Assorted i18n bugs with bfd, round 2

[Christian Rose]

Here's the next round of bugs.


BUG 4
-----
bfd.c:730

    (*_bfd_error_handler)
      (_("BFD %sinternal error, aborting at %s line %d\n"),
       BFD_VERSION_STRING, file, line);

Not an i18n bug, but I think there should be a space between %s and
"internal" here.


BUG 5
-----
elf32-arm.h:2303

          if (in_flags & EF_ARM_VFP_FLOAT)
            _bfd_error_handler (_("\
Error: %s uses VFP instructions, whereas %s FPA instructions"),
                                bfd_archive_filename (ibfd),
                                bfd_get_filename (obfd));
          else
            _bfd_error_handler (_("\
Error: %s uses FPA instructions, whereas %s VFP instructions"),
                                bfd_archive_filename (ibfd),
                                bfd_get_filename (obfd));

Not an i18n bug either. I think it should be "whereas %s uses".


BUG 6
-----
elf64-mmix.c:1268

          (*_bfd_error_handler)
            (_("%s: register relocation against non-register symbol: %s in %s"),
             bfd_get_filename (input_section->owner),
             symname == NULL || *symname == 0 ? _("(unknown)") : symname,
             bfd_get_section_name (symsec->owner, symsec));

"unknown" is an adjective, and in many languages the form of the
translation depends on the form of the word it is referring to. For this
reason, it would be good to make these cases two seperate messages.


BUG 7,8,9
---------
elf64-sparc.c:1334
elf64-sparc.c:1362
elf64-sparc.c:1411

          (*_bfd_error_handler)
            (_("Register %%g%d used incompatibly: %s in %s"),
             (int) sym->st_value,
             **namep ? *namep : "#scratch", bfd_archive_filename (abfd));
          (*_bfd_error_handler)
            (_("  previously %s in %s"),
             *p->name ? p->name : "#scratch", bfd_archive_filename (p->abfd));

Couldn't these be made one message, since they depend on each other in
that they are the same sentence? Similarily for the other two occasions.


BUG 10
------
ieee.c:877

                    (*_bfd_error_handler)
                      (_("%s: unimplemented ATI record  %u for symbol %u"),
                       bfd_archive_filename (abfd), symbol_attribute_def,
                       symbol_name_index);

There's an extra space in this message.


BUG 11
------
ihex.c:863

                      (*_bfd_error_handler)
                        (_("%s: address 0x%s out of range for Intex Hex file"),
                         bfd_get_filename (abfd), buf);

A typo, I assume this should be "Intel".


BUG 12
------
libbfd.c:1463

      fprintf (stderr, _("Deprecated %s called"), what);
      if (func)
        fprintf (stderr, _(" at %s line %d in %s\n"), file, line, func);
      else
        fprintf (stderr, "\n");

Please make these messages two separate and complete ones (move the if
statement out of the message).


BUG 13,14
---------
vms-tir.c:443
vms-tir.c:465

            (*_bfd_error_handler) (_("ETIR_S_C_STO_GBL: no symbol \"%s\""),
                                   name);

If you replace the ETIR_S_C_STO_GBL with %s and add the name as another
parameter, this message doesn't have to be translated uniquely for every
occasion (the ETIR_S_C_STO_GBL shouldn't be translated anyway):

            (*_bfd_error_handler) (_("%s: no symbol \"%s\""),
                                   fname, name);

Similarily on the other occasions.


BUG 15,16,17,18,19,20,21,22,23,24,25,26
---------------------------------------
vms-tir.c:478
vms-tir.c:538
vms-tir.c:667
vms-tir.c:685
vms-tir.c:691
vms-tir.c:710
vms-tir.c:716
vms-tir.c:816
vms-tir.c:834
vms-tir.c:842
vms-tir.c:851
vms-tir.c:1664


      (*_bfd_error_handler) (_("ETIR_S_C_STO_RB/AB: Not supported"));

Similarily. My suggestion:

      (*_bfd_error_handler) (_("%s: Not supported"), fname);

on all occasions.


BUG 27,28,29
------------
vms-tir.c:544
vms-tir.c:550
vms-tir.c:1199

      (*_bfd_error_handler) (_("ETIR_S_C_STO_HINT_GBL: not implemented"));

Similarily. My suggestion:

      (*_bfd_error_handler) (_("%s: not implemented"), fname);

on both occasions.


BUG 30,31,32,33,34,35,36,37
---------------------------
vms-tir.c:1219
vms-tir.c:1235
vms-tir.c:1256
vms-tir.c:1271
vms-tir.c:1294
vms-tir.c:1718
vms-tir.c:1726
vms-tir.c:1734


        (*_bfd_error_handler) (_("Stack-entry-mask not fully implemented"));

Simiarily. My suggestion:

        (*_bfd_error_handler) (_("%s is not fully implemented"), fname);

on all occasions.


BUG 38,39,40
------------
vms-tir.c:1608
vms-tir.c:1622
vms-tir.c:1636

      (*_bfd_error_handler) (_("TIR_S_C_OPR_ROT incomplete"));

Similarily. My suggestion:

      (*_bfd_error_handler) (_("%s is incomplete"), fname);

on all occasions.


BUG 41,42
---------
peigen.c:1132
peigen.c:1543

  fprintf (file,
           _(" vma:\t\tBegin    End      EH       EH       PrologEnd  
Exception\n"));
  fprintf (file,
           _("     \t\tAddress  Address  Handler  Data     Address    Mask\n"));

This is a serious i18n problem. You are dividing words in two separate
halfs, marked seperately for translation. This plain doesn't work in
many languages; the beginning of a word and the end of it cannot be
translated seperately. As an example, "Begin Address" is one word in
Swedish, "Startadress".
Please make this one single message (one single call to _()):

  fprintf (file,
           _(" vma:\t\tBegin    End      EH       EH       PrologEnd  
Exception\n"
             "     \t\tAddress  Address  Handler  Data     Address    Mask\n"));

Similarily on the other occasion.


Thanks for your help!

Christian