Hello. I'm trying to identify the organization of some MicroFocus data files.
Those files are associated with the following information:
- indexed organization
- sequential access
- VSAM dataset organization
- codeset ASCII
- record format KS
- it has duplicate alternate keys
There is a single file so no separate .idx file. I assume it contains information of both data and indexes. Here is the result of hexdump -n 100:
If possible send me the file.
Redacted if you can.
A subset if you can.
Zipped if you can.
Do you have Micro Focus Enterprise Developer or MFE or any other MF COBOL incarnation.
I can run your file thru some of my MF tools.
The hex dump of the file is not sufficient for me.
GnuCOBOL, via C-ISAM, only supports MF IDXFORMAT"1"
If it would help I can use IDCAMS to REPRO the ISAM dataset to a sequential file.
😄
1
Last edit: Ralph Linkletter 2024-08-02
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Do you have a copy of the program that created this file ?
Do you have the record layout ?
If you do the simplest and primary way is to create a sequential file using the ISAM file the ina very small PD do :
Open Input ISAM-FILE Output Seq-file.
perform until exit
read ISAM-FILE at end exit perform
write Seq-Record from ISAM-Record
exit perform cycle
end-perform
close ISAM-File SEQ-File.
goback.
Then copy the seq file to GC and read it in what ever way you need.
The normal method is above for migrating ANY ISAM type file to another platform.
Last edit: Vincent (Bryan) Coen 2024-08-06
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Boris does not have any Micro Focus (MF) products.
The file he is using is an ISAM / VSAM file from the MF bank demo suite.
H e downloaded that file and is trying to read it with GnuCOBOL.
He did not download the MF components,
It is a MF IDXFORMAT"8" ISAM / VSAM file.
Your suggestion would work if he had a license to the MF product.
He does not. Reading an MF IDXFORMAT"8" file will not work stand alone with GnuCOBOL.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
According to Ralph's analysis, it looks like it was IDXFORMAT"8". But it doesn't seem to match the information I have about structures. I might be missing something.
You could write your own tool to dump the structure (keys etc.), but that is not a trivial task.
Aren't file structures known already? So I guess it should not be too difficult? I may be wrong.
Thank you for your answers anyway!
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
It is IDXFORMAT"8"
What is it that you are attempting to do ?
What documentation are you looking at that defines the internal composition of a MF IDXFORMAT"8" ISAM / VSAM file ?
If I recall properly, records can contain data and index references.
There is no separate area of the data file dedicated only to the indices.
You could decode the first 128 bytes of the record to determine some of the file characteristics. Each recent MF VSAM file has a 128 byte header. Beyond the 128 header are data and index pointers.
Albeit you can probably discern "the file structure" to read the records. I doubt that you could support the integrity of the file to implement write, rewrite, delete support.
Somewhere within the MF paradigm is a process akin to an IBM control interval splitting.
The record length in the example is 200 bytes - but the MF I-O system presents the 200 bytes to the application program - the actual physical file is a collection of VSAM leaf nodes and a double linked list.
The attachment is a zone over decimal hex dump of parts of the file. Notice the mix of dtaa records and index entries.
Hello. I'm trying to identify the organization of some MicroFocus data files.
Those files are associated with the following information:
- indexed organization
- sequential access
- VSAM dataset organization
- codeset ASCII
- record format KS
- it has duplicate alternate keys
There is a single file so no separate .idx file. I assume it contains information of both data and indexes. Here is the result of
hexdump -n 100:I assume it is an IDXFORMAT unsupported by GnuCOBOL. I tried to read it with GC3, GC4, BDB, V(B)ISAM, MF file format and I also tried to manually and naively separate the file into a .dat and an .idx file but nothing works.
According to this:
https://www.microfocus.com/documentation/visual-cobol/vc60/DevHub/HRFLRHFILE0B.html
https://www.microfocus.com/documentation/server-express/sx20books/fhfile.htm
It should be IDXFORMAT(8) since there's no separate index file but the structure of the header doesn't seem to match.
Does anyone has an idea of what kind of file organization it is? Thank you!
Well I think if there is only one .idx file, then that, to me, means there are no secondary indexes. Have you tried the file command ?
It just says "data". I believe the file contains both data and indexes (as for BDB files except that it's not a BDB file).
If possible send me the file.
Redacted if you can.
A subset if you can.
Zipped if you can.
Do you have Micro Focus Enterprise Developer or MFE or any other MF COBOL incarnation.
I can run your file thru some of my MF tools.
The hex dump of the file is not sufficient for me.
GnuCOBOL, via C-ISAM, only supports MF IDXFORMAT"1"
If it would help I can use IDCAMS to REPRO the ISAM dataset to a sequential file.
Last edit: Ralph Linkletter 2024-08-02
Do you have a copy of the program that created this file ?
Do you have the record layout ?
If you do the simplest and primary way is to create a sequential file using the ISAM file the ina very small PD do :
Open Input ISAM-FILE Output Seq-file.
perform until exit
read ISAM-FILE at end exit perform
write Seq-Record from ISAM-Record
exit perform cycle
end-perform
close ISAM-File SEQ-File.
goback.
Then copy the seq file to GC and read it in what ever way you need.
The normal method is above for migrating ANY ISAM type file to another platform.
Last edit: Vincent (Bryan) Coen 2024-08-06
Boris does not have any Micro Focus (MF) products.
The file he is using is an ISAM / VSAM file from the MF bank demo suite.
H e downloaded that file and is trying to read it with GnuCOBOL.
He did not download the MF components,
It is a MF IDXFORMAT"8" ISAM / VSAM file.
Your suggestion would work if he had a license to the MF product.
He does not. Reading an MF IDXFORMAT"8" file will not work stand alone with GnuCOBOL.
Did you check the file header to make sure that it is in fact a MF indexed file.
Index File - File Header
- The first 4 bits are always set to 3 (0011 in binary) ...
- Don't forget about endianness when dealing with binary data
I am not aware of any tools available which can read or dump the structure for MF IDXFORMAT"8".
You could write your own tool to dump the structure (keys etc.), but that is not a trivial task.
Consequently, the best bet is to write a COBOL program, compiled with a MF compiler, and output the data in sequential format.
According to Ralph's analysis, it looks like it was IDXFORMAT"8". But it doesn't seem to match the information I have about structures. I might be missing something.
Aren't file structures known already? So I guess it should not be too difficult? I may be wrong.
Thank you for your answers anyway!
It is IDXFORMAT"8"
What is it that you are attempting to do ?
What documentation are you looking at that defines the internal composition of a MF IDXFORMAT"8" ISAM / VSAM file ?
If I recall properly, records can contain data and index references.
There is no separate area of the data file dedicated only to the indices.
You could decode the first 128 bytes of the record to determine some of the file characteristics. Each recent MF VSAM file has a 128 byte header. Beyond the 128 header are data and index pointers.
Albeit you can probably discern "the file structure" to read the records. I doubt that you could support the integrity of the file to implement write, rewrite, delete support.
Somewhere within the MF paradigm is a process akin to an IBM control interval splitting.
The record length in the example is 200 bytes - but the MF I-O system presents the 200 bytes to the application program - the actual physical file is a collection of VSAM leaf nodes and a double linked list.
The attachment is a zone over decimal hex dump of parts of the file. Notice the mix of dtaa records and index entries.
Last edit: Ralph Linkletter 2024-08-05
I'm mostly learning about indexed files and how they work. The references I have are the following ones:
- https://www.microfocus.com/documentation/visual-cobol/vc60/DevHub/-HRFLRHFILE0B.html
- https://www.microfocus.com/documentation/server-express/sx20books/fhfile.htm
But if it looks like if I truly want to read such files, the best way would be to use MF tools to convert to C-ISAM as suggested in other answers.
Boris,
do you have access to MF Visual COBOL ?
if so the REBUILD utility might be able to convert the files to C-ISAM after which you may be able to access it with GNUCOBOL using VISAM.
I currently don't but I'll think about it. Thank you!
Removed defect detected
Last edit: Ralph Linkletter 2024-08-10
On 09/08/2024 21:37, Ralph Linkletter wrote:
See the MF or Fujitsu PG for calling syntax.
Both COBOL compilers implemented CBL_SPLIT_ FILE and CBL_JOIN_FILE as ease of use subroutines.
Granted Unstring could be used - but why ? when there is a specific service available to do so.
The CBL_SPLIT parameter list:
Chuck implemented CBL_SPLIT_FILE for GnuCOBOL.
Probably hiding somewhere in "contributions"
Thanks for the feedback
Last edit: Ralph Linkletter 2024-08-09