[Bug locale/30694] New: The iconv program no longer tells the user which given encoding name was wrong

[Bug locale/30694] New: The iconv program no longer tells the user which given encoding name was wrong

[bruno at clisp dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=30694

            Bug ID: 30694
           Summary: The iconv program no longer tells the user which given
                    encoding name was wrong
           Product: glibc
           Version: 2.32
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: locale
          Assignee: unassigned at sourceware dot org
          Reporter: bruno at clisp dot org
  Target Milestone: ---

When the user invokes the iconv program with an invalid FROM_ENCODING or
invalid TO_ENCODING, the error message no longer indicates which of the two was
wrong. Which leaves the user puzzled. Thus, it is a bad usability experience.

How to reproduce:

In glibc 2.32:
$ echo abc | iconv -f ARMSCII-8 -t UCS_4BE
iconv: failed to start conversion processing

In glibc 2.30:
$ echo abc | iconv -f ARMSCII-8 -t UCS_4BE
iconv: conversion to `UCS_4BE' is not supported
Try `iconv --help' of `iconv --usage' for more information.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug locale/30694] The iconv program no longer tells the user which given encoding name was wrong

[bruno at clisp dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=30694

Bruno Haible <bruno at clisp dot org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
               Host|                            |x86_64-linux-gnu
             Target|                            |x86_64-linux-gnu
              Build|                            |x86_64-linux-gnu

--- Comment #1 from Bruno Haible <bruno at clisp dot org> ---
This is a regression. Most likely caused by commit
91927b7c76437db860cd86a7714476b56bb39d07.

The code for producing the useful error message is still there, in
glibc/iconv/iconv_prog.c. It is no longer reached, because __gconv_create_spec
fails before the code even gets to invoke __gconv_open.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug locale/30694] The iconv program no longer tells the user which given encoding name was wrong

[schwab@linux-m68k.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=30694

--- Comment #2 from Andreas Schwab <schwab@linux-m68k.org> ---
Looks like __gconv_open loses errno.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug locale/30694] The iconv program no longer tells the user which given encoding name was wrong

[bruno at clisp dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=30694

--- Comment #3 from Bruno Haible <bruno at clisp dot org> ---
(In reply to Andreas Schwab from comment #2)
> Looks like __gconv_open loses errno.

Indeed, that's more likely than what I wrote in comment #1.

In fact, on glibc 2.31 systems I observe the same result as on glibc 2.32;
therefore most likely, the regression appeared between glibc 2.30 and 2.31,
that is, *before* commit 91927b7c76437db860cd86a7714476b56bb39d07 .

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug locale/30694] The iconv program no longer tells the user which given encoding name was wrong

[schwab@linux-m68k.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=30694

Andreas Schwab <schwab@linux-m68k.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |2.39
             Target|x86_64-linux-gnu            |
         Resolution|---                         |FIXED
               Host|x86_64-linux-gnu            |
             Status|UNCONFIRMED                 |RESOLVED
              Build|x86_64-linux-gnu            |

--- Comment #4 from Andreas Schwab <schwab@linux-m68k.org> ---
Fixed by commit fc72b6d7d8.

-- 
You are receiving this mail because:
You are on the CC list for the bug.