diff options
Diffstat (limited to 'binutils/strings.c')
| -rw-r--r-- | binutils/strings.c | 757 |
1 files changed, 695 insertions, 62 deletions
diff --git a/binutils/strings.c b/binutils/strings.c index 44a8e1d..f50badf 100644 --- a/binutils/strings.c +++ b/binutils/strings.c @@ -55,6 +55,19 @@ -T {bfdname} Specify a non-default object file format. + --unicode={default|locale|invalid|hex|escape|highlight} + -u {d|l|i|x|e|h} + Determine how to handle UTF-8 unicode characters. The default + is no special treatment. All other versions of this option + only apply if the encoding is valid and enabling the option + implies --encoding=S. + The 'locale' option displays the characters according to the + current locale. The 'invalid' option treats them as + non-string characters. The 'hex' option displays them as hex + byte sequences. The 'escape' option displays them as escape + sequences and the 'highlight' option displays them as + coloured escape sequences. + --output-separator=sep_string -s sep_string String used to separate parsed strings in output. Default is newline. @@ -76,6 +89,22 @@ #include "safe-ctype.h" #include "bucomm.h" +#ifndef streq +#define streq(a,b) (strcmp ((a),(b)) == 0) +#endif + +typedef enum unicode_display_type +{ + unicode_default = 0, + unicode_locale, + unicode_escape, + unicode_hex, + unicode_highlight, + unicode_invalid +} unicode_display_type; + +static unicode_display_type unicode_display = unicode_default; + #define STRING_ISGRAPHIC(c) \ ( (c) >= 0 \ && (c) <= 255 \ @@ -94,7 +123,7 @@ extern int errno; static int address_radix; /* Minimum length of sequence of graphic chars to trigger output. */ -static int string_min; +static uint string_min; /* Whether or not we include all whitespace as a graphic char. */ static bool include_all_whitespace; @@ -121,21 +150,22 @@ static char *output_separator; static struct option long_options[] = { {"all", no_argument, NULL, 'a'}, + {"bytes", required_argument, NULL, 'n'}, {"data", no_argument, NULL, 'd'}, + {"encoding", required_argument, NULL, 'e'}, + {"help", no_argument, NULL, 'h'}, + {"include-all-whitespace", no_argument, NULL, 'w'}, + {"output-separator", required_argument, NULL, 's'}, {"print-file-name", no_argument, NULL, 'f'}, - {"bytes", required_argument, NULL, 'n'}, {"radix", required_argument, NULL, 't'}, - {"include-all-whitespace", no_argument, NULL, 'w'}, - {"encoding", required_argument, NULL, 'e'}, {"target", required_argument, NULL, 'T'}, - {"output-separator", required_argument, NULL, 's'}, - {"help", no_argument, NULL, 'h'}, + {"unicode", required_argument, NULL, 'U'}, {"version", no_argument, NULL, 'v'}, {NULL, 0, NULL, 0} }; static bool strings_file (char *); -static void print_strings (const char *, FILE *, file_ptr, int, int, char *); +static void print_strings (const char *, FILE *, file_ptr, int, char *); static void usage (FILE *, int) ATTRIBUTE_NORETURN; int main (int, char **); @@ -171,7 +201,7 @@ main (int argc, char **argv) encoding = 's'; output_separator = NULL; - while ((optc = getopt_long (argc, argv, "adfhHn:wot:e:T:s:Vv0123456789", + while ((optc = getopt_long (argc, argv, "adfhHn:wot:e:T:s:U:Vv0123456789", long_options, (int *) 0)) != EOF) { switch (optc) @@ -244,6 +274,23 @@ main (int argc, char **argv) output_separator = optarg; break; + case 'U': + if (streq (optarg, "default") || streq (optarg, "d")) + unicode_display = unicode_default; + else if (streq (optarg, "locale") || streq (optarg, "l")) + unicode_display = unicode_locale; + else if (streq (optarg, "escape") || streq (optarg, "e")) + unicode_display = unicode_escape; + else if (streq (optarg, "invalid") || streq (optarg, "i")) + unicode_display = unicode_invalid; + else if (streq (optarg, "hex") || streq (optarg, "x")) + unicode_display = unicode_hex; + else if (streq (optarg, "highlight") || streq (optarg, "h")) + unicode_display = unicode_highlight; + else + fatal (_("invalid argument to -U/--unicode: %s"), optarg); + break; + case 'V': case 'v': print_version ("strings"); @@ -258,6 +305,9 @@ main (int argc, char **argv) } } + if (unicode_display != unicode_default) + encoding = 'S'; + if (numeric_opt != 0) { string_min = (int) strtoul (argv[numeric_opt - 1] + 1, &s, 0); @@ -293,14 +343,14 @@ main (int argc, char **argv) { datasection_only = false; SET_BINARY (fileno (stdin)); - print_strings ("{standard input}", stdin, 0, 0, 0, (char *) NULL); + print_strings ("{standard input}", stdin, 0, 0, (char *) NULL); files_given = true; } else { for (; optind < argc; ++optind) { - if (strcmp (argv[optind], "-") == 0) + if (streq (argv[optind], "-")) datasection_only = false; else { @@ -342,7 +392,7 @@ strings_a_section (bfd *abfd, asection *sect, const char *filename, } *got_a_section = true; - print_strings (filename, NULL, sect->filepos, 0, sectsize, (char *) mem); + print_strings (filename, NULL, sect->filepos, sectsize, (char *) mem); free (mem); } @@ -427,7 +477,7 @@ strings_file (char *file) return false; } - print_strings (file, stream, (file_ptr) 0, 0, 0, (char *) 0); + print_strings (file, stream, (file_ptr) 0, 0, (char *) NULL); if (fclose (stream) == EOF) { @@ -551,11 +601,627 @@ unget_part_char (long c, file_ptr *address, int *magiccount, char **magic) } } } + +static void +print_filename_and_address (const char * filename, file_ptr address) +{ + if (print_filenames) + printf ("%s: ", filename); + + if (! print_addresses) + return; + + switch (address_radix) + { + case 8: + if (sizeof (address) > sizeof (long)) + { +#ifndef __MSVCRT__ + printf ("%7llo ", (unsigned long long) address); +#else + printf ("%7I64o ", (unsigned long long) address); +#endif + } + else + printf ("%7lo ", (unsigned long) address); + break; + + case 10: + if (sizeof (address) > sizeof (long)) + { +#ifndef __MSVCRT__ + printf ("%7llu ", (unsigned long long) address); +#else + printf ("%7I64d ", (unsigned long long) address); +#endif + } + else + printf ("%7ld ", (long) address); + break; + + case 16: + if (sizeof (address) > sizeof (long)) + { +#ifndef __MSVCRT__ + printf ("%7llx ", (unsigned long long) address); +#else + printf ("%7I64x ", (unsigned long long) address); +#endif + } + else + printf ("%7lx ", (unsigned long) address); + break; + } +} + +/* Return non-zero if the bytes starting at BUFFER form a valid UTF-8 encoding. + If the encoding is valid then returns the number of bytes it uses. */ + +static unsigned int +is_valid_utf8 (const unsigned char * buffer, unsigned long buflen) +{ + if (buffer[0] < 0xc0) + return 0; + + if (buflen < 2) + return 0; + + if ((buffer[1] & 0xc0) != 0x80) + return 0; + + if ((buffer[0] & 0x20) == 0) + return 2; + + if (buflen < 3) + return 0; + + if ((buffer[2] & 0xc0) != 0x80) + return 0; + + if ((buffer[0] & 0x10) == 0) + return 3; + + if (buflen < 4) + return 0; + + if ((buffer[3] & 0xc0) != 0x80) + return 0; + + return 4; +} + +/* Display a UTF-8 encoded character in BUFFER according to the setting + of unicode_display. The character is known to be valid. + Returns the number of bytes consumed. */ + +static uint +display_utf8_char (const unsigned char * buffer) +{ + uint j; + uint utf8_len; + + switch (buffer[0] & 0x30) + { + case 0x00: + case 0x10: + utf8_len = 2; + break; + case 0x20: + utf8_len = 3; + break; + default: + utf8_len = 4; + } + + switch (unicode_display) + { + default: + fprintf (stderr, "ICE: unexpected unicode display type\n"); + break; + + case unicode_escape: + case unicode_highlight: + if (unicode_display == unicode_highlight && isatty (1)) + printf ("\x1B[31;47m"); /* Red. */ + + switch (utf8_len) + { + case 2: + printf ("\\u%02x%02x", + ((buffer[0] & 0x1c) >> 2), + ((buffer[0] & 0x03) << 6) | (buffer[1] & 0x3f)); + break; + + case 3: + printf ("\\u%02x%02x", + ((buffer[0] & 0x0f) << 4) | ((buffer[1] & 0x3c) >> 2), + ((buffer[1] & 0x03) << 6) | ((buffer[2] & 0x3f))); + break; + + case 4: + printf ("\\u%02x%02x%02x", + ((buffer[0] & 0x07) << 6) | ((buffer[1] & 0x3c) >> 2), + ((buffer[1] & 0x03) << 6) | ((buffer[2] & 0x3c) >> 2), + ((buffer[2] & 0x03) << 6) | ((buffer[3] & 0x3f))); + break; + default: + /* URG. */ + break; + } + + if (unicode_display == unicode_highlight && isatty (1)) + printf ("\033[0m"); /* Default colour. */ + break; + + case unicode_hex: + putchar ('<'); + printf ("0x"); + for (j = 0; j < utf8_len; j++) + printf ("%02x", buffer [j]); + putchar ('>'); + break; + + case unicode_locale: + printf ("%.1s", buffer); + break; + } + + return utf8_len; +} + +/* Display strings in BUFFER. Treat any UTF-8 encoded characters encountered + according to the setting of the unicode_display variable. The buffer + contains BUFLEN bytes. + + Display the characters as if they started at ADDRESS and are contained in + FILENAME. */ + +static void +print_unicode_buffer (const char * filename, + file_ptr address, + const unsigned char * buffer, + unsigned long buflen) +{ + /* Paranoia checks... */ + if (filename == NULL + || buffer == NULL + || unicode_display == unicode_default + || encoding != 'S' + || encoding_bytes != 1) + { + fprintf (stderr, "ICE: bad arguments to print_unicode_buffer\n"); + return; + } + + if (buflen == 0) + return; + + /* We must only display strings that are at least string_min *characters* + long. So we scan the buffer in two stages. First we locate the start + of a potential string. Then we walk along it until we have found + string_min characters. Then we go back to the start point and start + displaying characters according to the unicode_display setting. */ + + unsigned long start_point = 0; + unsigned long i = 0; + unsigned int char_len = 1; + unsigned int num_found = 0; + + for (i = 0; i < buflen; i += char_len) + { + int c = buffer[i]; + + char_len = 1; + + /* Find the first potential character of a string. */ + if (! STRING_ISGRAPHIC (c)) + { + num_found = 0; + continue; + } + + if (c > 126) + { + if (c < 0xc0) + { + num_found = 0; + continue; + } + + if ((char_len = is_valid_utf8 (buffer + i, buflen - i)) == 0) + { + char_len = 1; + num_found = 0; + continue; + } + + if (unicode_display == unicode_invalid) + { + /* We have found a valid UTF-8 character, but we treat it as non-graphic. */ + num_found = 0; + continue; + } + } + + if (num_found == 0) + /* We have found a potential starting point for a string. */ + start_point = i; + + ++ num_found; + + if (num_found >= string_min) + break; + } + + if (num_found < string_min) + return; + + print_filename_and_address (filename, address + start_point); + + /* We have found string_min characters. Display them and any + more that follow. */ + for (i = start_point; i < buflen; i += char_len) + { + int c = buffer[i]; + + char_len = 1; + + if (! STRING_ISGRAPHIC (c)) + break; + else if (c < 127) + putchar (c); + else if (! is_valid_utf8 (buffer + i, buflen - i)) + break; + else if (unicode_display == unicode_invalid) + break; + else + char_len = display_utf8_char (buffer + i); + } + + if (output_separator) + fputs (output_separator, stdout); + else + putchar ('\n'); + + /* FIXME: Using tail recursion here is lazy programming... */ + print_unicode_buffer (filename, address + i, buffer + i, buflen - i); +} + +static int +get_unicode_byte (FILE * stream, unsigned char * putback, uint * num_putback, uint * num_read) +{ + if (* num_putback > 0) + { + * num_putback = * num_putback - 1; + return putback [* num_putback]; + } + + * num_read = * num_read + 1; + +#if defined(HAVE_GETC_UNLOCKED) && HAVE_DECL_GETC_UNLOCKED + return getc_unlocked (stream); +#else + return getc (stream); +#endif +} + +/* Helper function for print_unicode_stream. */ + +static void +print_unicode_stream_body (const char * filename, + file_ptr address, + FILE * stream, + unsigned char * putback_buf, + uint num_putback, + unsigned char * print_buf) +{ + /* It would be nice if we could just read the stream into a buffer + and then process if with print_unicode_buffer. But the input + might be huge or it might time-locked (eg stdin). So instead + we go one byte at a time... */ + + file_ptr start_point = 0; + uint num_read = 0; + uint num_chars = 0; + uint num_print = 0; + int c; + + /* Find a series of string_min characters. Put them into print_buf. */ + do + { + if (num_chars >= string_min) + break; + + c = get_unicode_byte (stream, putback_buf, & num_putback, & num_read); + if (c == EOF) + break; + + if (! STRING_ISGRAPHIC (c)) + { + num_chars = num_print = 0; + continue; + } + + if (num_chars == 0) + start_point = num_read - 1; + + if (c < 127) + { + print_buf[num_print] = c; + num_chars ++; + num_print ++; + continue; + } + + if (c < 0xc0) + { + num_chars = num_print = 0; + continue; + } + + /* We *might* have a UTF-8 sequence. Time to start peeking. */ + char utf8[4]; + + utf8[0] = c; + c = get_unicode_byte (stream, putback_buf, & num_putback, & num_read); + if (c == EOF) + break; + utf8[1] = c; + + if ((utf8[1] & 0xc0) != 0x80) + { + /* Invalid UTF-8. */ + putback_buf[num_putback++] = utf8[1]; + num_chars = num_print = 0; + continue; + } + else if ((utf8[0] & 0x20) == 0) + { + /* A valid 2-byte UTF-8 encoding. */ + if (unicode_display == unicode_invalid) + { + putback_buf[num_putback++] = utf8[1]; + num_chars = num_print = 0; + } + else + { + print_buf[num_print ++] = utf8[0]; + print_buf[num_print ++] = utf8[1]; + num_chars ++; + } + continue; + } + + c = get_unicode_byte (stream, putback_buf, & num_putback, & num_read); + if (c == EOF) + break; + utf8[2] = c; + + if ((utf8[2] & 0xc0) != 0x80) + { + /* Invalid UTF-8. */ + putback_buf[num_putback++] = utf8[2]; + putback_buf[num_putback++] = utf8[1]; + num_chars = num_print = 0; + continue; + } + else if ((utf8[0] & 0x10) == 0) + { + /* A valid 3-byte UTF-8 encoding. */ + if (unicode_display == unicode_invalid) + { + putback_buf[num_putback++] = utf8[2]; + putback_buf[num_putback++] = utf8[1]; + num_chars = num_print = 0; + } + else + { + print_buf[num_print ++] = utf8[0]; + print_buf[num_print ++] = utf8[1]; + print_buf[num_print ++] = utf8[2]; + num_chars ++; + } + continue; + } + + c = get_unicode_byte (stream, putback_buf, & num_putback, & num_read); + if (c == EOF) + break; + utf8[3] = c; + + if ((utf8[3] & 0xc0) != 0x80) + { + /* Invalid UTF-8. */ + putback_buf[num_putback++] = utf8[3]; + putback_buf[num_putback++] = utf8[2]; + putback_buf[num_putback++] = utf8[1]; + num_chars = num_print = 0; + } + /* We have a valid 4-byte UTF-8 encoding. */ + else if (unicode_display == unicode_invalid) + { + putback_buf[num_putback++] = utf8[3]; + putback_buf[num_putback++] = utf8[1]; + putback_buf[num_putback++] = utf8[2]; + num_chars = num_print = 0; + } + else + { + print_buf[num_print ++] = utf8[0]; + print_buf[num_print ++] = utf8[1]; + print_buf[num_print ++] = utf8[2]; + print_buf[num_print ++] = utf8[3]; + num_chars ++; + } + } + while (1); + + if (num_chars >= string_min) + { + /* We know that we have string_min valid characters in print_buf, + and there may be more to come in the stream. Start displaying + them. */ + + print_filename_and_address (filename, address + start_point); + + uint i; + for (i = 0; i < num_print;) + { + if (print_buf[i] < 127) + putchar (print_buf[i++]); + else + i += display_utf8_char (print_buf + i); + } + + /* OK so now we have to start read unchecked bytes. */ + + /* Find a series of string_min characters. Put them into print_buf. */ + do + { + c = get_unicode_byte (stream, putback_buf, & num_putback, & num_read); + if (c == EOF) + break; + + if (! STRING_ISGRAPHIC (c)) + break; + + if (c < 127) + { + putchar (c); + continue; + } + + if (c < 0xc0) + break; + + /* We *might* have a UTF-8 sequence. Time to start peeking. */ + unsigned char utf8[4]; + + utf8[0] = c; + c = get_unicode_byte (stream, putback_buf, & num_putback, & num_read); + if (c == EOF) + break; + utf8[1] = c; + + if ((utf8[1] & 0xc0) != 0x80) + { + /* Invalid UTF-8. */ + putback_buf[num_putback++] = utf8[1]; + break; + } + else if ((utf8[0] & 0x20) == 0) + { + /* Valid 2-byte UTF-8. */ + if (unicode_display == unicode_invalid) + { + putback_buf[num_putback++] = utf8[1]; + break; + } + else + { + (void) display_utf8_char (utf8); + continue; + } + } + + c = get_unicode_byte (stream, putback_buf, & num_putback, & num_read); + if (c == EOF) + break; + utf8[2] = c; + + if ((utf8[2] & 0xc0) != 0x80) + { + /* Invalid UTF-8. */ + putback_buf[num_putback++] = utf8[2]; + putback_buf[num_putback++] = utf8[1]; + break; + } + else if ((utf8[0] & 0x10) == 0) + { + /* Valid 3-byte UTF-8. */ + if (unicode_display == unicode_invalid) + { + putback_buf[num_putback++] = utf8[2]; + putback_buf[num_putback++] = utf8[1]; + break; + } + else + { + (void) display_utf8_char (utf8); + continue; + } + } + + c = get_unicode_byte (stream, putback_buf, & num_putback, & num_read); + if (c == EOF) + break; + utf8[3] = c; + + if ((utf8[3] & 0xc0) != 0x80) + { + /* Invalid UTF-8. */ + putback_buf[num_putback++] = utf8[3]; + putback_buf[num_putback++] = utf8[2]; + putback_buf[num_putback++] = utf8[1]; + break; + } + else if (unicode_display == unicode_invalid) + { + putback_buf[num_putback++] = utf8[3]; + putback_buf[num_putback++] = utf8[2]; + putback_buf[num_putback++] = utf8[1]; + break; + } + else + /* A valid 4-byte UTF-8 encoding. */ + (void) display_utf8_char (utf8); + } + while (1); + + if (output_separator) + fputs (output_separator, stdout); + else + putchar ('\n'); + } + + if (c != EOF) + /* FIXME: Using tail recursion here is lazy, but it works. */ + print_unicode_stream_body (filename, address + num_read, stream, putback_buf, num_putback, print_buf); +} + +/* Display strings read in from STREAM. Treat any UTF-8 encoded characters + encountered according to the setting of the unicode_display variable. + The stream is positioned at ADDRESS and is attached to FILENAME. */ + +static void +print_unicode_stream (const char * filename, + file_ptr address, + FILE * stream) +{ + /* Paranoia checks... */ + if (filename == NULL + || stream == NULL + || unicode_display == unicode_default + || encoding != 'S' + || encoding_bytes != 1) + { + fprintf (stderr, "ICE: bad arguments to print_unicode_stream\n"); + return; + } + + /* Allocate space for string_min 4-byte utf-8 characters. */ + unsigned char * print_buf = xmalloc ((4 * string_min) + 1); + /* We should never have to put back more than 4 bytes. */ + unsigned char putback_buf[5]; + uint num_putback = 0; + + print_unicode_stream_body (filename, address, stream, putback_buf, num_putback, print_buf); + free (print_buf); +} /* Find the strings in file FILENAME, read from STREAM. Assume that STREAM is positioned so that the next byte read is at address ADDRESS in the file. - Stop reading at address STOP_POINT in the file, if nonzero. If STREAM is NULL, do not read from it. The caller can supply a buffer of characters @@ -566,20 +1232,29 @@ unget_part_char (long c, file_ptr *address, int *magiccount, char **magic) static void print_strings (const char *filename, FILE *stream, file_ptr address, - int stop_point, int magiccount, char *magic) + int magiccount, char *magic) { + if (unicode_display != unicode_default) + { + if (magic != NULL) + print_unicode_buffer (filename, address, + (const unsigned char *) magic, magiccount); + + if (stream != NULL) + print_unicode_stream (filename, address, stream); + return; + } + char *buf = (char *) xmalloc (sizeof (char) * (string_min + 1)); while (1) { file_ptr start; - int i; + uint i; long c; /* See if the next `string_min' chars are all graphic chars. */ tryline: - if (stop_point && address >= stop_point) - break; start = address; for (i = 0; i < string_min; i++) { @@ -601,51 +1276,7 @@ print_strings (const char *filename, FILE *stream, file_ptr address, /* We found a run of `string_min' graphic characters. Print up to the next non-graphic character. */ - - if (print_filenames) - printf ("%s: ", filename); - if (print_addresses) - switch (address_radix) - { - case 8: - if (sizeof (start) > sizeof (long)) - { -#ifndef __MSVCRT__ - printf ("%7llo ", (unsigned long long) start); -#else - printf ("%7I64o ", (unsigned long long) start); -#endif - } - else - printf ("%7lo ", (unsigned long) start); - break; - - case 10: - if (sizeof (start) > sizeof (long)) - { -#ifndef __MSVCRT__ - printf ("%7llu ", (unsigned long long) start); -#else - printf ("%7I64d ", (unsigned long long) start); -#endif - } - else - printf ("%7ld ", (long) start); - break; - - case 16: - if (sizeof (start) > sizeof (long)) - { -#ifndef __MSVCRT__ - printf ("%7llx ", (unsigned long long) start); -#else - printf ("%7I64x ", (unsigned long long) start); -#endif - } - else - printf ("%7lx ", (unsigned long) start); - break; - } + print_filename_and_address (filename, start); buf[i] = '\0'; fputs (buf, stdout); @@ -697,6 +1328,8 @@ usage (FILE *stream, int status) -T --target=<BFDNAME> Specify the binary file format\n\ -e --encoding={s,S,b,l,B,L} Select character size and endianness:\n\ s = 7-bit, S = 8-bit, {b,l} = 16-bit, {B,L} = 32-bit\n\ + --unicode={default|show|invalid|hex|escape|highlight}\n\ + -u {d|s|i|x|e|h} Specify how to treat UTF-8 encoded unicode characters\n\ -s --output-separator=<string> String used to separate strings in output.\n\ @<file> Read options from <file>\n\ -h --help Display this information\n\ |