Marc8 Escape sequences for G1

I am using siconv.c version 1.29.  I have found the
following problem:

This code only seems to handle MARC8 Escape sequences using
the 
characters "(,$!" which are great for handling G0,
but I can't see where 
the code handles ")-" which specify the G1 code
page.  Does YAZ provide 
functionality to translate the G1 code page?  The specific
ESC sequence 
I am trying th handle comes from LC authority record
sh2006006361 in the 
2nd 670 field.  The code is:
ESC(3ESC)4zESC(BEarthWorksESC(3ESC)4yESC(B  
[where ESC is the Escape character].  Siconv seems to handle
the first 
sequence [ESC ( 3] fine; but since it doesn't recognize the
) in the 2nd 
Escape sequence, the marc8_esc_mode value gets set to 0x29.

Thanks for any help
Gary
_______________________________________________
Yazlist mailing list
Yazlistlists.indexdata.dk
http://lists.indexdata.dk/cgi-bin/mailman/listinfo/yaz
list

Gary,
While this doesn't address your question, I thought I would
mention that I think this record should be corrected by LC.
This portion of the 670 field is using a Basic Arabic escape
sequence in order to insert double left and double right 
quotation marks which are not present in the MARC-8 (Basic
Latin) character set.  I believe that the Basic ASCII
quotation
mark (i.e., hex'22) should be used instead.

I have reported this record to our Cataloging Policy and
Support Office for correction.
Larry



------------------------------------------------------------
------

1B 28 33 = Basic Arabic
1B 28 42 = Basic Latin

22  =  "  QUOTATION MARK (Basic Latin)
79  =  RIGHT DOUBLE QUOTATION MARK (Basic Arabic)
7A  =  LEFT DOUBLE QUOTATION MARK (Basic Arabic)

                      E  a  r  t  h  W  o  r  k  s          
          
1B 28 33 7A 1B 28 42 45 61 72 74 68 57 6F 72 6B 73 1B 28 33
79 1B 28 42


00670cz   2200169n  4500
001 sh2006006361
003 DLC
005 20061006110632.0
008 061006|| anannbabn          |a ana     c
010    $a sh2006006361
040    $a OC $b eng $c DLC
151    $a Little Miami River Valley (Ohio)
451    $a Little Miami Valley (Ohio)
550    $w g $a Valleys $z Ohio
670    $a Work cat.: The pioneer settlers of the Lower
Little Miami
Valley, 1897.
670    $a University of Cincinnati, Inventory of Sites
Rebuilt or
Otherwise Included in (3z(BEarthWorks(3y(B, Aug. 31, 2006 $b
(Little
Miami River Valley, Warren County and Greene County, Ohio)
781  0 $z Ohio $z Little Miami River Valley


On Mon, 18 Dec 2006, Gary Anderson wrote:

> I am using siconv.c version 1.29.  I have found the
following problem:
> 
> This code only seems to handle MARC8 Escape sequences
using the 
> characters "(,$!" which are great for
handling G0, but I can't see where 
> the code handles ")-" which specify the G1
code page.  Does YAZ provide 
> functionality to translate the G1 code page?  The
specific ESC sequence 
> I am trying th handle comes from LC authority record
sh2006006361 in the 
> 2nd 670 field.  The code is:
ESC(3ESC)4zESC(BEarthWorksESC(3ESC)4yESC(B  
> [where ESC is the Escape character].  Siconv seems to
handle the first 
> sequence [ESC ( 3] fine; but since it doesn't recognize
the ) in the 2nd 
> Escape sequence, the marc8_esc_mode value gets set to
0x29.
> 
> Thanks for any help
> Gary
> 


_______________________________________________
Yazlist mailing list
Yazlistlists.indexdata.dk
http://lists.indexdata.dk/cgi-bin/mailman/listinfo/yaz
list

Gary Anderson wrote:
> I am using siconv.c version 1.29.  I have found the
following problem:
> 
> This code only seems to handle MARC8 Escape sequences
using the 
> characters "(,$!" which are great for
handling G0, but I can't see where 
> the code handles ")-" which specify the G1
code page.  Does YAZ provide 
> functionality to translate the G1 code page?  The
specific ESC sequence 
> I am trying th handle comes from LC authority record
sh2006006361 in the 
> 2nd 670 field.  The code is:
ESC(3ESC)4zESC(BEarthWorksESC(3ESC)4yESC(B  
> [where ESC is the Escape character].  Siconv seems to
handle the first 
> sequence [ESC ( 3] fine; but since it doesn't recognize
the ) in the 2nd 
> Escape sequence, the marc8_esc_mode value gets set to
0x29.

Correct. The G1 sequences are not handled.

Can we get access to that Authority record via Z39.50 so we
have 
something to work from?

/ Adam

> Thanks for any help
> Gary
> 
> _______________________________________________
> Yazlist mailing list
> Yazlistlists.indexdata.dk
> http://lists.indexdata.dk/cgi-bin/mailman/listinfo/yaz
list


_______________________________________________
Yazlist mailing list
Yazlistlists.indexdata.dk
http://lists.indexdata.dk/cgi-bin/mailman/listinfo/yaz
list

I received the record in the LOC update file v21.i41.records
available 
from their FTP download site.  I have extracted that single
record from 
the file and attached it.  BTW - I just added the two G1
escape 
characters to your comparison string of G0 characters, and
that seemed 
to fix the problem.  The switch statement already had
ExtendedArabic 
defined.

Adam Dickmeiss wrote:

> Gary Anderson wrote:
>
>> I am using siconv.c version 1.29.  I have found the
following problem:
>>
>> This code only seems to handle MARC8 Escape
sequences using the 
>> characters "(,$!" which are great for
handling G0, but I can't see 
>> where the code handles ")-" which specify
the G1 code page.  Does YAZ 
>> provide functionality to translate the G1 code
page?  The specific 
>> ESC sequence I am trying th handle comes from LC
authority record 
>> sh2006006361 in the 2nd 670 field.  The code is: 
>> ESC(3ESC)4zESC(BEarthWorksESC(3ESC)4yESC(B  [where
ESC is the Escape 
>> character].  Siconv seems to handle the first
sequence [ESC ( 3] 
>> fine; but since it doesn't recognize the ) in the
2nd Escape 
>> sequence, the marc8_esc_mode value gets set to
0x29.
>
>
> Correct. The G1 sequences are not handled.
>
> Can we get access to that Authority record via Z39.50
so we have 
> something to work from?
>
> / Adam
>
>> Thanks for any help
>> Gary
>>
>> _______________________________________________
>> Yazlist mailing list
>> Yazlistlists.indexdata.dk
>> http://lists.indexdata.dk/cgi-bin/mailman/listinfo/yaz
list
>
>
>
> _______________________________________________
> Yazlist mailing list
> Yazlistlists.indexdata.dk
> http://lists.indexdata.dk/cgi-bin/mailman/listinfo/yaz
list
>

_______________________________________________
Yazlist mailing list
Yazlistlists.indexdata.dk
http://lists.indexdata.dk/cgi-bin/mailman/listinfo/yaz
list

Gary Anderson wrote:
> I received the record in the LOC update file
v21.i41.records available 
> from their FTP download site.  I have extracted that
single record from 
> the file and attached it.  BTW - I just added the two
G1 escape 
> characters to your comparison string of G0 characters,
and that seemed 

So
  strchr("(,$!)-", *inp)
did it?

> to fix the problem.  The switch statement already had
ExtendedArabic 
> defined.

Yes. I think all the cases - F final char - is implemented.

http://www.loc.gov/marc/specifications/speccharmarc
8.html#alt

I have always wonderded why there were both a G0 and G1 when
one seemed 
enough.

Thanks for the fix and sample record.

/ Adam


> Adam Dickmeiss wrote:
> 
>> Gary Anderson wrote:
>>
>>> I am using siconv.c version 1.29.  I have found
the following problem:
>>>
>>> This code only seems to handle MARC8 Escape
sequences using the 
>>> characters "(,$!" which are great for
handling G0, but I can't see 
>>> where the code handles ")-" which
specify the G1 code page.  Does YAZ 
>>> provide functionality to translate the G1 code
page?  The specific 
>>> ESC sequence I am trying th handle comes from
LC authority record 
>>> sh2006006361 in the 2nd 670 field.  The code
is: 
>>> ESC(3ESC)4zESC(BEarthWorksESC(3ESC)4yESC(B 
[where ESC is the Escape 
>>> character].  Siconv seems to handle the first
sequence [ESC ( 3] 
>>> fine; but since it doesn't recognize the ) in
the 2nd Escape 
>>> sequence, the marc8_esc_mode value gets set to
0x29.
>>
>>
>> Correct. The G1 sequences are not handled.
>>
>> Can we get access to that Authority record via
Z39.50 so we have 
>> something to work from?
>>
>> / Adam
>>
>>> Thanks for any help
>>> Gary
>>>
>>> _______________________________________________
>>> Yazlist mailing list
>>> Yazlistlists.indexdata.dk
>>> http://lists.indexdata.dk/cgi-bin/mailman/listinfo/yaz
list
>>
>>
>>
>> _______________________________________________
>> Yazlist mailing list
>> Yazlistlists.indexdata.dk
>> http://lists.indexdata.dk/cgi-bin/mailman/listinfo/yaz
list
>>
> 
> _______________________________________________
> Yazlist mailing list
> Yazlistlists.indexdata.dk
> http://lists.indexdata.dk/cgi-bin/mailman/listinfo/yaz
list


_______________________________________________
Yazlist mailing list
Yazlistlists.indexdata.dk
http://lists.indexdata.dk/cgi-bin/mailman/listinfo/yaz
list

That seemed to make it operate properly.  However, this
particular ESC 
sequence didn't really access a mapping in the G1 code page,
so I still 
don't know if that works.  Looking through the marc8.c file
indicates 
that it probably will work, but I haven't tried characters
in the 
Extended Arabic code page.

Adam Dickmeiss wrote:

> Gary Anderson wrote:
>
>> I received the record in the LOC update file
v21.i41.records 
>> available from their FTP download site.  I have
extracted that single 
>> record from the file and attached it.  BTW - I just
added the two G1 
>> escape characters to your comparison string of G0
characters, and 
>> that seemed 
>
>
> So
>  strchr("(,$!)-", *inp)
> did it?
>
>> to fix the problem.  The switch statement already
had ExtendedArabic 
>> defined.
>
>
> Yes. I think all the cases - F final char - is
implemented.
>
> http://www.loc.gov/marc/specifications/speccharmarc
8.html#alt
>
> I have always wonderded why there were both a G0 and G1
when one 
> seemed enough.
>
> Thanks for the fix and sample record.
>
> / Adam
>
>
>> Adam Dickmeiss wrote:
>>
>>> Gary Anderson wrote:
>>>
>>>> I am using siconv.c version 1.29.  I have
found the following problem:
>>>>
>>>> This code only seems to handle MARC8 Escape
sequences using the 
>>>> characters "(,$!" which are great
for handling G0, but I can't see 
>>>> where the code handles ")-" which
specify the G1 code page.  Does 
>>>> YAZ provide functionality to translate the
G1 code page?  The 
>>>> specific ESC sequence I am trying th handle
comes from LC authority 
>>>> record sh2006006361 in the 2nd 670 field. 
The code is: 
>>>> ESC(3ESC)4zESC(BEarthWorksESC(3ESC)4yESC(B 
[where ESC is the 
>>>> Escape character].  Siconv seems to handle
the first sequence [ESC 
>>>> ( 3] fine; but since it doesn't recognize
the ) in the 2nd Escape 
>>>> sequence, the marc8_esc_mode value gets set
to 0x29.
>>>
>>>
>>>
>>> Correct. The G1 sequences are not handled.
>>>
>>> Can we get access to that Authority record via
Z39.50 so we have 
>>> something to work from?
>>>
>>> / Adam
>>>
>>>> Thanks for any help
>>>> Gary
>>>>
>>>>
_______________________________________________
>>>> Yazlist mailing list
>>>> Yazlistlists.indexdata.dk
>>>> http://lists.indexdata.dk/cgi-bin/mailman/listinfo/yaz
list
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> Yazlist mailing list
>>> Yazlistlists.indexdata.dk
>>> http://lists.indexdata.dk/cgi-bin/mailman/listinfo/yaz
list
>>>
>>
>> _______________________________________________
>> Yazlist mailing list
>> Yazlistlists.indexdata.dk
>> http://lists.indexdata.dk/cgi-bin/mailman/listinfo/yaz
list
>
>
>
> _______________________________________________
> Yazlist mailing list
> Yazlistlists.indexdata.dk
> http://lists.indexdata.dk/cgi-bin/mailman/listinfo/yaz
list
>

_______________________________________________
Yazlist mailing list
Yazlistlists.indexdata.dk
http://lists.indexdata.dk/cgi-bin/mailman/listinfo/yaz
list