It allows string replacements
It handles repeated fields.
It allows you to limit to number of occurrences of a repeatable field you want to convert.
GTZ GmbH
Information und Dokumentation
Postfach 5180
6236 Eschborn 1
Germany
KNVRT131 comes with an English manual, in form of a ".DOC"
text. The source code, containing many comments in German, is
also available.
It allows string replacements
It handles repeated fields.
It allows you to limit to number of occurrences of a repeatable field you want to convert.
KNVRT131 requires that the user knows the maximum length of the records and the fields.
The syntax for the conversion table is not very transparant.
It is difficult to locate errors in the conversion table.
It only allows one character as occurrence delimiter.
It does not support path statements
The first line identifies the record, e.g.
ISO5000CLB0245means:
- ISO : it will be a conversion to the
ISO 2709 format
- 5000 : the maximum record length is
5000 characters
- CL : every line ends with a carriage
return and a line feed (the
alternative would be CR, which
means: only a carriage return)
- B : insert a blank at the end of
every line
- 0245 : some characteristics of the ISO
2709 format (it is best to take
always 0245)
The second line describes how a record can be recognized. The
safest way is to use the first field:
ETI:The E stands for "entry"; the TI: for the first tag, nl.
"TI:".After this you must describe every field. Non-repeatable fields are described like this:
FTI: 0010500The second line has to be read as:
001 0500This means:
0000100Repeatable fields are characterized like this:
FAU: 0030200R;0100150This may be read more easily as
003 0200 R ; 010 0200This means:
CHA '-' ' 'which means: change '-' into ' ', or
CHA '.' ''which means: change '.' into nothing (i.e. omit the full stop).
After you have finished the conversion tabel, save it under the name of the input file, followed by ".cmd", e.g.
downl.cmdThis is an example of a complete conversion table:
5000CLB0245
ETI:
FTI:
0010200
FTO:
0020200
FAU:
0030200R;0100200
CHA ',-' ' '
FSO:
0040200
CHA '; ' '^b'
CHA '(' '^c'
CHA ') ' '^d'
CHA ', ' '^e'
CHA '.' '^f'
FPY:
0050004
FLA:
0060020
FAB:
0071000
FFH:
0000200
FDE:
0080500R;0200500
CHA '-' ' '
FCC:
0000100
FDA:
0000100
FAN:
0000100
knvrt131which produces this screen:
Conversion of bibliographic data Version 1.3
============================================
Copyright GTZ GmbH, Information und Dokumentation
(1990) Postfach 5180, D-6236 Eschborn 1
Input File:
Now you must fill in the name of the download, e.g.
Input File: downl.docNow, the program will ask if you want to confirm errors by hitting <return< or not. You must answer with 'j' or 'n', which are abbreviations for the German 'ja' (yes) and 'nein' (no).
Errors will be stored in an error file. This means that the program will write that part of a record in this file up to where it was not able continue the conversion of it. For this example the error file will be called "downl.err". If an error file with this name already exists, the program will ask you if it is OK to overwrite it. And if a message file, called "downl.msg" already exists, the program will also ask if it can by overwritten or not. It is not clear what purpose the message file has. I could not provoke the program to write anything in it:
Conversion of bibliographic data Version 1.3
============================================
Copyright GTZ GmbH, Information und Dokumentation
(1990) Postfach 5180, D-6236 Eschborn 1
Input File: downl.doc
Confirm errors by hitting return (j/n) j
Error file downl.err already exists. Overwrite? (J/N) j
Message file downl.msg already exists. Overwrite (J/N) j
Now the program will ask for the record and the field
separator. The best thing to fill in is the # sign CDS/ISIS
normally uses. After this, the program displays some of the
parameters you have defined in the specification file. Then it
warns you that it may crash due to an error in the
specification file... The whole screen looks like this:
Conversion of bibliographic data Version 1.3
============================================
Copyright GTZ GmbH, Information und Dokumentation
(1990) Postfach 5180, D-6236 Eschborn 1
Input File: downl.doc
Confirm errors by hitting return (j/n) j
Error file downl.err already exists. Overwrite? (J/N) j
Message file downl.msg already exists. Overwrite (J/N) j
output file type: ISO
_____________________________________________________________
| Conversion into ISO2709-Format |
_____________________________________________________________
Field separator: #
Record separator: #
maximum record length: 5000
line end in input file: CR LF
line end will be replaced by Blank
ISO Spezifikationen: 0245
first line of .CMD file read
String identifying start of entry ->TI:<-
if the programm crashes at this point, you probably
will have to change the line end specification in the
.CMD file
Continue (j/n)The second screen briefly displays a part of the first record. If the conversion is succesfull you will see this message:
All entries processed.
Conversion of another file (j/n) ?
Now you can start a new conversion or you can stop.
AU: P. Jones and L. SmithYou cannot fill in the field description this way:
FAU: 0030200Rand0100200Normally the program gives clear error messages:
downl.cmd: Field specification incomplete -> 002
Hit Return to quit
or
downl.cmd: First Line in error. Specification incomplete -> 5000CLB0245
Hit Return to quit
But in some cases, it may be very difficult to detect a
specific error in the conversion table. In the worst case the
conversion will stop; you will then try to correct the
conversion table. If you do not see the error, you will be
doing this over and over again. The only solution will
probably be to test one field specification after another,
which can take a lot of time. The conversion will stop each
time because the last field is too long (as, in fact, the
program assumes that it is the rest of the record).
The program will accept certain errors in the conversion table, which leads to an incorrect conversion. If we want to convert:
SO: Library news; 51 (3-4) July-Dec 75, 134-140. 5 illus. 3 refsinto a field number 4, containing this subfields:
Library news^b51^c3-4^dJuly-Dec 75^e134-140^f 5 illus^f 3 refs(Only the first subfield and subfields ^b, ^c, ^d and ^e are of any importance; the subfields ^f will be rejected while importing the ISO 2709 file into the database by means of a FST.)
To accomplish this we can use this field description:
FSO:
0040200
CHA '; ' '^b'
CHA '(' '^c'
CHA ') ' '^d'
CHA ', ' '^e'
CHA '.' '^f'
If we forget the closing brackets, e.g. in the first CHA-line:
CHA ';' '^bthe program will accept the specifications without any warning, but the conversion will not be correct. We would get this:
Library news51etc.
Make sure the program and the input file are in the same (sub)directory. KNVRT131 does not support path statements. If you try to convert a file in another (sub)directory you will get an error message:
Conversion of bibliographic data Version 1.3
============================================
Copyright GTZ GmbH, Information und Dokumentation
(1990) Postfach 5180, D-6236 Eschborn 1
Input File: a:downl.doc
Error inputting file name:
File names may be up to 8 characters long.
They may not contain special characters.
An Extension may be up to 3 characters long.
.TXT is assumed as default
Input File:
Now, if you try to get out of this situation by hitting
<return> or <Esc>, you will get more and more of these error
messages:
File names may be up to 8 characters long.
They may not contain special characters.
An Extension may be up to 3 characters long.
.TXT is assumed as default
Input File:
Error inputting file name:
File names may be up to 8 characters long.
They may not contain special characters.
An Extension may be up to 3 characters long.
.TXT is assumed as default
Input File:
Error inputting file name:
File names may be up to 8 characters long.
They may not contain special characters.
An Extension may be up to 3 characters long.
.TXT is assumed as default
The only way to get out of this situation is to use
<Ctrl><Break> or another drastic key combination.
KNVRT131 may be better if you have to change a lot of strings in the input file, although I could not detect the limits of the program in this. As goes for most conversion programs, KNVRT131 allows you to specify the tags of your target database. It may be better to use tags which reflect the order of the input file. You can change this into the appropriate tag number by means of a FST which you will use to import the ISO2709 file in your test database.