Mark ChangeLog
[official-gcc.git] / gcc / java / jcf-io.c
blobd4fbe6b25f35bec1560224ad92557c291b58768f
1 /* Utility routines for finding and reading Java(TM) .class files.
2 Copyright (C) 1996, 1997, 1998, 1999, 2000, 2002, 2003, 2004
3 Free Software Foundation, Inc.
5 This file is part of GCC.
7 GCC is free software; you can redistribute it and/or modify
8 it under the terms of the GNU General Public License as published by
9 the Free Software Foundation; either version 2, or (at your option)
10 any later version.
12 GCC is distributed in the hope that it will be useful,
13 but WITHOUT ANY WARRANTY; without even the implied warranty of
14 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 GNU General Public License for more details.
17 You should have received a copy of the GNU General Public License
18 along with GCC; see the file COPYING. If not, write to
19 the Free Software Foundation, 59 Temple Place - Suite 330,
20 Boston, MA 02111-1307, USA.
22 Java and all Java-based marks are trademarks or registered trademarks
23 of Sun Microsystems, Inc. in the United States and other countries.
24 The Free Software Foundation is independent of Sun Microsystems, Inc. */
26 /* Written by Per Bothner <bothner@cygnus.com>, February 1996. */
28 #include "config.h"
29 #include "system.h"
30 #include "coretypes.h"
31 #include "tm.h"
33 #include "jcf.h"
34 #include "tree.h"
35 #include "toplev.h"
36 #include "java-tree.h"
37 #include "hashtab.h"
38 #if JCF_USE_SCANDIR
39 #include <dirent.h>
40 #include <fnmatch.h>
41 #endif
43 #include "zlib.h"
45 /* DOS brain-damage */
46 #ifndef O_BINARY
47 #define O_BINARY 0 /* MS-DOS brain-damage */
48 #endif
50 int
51 jcf_unexpected_eof (JCF *jcf, int count ATTRIBUTE_UNUSED)
53 if (jcf->filename)
54 fprintf (stderr, "Premature end of .class file %s.\n", jcf->filename);
55 else
56 fprintf (stderr, "Premature end of .class file <stdin>.\n");
57 exit (-1);
60 void
61 jcf_trim_old_input (JCF *jcf)
63 int count = jcf->read_ptr - jcf->buffer;
64 if (count > 0)
66 memmove (jcf->buffer, jcf->read_ptr, jcf->read_end - jcf->read_ptr);
67 jcf->read_ptr -= count;
68 jcf->read_end -= count;
72 int
73 jcf_filbuf_from_stdio (JCF *jcf, int count)
75 FILE *file = (FILE*) (jcf->read_state);
76 if (count > jcf->buffer_end - jcf->read_ptr)
78 JCF_u4 old_read_ptr = jcf->read_ptr - jcf->buffer;
79 JCF_u4 old_read_end = jcf->read_end - jcf->buffer;
80 JCF_u4 old_size = jcf->buffer_end - jcf->buffer;
81 JCF_u4 new_size = (old_size == 0 ? 2000 : 2 * old_size) + count;
82 unsigned char *new_buffer = jcf->buffer == NULL ? ALLOC (new_size)
83 : REALLOC (jcf->buffer, new_size);
84 jcf->buffer = new_buffer;
85 jcf->buffer_end = new_buffer + new_size;
86 jcf->read_ptr = new_buffer + old_read_ptr;
87 jcf->read_end = new_buffer + old_read_end;
89 count -= jcf->read_end - jcf->read_ptr;
90 if (count <= 0)
91 return 0;
92 if ((int) fread (jcf->read_end, 1, count, file) != count)
93 jcf_unexpected_eof (jcf, count);
94 jcf->read_end += count;
95 return 0;
98 #include "zipfile.h"
100 struct ZipFile *SeenZipFiles = NULL;
102 /* Open a zip file with the given name, and cache directory and file
103 descriptor. If the file is missing, treat it as an empty archive.
104 Return NULL if the .zip file is malformed.
107 ZipFile *
108 opendir_in_zip (const char *zipfile, int is_system)
110 struct ZipFile* zipf;
111 char magic [4];
112 int fd;
113 for (zipf = SeenZipFiles; zipf != NULL; zipf = zipf->next)
115 if (strcmp (zipf->name, zipfile) == 0)
116 return zipf;
119 zipf = ALLOC (sizeof (struct ZipFile) + strlen (zipfile) + 1);
120 zipf->next = SeenZipFiles;
121 zipf->name = (char*)(zipf+1);
122 strcpy (zipf->name, zipfile);
123 fd = open (zipfile, O_RDONLY | O_BINARY);
124 zipf->fd = fd;
125 if (fd < 0)
127 /* A missing zip file is not considered an error.
128 We may want to re-consider that. FIXME. */
129 zipf->count = 0;
130 zipf->dir_size = 0;
131 zipf->central_directory = NULL;
133 else
135 jcf_dependency_add_file (zipfile, is_system);
136 if (read (fd, magic, 4) != 4 || GET_u4 (magic) != (JCF_u4)ZIPMAGIC)
137 return NULL;
138 lseek (fd, 0L, SEEK_SET);
139 if (read_zip_archive (zipf) != 0)
140 return NULL;
143 SeenZipFiles = zipf;
144 return zipf;
147 /* Returns:
148 0: OK - zipmember found.
149 -1: Not found.
150 -2: Malformed archive.
154 open_in_zip (JCF *jcf, const char *zipfile, const char *zipmember,
155 int is_system)
157 ZipDirectory *zipd;
158 int i, len;
159 ZipFile *zipf = opendir_in_zip (zipfile, is_system);
161 if (zipf == NULL)
162 return -2;
164 if (!zipmember)
165 return 0;
167 len = strlen (zipmember);
169 zipd = (struct ZipDirectory*) zipf->central_directory;
170 for (i = 0; i < zipf->count; i++, zipd = ZIPDIR_NEXT (zipd))
172 if (len == zipd->filename_length &&
173 strncmp (ZIPDIR_FILENAME (zipd), zipmember, len) == 0)
175 JCF_ZERO (jcf);
177 jcf->filename = xstrdup (zipfile);
178 jcf->classname = xstrdup (zipmember);
179 return read_zip_member(jcf, zipd, zipf);
182 return -1;
185 /* Read data from zip archive member. */
188 read_zip_member (JCF *jcf, ZipDirectory *zipd, ZipFile *zipf)
190 jcf->filbuf = jcf_unexpected_eof;
191 jcf->zipd = (void *)zipd;
193 if (zipd->compression_method == Z_NO_COMPRESSION)
195 jcf->buffer = ALLOC (zipd->size);
196 jcf->buffer_end = jcf->buffer + zipd->size;
197 jcf->read_ptr = jcf->buffer;
198 jcf->read_end = jcf->buffer_end;
199 if (lseek (zipf->fd, zipd->filestart, 0) < 0
200 || read (zipf->fd, jcf->buffer, zipd->size) != (long) zipd->size)
201 return -2;
203 else
205 char *buffer;
206 z_stream d_stream; /* decompression stream */
207 d_stream.zalloc = (alloc_func) 0;
208 d_stream.zfree = (free_func) 0;
209 d_stream.opaque = (voidpf) 0;
211 jcf->buffer = ALLOC (zipd->uncompressed_size);
212 d_stream.next_out = jcf->buffer;
213 d_stream.avail_out = zipd->uncompressed_size;
214 jcf->buffer_end = jcf->buffer + zipd->uncompressed_size;
215 jcf->read_ptr = jcf->buffer;
216 jcf->read_end = jcf->buffer_end;
217 buffer = ALLOC (zipd->size);
218 d_stream.next_in = (unsigned char *) buffer;
219 d_stream.avail_in = zipd->size;
220 if (lseek (zipf->fd, zipd->filestart, 0) < 0
221 || read (zipf->fd, buffer, zipd->size) != (long) zipd->size)
222 return -2;
223 /* Handle NO_HEADER using undocumented zlib feature.
224 This is a very common hack. */
225 inflateInit2 (&d_stream, -MAX_WBITS);
226 inflate (&d_stream, Z_NO_FLUSH);
227 inflateEnd (&d_stream);
228 FREE (buffer);
231 return 0;
234 const char *
235 open_class (const char *filename, JCF *jcf, int fd, const char *dep_name)
237 if (jcf)
239 struct stat stat_buf;
240 if (fstat (fd, &stat_buf) != 0
241 || ! S_ISREG (stat_buf.st_mode))
243 perror ("Could not figure length of .class file");
244 return NULL;
246 if (dep_name != NULL)
247 jcf_dependency_add_file (dep_name, 0);
248 JCF_ZERO (jcf);
249 jcf->buffer = ALLOC (stat_buf.st_size);
250 jcf->buffer_end = jcf->buffer + stat_buf.st_size;
251 jcf->read_ptr = jcf->buffer;
252 jcf->read_end = jcf->buffer_end;
253 jcf->read_state = NULL;
254 jcf->filename = filename;
255 if (read (fd, jcf->buffer, stat_buf.st_size) != stat_buf.st_size)
257 perror ("Failed to read .class file");
258 return NULL;
260 close (fd);
261 jcf->filbuf = jcf_unexpected_eof;
263 else
264 close (fd);
265 return filename;
269 const char *
270 find_classfile (char *filename, JCF *jcf, const char *dep_name)
272 int fd = open (filename, O_RDONLY | O_BINARY);
273 if (fd < 0)
274 return NULL;
275 return open_class (filename, jcf, fd, dep_name);
278 #if JCF_USE_SCANDIR
280 /* A comparison function (as for qsort) that compares KEY (a char *
281 giving the basename of a file) with the name stored in ENTRY (a
282 dirent **). */
284 static int
285 compare_path (const void *key, const void *entry)
287 return strcmp ((const char *) key,
288 (*((const struct dirent **) entry))->d_name);
291 /* Returns nonzero if ENTRY names a .java or .class file. */
293 static int
294 java_or_class_file (const struct dirent *entry)
296 const char *base = lbasename (entry->d_name);
297 return (fnmatch ("*.java", base, 0) == 0 ||
298 fnmatch ("*.class", base, 0) == 0);
301 /* Information about the files present in a particular directory. */
302 typedef struct memoized_dirlist_entry
304 /* The name of the directory. */
305 const char *dir;
306 /* The number of .java and .class files present, or -1 if we could
307 not, for some reason, obtain the list. */
308 int num_files;
309 /* The .java and .class files in the directory, in alphabetical
310 order. */
311 struct dirent **files;
312 } memoized_dirlist_entry;
314 /* A hash function for a memoized_dirlist_entry. */
315 static hashval_t
316 memoized_dirlist_hash (const void *entry)
318 const memoized_dirlist_entry *mde = (const memoized_dirlist_entry *) entry;
319 return htab_hash_string (mde->dir);
322 /* Returns true if ENTRY (a memoized_dirlist_entry *) corresponds to
323 the directory given by KEY (a char *) giving the directory
324 name. */
326 static int
327 memoized_dirlist_lookup_eq (const void *entry, const void *key)
329 return strcmp ((const char *) key,
330 ((const memoized_dirlist_entry *) entry)->dir) == 0;
333 /* A hash table mapping directory names to the lists of .java and
334 .class files in that directory. */
336 static htab_t memoized_dirlists;
338 #endif
340 /* Like stat, but avoids actually making the stat system call if we
341 know that it cannot succeed. FILENAME and BUF are as for stat. */
343 static int
344 caching_stat (char *filename, struct stat *buf)
346 #if JCF_USE_SCANDIR
347 char *sep;
348 char origsep = 0;
349 char *base;
350 memoized_dirlist_entry *dent;
351 void **slot;
352 struct memoized_dirlist_entry temp;
354 /* If the hashtable has not already been created, create it now. */
355 if (!memoized_dirlists)
356 memoized_dirlists = htab_create (37,
357 memoized_dirlist_hash,
358 memoized_dirlist_lookup_eq,
359 NULL);
361 /* Get the name of the directory. */
362 sep = strrchr (filename, DIR_SEPARATOR);
363 #ifdef DIR_SEPARATOR_2
364 if (! sep)
365 sep = strrchr (filename, DIR_SEPARATOR_2);
366 #endif
367 if (sep)
369 origsep = *sep;
370 *sep = '\0';
371 base = sep + 1;
373 else
374 base = filename;
376 /* Obtain the entry for this directory from the hash table. This
377 approach is ok since we know that the hash function only looks at
378 the directory name. */
379 temp.dir = filename;
380 temp.num_files = 0;
381 temp.files = NULL;
382 slot = htab_find_slot (memoized_dirlists, &temp, INSERT);
383 if (!*slot)
385 /* We have not already scanned this directory; scan it now. */
386 dent = ((memoized_dirlist_entry *)
387 ALLOC (sizeof (memoized_dirlist_entry)));
388 dent->dir = xstrdup (filename);
389 /* Unfortunately, scandir is not fully standardized. In
390 particular, the type of the function pointer passed as the
391 third argument sometimes takes a "const struct dirent *"
392 parameter, and sometimes just a "struct dirent *". We cast
393 to (void *) and use __extension__ so that either way it is
394 quietly accepted. FIXME: scandir is not in POSIX. */
395 dent->num_files = __extension__ scandir (filename, &dent->files,
396 (void *) java_or_class_file,
397 alphasort);
398 *slot = dent;
400 else
401 dent = *((memoized_dirlist_entry **) slot);
403 /* Put the separator back. */
404 if (sep)
405 *sep = origsep;
407 /* If the file is not in the list, there is no need to stat it; it
408 does not exist. */
409 if (dent->num_files != -1
410 && !bsearch (base, dent->files, dent->num_files,
411 sizeof (struct dirent *), compare_path))
412 return -1;
413 #endif
415 return stat (filename, buf);
418 /* Returns 1 if the CLASSNAME (really a char *) matches the name
419 stored in TABLE_ENTRY (also a char *). */
421 static int
422 memoized_class_lookup_eq (const void *table_entry, const void *classname)
424 return strcmp ((const char *)classname, (const char *)table_entry) == 0;
427 /* A hash table keeping track of class names that were not found
428 during class lookup. (There is no need to cache the values
429 associated with names that were found; they are saved in
430 IDENTIFIER_CLASS_VALUE.) */
431 static htab_t memoized_class_lookups;
433 /* Returns a freshly malloc'd string with the fully qualified pathname
434 of the .class file for the class CLASSNAME. CLASSNAME must be
435 allocated in permanent storage; this function may retain a pointer
436 to it. Returns NULL on failure. If JCF != NULL, it is suitably
437 initialized. SOURCE_OK is true if we should also look for .java
438 file. */
440 const char *
441 find_class (const char *classname, int classname_length, JCF *jcf,
442 int source_ok)
444 int fd;
445 int i, k, java = -1, class = -1;
446 struct stat java_buf, class_buf;
447 char *dep_file;
448 void *entry;
449 char *java_buffer;
450 int buflen;
451 char *buffer;
452 hashval_t hash;
454 /* Create the hash table, if it does not already exist. */
455 if (!memoized_class_lookups)
456 memoized_class_lookups = htab_create (37,
457 htab_hash_string,
458 memoized_class_lookup_eq,
459 NULL);
461 /* Loop for this class in the hashtable. If it is present, we've
462 already looked for this class and failed to find it. */
463 hash = htab_hash_string (classname);
464 if (htab_find_with_hash (memoized_class_lookups, classname, hash))
465 return NULL;
467 /* Allocate and zero out the buffer, since we don't explicitly put a
468 null pointer when we're copying it below. */
469 buflen = jcf_path_max_len () + classname_length + 10;
470 buffer = ALLOC (buflen);
471 memset (buffer, 0, buflen);
473 java_buffer = alloca (buflen);
475 jcf->java_source = 0;
477 for (entry = jcf_path_start (); entry != NULL; entry = jcf_path_next (entry))
479 const char *path_name = jcf_path_name (entry);
480 if (class != 0)
482 int dir_len;
484 strcpy (buffer, path_name);
485 i = strlen (buffer);
487 /* This is right because we know that `.zip' entries will have a
488 trailing slash. See jcf-path.c. */
489 dir_len = i - 1;
491 for (k = 0; k < classname_length; k++, i++)
493 char ch = classname[k];
494 buffer[i] = ch == '.' ? '/' : ch;
496 strcpy (buffer+i, ".class");
498 if (jcf_path_is_zipfile (entry))
500 int err_code;
501 JCF _jcf;
502 buffer[dir_len] = '\0';
503 SOURCE_FRONTEND_DEBUG
504 (("Trying [...%s]:%s",
505 &buffer[dir_len-(dir_len > 15 ? 15 : dir_len)],
506 buffer+dir_len+1));
507 if (jcf == NULL)
508 jcf = &_jcf;
509 err_code = open_in_zip (jcf, buffer, buffer+dir_len+1,
510 jcf_path_is_system (entry));
511 if (err_code == 0)
513 /* Should we check if .zip is out-of-date wrt .java? */
514 buffer[dir_len] = '(';
515 strcpy (buffer+i, ".class)");
516 if (jcf == &_jcf)
517 JCF_FINISH (jcf);
518 return buffer;
520 else
521 continue;
523 class = caching_stat(buffer, &class_buf);
526 if (source_ok)
528 /* Compute name of .java file. */
529 int l, m;
530 strcpy (java_buffer, path_name);
531 l = strlen (java_buffer);
532 for (m = 0; m < classname_length; ++m)
533 java_buffer[m + l] = (classname[m] == '.'
534 ? DIR_SEPARATOR : classname[m]);
535 strcpy (java_buffer + m + l, ".java");
536 java = caching_stat (java_buffer, &java_buf);
537 if (java == 0)
538 break;
542 /* We preferably pick a class file if we have a chance. If the source
543 file is newer than the class file, we issue a warning and parse the
544 source file instead.
545 There should be a flag to allow people have the class file picked
546 up no matter what. FIXME. */
547 if (! java && ! class && java_buf.st_mtime > class_buf.st_mtime)
549 if (flag_newer)
550 warning ("source file for class %qs is newer than its matching class file. Source file %qs used instead", classname, java_buffer);
551 class = -1;
554 if (! java)
555 dep_file = java_buffer;
556 else
557 dep_file = buffer;
558 if (!class)
560 SOURCE_FRONTEND_DEBUG ((stderr, "[Class selected: %s]\n",
561 classname+classname_length-
562 (classname_length <= 30 ?
563 classname_length : 30)));
564 fd = JCF_OPEN_EXACT_CASE (buffer, O_RDONLY | O_BINARY);
565 if (fd >= 0)
566 goto found;
568 /* Give .java a try, if necessary */
569 if (!java)
571 strcpy (buffer, java_buffer);
572 SOURCE_FRONTEND_DEBUG ((stderr, "[Source selected: %s]\n",
573 classname+classname_length-
574 (classname_length <= 30 ?
575 classname_length : 30)));
576 fd = JCF_OPEN_EXACT_CASE (buffer, O_RDONLY);
577 if (fd >= 0)
579 jcf->java_source = 1;
580 goto found;
584 free (buffer);
586 /* Remember that this class could not be found so that we do not
587 have to look again. */
588 *htab_find_slot_with_hash (memoized_class_lookups, classname, hash, INSERT)
589 = (void *) classname;
591 return NULL;
592 found:
593 if (jcf->java_source)
595 JCF_ZERO (jcf); /* JCF_FINISH relies on this */
596 jcf->java_source = 1;
597 jcf->filename = xstrdup (buffer);
598 close (fd); /* We use STDIO for source file */
600 else
601 buffer = (char *) open_class (buffer, jcf, fd, dep_file);
602 jcf->classname = xstrdup (classname);
603 return buffer;
606 void
607 jcf_print_char (FILE *stream, int ch)
609 switch (ch)
611 case '\'':
612 case '\\':
613 case '\"':
614 fprintf (stream, "\\%c", ch);
615 break;
616 case '\n':
617 fprintf (stream, "\\n");
618 break;
619 case '\t':
620 fprintf (stream, "\\t");
621 break;
622 case '\r':
623 fprintf (stream, "\\r");
624 break;
625 default:
626 if (ch >= ' ' && ch < 127)
627 putc (ch, stream);
628 else if (ch < 256)
629 fprintf (stream, "\\%03x", ch);
630 else
631 fprintf (stream, "\\u%04x", ch);
635 /* Print UTF8 string at STR of length LENGTH bytes to STREAM. */
637 void
638 jcf_print_utf8 (FILE *stream, const unsigned char *str, int length)
640 const unsigned char * limit = str + length;
641 while (str < limit)
643 int ch = UTF8_GET (str, limit);
644 if (ch < 0)
646 fprintf (stream, "\\<invalid>");
647 return;
649 jcf_print_char (stream, ch);
653 /* Same as jcf_print_utf8, but print IN_CHAR as OUT_CHAR. */
655 void
656 jcf_print_utf8_replace (FILE *stream, const unsigned char *str, int length,
657 int in_char, int out_char)
659 const unsigned char *limit = str + length;
660 while (str < limit)
662 int ch = UTF8_GET (str, limit);
663 if (ch < 0)
665 fprintf (stream, "\\<invalid>");
666 return;
668 jcf_print_char (stream, ch == in_char ? out_char : ch);
672 /* Check that all the cross-references in the constant pool are
673 valid. Returns 0 on success.
674 Otherwise, returns the index of the (first) invalid entry.
675 Only checks internal consistency, but does not check that
676 any classes, fields, or methods are valid.*/
679 verify_constant_pool (JCF *jcf)
681 int i, n;
682 for (i = 1; i < JPOOL_SIZE (jcf); i++)
684 switch (JPOOL_TAG (jcf, i))
686 case CONSTANT_NameAndType:
687 n = JPOOL_USHORT2 (jcf, i);
688 if (n <= 0 || n >= JPOOL_SIZE(jcf)
689 || JPOOL_TAG (jcf, n) != CONSTANT_Utf8)
690 return i;
691 /* ... fall through ... */
692 case CONSTANT_Class:
693 case CONSTANT_String:
694 n = JPOOL_USHORT1 (jcf, i);
695 if (n <= 0 || n >= JPOOL_SIZE(jcf)
696 || JPOOL_TAG (jcf, n) != CONSTANT_Utf8)
697 return i;
698 break;
699 case CONSTANT_Fieldref:
700 case CONSTANT_Methodref:
701 case CONSTANT_InterfaceMethodref:
702 n = JPOOL_USHORT1 (jcf, i);
703 if (n <= 0 || n >= JPOOL_SIZE(jcf)
704 || JPOOL_TAG (jcf, n) != CONSTANT_Class)
705 return i;
706 n = JPOOL_USHORT2 (jcf, i);
707 if (n <= 0 || n >= JPOOL_SIZE(jcf)
708 || JPOOL_TAG (jcf, n) != CONSTANT_NameAndType)
709 return i;
710 break;
711 case CONSTANT_Long:
712 case CONSTANT_Double:
713 i++;
714 break;
715 case CONSTANT_Float:
716 case CONSTANT_Integer:
717 case CONSTANT_Utf8:
718 case CONSTANT_Unicode:
719 break;
720 default:
721 return i;
724 return 0;
727 void
728 format_uint (char *buffer, uint64 value, int base)
730 #define WRITE_BUF_SIZE (4 + sizeof(uint64) * 8)
731 char buf[WRITE_BUF_SIZE];
732 char *buf_ptr = buf+WRITE_BUF_SIZE; /* End of buf. */
733 int chars_written;
734 int i;
736 /* Now do the actual conversion, placing the result at the *end* of buf. */
737 /* Note this code does not pretend to be optimized. */
738 do {
739 int digit = value % base;
740 static const char digit_chars[] = "0123456789abcdefghijklmnopqrstuvwxyz";
741 *--buf_ptr = digit_chars[digit];
742 value /= base;
743 } while (value != 0);
745 chars_written = buf+WRITE_BUF_SIZE - buf_ptr;
746 for (i = 0; i < chars_written; i++)
747 buffer[i] = *buf_ptr++;
748 buffer[i] = 0;
751 void
752 format_int (char *buffer, jlong value, int base)
754 uint64 abs_value;
755 if (value < 0)
757 abs_value = -(uint64)value;
758 *buffer++ = '-';
760 else
761 abs_value = (uint64) value;
762 format_uint (buffer, abs_value, base);