-# Copyright 2001 Free Software Foundation, Inc.
+# This testcase is part of GDB, the GNU debugger.
+
+# Copyright 2001, 2004, 2007, 2008, 2009 Free Software Foundation, Inc.
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
-# the Free Software Foundation; either version 2 of the License, or
+# the Free Software Foundation; either version 3 of the License, or
# (at your option) any later version.
-#
+#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
-#
+#
# You should have received a copy of the GNU General Public License
-# along with this program; if not, write to the Free Software
-# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+# along with this program. If not, see <http://www.gnu.org/licenses/>.
# Please email any bugs, comments, and/or additions to this file to:
-# bug-gdb@prep.ai.mit.edu
+# bug-gdb@gnu.org
# Test GDB's character set support.
set srcfile ${testfile}.c
set binfile ${objdir}/${subdir}/${testfile}
if { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable {debug}] != "" } {
- gdb_suppress_entire_file "Testcase compile failed, so all tests in this file will automatically fail."
+ untested "couldn't compile ${srcdir}/${subdir}/${srcfile}"
+ return -1
}
# Start with a fresh gdb.
global gdb_prompt
gdb_expect {
- -re "The current host and target character set is `(.*)'\\.\[\r\n\]+$gdb_prompt $" {
+ -re "The host character set is \"(.*)\"\\.\[\r\n\]+The target character set is \"(.*)\"\\.\[\r\n\]+The target wide character set is \"(.*)\"\\.\[\r\n\]+$gdb_prompt $" {
set host_charset $expect_out(1,string)
- set target_charset $expect_out(1,string)
+ set target_charset $expect_out(2,string)
+ set retlist [list $host_charset $target_charset]
pass $testname
}
- -re "The current host character set is `(.*)'\\.\[\r\n\]+The current target character set is `(.*)'\\.\[\r\n\]+$gdb_prompt $" {
+ -re "The host character set is \"(.*)\"\\.\[\r\n\]+$gdb_prompt $" {
set host_charset $expect_out(1,string)
- set target_charset $expect_out(2,string)
+ set retlist [list $host_charset]
+ pass $testname
+ }
+ -re "The target character set is \"(.*)\"\\.\[\r\n\]+$gdb_prompt $" {
+ set target_charset $expect_out(1,string)
+ set retlist [list $target_charset]
pass $testname
}
-re ".*$gdb_prompt $" {
}
}
- return [list $host_charset $target_charset]
+ return $retlist
}
-# Try the various `show charset' commands. These are all aliases of each
-# other; `show target-charset' and `show host-charset' actually print
-# both the host and target charsets.
+# Try the various `show charset' commands.
send_gdb "show charset\n"
set show_charset [parse_show_charset_output "show charset"]
send_gdb "show target-charset\n"
-set show_target_charset [parse_show_charset_output "show target-charset"]
+set show_target_charset \
+ [lindex [parse_show_charset_output "show target-charset"] 0]
-if {! [string compare $show_charset $show_target_charset]} {
+if {[lsearch -exact $show_charset $show_target_charset] >= 0} {
pass "check `show target-charset' against `show charset'"
} else {
fail "check `show target-charset' against `show charset'"
}
send_gdb "show host-charset\n"
-set show_host_charset [parse_show_charset_output "show host-charset"]
+set show_host_charset \
+ [lindex [parse_show_charset_output "show host-charset"] 0]
-if {! [string compare $show_charset $show_host_charset]} {
+if {[lsearch -exact $show_charset $show_host_charset] >= 0} {
pass "check `show host-charset' against `show charset'"
} else {
fail "check `show host-charset' against `show charset'"
}
+# Try a malformed `set charset'.
+gdb_test "set charset" \
+ "Requires an argument. Valid arguments are.*" \
+ "try malformed `set charset'"
-# Get the list of supported charsets.
-send_gdb "set charset\n"
-
-# True iff we've seen the "Valid character sets are:" message.
-set seen_valid 0
+# Try using `set host-charset' on an invalid character set.
+gdb_test "set host-charset my_grandma_bonnie" \
+ "Undefined item: \"my_grandma_bonnie\"." \
+ "try `set host-charset' with invalid charset"
-# True iff we've seen the "can be used as a host character set" message.
-set seen_can_host 0
+# Try using `set target-charset' on an invalid character set.
+gdb_test "set target-charset my_grandma_bonnie" \
+ "Undefined item: \"my_grandma_bonnie\"." \
+ "try `set target-charset' with invalid charset"
# A Tcl array mapping the names of all the character sets we've seen
# to "1" if the character set can be used as a host character set, or
return [array names charsets]
}
-proc charset_exists {charset} {
+proc valid_host_charset {charset} {
global charsets
- return [info exists charsets($charset)]
+ return [expr {[info exists charsets($charset)] && $charsets($charset)}]
}
-proc valid_host_charset {charset} {
+proc valid_target_charset {charset} {
global charsets
- return $charsets($charset)
+ return [info exists charsets($charset)]
}
+send_gdb "set host-charset\n"
gdb_expect {
- -re "Valid character sets are:\[\r\n\]+" {
- # There's no ^ at the beginning of the pattern above, so that
- # expect can skip the echoed `set charset' command.
- set seen_valid 1
- exp_continue
+ -re "Requires an argument. Valid arguments are (\[^ \t\n\r,.\]*)" {
+ #set host_charset_list $expect_out(1,string)
+ set charsets($expect_out(1,string)) 1
+ exp_continue
+ #pass "capture valid host charsets"
}
- -re "^ (\[^ \t\n\]*) \\*\[\r\n\]+" {
- set charsets($expect_out(1,string)) 1
- exp_continue
- }
- -re "^ (\[^ \t\n\]*)\[ \t\]*\[\r\n\]+" {
- set charsets($expect_out(1,string)) 0
- exp_continue
+
+ -re ", (\[^ \t\n\r,.\]*)" {
+ #set host_charset_list $expect_out(1,string)
+ set charsets($expect_out(1,string)) 1
+ exp_continue
+ #pass "capture valid host charsets"
}
- -re "^\\* - can be used as a host character set\[\r\n\]+" {
- set seen_can_host 1
- exp_continue
+
+ -re "\\.\r\n$gdb_prompt $" {
+ #set host_charset_list $expect_out(1,string)
+ pass "capture valid host charsets"
}
- -re ".*${gdb_prompt} $" {
- # We don't do an exp_continue here.
+
+ -re ".*$gdb_prompt $" {
+ fail "capture valid host charsets"
}
timeout {
- fail "get valid character sets (timeout)"
+ fail "(timeout) capture valid host charsets"
}
}
+# If gdb was built with a phony iconv, it will only have two character
+# sets: "auto" and the default. In this situation, this set of tests
+# is pointless.
+if {[llength [array names charsets]] < 3} {
+ untested charset.exp
+ return -1
+}
-# Check that we've seen all the right pieces of the output, and that
-# we can at least use ASCII as a host character set.
-if {$seen_valid && $seen_can_host && [charset_exists ascii]} {
- # We can't do the below as part of the test above, since all the
- # [] substitution takes place before any expression evaluation
- # takes place; && doesn't really short circuit things the way
- # you'd like. We'd get an "can't read $charsets(ascii)" error
- # even when `info exists' had returned zero.
- if {[valid_host_charset ascii]} {
- pass "get valid character sets"
- } else {
- fail "get valid character sets"
+send_gdb "set target-charset\n"
+gdb_expect {
+ -re "Requires an argument. Valid arguments are (\[^ \t\n\r,.\]*)" {
+ set target_charset $expect_out(1,string)
+ if {! [info exists charsets($target_charset)]} {
+ set charsets($target_charset) 0
+ }
+ exp_continue
}
-} else {
- fail "get valid character sets (no ascii charset)"
-}
+ -re ", (\[^ \t\n\r,.\]*)" {
+ set target_charset $expect_out(1,string)
+ if {! [info exists charsets($target_charset)]} {
+ set charsets($target_charset) 0
+ }
+ exp_continue
+ }
-# Try using `set host-charset' on an invalid character set.
-gdb_test "set host-charset my_grandma_bonnie" \
- "GDB doesn't know of any character set named `my_grandma_bonnie'." \
- "try `set host-charset' with invalid charset"
+ -re "\\.\r\n$gdb_prompt $" {
+ pass "capture valid target charsets"
+ }
-# Try using `set target-charset' on an invalid character set.
-gdb_test "set target-charset my_grandma_bonnie" \
- "GDB doesn't know of any character set named `my_grandma_bonnie'." \
- "try `set target-charset' with invalid charset"
+ -re ".*$gdb_prompt $" {
+ fail "capture valid target charsets"
+ }
+ timeout {
+ fail "(timeout) capture valid target charsets"
+ }
+}
-# Make sure that GDB supports every host/target charset combination.
-foreach host_charset [all_charset_names] {
+# We don't want to test all the charset names here, since that would
+# be too many combinations. We we pick a subset.
+set charset_subset {ASCII ISO-8859-1 EBCDIC-US IBM1047}
+foreach host_charset $charset_subset {
if {[valid_host_charset $host_charset]} {
set testname "try `set host-charset $host_charset'"
# Now try setting every possible target character set,
# given that host charset.
- foreach target_charset [all_charset_names] {
+ foreach target_charset $charset_subset {
+ if {![valid_target_charset $target_charset]} {
+ continue
+ }
set testname "try `set target-charset $target_charset'"
send_gdb "set target-charset $target_charset\n"
gdb_expect {
# some strings in various target character sets. We need to run the
# test program to the point at which the strings have been
# initialized.
-gdb_test "break [gdb_get_line_number "all strings initialized"]" \
+gdb_test "break ${srcfile}:[gdb_get_line_number "all strings initialized"]" \
".*Breakpoint.* at .*" \
"set breakpoint after all strings have been initialized"
gdb_run_cmd
}
-gdb_test "set host-charset ascii" ""
-foreach target_charset [all_charset_names] {
- send_gdb "set target-charset $target_charset\n"
+# We only try the wide character tests on machines where the wchar_t
+# typedef in the test case has the right size.
+set wchar_size [get_sizeof wchar_t 99]
+set wchar_ok 0
+if {$wchar_size == 2} {
+ lappend charset_subset UCS-2
+ set wchar_ok 1
+} elseif {$wchar_size == 4} {
+ lappend charset_subset UCS-4
+ set wchar_ok 1
+}
+
+gdb_test "set host-charset ASCII" ""
+foreach target_charset $charset_subset {
+ if {![valid_target_charset $target_charset]} {
+ continue
+ }
+
+ if {$target_charset == "UCS-4" || $target_charset == "UCS-2"} {
+ set param target-wide-charset
+ set L L
+ } else {
+ set param target-charset
+ set L ""
+ }
+ send_gdb "set $param $target_charset\n"
gdb_expect {
-re "$gdb_prompt $" {
- pass "set target-charset $target_charset"
+ pass "set $param $target_charset"
}
timeout {
- fail "set target-charset $target_charset (timeout)"
+ fail "set $param $target_charset (timeout)"
}
}
# Try printing the null character. There seems to be a bug in
# gdb_test that requires us to use gdb_expect here.
- send_gdb "print '\\0'\n"
+ send_gdb "print $L'\\0'\n"
gdb_expect {
- -re "\\\$${decimal} = 0 '\\\\0'\[\r\n\]+$gdb_prompt $" {
+ -re "\\\$${decimal} = 0 $L'\\\\0'\[\r\n\]+$gdb_prompt $" {
pass "print the null character in ${target_charset}"
}
-re "$gdb_prompt $" {
# a string in $target_charset. The variable's name is the
# character set's name, in lower-case, with all non-identifier
# characters replaced with '_', with "_string" stuck on the end.
- set var_name [string tolower "${target_charset}_string"]
- regsub -all -- "\[^a-z0-9_\]" $var_name "_" var_name
+ if {$target_charset == "UCS-2"} {
+ # We still use the ucs_4_string variable -- but the size is
+ # correct for UCS-2.
+ set var_name ucs_4_string
+ } else {
+ set var_name [string tolower "${target_charset}_string"]
+ regsub -all -- "\[^a-z0-9_\]" $var_name "_" var_name
+ }
# Compute a regexp matching the results we expect. This is static,
# but it's easier than writing it out.
- regsub -all "." "abefnrtv" "(\\\\&|x)" escapes
+ regsub -all "." "abfnrtv" "(\\\\&|x)" escapes
set uppercase "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
set lowercase "abcdefghijklmnopqrstuvwxyz"
set digits "0123456789"
- set octal_escape "\\\\\[0-9\]\[0-9\]\[0-9\]"
+ set octal_escape "\\\\\[0-9\]+"
send_gdb "print $var_name\n"
# ${escapes}${uppercase}${lowercase}${digits}${octal}${octal}
gdb_expect {
- -re ".* = \"(\\\\a|x)(\\\\b|x)(\\\\e|x)(\\\\f|x)(\\\\n|x)(\\\\r|x)(\\\\t|x)(\\\\v|x)${uppercase}${lowercase}${digits}(\\\\\[0-9\]\[0-9\]\[0-9\]|x)(\\\\\[0-9\]\[0-9\]\[0-9\]|x).*\"\[\r\n\]+$gdb_prompt $" {
+ -re ".* = $L\"(\\\\a|x)(\\\\b|x)(\\\\f|x)(\\\\n|x)(\\\\r|x)(\\\\t|x)(\\\\v|x)${uppercase}${lowercase}${digits}(${octal_escape}|x)+\"\[\r\n\]+$gdb_prompt $" {
pass "print string in $target_charset"
}
-re "$gdb_prompt $" {
}
# Try entering a character literal, and see if it comes back unchanged.
- gdb_test "print 'A'" \
- " = \[0-9-\]+ 'A'" \
+ gdb_test "print $L'A'" \
+ " = \[0-9-\]+ $L'A'" \
"parse character literal in ${target_charset}"
# Check that the character literal was encoded correctly.
- gdb_test "print 'A' == $var_name\[8\]" \
+ gdb_test "print $L'A' == $var_name\[7\]" \
" = 1" \
"check value of parsed character literal in ${target_charset}"
# Try entering a string literal, and see if it comes back unchanged.
- gdb_test "print \"abcdefABCDEF012345\"" \
- " = \"abcdefABCDEF012345\"" \
+ gdb_test "print $L\"abcdefABCDEF012345\"" \
+ " = $L\"abcdefABCDEF012345\"" \
"parse string literal in ${target_charset}"
# Check that the string literal was encoded correctly.
- gdb_test "print \"q\"\[0\] == $var_name\[50\]" \
+ gdb_test "print $L\"q\"\[0\] == $var_name\[49\]" \
" = 1" \
"check value of parsed string literal in ${target_charset}"
# Test handling of characters in the target charset which
# can't be translated into the host charset.
if {! [string compare $target_charset iso-8859-1]} {
- gdb_test "print iso_8859_1_string\[70\]" \
+ gdb_test "print iso_8859_1_string\[69\]" \
" = \[0-9-\]+ '\\\\242'" \
"print character with no equivalent in host character set"
gdb_test "print iso_8859_1_string + 70" \
# Make sure that we don't apply the ISO-8859-1 `print_literally'
# function to ASCII.
if {! [string compare $target_charset ascii]} {
- gdb_test "print iso_8859_1_string\[70\]" \
+ gdb_test "print iso_8859_1_string\[69\]" \
" = \[0-9-\]+ '\\\\242'" \
"print ASCII unprintable character"
gdb_test "print iso_8859_1_string + 70" \
}
# Try printing characters with backslash escape equivalents.
- set escapees {a b e f n r t v}
+ set escapees {a b f n r t v}
for {set i 0} {$i < [llength $escapees]} {incr i} {
set escape [lindex $escapees $i]
send_gdb "print $var_name\[$i\]\n"
set have_escape 1
gdb_expect {
- -re "= \[0-9-\]+ '\\\\${escape}'\[\r\n\]+$gdb_prompt $" {
+ -re "= \[0-9-\]+ $L'\\\\${escape}'\[\r\n\]+$gdb_prompt $" {
pass "try printing '\\${escape}' in ${target_charset}"
}
-re "= \[0-9-\]+ 'x'\[\r\n\]+$gdb_prompt $" {
if {$have_escape} {
# Try parsing a backslash escape in a character literal.
- gdb_test "print '\\${escape}' == $var_name\[$i\]" \
+ gdb_test "print $L'\\${escape}' == $var_name\[$i\]" \
" = 1" \
"check value of '\\${escape}' in ${target_charset}"
# Try parsing a backslash escape in a string literal.
- gdb_test "print \"\\${escape}\"\[0\] == $var_name\[$i\]" \
+ gdb_test "print $L\"\\${escape}\"\[0\] == $var_name\[$i\]" \
" = 1" \
"check value of \"\\${escape}\" in ${target_charset}"
}
# Try printing a character escape that doesn't exist. We should
# get the unescaped character, in the target character set.
- gdb_test "print '\\q'" " = \[0-9-\]+ 'q'" \
+ gdb_test "print $L'\\q'" " = \[0-9-\]+ $L'q'" \
"print escape that doesn't exist in $target_charset"
- gdb_test "print '\\q' == $var_name\[50\]" " = 1" \
+ gdb_test "print $L'\\q' == $var_name\[49\]" " = 1" \
"check value of escape that doesn't exist in $target_charset"
}
+# Reset the target charset.
+gdb_test "set target-charset UTF-8" ""
+
+# \242 is not a valid UTF-8 character.
+gdb_test "print \"\\242\"" " = \"\\\\242\"" \
+ "non-representable target character"
+
+gdb_test "print '\\x'" "\\\\x escape without a following hex digit."
+gdb_test "print '\\u'" "\\\\u escape without a following hex digit."
+gdb_test "print '\\9'" " = \[0-9\]+ '9'"
+
+# Tests for wide- or unicode- strings. L is the prefix letter to use,
+# either "L" (for wide strings), "u" (for UCS-2), or "U" (for UCS-4).
+# NAME is used in the test names and should be related to the prefix
+# letter in some easy-to-undestand way.
+proc test_wide_or_unicode {L name} {
+ gdb_test "print $L\"ab\" $L\"c\"" " = $L\"abc\"" \
+ "basic $name string concatenation"
+ gdb_test "print $L\"ab\" \"c\"" " = $L\"abc\"" \
+ "narrow and $name string concatenation"
+ gdb_test "print \"ab\" $L\"c\"" " = $L\"abc\"" \
+ "$name and narrow string concatenation"
+ gdb_test "print $L\"\\xe\" $L\"c\"" " = $L\"\\\\16c\"" \
+ "$name string concatenation with escape"
+ gdb_test "print $L\"\" \"abcdef\" \"g\"" \
+ "$L\"abcdefg\"" \
+ "concatenate three strings with empty $name string"
+
+ gdb_test "print $L'a'" "= \[0-9\]+ $L'a'" \
+ "basic $name character"
+}
+
+if {$wchar_ok} {
+ test_wide_or_unicode L wide
+}
+
+set ucs2_ok [expr {[get_sizeof char16_t 99] == 2}]
+if {$ucs2_ok} {
+ test_wide_or_unicode u UCS-2
+}
+
+set ucs4_ok [expr {[get_sizeof char32_t 99] == 4}]
+if {$ucs4_ok} {
+ test_wide_or_unicode U UCS-4
+}
+
+# Test an invalid string combination.
+proc test_combination {L1 name1 L2 name2} {
+ gdb_test "print $L1\"abc\" $L2\"def\"" \
+ "Undefined string concatenation." \
+ "undefined concatenation of $name1 and $name2"
+}
+
+if {$wchar_ok && $ucs2_ok} {
+ test_combination L wide u UCS-2
+}
+if {$wchar_ok && $ucs4_ok} {
+ test_combination L wide U UCS-4
+}
+if {$ucs2_ok && $ucs4_ok} {
+ test_combination u UCS-2 U UCS-4
+}
+
gdb_exit