Image files with IPTC data using Japanese Shift-JIS

[Mac2]

I think ExifTool cannot handle the Japanese Shift-JIS character set but I wanted to make sure.

In my old application I supported this character set (and others like Arabian) so users could properly convert their IPTC data into UTF-8 and finally convert to XMP. Which was easy because Windows provides the corresponding routines for character set conversion. I guess Perl/Windows does not give Phil this flexibility and this is why ExifTool can support only the character sets listed under -CHARSET.

I use the iptc2xmp.arg files to perform the conversion. Is there a work around to get the Japanese descriptions etc. into XMP that way?

[Phil Harvey]

ExifTool handles Shift-JIS in EXIF.  However, ExifTool does not support ISO 2022 shift codes in IPTC for 2 reasons:

1. I have never seen them used.

2. It would slow processing of IPTC

If you post an example and tell me exactly what you want to do I will see if I can help.

XMP is UTF-8, so it supports essentially all characters.

- Phil

[Phil Harvey]

I got the sample, thanks.

The EXIF ImageDescription and Artist are stored as UTF-8, and extracted properly by ExifTool.

The IPTC Caption-Abstract, By-line and ObjectName are garbage (all binary 0x3f, which is an ASCII question mark), as you can see from the -v3 output:

Code Select Expand

  | + [IPTC directory, 72 bytes]
  | | CurrentIPTCDigest = .a.B/.x...rj....
  | | -- IPTCApplication record --
  | | ApplicationRecordVersion = 2
  | | - Tag 0x0000, IPTCApplication record (2 bytes, int16u):
  | |     257b: 00 02                                           [..]
  | | Caption-Abstract = ????????????
  | | - Tag 0x0078, IPTCApplication record (12 bytes, string[0,2000]):
  | |     2582: 3f 3f 3f 3f 3f 3f 3f 3f 3f 3f 3f 3f             [????????????]
  | | By-line = ????
  | | - Tag 0x0050, IPTCApplication record (4 bytes, string[0,32]):
  | |     2593: 3f 3f 3f 3f                                     [????]
  | | ObjectName = ?????
  | | - Tag 0x0005, IPTCApplication record (5 bytes, string[0,64]):
  | |     259c: 3f 3f 3f 3f 3f                                  [?????]
  | | DigitalCreationDate = 20050807
  | | - Tag 0x003e, IPTCApplication record (8 bytes, digits[8]):
  | |     25a6: 32 30 30 35 30 38 30 37                         [20050807]
  | | DigitalCreationTime = 210614+0900
  | | - Tag 0x003f, IPTCApplication record (11 bytes, string[11]):
  | |     25b3: 32 31 30 36 31 34 2b 30 39 30 30                [210614+0900]


- Phil

Edit: If you wrote this IPTC with ExifTool, then reading FAQ number 10 may help you to understand what needs to be done.