Project

General

Profile

Actions
Copy link

Bug #3154

closed

Nonconforming tolower and toupper with UTF-8 locales

Added by Bernardo Signori over 9 years ago. Updated over 9 years ago.

Status:
Resolved
Priority:
Normal
Assignee:
-
Category:
cmd - userland programs
Start date:
2012-09-05
Due date:
% Done:

100%

Estimated time:
Difficulty:
Medium
Tags:
Gerrit CR:

Description

With any UTF-8 locale, tolower and toupper return null when passed an argument that is an invalid UTF-8 character. My understanding of the c standard is than any "arguments in the domain" that have no corresponding lowercase (or uppercase) letter should be returned unchanged.

Files

Download all files
test.c (395 Bytes) test.c test case Bernardo Signori, 2012-09-05 02:06 PM
test2.c (525 Bytes) test2.c better test case Bernardo Signori, 2012-09-05 02:32 PM
Actions
Copy link
 #1

Updated by Yuri Pankov over 9 years ago

What's the testcase?

Actions
Copy link
 #2

Updated by Igor Pashev over 9 years ago

"an invalid UTF-8 character" != "have no corresponding lowercase (or uppercase) letter"

Actions
Copy link
 #3

Updated by Bernardo Signori over 9 years ago

Actions
Copy link
 #4

Updated by Yuri Pankov over 9 years ago

Indeed.

debian:yuri:~$ uname -a
Linux debian 2.6.32-5-amd64 #1 SMP Sun May 6 04:00:17 UTC 2012 x86_64 GNU/Linux
debian:yuri:~$ ./test
tolower test PASS
toupper test PASS

phoenix:yuri:~$ uname -a
FreeBSD phoenix.xvoid.org 10.0-CURRENT FreeBSD 10.0-CURRENT #0 r240078: Tue Sep  4 07:38:03 MSK 2012     yuri@phoenix.xvoid.org:/usr/obj/data/src/freebsd/head/sys/PHOENIX  amd64
phoenix:yuri:~$ ./test
tolower test PASS
toupper test PASS

sirius:yuri:~$ uname -a
SunOS sirius 5.11 tzdata2012e i86pc i386 i86pc
sirius:yuri:~$ ./test
tolower test FAIL
toupper test FAIL

Actions
Copy link
 #5

Updated by Bernardo Signori over 9 years ago

Updated the test to make sure the locale changed, otherwise the tests may pass because the locale is still C.

Actions
Copy link
 #6

Updated by David Höppner over 9 years ago

Seems to be a bug in localedef.

Could it be that ctypes contains not all values < _CACHED_RUNES? And
the identity map will so not be complete?

for (ctn = avl_first(&ctypes); ctn; ctn = AVL_NEXT(&ctypes, ctn)) {

Actions
Copy link
 #7

Updated by David Höppner over 9 years ago

  • Status changed from New to In Progress
  • % Done changed from 0 to 70
Actions
Copy link
 #8

Updated by Rich Lowe over 9 years ago

  • Category set to cmd - userland programs
  • Status changed from In Progress to Resolved
  • % Done changed from 70 to 100
  • Tags deleted (needs-triage)

Resolved in r13844 commit:90142813f484

Actions
Copy link

Also available in: Atom PDF