1 /* strings -- print the strings of printable characters in files
2 Copyright (C) 1993-2017 Free Software Foundation, Inc.
4 This program is free software; you can redistribute it and/or modify
5 it under the terms of the GNU General Public License as published by
6 the Free Software Foundation; either version 3, or (at your option)
9 This program is distributed in the hope that it will be useful,
10 but WITHOUT ANY WARRANTY; without even the implied warranty of
11 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
12 GNU General Public License for more details.
14 You should have received a copy of the GNU General Public License
15 along with this program; if not, write to the Free Software
16 Foundation, Inc., 51 Franklin Street - Fifth Floor, Boston, MA
19 /* Usage: strings [options] file...
24 - Scan each file in its entirety.
27 -d Scan only the initialized data section(s) of object files.
30 -f Print the name of the file before each string.
34 -min-len Print graphic char sequences, MIN-LEN or more bytes long,
35 that are followed by a NUL or a newline. Default is 4.
38 -t {o,x,d} Print the offset within the file before each string,
41 --include-all-whitespace
42 -w By default tab and space are the only whitepace included in graphic
43 char sequences. This option considers all of isspace() valid.
45 -o Like -to. (Some other implementations have -o like -to,
46 others like -td. We chose one arbitrarily.)
48 --encoding={s,S,b,l,B,L}
50 Select character encoding: 7-bit-character, 8-bit-character,
51 bigendian 16-bit, littleendian 16-bit, bigendian 32-bit,
56 Specify a non-default object file format.
58 --output-separator=sep_string
59 -s sep_string String used to separate parsed strings in output.
63 -h Print the usage message on the standard output.
67 -v Print the program version number.
69 Written by Richard Stallman <rms@gnu.ai.mit.edu>
70 and David MacKenzie <djm@gnu.ai.mit.edu>. */
75 #include "libiberty.h"
76 #include "safe-ctype.h"
79 #define STRING_ISGRAPHIC(c) \
82 && ((c) == '\t' || ISPRINT (c) || (encoding == 'S' && (c) > 127) \
83 || (include_all_whitespace && ISSPACE (c))) \
90 /* The BFD section flags that identify an initialized data section. */
91 #define DATA_FLAGS (SEC_ALLOC | SEC_LOAD | SEC_HAS_CONTENTS)
93 /* Radix for printing addresses (must be 8, 10 or 16). */
94 static int address_radix;
96 /* Minimum length of sequence of graphic chars to trigger output. */
97 static int string_min;
99 /* Whether or not we include all whitespace as a graphic char. */
100 static bfd_boolean include_all_whitespace;
102 /* TRUE means print address within file for each string. */
103 static bfd_boolean print_addresses;
105 /* TRUE means print filename for each string. */
106 static bfd_boolean print_filenames;
108 /* TRUE means for object files scan only the data section. */
109 static bfd_boolean datasection_only;
111 /* TRUE if we found an initialized data section in the current file. */
112 static bfd_boolean got_a_section;
114 /* The BFD object file format. */
117 /* The character encoding format. */
118 static char encoding;
119 static int encoding_bytes;
121 /* Output string used to separate parsed strings */
122 static char *output_separator;
124 static struct option long_options[] =
126 {"all", no_argument, NULL, 'a'},
127 {"data", no_argument, NULL, 'd'},
128 {"print-file-name", no_argument, NULL, 'f'},
129 {"bytes", required_argument, NULL, 'n'},
130 {"radix", required_argument, NULL, 't'},
131 {"include-all-whitespace", required_argument, NULL, 'w'},
132 {"encoding", required_argument, NULL, 'e'},
133 {"target", required_argument, NULL, 'T'},
134 {"output-separator", required_argument, NULL, 's'},
135 {"help", no_argument, NULL, 'h'},
136 {"version", no_argument, NULL, 'v'},
140 /* Records the size of a named file so that we
141 do not repeatedly run bfd_stat() on it. */
145 const char * filename;
146 bfd_size_type filesize;
147 } filename_and_size_t;
149 static bfd_boolean strings_file (char *);
150 static void print_strings (const char *, FILE *, file_ptr, int, int, char *);
151 static void usage (FILE *, int) ATTRIBUTE_NORETURN;
153 int main (int, char **);
156 main (int argc, char **argv)
160 bfd_boolean files_given = FALSE;
164 #if defined (HAVE_SETLOCALE)
165 setlocale (LC_ALL, "");
167 bindtextdomain (PACKAGE, LOCALEDIR);
168 textdomain (PACKAGE);
170 program_name = argv[0];
171 xmalloc_set_program_name (program_name);
172 bfd_set_error_program_name (program_name);
174 expandargv (&argc, &argv);
177 include_all_whitespace = FALSE;
178 print_addresses = FALSE;
179 print_filenames = FALSE;
180 if (DEFAULT_STRINGS_ALL)
181 datasection_only = FALSE;
183 datasection_only = TRUE;
186 output_separator = NULL;
188 while ((optc = getopt_long (argc, argv, "adfhHn:wot:e:T:s:Vv0123456789",
189 long_options, (int *) 0)) != EOF)
194 datasection_only = FALSE;
198 datasection_only = TRUE;
202 print_filenames = TRUE;
210 string_min = (int) strtoul (optarg, &s, 0);
211 if (s != NULL && *s != 0)
212 fatal (_("invalid integer argument %s"), optarg);
216 include_all_whitespace = TRUE;
220 print_addresses = TRUE;
225 print_addresses = TRUE;
226 if (optarg[1] != '\0')
252 if (optarg[1] != '\0')
254 encoding = optarg[0];
258 output_separator = optarg;
263 print_version ("strings");
270 numeric_opt = optind;
275 if (numeric_opt != 0)
277 string_min = (int) strtoul (argv[numeric_opt - 1] + 1, &s, 0);
278 if (s != NULL && *s != 0)
279 fatal (_("invalid integer argument %s"), argv[numeric_opt - 1] + 1);
282 fatal (_("invalid minimum string length %d"), string_min);
303 set_default_bfd_target ();
307 datasection_only = FALSE;
308 SET_BINARY (fileno (stdin));
309 print_strings ("{standard input}", stdin, 0, 0, 0, (char *) NULL);
314 for (; optind < argc; ++optind)
316 if (strcmp (argv[optind], "-") == 0)
317 datasection_only = FALSE;
321 exit_status |= !strings_file (argv[optind]);
329 return (exit_status);
332 /* Scan section SECT of the file ABFD, whose printable name is in
333 ARG->filename and whose size might be in ARG->filesize. If it
334 contains initialized data set `got_a_section' and print the
337 FIXME: We ought to be able to return error codes/messages for
338 certain conditions. */
341 strings_a_section (bfd *abfd, asection *sect, void *arg)
343 filename_and_size_t * filename_and_sizep;
344 bfd_size_type *filesizep;
345 bfd_size_type sectsize;
348 if ((sect->flags & DATA_FLAGS) != DATA_FLAGS)
351 sectsize = bfd_get_section_size (sect);
356 /* Get the size of the file. This might have been cached for us. */
357 filename_and_sizep = (filename_and_size_t *) arg;
358 filesizep = & filename_and_sizep->filesize;
364 if (bfd_stat (abfd, &st))
367 /* Cache the result so that we do not repeatedly stat this file. */
368 *filesizep = st.st_size;
371 /* Compare the size of the section against the size of the file.
372 If the section is bigger then the file must be corrupt and
373 we should not try dumping it. */
374 if (sectsize >= *filesizep)
377 mem = xmalloc (sectsize);
379 if (bfd_get_section_contents (abfd, sect, mem, (file_ptr) 0, sectsize))
381 got_a_section = TRUE;
383 print_strings (filename_and_sizep->filename, NULL, sect->filepos,
384 0, sectsize, (char *) mem);
390 /* Scan all of the sections in FILE, and print the strings
391 in the initialized data section(s).
393 Return TRUE if successful,
394 FALSE if not (such as if FILE is not an object file). */
397 strings_object_file (const char *file)
399 filename_and_size_t filename_and_size;
402 abfd = bfd_openr (file, target);
405 /* Treat the file as a non-object file. */
408 /* This call is mainly for its side effect of reading in the sections.
409 We follow the traditional behavior of `strings' in that we don't
410 complain if we don't recognize a file to be an object file. */
411 if (!bfd_check_format (abfd, bfd_object))
417 got_a_section = FALSE;
418 filename_and_size.filename = file;
419 filename_and_size.filesize = 0;
420 bfd_map_over_sections (abfd, strings_a_section, & filename_and_size);
422 if (!bfd_close (abfd))
428 return got_a_section;
431 /* Print the strings in FILE. Return TRUE if ok, FALSE if an error occurs. */
434 strings_file (char *file)
438 /* get_file_size does not support non-S_ISREG files. */
440 if (stat (file, &st) < 0)
443 non_fatal (_("'%s': No such file"), file);
445 non_fatal (_("Warning: could not locate '%s'. reason: %s"),
446 file, strerror (errno));
449 else if (S_ISDIR (st.st_mode))
451 non_fatal (_("Warning: '%s' is a directory"), file);
455 /* If we weren't told to scan the whole file,
456 try to open it as an object file and only look at
457 initialized data sections. If that fails, fall back to the
459 if (!datasection_only || !strings_object_file (file))
463 stream = fopen (file, FOPEN_RB);
466 fprintf (stderr, "%s: ", program_name);
471 print_strings (file, stream, (file_ptr) 0, 0, 0, (char *) 0);
473 if (fclose (stream) == EOF)
475 fprintf (stderr, "%s: ", program_name);
484 /* Read the next character, return EOF if none available.
485 Assume that STREAM is positioned so that the next byte read
486 is at address ADDRESS in the file.
488 If STREAM is NULL, do not read from it.
489 The caller can supply a buffer of characters
490 to be processed before the data in STREAM.
491 MAGIC is the address of the buffer and
492 MAGICCOUNT is how many characters are in it. */
495 get_char (FILE *stream, file_ptr *address, int *magiccount, char **magic)
500 for (i = 0; i < encoding_bytes; i++)
512 /* Only use getc_unlocked if we found a declaration for it.
513 Otherwise, libc is not thread safe by default, and we
514 should not use it. */
516 #if defined(HAVE_GETC_UNLOCKED) && HAVE_DECL_GETC_UNLOCKED
517 c = getc_unlocked (stream);
526 r = (r << 8) | (c & 0xff);
534 r = ((r & 0xff) << 8) | ((r & 0xff00) >> 8);
537 r = (((r & 0xff) << 24) | ((r & 0xff00) << 8)
538 | ((r & 0xff0000) >> 8) | ((r & 0xff000000) >> 24));
545 /* Find the strings in file FILENAME, read from STREAM.
546 Assume that STREAM is positioned so that the next byte read
547 is at address ADDRESS in the file.
548 Stop reading at address STOP_POINT in the file, if nonzero.
550 If STREAM is NULL, do not read from it.
551 The caller can supply a buffer of characters
552 to be processed before the data in STREAM.
553 MAGIC is the address of the buffer and
554 MAGICCOUNT is how many characters are in it.
555 Those characters come at address ADDRESS and the data in STREAM follow. */
558 print_strings (const char *filename, FILE *stream, file_ptr address,
559 int stop_point, int magiccount, char *magic)
561 char *buf = (char *) xmalloc (sizeof (char) * (string_min + 1));
569 /* See if the next `string_min' chars are all graphic chars. */
571 if (stop_point && address >= stop_point)
574 for (i = 0; i < string_min; i++)
576 c = get_char (stream, &address, &magiccount, &magic);
582 if (! STRING_ISGRAPHIC (c))
583 /* Found a non-graphic. Try again starting with next char. */
588 /* We found a run of `string_min' graphic characters. Print up
589 to the next non-graphic character. */
592 printf ("%s: ", filename);
594 switch (address_radix)
597 #ifdef HAVE_LONG_LONG
598 if (sizeof (start) > sizeof (long))
601 printf ("%7llo ", (unsigned long long) start);
603 printf ("%7I64o ", (unsigned long long) start);
607 #elif !BFD_HOST_64BIT_LONG
608 if (start != (unsigned long) start)
609 printf ("++%7lo ", (unsigned long) start);
612 printf ("%7lo ", (unsigned long) start);
616 #ifdef HAVE_LONG_LONG
617 if (sizeof (start) > sizeof (long))
620 printf ("%7lld ", (unsigned long long) start);
622 printf ("%7I64d ", (unsigned long long) start);
626 #elif !BFD_HOST_64BIT_LONG
627 if (start != (unsigned long) start)
628 printf ("++%7ld ", (unsigned long) start);
631 printf ("%7ld ", (long) start);
635 #ifdef HAVE_LONG_LONG
636 if (sizeof (start) > sizeof (long))
639 printf ("%7llx ", (unsigned long long) start);
641 printf ("%7I64x ", (unsigned long long) start);
645 #elif !BFD_HOST_64BIT_LONG
646 if (start != (unsigned long) start)
647 printf ("%lx%8.8lx ", (unsigned long) (start >> 32),
648 (unsigned long) (start & 0xffffffff));
651 printf ("%7lx ", (unsigned long) start);
660 c = get_char (stream, &address, &magiccount, &magic);
663 if (! STRING_ISGRAPHIC (c))
668 if (output_separator)
669 fputs (output_separator, stdout);
677 usage (FILE *stream, int status)
679 fprintf (stream, _("Usage: %s [option(s)] [file(s)]\n"), program_name);
680 fprintf (stream, _(" Display printable strings in [file(s)] (stdin by default)\n"));
681 fprintf (stream, _(" The options are:\n"));
683 if (DEFAULT_STRINGS_ALL)
684 fprintf (stream, _("\
685 -a - --all Scan the entire file, not just the data section [default]\n\
686 -d --data Only scan the data sections in the file\n"));
688 fprintf (stream, _("\
689 -a - --all Scan the entire file, not just the data section\n\
690 -d --data Only scan the data sections in the file [default]\n"));
692 fprintf (stream, _("\
693 -f --print-file-name Print the name of the file before each string\n\
694 -n --bytes=[number] Locate & print any NUL-terminated sequence of at\n\
695 -<number> least [number] characters (default 4).\n\
696 -t --radix={o,d,x} Print the location of the string in base 8, 10 or 16\n\
697 -w --include-all-whitespace Include all whitespace as valid string characters\n\
698 -o An alias for --radix=o\n\
699 -T --target=<BFDNAME> Specify the binary file format\n\
700 -e --encoding={s,S,b,l,B,L} Select character size and endianness:\n\
701 s = 7-bit, S = 8-bit, {b,l} = 16-bit, {B,L} = 32-bit\n\
702 -s --output-separator=<string> String used to separate strings in output.\n\
703 @<file> Read options from <file>\n\
704 -h --help Display this information\n\
705 -v -V --version Print the program's version number\n"));
706 list_supported_targets (program_name, stream);
707 if (REPORT_BUGS_TO[0] && status == 0)
708 fprintf (stream, _("Report bugs to %s\n"), REPORT_BUGS_TO);