctx->state == IN_COMMENT) {
217 while (ch != '\n' && ch != '\r') {
218 nextc++;
219 if (nextc > eob) {
220 return NULL;
221 }
222 ch = *nextc;
223 }
224 pctx->state = FIND_NEXT;
225 continue;
226 }
227
228 assert(pctx->state != IN_ESCAPE);
229 assert(pctx->state != FIND_NEXT);
230 assert(pctx->state != SKIP_LEAD_WS);
231 assert(pctx->state != IN_COMMENT);
232
233 switch(ch) {
234 case ' ':
235 case '\t':
236 case '\f':
237 if (pctx->state == IN_QUOTE) {
238 continue;
239 }
240 // fall through
241 case '\n':
242 case '\r':
243 if (pctx->parts->size == 0) {
244 token = clone_substring(anchor, nextc - anchor);
245 } else {
246 JLI_List_addSubstring(pctx->parts, anchor, nextc - anchor);
247 token = JLI_List_combine(pctx->parts);
248 JLI_List_free(pctx->parts);
249 pctx->parts = JLI_List_new(4);
250 }
251 pctx->cptr = nextc + 1;
252 pctx->state = FIND_NEXT;
253 return token;
254 case '#':
255 if (pctx->state == IN_QUOTE) {
256 continue;
257 }
258 pctx->state = IN_COMMENT;
259 break;
260 case '\\':
261 if (pctx->state != IN_QUOTE) {
262 continue;
263 }
264 JLI_List_addSubstring(pctx->parts, anchor, nextc - anchor);
265 pctx->state = IN_ESCAPE;
266 // anchor after backslash character
267 anchor = nextc + 1;
268 break;
269 case '\'':
270 case '"':
271 if (pctx->state == IN_QUOTE && pctx->quote_char != ch) {
272 // not matching quote
273 continue;
274 }
275 // partial before quote
276 if (anchor != nextc) {
277 JLI_List_addSubstring(pctx->parts, anchor, nextc - anchor);
278 }
279 // anchor after quote character
280 anchor = nextc + 1;
281 if (pctx->state == IN_TOKEN) {
282 pctx->quote_char = ch;
283 pctx->state = IN_QUOTE;
284 } else {
285 pctx->state = IN_TOKEN;
286 }
287 break;
288 default:
289 break;
290 }
291 }
292
293 assert(nextc == eob);
294 if (anchor != nextc) {
295 // not yet return until end of stream, we have part of a token.
296 JLI_List_addSubstring(pctx->parts, anchor, nextc - anchor);
297 }
298 return NULL;
299 }
300
301 static JLI_List readArgFile(FILE *file) {
302 char buf[4096];
303 JLI_List rv;
304 __ctx_args ctx;
305 size_t size;
306 char *token;
307
308 ctx.state = FIND_NEXT;
309 ctx.parts = JLI_List_new(4);
310 // initialize to avoid -Werror=maybe-uninitialized issues from gcc 7.3 onwards.
311 ctx.quote_char = '"';
312
313 /* arbitrarily pick 8, seems to be a reasonable number of arguments */
314 rv = JLI_List_new(8);
315
316 while (!feof(file)) {
317 size = fread(buf, sizeof(char), sizeof(buf), file);
318 if (ferror(file)) {
319 JLI_List_free(rv);
320 return NULL;
321 }
322
323 /* nextc is next character to read from the buffer
324 * eob is the end of input
325 * token is the copied token value, NULL if no a complete token
326 */
327 ctx.cptr = buf;
328 ctx.eob = buf + size;
329 token = nextToken(&ctx);
330 while (token != NULL) {
331 checkArg(token);
332 JLI_List_add(rv, token);
333 token = nextToken(&ctx);
334 }
335 }
336
337 // remaining partial token
338 if (ctx.state == IN_TOKEN || ctx.state == IN_QUOTE) {
339 if (ctx.parts->size != 0) {
340 token = JLI_List_combine(ctx.parts);
341 checkArg(token);
342 JLI_List_add(rv, token);
343 }
344 }
345 JLI_List_free(ctx.parts);
346
347 return rv;
348 }
349
350 /*
351 * if the arg represent a file, that is, prefix with a single '@',
352 * return a list of arguments from the file.
353 * otherwise, return NULL.
354 */
355 static JLI_List expandArgFile(const char *arg) {
356 FILE *fptr;
357 struct stat st;
358 JLI_List rv;
359
360 /* failed to access the file */
361 if (stat(arg, &st) != 0) {
362 JLI_ReportMessage(CFG_ERROR6, arg);
363 exit(1);
364 }
365
366 if (st.st_size > MAX_ARGF_SIZE) {
367 JLI_ReportMessage(CFG_ERROR10, MAX_ARGF_SIZE);
368 exit(1);
369 }
370
371 fptr = fopen(arg, "r");
372 /* arg file cannot be openned */
373 if (fptr == NULL) {
374 JLI_ReportMessage(CFG_ERROR6, arg);
375 exit(1);
376 }
377
378 rv = readArgFile(fptr);
379 fclose(fptr);
380
381 /* error occurred reading the file */
382 if (rv == NULL) {
383 JLI_ReportMessage(DLL_ERROR4, arg);
384 exit(1);
385 }
386
387 return rv;
388 }
389
390 /*
391 * expand a string into a list of words separated by whitespace.
392 */
393 static JLI_List expandArg(const char *arg) {
394 JLI_List rv;
395
396 /* arbitrarily pick 8, seems to be a reasonable number of arguments */
397 rv = JLI_List_new(8);
398
399 expand(rv, arg, NULL);
400
401 return rv;
402 }
403
404 JNIEXPORT JLI_List JNICALL
405 JLI_PreprocessArg(const char *arg, jboolean expandSourceOpt) {
406 JLI_List rv;
407
408 if (firstAppArgIndex > 0) {
409 // In user application arg, no more work.
410 return NULL;
411 }
412
413 if (stopExpansion) {
414 // still looking for user application arg
415 checkArg(arg);
416 return NULL;
417 }
418
419 if (expandSourceOpt
420 && JLI_StrCCmp(arg, "--source") == 0
421 && JLI_StrChr(arg, ' ') != NULL) {
422 return expandArg(arg);
423 }
424
425 if (arg[0] != '@') {
426 checkArg(arg);
427 return NULL;
428 }
429
430 if (arg[1] == '\0') {
431 // @ by itself is an argument
432 checkArg(arg);
433 return NULL;
434 }
435
436 arg++;
437 if (arg[0] == '@') {
438 // escaped @argument
439 rv = JLI_List_new(1);
440 checkArg(arg);
441 JLI_List_add(rv, JLI_StringDup(arg));
442 } else {
443 rv = expandArgFile(arg);
444 }
445 return rv;
446 }
447
448 int isTerminalOpt(char *arg) {
449 return JLI_StrCmp(arg, "-jar") == 0 ||
450 JLI_StrCmp(arg, "-m") == 0 ||
451 JLI_StrCmp(arg, "--module") == 0 ||
452 JLI_StrCmp(arg, "--dry-run") == 0 ||
453 JLI_StrCmp(arg, "-h") == 0 ||
454 JLI_StrCmp(arg, "-?") == 0 ||
455 JLI_StrCmp(arg, "-help") == 0 ||
456 JLI_StrCmp(arg, "--help") == 0 ||
457 JLI_StrCmp(arg, "-X") == 0 ||
458 JLI_StrCmp(arg, "--help-extra") == 0 ||
459 JLI_StrCmp(arg, "-version") == 0 ||
460 JLI_StrCmp(arg, "--version") == 0 ||
461 JLI_StrCmp(arg, "-fullversion") == 0 ||
462 JLI_StrCmp(arg, "--full-version") == 0;
463 }
464
465 JNIEXPORT jboolean JNICALL
466 JLI_AddArgsFromEnvVar(JLI_List args, const char *var_name) {
467 char *env = getenv(var_name);
468
469 if (firstAppArgIndex == 0) {
470 // Not 'java', return
471 return JNI_FALSE;
472 }
473
474 if (relaunch) {
475 return JNI_FALSE;
476 }
477
478 if (NULL == env) {
479 return JNI_FALSE;
480 }
481
482 JLI_ReportMessage(ARG_INFO_ENVVAR, var_name, env);
483 return expand(args, env, var_name);
484 }
485
486 /*
487 * Expand a string into a list of args.
488 * If the string is the result of looking up an environment variable,
489 * var_name should be set to the name of that environment variable,
490 * for use if needed in error messages.
491 */
492
493 static jboolean expand(JLI_List args, const char *str, const char *var_name) {
494 jboolean inEnvVar = (var_name != NULL);
495
496 char *p, *arg;
497 char quote;
498 JLI_List argsInFile;
499
500 // This is retained until the process terminates as it is saved as the args
501 p = JLI_MemAlloc(JLI_StrLen(str) + 1);
502 while (*str != '\0') {
503 while (*str != '\0' && isspace(*str)) {
504 str++;
505 }
506
507 // Trailing space
508 if (*str == '\0') {
509 break;
510 }
511
512 arg = p;
513 while (*str != '\0' && !isspace(*str)) {
514 if (inEnvVar && (*str == '"' || *str == '\'')) {
515 quote = *str++;
516 while (*str != quote && *str != '\0') {
517 *p++ = *str++;
518 }
519
520 if (*str == '\0') {
521 JLI_ReportMessage(ARG_ERROR8, var_name);
522 exit(1);
523 }
524 str++;
525 } else {
526 *p++ = *str++;
527 }
528 }
529
530 *p++ = '\0';
531
532 argsInFile = JLI_PreprocessArg(arg, JNI_FALSE);
533
534 if (NULL == argsInFile) {
535 if (isTerminalOpt(arg)) {
536 if (inEnvVar) {
537 JLI_ReportMessage(ARG_ERROR9, arg, var_name);
538 } else {
539 JLI_ReportMessage(ARG_ERROR15, arg);
540 }
541 exit(1);
542 }
543 JLI_List_add(args, arg);
544 } else {
545 size_t cnt, idx;
546 char *argFile = arg;
547 cnt = argsInFile->size;
548 for (idx = 0; idx < cnt; idx++) {
549 arg = argsInFile->elements[idx];
550 if (isTerminalOpt(arg)) {
551 if (inEnvVar) {
552 JLI_ReportMessage(ARG_ERROR10, arg, argFile, var_name);
553 } else {
554 JLI_ReportMessage(ARG_ERROR16, arg, argFile);
555 }
556 exit(1);
557 }
558 JLI_List_add(args, arg);
559 }
560 // Shallow free, we reuse the string to avoid copy
561 JLI_MemFree(argsInFile->elements);
562 JLI_MemFree(argsInFile);
563 }
564 /*
565 * Check if main-class is specified after argument being checked. It
566 * must always appear after expansion, as a main-class could be specified
567 * indirectly into environment variable via an @argfile, and it must be
568 * caught now.
569 */
570 if (firstAppArgIndex != NOT_FOUND) {
571 if (inEnvVar) {
572 JLI_ReportMessage(ARG_ERROR11, var_name);
573 } else {
574 JLI_ReportMessage(ARG_ERROR17);
575 }
576 exit(1);
577 }
578
579 assert (*str == '\0' || isspace(*str));
580 }
581
582 return JNI_TRUE;
583 }
584
585 #ifdef DEBUG_ARGFILE
586 /*
587 * Stand-alone sanity test, build with following command line
588 * $ CC -DDEBUG_ARGFILE -DNO_JNI -g args.c jli_util.c
589 */
590
591 void fail(char *expected, char *actual, size_t idx) {
592 printf("FAILED: Token[%lu] expected to be <%s>, got <%s>\n", idx, expected, actual);
593 exit(1);
594 }
595
596 void test_case(char *case_data, char **tokens, size_t cnt_tokens) {
597 size_t actual_cnt;
598 char *token;
599 __ctx_args ctx;
600
601 actual_cnt = 0;
602
603 ctx.state = FIND_NEXT;
604 ctx.parts = JLI_List_new(4);
605 ctx.cptr = case_data;
606 ctx.eob = case_data + strlen(case_data);
607
608 printf("Test case: <%s>, expected %lu tokens.\n", case_data, cnt_tokens);
609
610 for (token = nextToken(&ctx); token != NULL; token = nextToken(&ctx)) {
611 // should not have more tokens than expected
612 if (actual_cnt >= cnt_tokens) {
613 printf("FAILED: Extra token detected: <%s>\n", token);
614 exit(2);
615 }
616 if (JLI_StrCmp(token, tokens[actual_cnt]) != 0) {
617 fail(tokens[actual_cnt], token, actual_cnt);
618 }
619 actual_cnt++;
620 }
621
622 char* last = NULL;
623 if (ctx.parts->size != 0) {
624 last = JLI_List_combine(ctx.parts);
625 }
626 JLI_List_free(ctx.parts);
627
628 if (actual_cnt >= cnt_tokens) {
629 // same number of tokens, should have nothing left to parse
630 if (last != NULL) {
631 if (*last != '#') {
632 printf("Leftover detected: %s", last);
633 exit(2);
634 }
635 }
636 } else {
637 if (JLI_StrCmp(last, tokens[actual_cnt]) != 0) {
638 fail(tokens[actual_cnt], last, actual_cnt);
639 }
640 actual_cnt++;
641 }
642 if (actual_cnt != cnt_tokens) {
643 printf("FAILED: Number of tokens not match, expected %lu, got %lu\n",
644 cnt_tokens, actual_cnt);
645 exit(3);
646 }
647
648 printf("PASS\n");
649 }
650
651 #define DO_CASE(name) \
652 test_case(name[0], name + 1, sizeof(name)/sizeof(char*) - 1)
653
654 int main(int argc, char** argv) {
655 size_t i, j;
656
657 char* case1[] = { "-version -cp \"c:\\\\java libs\\\\one.jar\" \n",
658 "-version", "-cp", "c:\\java libs\\one.jar" };
659 DO_CASE(case1);
660
661 // note the open quote at the end
662 char* case2[] = { "com.foo.Panda \"Furious 5\"\fand\t'Shi Fu' \"escape\tprison",
663 "com.foo.Panda", "Furious 5", "and", "Shi Fu", "escape\tprison"};
664 DO_CASE(case2);
665
666 char* escaped_chars[] = { "escaped chars testing \"\\a\\b\\c\\f\\n\\r\\t\\v\\9\\6\\23\\82\\28\\377\\477\\278\\287\"",
667 "escaped", "chars", "testing", "abc\f\n\r\tv96238228377477278287"};
668 DO_CASE(escaped_chars);
669
670 char* mixed_quote[] = { "\"mix 'single quote' in double\" 'mix \"double quote\" in single' partial\"quote me\"this",
671 "mix 'single quote' in double", "mix \"double quote\" in single", "partialquote methis"};
672 DO_CASE(mixed_quote);
673
674 char* comments[] = { "line one #comment\n'line #2' #rest are comment\r\n#comment on line 3\nline 4 #comment to eof",
675 "line", "one", "line #2", "line", "4"};
676 DO_CASE(comments);
677
678 char* open_quote[] = { "This is an \"open quote \n across line\n\t, note for WS.",
679 "This", "is", "an", "open quote ", "across", "line", ",", "note", "for", "WS." };
680 DO_CASE(open_quote);
681
682 char* escape_in_open_quote[] = { "Try \"this \\\\\\\\ escape\\n double quote \\\" in open quote",
683 "Try", "this \\\\ escape\n double quote \" in open quote" };
684 DO_CASE(escape_in_open_quote);
685
686 char* quote[] = { "'-Dmy.quote.single'='Property in single quote. Here a double quote\" Add some slashes \\\\/'",
687 "-Dmy.quote.single=Property in single quote. Here a double quote\" Add some slashes \\/" };
688 DO_CASE(quote);
689
690 char* multi[] = { "\"Open quote to \n new \"line \\\n\r third\\\n\r\\\tand\ffourth\"",
691 "Open quote to ", "new", "line third\tand\ffourth" };
692 DO_CASE(multi);
693
694 char* escape_quote[] = { "c:\\\"partial quote\"\\lib",
695 "c:\\partial quote\\lib" };
696 DO_CASE(escape_quote);
697
698 if (argc > 1) {
699 for (i = 0; i < argc; i++) {
700 JLI_List tokens = JLI_PreprocessArg(argv[i], JNI_FALSE);
701 if (NULL != tokens) {
702 for (j = 0; j < tokens->size; j++) {
703 printf("Token[%lu]: <%s>\n", (unsigned long) j, tokens->elements[j]);
704 }
705 }
706 }
707 }
708 }
709
710 #endif // DEBUG_ARGFILE