From b3aa80b45c4f46029efeb204bb9f2d2c4278a0e5 Mon Sep 17 00:00:00 2001 From: Nick Clifton Date: Tue, 9 Nov 2021 13:25:42 +0000 Subject: [PATCH] Add --unicode option to control how unicode characters are handled by display tools. * nm.c: Add --unicode option to control how unicode characters are handled. * objdump.c: Likewise. * readelf.c: Likewise. * strings.c: Likewise. * binutils.texi: Document the new feature. * NEWS: Document the new feature. * testsuite/binutils-all/unicode.exp: New file. * testsuite/binutils-all/nm.hex.unicode * testsuite/binutils-all/strings.escape.unicode * testsuite/binutils-all/objdump.highlight.unicode * testsuite/binutils-all/readelf.invalid.unicode Conflict:Context adaptation --- binutils/NEWS | 9 + binutils/doc/binutils.texi | 78 +++++ binutils/nm.c | 225 ++++++++++++- binutils/objdump.c | 190 ++++++++++- binutils/readelf.c | 190 ++++++++++- binutils/strings.c | 766 ++++++++++++++++++++++++++++++++++++++++----- 6 files changed, 1363 insertions(+), 95 deletions(-) diff --git a/binutils/NEWS b/binutils/NEWS index 1650a3a..1659a99 100644 --- a/binutils/NEWS +++ b/binutils/NEWS @@ -712,6 +712,15 @@ Changes in binutils 2.1: * The strip and copy programs now have options to remove debug-symbols only and/or local symbols only. They now also support long options. +* Tools which display symbols or strings (readelf, strings, nm, objdump) + have a new command line option which controls how unicode characters are + handled. By default they are treated as normal for the tool. Using + --unicode=locale will display them according to the current locale. + Using --unicode=hex will display them as hex byte values, whilst + --unicode=escape will display them as escape sequences. In addition + using --unicode=highlight will display them as unicode escape sequences + highlighted in red (if supported by the output device). + Copyright (C) 2012-2020 Free Software Foundation, Inc. diff --git a/binutils/doc/binutils.texi b/binutils/doc/binutils.texi index 669bee9..80fc864 100644 --- a/binutils/doc/binutils.texi +++ b/binutils/doc/binutils.texi @@ -792,6 +792,7 @@ nm [@option{-A}|@option{-o}|@option{--print-file-name}] [@option{-a}|@option{--d [@option{-r}|@option{--reverse-sort}] [@option{-S}|@option{--print-size}] [@option{-s}|@option{--print-armap}] [@option{-t} @var{radix}|@option{--radix=}@var{radix}] [@option{-u}|@option{--undefined-only}] [@option{-V}|@option{--version}] + [@option{-U} @var{method}] [@option{--unicode=}@var{method}] [@option{-X 32_64}] [@option{--defined-only}] [@option{--no-demangle}] [@option{--plugin} @var{name}] [@option{--no-recurse-limit}|@option{--recurse-limit}]] @@ -1075,6 +1076,21 @@ Use @var{radix} as the radix for printing the symbol values. It must be @cindex undefined symbols Display only undefined symbols (those external to each object file). +@item -U @var{[d|i|l|e|x|h]} +@itemx --unicode=@var{[default|invalid|locale|escape|hex|highlight]} +Controls the display of UTF-8 encoded mulibyte characters in strings. +The default (@option{--unicode=default}) is to give them no special +treatment. The @option{--unicode=locale} option displays the sequence +in the current locale, which may or may not support them. The options +@option{--unicode=hex} and @option{--unicode=invalid} display them as +hex byte sequences enclosed by either angle brackets or curly braces. + +The @option{--unicode=escape} option displays them as escape sequences +(@var{\uxxxx}) and the @option{--unicode=highlight} option displays +them as escape sequences highlighted in red (if supported by the +output device). The colouring is intended to draw attention to the +presence of unicode sequences where they might not be expected. + @item -V @itemx --version Show the version number of @command{nm} and exit. @@ -2170,6 +2186,7 @@ objdump [@option{-a}|@option{--archive-headers}] [@option{--prefix-strip=}@var{level}] [@option{--insn-width=}@var{width}] [@option{--visualize-jumps[=color|=extended-color|=off]} + [@option{-U} @var{method}] [@option{--unicode=}@var{method}] [@option{-V}|@option{--version}] [@option{-H}|@option{--help}] @var{objfile}@dots{} @@ -2831,6 +2848,21 @@ When displaying symbols include those which the target considers to be special in some way and which would not normally be of interest to the user. +@item -U @var{[d|i|l|e|x|h]} +@itemx --unicode=@var{[default|invalid|locale|escape|hex|highlight]} +Controls the display of UTF-8 encoded mulibyte characters in strings. +The default (@option{--unicode=default}) is to give them no special +treatment. The @option{--unicode=locale} option displays the sequence +in the current locale, which may or may not support them. The options +@option{--unicode=hex} and @option{--unicode=invalid} display them as +hex byte sequences enclosed by either angle brackets or curly braces. + +The @option{--unicode=escape} option displays them as escape sequences +(@var{\uxxxx}) and the @option{--unicode=highlight} option displays +them as escape sequences highlighted in red (if supported by the +output device). The colouring is intended to draw attention to the +presence of unicode sequences where they might not be expected. + @item -V @itemx --version Print the version number of @command{objdump} and exit. @@ -3107,6 +3139,7 @@ strings [@option{-afovV}] [@option{-}@var{min-len}] [@option{-n} @var{min-len}] [@option{--bytes=}@var{min-len}] [@option{-t} @var{radix}] [@option{--radix=}@var{radix}] [@option{-e} @var{encoding}] [@option{--encoding=}@var{encoding}] + [@option{-U} @var{method}] [@option{--unicode=}@var{method}] [@option{-}] [@option{--all}] [@option{--print-file-name}] [@option{-T} @var{bfdname}] [@option{--target=}@var{bfdname}] [@option{-w}] [@option{--include-all-whitespace}] @@ -3198,6 +3231,28 @@ single-8-bit-byte characters, @samp{b} = 16-bit bigendian, @samp{l} = littleendian. Useful for finding wide character strings. (@samp{l} and @samp{b} apply to, for example, Unicode UTF-16/UCS-2 encodings). +@item -U @var{[d|i|l|e|x|h]} +@itemx --unicode=@var{[default|invalid|locale|escape|hex|highlight]} +Controls the display of UTF-8 encoded mulibyte characters in strings. +The default (@option{--unicode=default}) is to give them no special +treatment, and instead rely upon the setting of the +@option{--encoding} option. The other values for this option +automatically enable @option{--encoding=S}. + +The @option{--unicode=invalid} option treats them as non-graphic +characters and hence not part of a valid string. All the remaining +options treat them as valid string characters. + +The @option{--unicode=locale} option displays them in the current +locale, which may or may not support UTF-8 encoding. The +@option{--unicode=hex} option displays them as hex byte sequences +enclosed between @var{<>} characters. The @option{--unicode=escape} +option displays them as escape sequences (@var{\uxxxx}) and the +@option{--unicode=highlight} option displays them as escape sequences +highlighted in red (if supported by the output device). The colouring +is intended to draw attention to the presence of unicode sequences +where they might not be expected. + @item -T @var{bfdname} @itemx --target=@var{bfdname} @cindex object code format @@ -4690,6 +4745,7 @@ readelf [@option{-a}|@option{--all}] [@option{-e}|@option{--headers}] [@option{-s}|@option{--syms}|@option{--symbols}] [@option{--dyn-syms}] + [@option{-U} @var{method}|@option{--unicode=}@var{method}] [@option{-n}|@option{--notes}] [@option{-r}|@option{--relocs}] [@option{-u}|@option{--unwind}] @@ -4803,6 +4859,28 @@ Displays the entries in dynamic symbol table section of the file, if it has one. The output format is the same as the format used by the @option{--syms} option. +@item -U @var{[d|i|l|e|x|h]} +@itemx --unicode=[default|invalid|locale|escape|hex|highlight] +Controls the display of non-ASCII characters in identifier names. +The default (@option{--unicode=locale} or @option{--unicode=default}) is +to treat them as multibyte characters and display them in the current +locale. All other versions of this option treat the bytes as UTF-8 +encoded values and attempt to interpret them. If they cannot be +interpreted or if the @option{--unicode=invalid} option is used then +they are displayed as a sequence of hex bytes, encloses in curly +parethesis characters. + +Using the @option{--unicode=escape} option will display the characters +as as unicode escape sequences (@var{\uxxxx}). Using the +@option{--unicode=hex} will display the characters as hex byte +sequences enclosed between angle brackets. + +Using the @option{--unicode=highlight} will display the characters as +unicode escape sequences but it will also highlighted them in red, +assuming that colouring is supported by the output device. The +colouring is intended to draw attention to the presence of unicode +sequences when they might not be expected. + @item -e @itemx --headers Display all the headers in the file. Equivalent to @option{-h -l -S}. diff --git a/binutils/nm.c b/binutils/nm.c index 1c3e866..e6e5b17 100644 --- a/binutils/nm.c +++ b/binutils/nm.c @@ -39,6 +39,11 @@ #include "demanguse.h" #include "plugin-api.h" #include "plugin.h" +#include "safe-ctype.h" + +#ifndef streq +#define streq(a,b) (strcmp ((a),(b)) == 0) +#endif /* When sorting by size, we use this structure to hold the size and a pointer to the minisymbol. */ @@ -188,6 +193,18 @@ static const char *plugin_target = NULL; static bfd *lineno_cache_bfd; static bfd *lineno_cache_rel_bfd; +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; + enum long_option_values { OPTION_TARGET = 200, @@ -227,6 +244,7 @@ static struct option long_options[] = {"target", required_argument, 0, OPTION_TARGET}, {"defined-only", no_argument, &defined_only, 1}, {"undefined-only", no_argument, &undefined_only, 1}, + {"unicode", required_argument, NULL, 'U'}, {"version", no_argument, &show_version, 1}, {"with-symbol-versions", no_argument, &with_symbol_versions, 1}, {0, no_argument, 0, 0} @@ -277,6 +295,8 @@ usage (FILE *stream, int status) -t, --radix=RADIX Use RADIX for printing symbol values\n\ --target=BFDNAME Specify the target object format as BFDNAME\n\ -u, --undefined-only Display only undefined symbols\n\ + -U {d|s|i|x|e|h} Specify how to treat UTF-8 encoded unicode characters\n\ + --unicode={default|show|invalid|hex|escape|highlight}\n\ --with-symbol-versions Display version strings after symbol names\n\ -X 32_64 (ignored)\n\ @FILE Read options from FILE\n\ @@ -392,6 +412,186 @@ get_coff_symbol_type (const struct internal_syment *sym) return bufp; } +/* Convert a potential UTF-8 encoded sequence in IN into characters in OUT. + The conversion format is controlled by the unicode_display variable. + Returns the number of characters added to OUT. + Returns the number of bytes consumed from IN in CONSUMED. + Always consumes at least one byte and displays at least one character. */ + +static unsigned int +display_utf8 (const unsigned char * in, char * out, unsigned int * consumed) +{ + char * orig_out = out; + unsigned int nchars = 0; + unsigned int j; + + if (unicode_display == unicode_default) + goto invalid; + + if (in[0] < 0xc0) + goto invalid; + + if ((in[1] & 0xc0) != 0x80) + goto invalid; + + if ((in[0] & 0x20) == 0) + { + nchars = 2; + goto valid; + } + + if ((in[2] & 0xc0) != 0x80) + goto invalid; + + if ((in[0] & 0x10) == 0) + { + nchars = 3; + goto valid; + } + + if ((in[3] & 0xc0) != 0x80) + goto invalid; + + nchars = 4; + + valid: + switch (unicode_display) + { + case unicode_locale: + /* Copy the bytes into the output buffer as is. */ + memcpy (out, in, nchars); + out += nchars; + break; + + case unicode_invalid: + case unicode_hex: + out += sprintf (out, "%c", unicode_display == unicode_hex ? '<' : '{'); + out += sprintf (out, "0x"); + for (j = 0; j < nchars; j++) + out += sprintf (out, "%02x", in [j]); + out += sprintf (out, "%c", unicode_display == unicode_hex ? '>' : '}'); + break; + + case unicode_highlight: + if (isatty (1)) + out += sprintf (out, "\x1B[31;47m"); /* Red. */ + /* Fall through. */ + case unicode_escape: + switch (nchars) + { + case 2: + out += sprintf (out, "\\u%02x%02x", + ((in[0] & 0x1c) >> 2), + ((in[0] & 0x03) << 6) | (in[1] & 0x3f)); + break; + + case 3: + out += sprintf (out, "\\u%02x%02x", + ((in[0] & 0x0f) << 4) | ((in[1] & 0x3c) >> 2), + ((in[1] & 0x03) << 6) | ((in[2] & 0x3f))); + break; + + case 4: + out += sprintf (out, "\\u%02x%02x%02x", + ((in[0] & 0x07) << 6) | ((in[1] & 0x3c) >> 2), + ((in[1] & 0x03) << 6) | ((in[2] & 0x3c) >> 2), + ((in[2] & 0x03) << 6) | ((in[3] & 0x3f))); + break; + default: + /* URG. */ + break; + } + + if (unicode_display == unicode_highlight && isatty (1)) + out += sprintf (out, "\033[0m"); /* Default colour. */ + break; + + default: + /* URG */ + break; + } + + * consumed = nchars; + return out - orig_out; + + invalid: + /* Not a valid UTF-8 sequence. */ + *out = *in; + * consumed = 1; + return 1; +} + +/* Convert any UTF-8 encoded characters in NAME into the form specified by + unicode_display. Also converts control characters. Returns a static + buffer if conversion was necessary. + Code stolen from objdump.c:sanitize_string(). */ + +static const char * +convert_utf8 (const char * in) +{ + static char * buffer = NULL; + static size_t buffer_len = 0; + const char * original = in; + char * out; + + /* Paranoia. */ + if (in == NULL) + return ""; + + /* See if any conversion is necessary. + In the majority of cases it will not be needed. */ + do + { + unsigned char c = *in++; + + if (c == 0) + return original; + + if (ISCNTRL (c)) + break; + + if (unicode_display != unicode_default && c >= 0xc0) + break; + } + while (1); + + /* Copy the input, translating as needed. */ + in = original; + if (buffer_len < (strlen (in) * 9)) + { + free ((void *) buffer); + buffer_len = strlen (in) * 9; + buffer = xmalloc (buffer_len + 1); + } + + out = buffer; + do + { + unsigned char c = *in++; + + if (c == 0) + break; + + if (ISCNTRL (c)) + { + *out++ = '^'; + *out++ = c + 0x40; + } + else if (unicode_display != unicode_default && c >= 0xc0) + { + unsigned int num_consumed; + + out += display_utf8 ((const unsigned char *)(in - 1), out, & num_consumed); + in += num_consumed - 1; + } + else + *out++ = c; + } + while (1); + + *out = 0; + return buffer; +} /* Print symbol name NAME, read from ABFD, with printf format FORM, demangling it if requested. */ @@ -410,6 +610,11 @@ print_symname (const char *form, const char *name, bfd *abfd) } } + if (unicode_display != unicode_default) + { + name = convert_utf8 (name); + } + printf (form, name); } @@ -1714,7 +1919,7 @@ main (int argc, char **argv) fatal (_("fatal error: libbfd ABI mismatch")); set_default_bfd_target (); - while ((c = getopt_long (argc, argv, "aABCDef:gHhlnopPrSst:uvVvX:", + while ((c = getopt_long (argc, argv, "aABCDef:gHhlnopPrSst:uU:vVvX:", long_options, (int *) 0)) != EOF) { switch (c) @@ -1801,6 +2006,24 @@ main (int argc, char **argv) case 'u': undefined_only = 1; 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': show_version = 1; break; diff --git a/binutils/objdump.c b/binutils/objdump.c index e99d663..531a093 100644 --- a/binutils/objdump.c +++ b/binutils/objdump.c @@ -205,6 +205,18 @@ static const struct objdump_private_desc * const objdump_private_vectors[] = /* The list of detected jumps inside a function. */ static struct jump_info *detected_jumps = NULL; + +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; static void usage (FILE *, int) ATTRIBUTE_NORETURN; static void @@ -269,6 +281,8 @@ usage (FILE *stream, int status) --recurse-limit Enable a limit on recursion whilst demangling. [Default]\n\ --no-recurse-limit Disable a limit on recursion whilst demangling\n\ -w, --wide Format output for more than 80 columns\n\ + -U[d|l|i|x|e|h] Controls the display of UTF-8 unicode characters\n\ + --unicode=[default|locale|invalid|hex|escape|highlight]\n\ -z, --disassemble-zeroes Do not skip blocks of zeroes when disassembling\n\ --start-address=ADDR Only process data whose address is >= ADDR\n\ --stop-address=ADDR Only process data whose address is < ADDR\n\ @@ -381,6 +395,7 @@ static struct option long_options[]= {"stop-address", required_argument, NULL, OPTION_STOP_ADDRESS}, {"syms", no_argument, NULL, 't'}, {"target", required_argument, NULL, 'b'}, + {"unicode", required_argument, NULL, 'U'}, {"version", no_argument, NULL, 'V'}, {"wide", no_argument, NULL, 'w'}, {"prefix", required_argument, NULL, OPTION_PREFIX}, @@ -401,9 +416,121 @@ nonfatal (const char *msg) exit_status = 1; } +/* Convert a potential UTF-8 encoded sequence in IN into characters in OUT. + The conversion format is controlled by the unicode_display variable. + Returns the number of characters added to OUT. + Returns the number of bytes consumed from IN in CONSUMED. + Always consumes at least one byte and displays at least one character. */ + +static unsigned int +display_utf8 (const unsigned char * in, char * out, unsigned int * consumed) +{ + char * orig_out = out; + unsigned int nchars = 0; + unsigned int j; + + if (unicode_display == unicode_default) + goto invalid; + + if (in[0] < 0xc0) + goto invalid; + + if ((in[1] & 0xc0) != 0x80) + goto invalid; + + if ((in[0] & 0x20) == 0) + { + nchars = 2; + goto valid; + } + + if ((in[2] & 0xc0) != 0x80) + goto invalid; + + if ((in[0] & 0x10) == 0) + { + nchars = 3; + goto valid; + } + + if ((in[3] & 0xc0) != 0x80) + goto invalid; + + nchars = 4; + + valid: + switch (unicode_display) + { + case unicode_locale: + /* Copy the bytes into the output buffer as is. */ + memcpy (out, in, nchars); + out += nchars; + break; + + case unicode_invalid: + case unicode_hex: + out += sprintf (out, "%c", unicode_display == unicode_hex ? '<' : '{'); + out += sprintf (out, "0x"); + for (j = 0; j < nchars; j++) + out += sprintf (out, "%02x", in [j]); + out += sprintf (out, "%c", unicode_display == unicode_hex ? '>' : '}'); + break; + + case unicode_highlight: + if (isatty (1)) + out += sprintf (out, "\x1B[31;47m"); /* Red. */ + /* Fall through. */ + case unicode_escape: + switch (nchars) + { + case 2: + out += sprintf (out, "\\u%02x%02x", + ((in[0] & 0x1c) >> 2), + ((in[0] & 0x03) << 6) | (in[1] & 0x3f)); + break; + + case 3: + out += sprintf (out, "\\u%02x%02x", + ((in[0] & 0x0f) << 4) | ((in[1] & 0x3c) >> 2), + ((in[1] & 0x03) << 6) | ((in[2] & 0x3f))); + break; + + case 4: + out += sprintf (out, "\\u%02x%02x%02x", + ((in[0] & 0x07) << 6) | ((in[1] & 0x3c) >> 2), + ((in[1] & 0x03) << 6) | ((in[2] & 0x3c) >> 2), + ((in[2] & 0x03) << 6) | ((in[3] & 0x3f))); + break; + default: + /* URG. */ + break; + } + + if (unicode_display == unicode_highlight && isatty (1)) + out += sprintf (out, "\033[0m"); /* Default colour. */ + break; + + default: + /* URG */ + break; + } + + * consumed = nchars; + return out - orig_out; + + invalid: + /* Not a valid UTF-8 sequence. */ + *out = *in; + * consumed = 1; + return 1; +} + /* Returns a version of IN with any control characters replaced by escape sequences. Uses a static buffer - if necessary. */ + if necessary. + + If unicode display is enabled, then also handles the + conversion of unicode characters. */ static const char * sanitize_string (const char * in) @@ -421,40 +548,50 @@ sanitize_string (const char * in) of cases it will not be needed. */ do { - char c = *in++; + unsigned char c = *in++; if (c == 0) return original; if (ISCNTRL (c)) break; + + if (unicode_display != unicode_default && c >= 0xc0) + break; } while (1); /* Copy the input, translating as needed. */ in = original; - if (buffer_len < (strlen (in) * 2)) + if (buffer_len < (strlen (in) * 9)) { free ((void *) buffer); - buffer_len = strlen (in) * 2; + buffer_len = strlen (in) * 9; buffer = xmalloc (buffer_len + 1); } out = buffer; do { - char c = *in++; + unsigned char c = *in++; if (c == 0) break; - if (!ISCNTRL (c)) - *out++ = c; - else + if (ISCNTRL (c)) { *out++ = '^'; *out++ = c + 0x40; } + else if (unicode_display != unicode_default && c >= 0xc0) + { + unsigned int num_consumed; + + out += display_utf8 ((const unsigned char *)(in - 1), out, & num_consumed); + in += num_consumed - 1; + } + else + *out++ = c; } while (1); @@ -4367,6 +4504,24 @@ dump_symbols (bfd *abfd ATTRIBUTE_UNUSED, bfd_boolean dynamic) free (alloc); } } + else if (unicode_display != unicode_default + && name != NULL && *name != '\0') + { + const char * sanitized_name; + + /* If we want to sanitize the name, we do it here, and + temporarily clobber it while calling bfd_print_symbol. + FIXME: This is a gross hack. */ + sanitized_name = sanitize_string (name); + if (sanitized_name != name) + (*current)->name = sanitized_name; + else + sanitized_name = NULL; + bfd_print_symbol (cur_bfd, stdout, *current, + bfd_print_symbol_all); + if (sanitized_name != NULL) + (*current)->name = name; + } else bfd_print_symbol (cur_bfd, stdout, *current, bfd_print_symbol_all); @@ -5048,7 +5203,7 @@ main (int argc, char **argv) set_default_bfd_target (); while ((c = getopt_long (argc, argv, - "pP:ib:m:M:VvCdDlfFaHhrRtTxsSI:j:wE:zgeGW::", + "pP:ib:m:M:U:VvCdDlfFaHhrRtTxsSI:j:wE:zgeGW::", long_options, (int *) 0)) != EOF) { @@ -5325,6 +5480,23 @@ main (int argc, char **argv) seenflag = TRUE; 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 'H': usage (stdout, 0); /* No need to set seenflag or to break - usage() does not return. */ diff --git a/binutils/readelf.c b/binutils/readelf.c index 8fc5bfa..f16fd5f 100644 --- a/binutils/readelf.c +++ b/binutils/readelf.c @@ -299,6 +299,18 @@ typedef enum print_mode } print_mode; +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; + /* Versioned symbol info. */ enum versioned_symbol_info { @@ -554,11 +566,18 @@ print_symbol (signed int width, const char *symbol) if (c == 0) break; - /* Do not print control characters directly as they can affect terminal - settings. Such characters usually appear in the names generated - by the assembler for local labels. */ - if (ISCNTRL (c)) + if (ISPRINT (c)) + { + putchar (c); + width_remaining --; + num_printed ++; + } + else if (ISCNTRL (c)) { + /* Do not print control characters directly as they can affect terminal + settings. Such characters usually appear in the names generated + by the assembler for local labels. */ + if (width_remaining < 2) break; @@ -566,11 +585,137 @@ print_symbol (signed int width, const char *symbol) width_remaining -= 2; num_printed += 2; } - else if (ISPRINT (c)) + else if (c == 0x7f) { - putchar (c); - width_remaining --; - num_printed ++; + if (width_remaining < 5) + break; + printf (""); + width_remaining -= 5; + num_printed += 5; + } + else if (unicode_display != unicode_locale + && unicode_display != unicode_default) + { + /* Display unicode characters as something else. */ + unsigned char bytes[4]; + bfd_boolean is_utf8; + unsigned int nbytes; + + bytes[0] = c; + + if (bytes[0] < 0xc0) + { + nbytes = 1; + is_utf8 = FALSE; + } + else + { + bytes[1] = *symbol++; + + if ((bytes[1] & 0xc0) != 0x80) + { + is_utf8 = FALSE; + /* Do not consume this character. It may only + be the first byte in the sequence that was + corrupt. */ + --symbol; + nbytes = 1; + } + else if ((bytes[0] & 0x20) == 0) + { + is_utf8 = TRUE; + nbytes = 2; + } + else + { + bytes[2] = *symbol++; + + if ((bytes[2] & 0xc0) != 0x80) + { + is_utf8 = FALSE; + symbol -= 2; + nbytes = 1; + } + else if ((bytes[0] & 0x10) == 0) + { + is_utf8 = TRUE; + nbytes = 3; + } + else + { + bytes[3] = *symbol++; + + nbytes = 4; + + if ((bytes[3] & 0xc0) != 0x80) + { + is_utf8 = FALSE; + symbol -= 3; + nbytes = 1; + } + else + is_utf8 = TRUE; + } + } + } + + if (unicode_display == unicode_invalid) + is_utf8 = FALSE; + + if (unicode_display == unicode_hex || ! is_utf8) + { + unsigned int i; + + if (width_remaining < (nbytes * 2) + 2) + break; + + putchar (is_utf8 ? '<' : '{'); + printf ("0x"); + for (i = 0; i < nbytes; i++) + printf ("%02x", bytes[i]); + putchar (is_utf8 ? '>' : '}'); + } + else + { + if (unicode_display == unicode_highlight && isatty (1)) + printf ("\x1B[31;47m"); /* Red. */ + + switch (nbytes) + { + case 2: + if (width_remaining < 6) + break; + printf ("\\u%02x%02x", + (bytes[0] & 0x1c) >> 2, + ((bytes[0] & 0x03) << 6) | (bytes[1] & 0x3f)); + break; + case 3: + if (width_remaining < 6) + break; + printf ("\\u%02x%02x", + ((bytes[0] & 0x0f) << 4) | ((bytes[1] & 0x3c) >> 2), + ((bytes[1] & 0x03) << 6) | (bytes[2] & 0x3f)); + break; + case 4: + if (width_remaining < 8) + break; + printf ("\\u%02x%02x%02x", + ((bytes[0] & 0x07) << 6) | ((bytes[1] & 0x3c) >> 2), + ((bytes[1] & 0x03) << 6) | ((bytes[2] & 0x3c) >> 2), + ((bytes[2] & 0x03) << 6) | (bytes[3] & 0x3f)); + + break; + default: + /* URG. */ + break; + } + + if (unicode_display == unicode_highlight && isatty (1)) + printf ("\033[0m"); /* Default colour. */ + } + + if (bytes[nbytes - 1] == 0) + break; } else { @@ -4449,6 +4594,7 @@ static struct option options[] = {"section-headers", no_argument, 0, 'S'}, {"section-groups", no_argument, 0, 'g'}, {"section-details", no_argument, 0, 't'}, + {"unicode", required_argument, NULL, 'U'}, {"full-section-name",no_argument, 0, 'N'}, {"symbols", no_argument, 0, 's'}, {"syms", no_argument, 0, 's'}, @@ -4526,6 +4672,12 @@ usage (FILE * stream) =addr,=cu_index,=links,=follow-links]\n\ Display the contents of DWARF debug sections\n")); fprintf (stream, _("\ + -U[dlexhi] --unicode=[default|locale|escape|hex|highlight|invalid]\n\ + Display unicode characters as determined by the current locale\n\ + (default), escape sequences, \"\", highlighted\n\ + escape sequences, or treat them as invalid and display as\n\ + \"{hex sequences}\"\n")); + fprintf (stream, _("\ --dwarf-depth=N Do not display DIEs at depth N or greater\n\ --dwarf-start=N Display DIEs starting with N, at the same depth\n\ or deeper\n")); @@ -4639,7 +4791,7 @@ parse_args (Filedata * filedata, int argc, char ** argv) usage (stderr); while ((c = getopt_long - (argc, argv, "ADHINR:SVWacdeghi:lnp:rstuvw::x:z", options, NULL)) != EOF) + (argc, argv, "ADHINR:SU:VWacdeghi:lnp:rstuvw::x:z", options, NULL)) != EOF) { switch (c) { @@ -4795,6 +4947,26 @@ parse_args (Filedata * filedata, int argc, char ** argv) case 'W': do_wide = TRUE; break; + + case 'U': + if (optarg == NULL) + error (_("Missing arg to -U/--unicode")); /* Can this happen ? */ + else 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 + error (_("invalid argument to -U/--unicode: %s"), optarg); + break; + default: /* xgettext:c-format */ error (_("Invalid option '-%c'\n"), c); diff --git a/binutils/strings.c b/binutils/strings.c index dc42a6f..c6b95dc 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 unsigned int string_min; /* Whether or not we include all whitespace as a graphic char. */ static bfd_boolean include_all_whitespace; @@ -130,12 +159,13 @@ static struct option long_options[] = {"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 bfd_boolean 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 **); @@ -173,7 +203,7 @@ 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) @@ -246,6 +276,23 @@ 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"); @@ -260,6 +307,9 @@ 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); @@ -295,14 +345,14 @@ 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 { @@ -344,7 +394,7 @@ 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); } @@ -429,7 +479,7 @@ 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) { @@ -553,11 +603,627 @@ 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 unsigned int +display_utf8_char (const unsigned char * buffer) +{ + unsigned int j; + unsigned int 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, unsigned int * 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, + unsigned int 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; + unsigned int num_read = 0; + unsigned int num_chars = 0; + unsigned int num_print = 0; + int c = 0; + + /* 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); + + unsigned int 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]; + unsigned int 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 @@ -568,20 +1234,29 @@ 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; + unsigned int 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++) { @@ -603,70 +1278,7 @@ 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: -#ifdef HAVE_LONG_LONG - if (sizeof (start) > sizeof (long)) - { -# ifndef __MSVCRT__ - printf ("%7llo ", (unsigned long long) start); -# else - printf ("%7I64o ", (unsigned long long) start); -# endif - } - else -#elif !BFD_HOST_64BIT_LONG - if (start != (unsigned long) start) - printf ("++%7lo ", (unsigned long) start); - else -#endif - printf ("%7lo ", (unsigned long) start); - break; - - case 10: -#ifdef HAVE_LONG_LONG - if (sizeof (start) > sizeof (long)) - { -# ifndef __MSVCRT__ - printf ("%7llu ", (unsigned long long) start); -# else - printf ("%7I64d ", (unsigned long long) start); -# endif - } - else -#elif !BFD_HOST_64BIT_LONG - if (start != (unsigned long) start) - printf ("++%7lu ", (unsigned long) start); - else -#endif - printf ("%7ld ", (long) start); - break; - - case 16: -#ifdef HAVE_LONG_LONG - if (sizeof (start) > sizeof (long)) - { -# ifndef __MSVCRT__ - printf ("%7llx ", (unsigned long long) start); -# else - printf ("%7I64x ", (unsigned long long) start); -# endif - } - else -#elif !BFD_HOST_64BIT_LONG - if (start != (unsigned long) start) - printf ("%lx%8.8lx ", (unsigned long) (start >> 32), - (unsigned long) (start & 0xffffffff)); - else -#endif - printf ("%7lx ", (unsigned long) start); - break; - } + print_filename_and_address (filename, start); buf[i] = '\0'; fputs (buf, stdout); @@ -718,6 +1330,8 @@ usage (FILE *stream, int status) -T --target= 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 used to separate strings in output.\n\ @ Read options from \n\ -h --help Display this information\n\ -- 1.8.3.1