gdb:
[deliverable/binutils-gdb.git] / gdb / testsuite / gdb.base / charset.exp
1 # This testcase is part of GDB, the GNU debugger.
2
3 # Copyright 2001, 2004, 2007, 2008, 2009 Free Software Foundation, Inc.
4
5 # This program is free software; you can redistribute it and/or modify
6 # it under the terms of the GNU General Public License as published by
7 # the Free Software Foundation; either version 3 of the License, or
8 # (at your option) any later version.
9 #
10 # This program is distributed in the hope that it will be useful,
11 # but WITHOUT ANY WARRANTY; without even the implied warranty of
12 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 # GNU General Public License for more details.
14 #
15 # You should have received a copy of the GNU General Public License
16 # along with this program. If not, see <http://www.gnu.org/licenses/>.
17
18 # Please email any bugs, comments, and/or additions to this file to:
19 # bug-gdb@gnu.org
20
21 # Test GDB's character set support.
22
23 if $tracelevel then {
24 strace $tracelevel
25 }
26
27 set prms_id 0
28 set bug_id 0
29
30 set testfile "charset"
31 set srcfile ${testfile}.c
32 set binfile ${objdir}/${subdir}/${testfile}
33 if { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable {debug}] != "" } {
34 untested "couldn't compile ${srcdir}/${subdir}/${srcfile}"
35 return -1
36 }
37
38 # Start with a fresh gdb.
39 gdb_exit
40 gdb_start
41 gdb_reinitialize_dir $srcdir/$subdir
42 gdb_load ${binfile}
43
44 # Parse the output from a `show charset' command. Return the host
45 # and target charset as a two-element list.
46 proc parse_show_charset_output {testname} {
47 global gdb_prompt
48
49 gdb_expect {
50 -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 $" {
51 set host_charset $expect_out(1,string)
52 set target_charset $expect_out(2,string)
53 set retlist [list $host_charset $target_charset]
54 pass $testname
55 }
56 -re "The host character set is \"(.*)\"\\.\[\r\n\]+$gdb_prompt $" {
57 set host_charset $expect_out(1,string)
58 set retlist [list $host_charset]
59 pass $testname
60 }
61 -re "The target character set is \"(.*)\"\\.\[\r\n\]+$gdb_prompt $" {
62 set target_charset $expect_out(1,string)
63 set retlist [list $target_charset]
64 pass $testname
65 }
66 -re ".*$gdb_prompt $" {
67 fail $testname
68 }
69 timeout {
70 fail "$testname (timeout)"
71 }
72 }
73
74 return $retlist
75 }
76
77
78 # Try the various `show charset' commands.
79
80 send_gdb "show charset\n"
81 set show_charset [parse_show_charset_output "show charset"]
82
83 send_gdb "show target-charset\n"
84 set show_target_charset \
85 [lindex [parse_show_charset_output "show target-charset"] 0]
86
87 if {[lsearch -exact $show_charset $show_target_charset] >= 0} {
88 pass "check `show target-charset' against `show charset'"
89 } else {
90 fail "check `show target-charset' against `show charset'"
91 }
92
93 send_gdb "show host-charset\n"
94 set show_host_charset \
95 [lindex [parse_show_charset_output "show host-charset"] 0]
96
97 if {[lsearch -exact $show_charset $show_host_charset] >= 0} {
98 pass "check `show host-charset' against `show charset'"
99 } else {
100 fail "check `show host-charset' against `show charset'"
101 }
102
103 # Try a malformed `set charset'.
104 gdb_test "set charset" \
105 "Requires an argument. Valid arguments are.*" \
106 "try malformed `set charset'"
107
108 # Try using `set host-charset' on an invalid character set.
109 gdb_test "set host-charset my_grandma_bonnie" \
110 "Undefined item: \"my_grandma_bonnie\"." \
111 "try `set host-charset' with invalid charset"
112
113 # Try using `set target-charset' on an invalid character set.
114 gdb_test "set target-charset my_grandma_bonnie" \
115 "Undefined item: \"my_grandma_bonnie\"." \
116 "try `set target-charset' with invalid charset"
117
118 # A Tcl array mapping the names of all the character sets we've seen
119 # to "1" if the character set can be used as a host character set, or
120 # "0" otherwise. We can use `array names charsets' just to get a list
121 # of all character sets.
122 array set charsets {}
123
124 proc all_charset_names {} {
125 global charsets
126 return [array names charsets]
127 }
128
129 proc valid_host_charset {charset} {
130 global charsets
131 return [expr {[info exists charsets($charset)] && $charsets($charset)}]
132 }
133
134 send_gdb "set host-charset\n"
135 gdb_expect {
136 -re "Requires an argument. Valid arguments are (\[^ \t\n\r,.\]*)" {
137 #set host_charset_list $expect_out(1,string)
138 set charsets($expect_out(1,string)) 1
139 exp_continue
140 #pass "capture valid host charsets"
141 }
142
143 -re ", (\[^ \t\n\r,.\]*)" {
144 #set host_charset_list $expect_out(1,string)
145 set charsets($expect_out(1,string)) 1
146 exp_continue
147 #pass "capture valid host charsets"
148 }
149
150 -re "\\.\r\n$gdb_prompt $" {
151 #set host_charset_list $expect_out(1,string)
152 pass "capture valid host charsets"
153 }
154
155 -re ".*$gdb_prompt $" {
156 fail "capture valid host charsets"
157 }
158 timeout {
159 fail "(timeout) capture valid host charsets"
160 }
161 }
162
163 # If gdb was built with a phony iconv, it will only have two character
164 # sets: "auto" and the default. In this situation, this set of tests
165 # is pointless.
166 if {[llength [array names charsets]] < 3} {
167 untested charset.exp
168 return -1
169 }
170
171 send_gdb "set target-charset\n"
172 gdb_expect {
173 -re "Requires an argument. Valid arguments are (\[^ \t\n\r,.\]*)" {
174 set target_charset $expect_out(1,string)
175 if {! [info exists charsets($target_charset)]} {
176 set charsets($target_charset) 0
177 }
178 exp_continue
179 }
180
181 -re ", (\[^ \t\n\r,.\]*)" {
182 set target_charset $expect_out(1,string)
183 if {! [info exists charsets($target_charset)]} {
184 set charsets($target_charset) 0
185 }
186 exp_continue
187 }
188
189 -re "\\.\r\n$gdb_prompt $" {
190 pass "capture valid target charsets"
191
192 }
193
194 -re ".*$gdb_prompt $" {
195 fail "capture valid target charsets"
196 }
197
198 timeout {
199 fail "(timeout) capture valid target charsets"
200 }
201 }
202
203 # We don't want to test all the charset names here, since that would
204 # be too many combinations. We we pick a subset.
205 set charset_subset {ASCII ISO-8859-1 EBCDIC-US IBM1047}
206 foreach host_charset $charset_subset {
207 if {[valid_host_charset $host_charset]} {
208
209 set testname "try `set host-charset $host_charset'"
210 send_gdb "set host-charset $host_charset\n"
211 gdb_expect {
212 -re "GDB doesn't know of any character set named.*\[\r\n]+${gdb_prompt} $" {
213 # How did it get into `charsets' then?
214 fail "$testname (didn't recognize name)"
215 }
216 -re "GDB can't use `.*' as its host character set\\.\[\r\n]+${gdb_prompt} $" {
217 # Well, then why does its `charsets' entry say it can?
218 fail $testname
219 }
220 -re "${gdb_prompt} $" {
221 pass $testname
222 }
223 timeout {
224 fail "$testname (timeout)"
225 }
226 }
227
228 # Check that the command actually had its intended effect:
229 # $host_charset should now be the host character set.
230 send_gdb "show charset\n"
231 set result [parse_show_charset_output "parse `show charset' after `set host-charset $host_charset'"]
232 if {! [string compare [lindex $result 0] $host_charset]} {
233 pass "check effect of `set host-charset $host_charset'"
234 } else {
235 fail "check effect of `set host-charset $host_charset'"
236 }
237
238 # Now try setting every possible target character set,
239 # given that host charset.
240 foreach target_charset $charset_subset {
241 set testname "try `set target-charset $target_charset'"
242 send_gdb "set target-charset $target_charset\n"
243 gdb_expect {
244 -re "GDB doesn't know of any character set named.*\[\r\n]+${gdb_prompt} $" {
245 fail "$testname (didn't recognize name)"
246 }
247 -re "GDB can't convert from the .* character set to .*\\.\[\r\n\]+${gdb_prompt} $" {
248 # This is a serious problem. GDB should be able to convert
249 # between any arbitrary pair of character sets.
250 fail "$testname (can't convert)"
251 }
252 -re "${gdb_prompt} $" {
253 pass $testname
254 }
255 timeout {
256 fail "$testname (timeout)"
257 }
258 }
259
260 # Check that the command actually had its intended effect:
261 # $target_charset should now be the target charset.
262 send_gdb "show charset\n"
263 set result [parse_show_charset_output "parse `show charset' after `set target-charset $target_charset'"]
264 if {! [string compare $result [list $host_charset $target_charset]]} {
265 pass "check effect of `set target-charset $target_charset'"
266 } else {
267 fail "check effect of `set target-charset $target_charset'"
268 }
269
270 # Test handling of characters in the host charset which
271 # can't be translated into the target charset. \xA2 is
272 # `cent' in ISO-8859-1, which has no equivalent in ASCII.
273 #
274 # On some systems, the pseudo-tty through which we
275 # communicate with GDB insists on stripping the high bit
276 # from input characters, meaning that `cent' turns into
277 # `"'. Since ISO-8859-1 and ASCII are identical in the
278 # lower 128 characters, it's tough to see how we can test
279 # this behavior on such systems, so we just xfail it.
280 #
281 # Note: the \x16 (Control-V) is an escape to allow \xA2 to
282 # get past readline.
283 if {! [string compare $host_charset iso-8859-1] && ! [string compare $target_charset ascii]} {
284
285 set testname "untranslatable character in character literal"
286 send_gdb "print '\x16\xA2'\n"
287 gdb_expect {
288 -re "There is no character corresponding to .* in the target character set .*\\.\[\r\n\]+$gdb_prompt $" {
289 pass $testname
290 }
291 -re " = 34 '\"'\[\r\n\]+$gdb_prompt $" {
292 xfail "$testname (DejaGNU's pseudo-tty strips eighth bit)"
293 }
294 -re "$gdb_prompt $" {
295 fail $testname
296 }
297 timeout {
298 fail "$testname (timeout)"
299 }
300 }
301
302 set testname "untranslatable character in string literal"
303 # If the PTTY zeros bit seven, then this turns into
304 # print """
305 # which gets us a syntax error. We don't care.
306 send_gdb "print \"\x16\xA2\"\n"
307 gdb_expect {
308 -re "There is no character corresponding to .* in the target character set .*\\.\[\r\n\]+$gdb_prompt $" {
309 pass $testname
310 }
311 -re "Unterminated string in expression.\[\r\n\]+$gdb_prompt $" {
312 xfail "$testname (DejaGNU's pseudo-tty strips eighth bit)"
313 }
314 -re "$gdb_prompt $" {
315 fail $testname
316 }
317 timeout {
318 fail "$testname (timeout)"
319 }
320 }
321
322 set testname "untranslatable characters in backslash escape"
323 send_gdb "print '\\\x16\xA2'\n"
324 gdb_expect {
325 -re "The escape sequence .* is equivalent to plain .*, which has no equivalent\[\r\n\]+in the .* character set\\.\[\r\n\]+$gdb_prompt $" {
326 pass $testname
327 }
328 -re " = 34 '\"'\[\r\n\]+$gdb_prompt $" {
329 xfail "$testname (DejaGNU's pseudo-tty strips eighth bit)"
330 }
331 -re "$gdb_prompt $" {
332 fail $testname
333 }
334 timeout {
335 fail "$testname (timeout)"
336 }
337 }
338 }
339 }
340 }
341 }
342
343
344 # Set the host character set to plain ASCII, and try actually printing
345 # some strings in various target character sets. We need to run the
346 # test program to the point at which the strings have been
347 # initialized.
348 gdb_test "break ${srcfile}:[gdb_get_line_number "all strings initialized"]" \
349 ".*Breakpoint.* at .*" \
350 "set breakpoint after all strings have been initialized"
351 gdb_run_cmd
352 gdb_expect {
353 -re "Breakpoint.*all strings initialized.*$gdb_prompt $" {
354 pass "run until all strings have been initialized"
355 }
356 -re "$gdb_prompt $" {
357 fail "run until all strings have been initialized"
358 }
359 timeout {
360 fail "run until all strings have been initialized (timeout)"
361 }
362 }
363
364
365 # We only try the wide character tests on machines where the wchar_t
366 # typedef in the test case has the right size.
367 set wchar_size [get_sizeof wchar_t 99]
368 set wchar_ok 0
369 if {$wchar_size == 2} {
370 lappend charset_subset UCS-2
371 set wchar_ok 1
372 } elseif {$wchar_size == 4} {
373 lappend charset_subset UCS-4
374 set wchar_ok 1
375 }
376
377 gdb_test "set host-charset ASCII" ""
378 foreach target_charset $charset_subset {
379 if {$target_charset == "UCS-4" || $target_charset == "UCS-2"} {
380 set param target-wide-charset
381 set L L
382 } else {
383 set param target-charset
384 set L ""
385 }
386 send_gdb "set $param $target_charset\n"
387 gdb_expect {
388 -re "$gdb_prompt $" {
389 pass "set $param $target_charset"
390 }
391 timeout {
392 fail "set $param $target_charset (timeout)"
393 }
394 }
395
396 # Try printing the null character. There seems to be a bug in
397 # gdb_test that requires us to use gdb_expect here.
398 send_gdb "print $L'\\0'\n"
399 gdb_expect {
400 -re "\\\$${decimal} = 0 $L'\\\\0'\[\r\n\]+$gdb_prompt $" {
401 pass "print the null character in ${target_charset}"
402 }
403 -re "$gdb_prompt $" {
404 fail "print the null character in ${target_charset}"
405 }
406 timeout {
407 fail "print the null character in ${target_charset} (timeout)"
408 }
409 }
410
411 # Compute the name of the variable in the test program that holds
412 # a string in $target_charset. The variable's name is the
413 # character set's name, in lower-case, with all non-identifier
414 # characters replaced with '_', with "_string" stuck on the end.
415 if {$target_charset == "UCS-2"} {
416 # We still use the ucs_4_string variable -- but the size is
417 # correct for UCS-2.
418 set var_name ucs_4_string
419 } else {
420 set var_name [string tolower "${target_charset}_string"]
421 regsub -all -- "\[^a-z0-9_\]" $var_name "_" var_name
422 }
423
424 # Compute a regexp matching the results we expect. This is static,
425 # but it's easier than writing it out.
426 regsub -all "." "abfnrtv" "(\\\\&|x)" escapes
427 set uppercase "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
428 set lowercase "abcdefghijklmnopqrstuvwxyz"
429 set digits "0123456789"
430 set octal_escape "\\\\\[0-9\]+"
431
432 send_gdb "print $var_name\n"
433 # ${escapes}${uppercase}${lowercase}${digits}${octal}${octal}
434 gdb_expect {
435 -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 $" {
436 pass "print string in $target_charset"
437 }
438 -re "$gdb_prompt $" {
439 fail "print string in $target_charset"
440 }
441 timeout {
442 fail "print string in $target_charset (timeout)"
443 }
444 }
445
446 # Try entering a character literal, and see if it comes back unchanged.
447 gdb_test "print $L'A'" \
448 " = \[0-9-\]+ $L'A'" \
449 "parse character literal in ${target_charset}"
450
451 # Check that the character literal was encoded correctly.
452 gdb_test "print $L'A' == $var_name\[7\]" \
453 " = 1" \
454 "check value of parsed character literal in ${target_charset}"
455
456 # Try entering a string literal, and see if it comes back unchanged.
457 gdb_test "print $L\"abcdefABCDEF012345\"" \
458 " = $L\"abcdefABCDEF012345\"" \
459 "parse string literal in ${target_charset}"
460
461 # Check that the string literal was encoded correctly.
462 gdb_test "print $L\"q\"\[0\] == $var_name\[49\]" \
463 " = 1" \
464 "check value of parsed string literal in ${target_charset}"
465
466 # Test handling of characters in the target charset which
467 # can't be translated into the host charset.
468 if {! [string compare $target_charset iso-8859-1]} {
469 gdb_test "print iso_8859_1_string\[69\]" \
470 " = \[0-9-\]+ '\\\\242'" \
471 "print character with no equivalent in host character set"
472 gdb_test "print iso_8859_1_string + 70" \
473 " = ${hex} \"\\\\242.*\"" \
474 "print string with no equivalent in host character set"
475 }
476
477 # Make sure that we don't apply the ISO-8859-1 `print_literally'
478 # function to ASCII.
479 if {! [string compare $target_charset ascii]} {
480 gdb_test "print iso_8859_1_string\[69\]" \
481 " = \[0-9-\]+ '\\\\242'" \
482 "print ASCII unprintable character"
483 gdb_test "print iso_8859_1_string + 70" \
484 " = ${hex} \"\\\\242.*\"" \
485 "print ASCII unprintable string"
486 }
487
488 # Try printing characters with backslash escape equivalents.
489 set escapees {a b f n r t v}
490 for {set i 0} {$i < [llength $escapees]} {incr i} {
491 set escape [lindex $escapees $i]
492 send_gdb "print $var_name\[$i\]\n"
493 set have_escape 1
494 gdb_expect {
495 -re "= \[0-9-\]+ $L'\\\\${escape}'\[\r\n\]+$gdb_prompt $" {
496 pass "try printing '\\${escape}' in ${target_charset}"
497 }
498 -re "= \[0-9-\]+ 'x'\[\r\n\]+$gdb_prompt $" {
499 xfail "try printing '\\${escape}' in ${target_charset} (no such escape)"
500 set have_escape 0
501 }
502 -re "$gdb_prompt $" {
503 fail "try printing '\\${escape}' in ${target_charset}"
504 }
505 timeout {
506 fail "try printing '\\${escape}' in ${target_charset} (timeout)"
507 }
508 }
509
510 if {$have_escape} {
511
512 # Try parsing a backslash escape in a character literal.
513 gdb_test "print $L'\\${escape}' == $var_name\[$i\]" \
514 " = 1" \
515 "check value of '\\${escape}' in ${target_charset}"
516
517 # Try parsing a backslash escape in a string literal.
518 gdb_test "print $L\"\\${escape}\"\[0\] == $var_name\[$i\]" \
519 " = 1" \
520 "check value of \"\\${escape}\" in ${target_charset}"
521 }
522 }
523
524 # Try printing a character escape that doesn't exist. We should
525 # get the unescaped character, in the target character set.
526 gdb_test "print $L'\\q'" " = \[0-9-\]+ $L'q'" \
527 "print escape that doesn't exist in $target_charset"
528 gdb_test "print $L'\\q' == $var_name\[49\]" " = 1" \
529 "check value of escape that doesn't exist in $target_charset"
530 }
531
532 # Reset the target charset.
533 gdb_test "set target-charset UTF-8" ""
534
535 # \242 is not a valid UTF-8 character.
536 gdb_test "print \"\\242\"" " = \"\\\\242\"" \
537 "non-representable target character"
538
539 gdb_test "print '\\x'" "\\\\x escape without a following hex digit."
540 gdb_test "print '\\u'" "\\\\u escape without a following hex digit."
541 gdb_test "print '\\9'" " = \[0-9\]+ '9'"
542
543 # Tests for wide- or unicode- strings. L is the prefix letter to use,
544 # either "L" (for wide strings), "u" (for UCS-2), or "U" (for UCS-4).
545 # NAME is used in the test names and should be related to the prefix
546 # letter in some easy-to-undestand way.
547 proc test_wide_or_unicode {L name} {
548 gdb_test "print $L\"ab\" $L\"c\"" " = $L\"abc\"" \
549 "basic $name string concatenation"
550 gdb_test "print $L\"ab\" \"c\"" " = $L\"abc\"" \
551 "narrow and $name string concatenation"
552 gdb_test "print \"ab\" $L\"c\"" " = $L\"abc\"" \
553 "$name and narrow string concatenation"
554 gdb_test "print $L\"\\xe\" $L\"c\"" " = $L\"\\\\16c\"" \
555 "$name string concatenation with escape"
556 gdb_test "print $L\"\" \"abcdef\" \"g\"" \
557 "$L\"abcdefg\"" \
558 "concatenate three strings with empty $name string"
559
560 gdb_test "print $L'a'" "= \[0-9\]+ $L'a'" \
561 "basic $name character"
562 }
563
564 if {$wchar_ok} {
565 test_wide_or_unicode L wide
566 }
567
568 set ucs2_ok [expr {[get_sizeof char16_t 99] == 2}]
569 if {$ucs2_ok} {
570 test_wide_or_unicode u UCS-2
571 }
572
573 set ucs4_ok [expr {[get_sizeof char32_t 99] == 4}]
574 if {$ucs4_ok} {
575 test_wide_or_unicode U UCS-4
576 }
577
578 # Test an invalid string combination.
579 proc test_combination {L1 name1 L2 name2} {
580 gdb_test "print $L1\"abc\" $L2\"def\"" \
581 "Undefined string concatenation." \
582 "undefined concatenation of $name1 and $name2"
583 }
584
585 if {$wchar_ok && $ucs2_ok} {
586 test_combination L wide u UCS-2
587 }
588 if {$wchar_ok && $ucs4_ok} {
589 test_combination L wide U UCS-4
590 }
591 if {$ucs2_ok && $ucs4_ok} {
592 test_combination u UCS-2 U UCS-4
593 }
594
595 gdb_exit
This page took 0.041246 seconds and 4 git commands to generate.