I need to be able to read Labview 7.1 history files (*.hst)

[vgathmann]

We have an obsolete data aquisition system that has archives of all monitored analyzers (CEM system). I have 3GB of these *.hst files that have a clear text header and then are followed by binary. I have the task of converting these files to *.csv files. The only thing I need to do this is to have the file format for these *.hst files, or a LabView utility (or *.vi program) to convert them. I can write my own conversion utility rather quickly if I have the file format. Does anybody know what this format is? A sample of the clear-text portion of the files is shown below:

CREATION DATE=20070702
TAG NAME=CO
TYPE=AI
TIME-HISTORY=1.000000
MX ERR-HISTORY=0.100000
ENGR UNITS=ppm
DESCRIPTION=
ORDER=2
End File Information

Any help would be greatly appreciated.

Vincent Gathmann
MIS/IT Director
Pictorial Offset Corporation
111 Amor Avenue
Carlstadt, NJ 07072
(201) 935-7101 x429
vgathmann@pictorialoffset.com

[tst]

I'm assuming that this is a proprietary format, so we can't really help you.

I suggest one of two things:

1. Find the original documetation which describes the format.
2. Get the original code for creating or reading the format.

Either of these will allow you to write a converter.

If you don't have them, you could try the following trick - if the executable was written in a version of LV new enough to enable VI server connections, you can try opening a connection to it and calling the reader in your own VI to load the data. Once you have the data, you can resave it.

If this doesn't help, can you upload a sample file and tell us which version the original program was built in? 

[vgathmann]

Thank you for your reply. The version of Labview that wrote the .HST files is Labview 7.1. We have a license for the application, but somebody missplaced the software disks (and license key) and we cannot locate it. The PC that had the application on it was repurposed. We have the original VI files that ran on the computer -- but missing the software to reinstall -- the VI files do no good. I will attach a sample file if you think this will help determine the format. Thanks again for your time.

Vincent Gathmann
MIS/IT Director
Pictorial Offset Corporation
111 Amor Avenue
Carlstadt, NJ 07072
(201) 935-7101 x429
vgathmann@pictorialoffset.com

[Dennis_Knutson]

If you have the original VIs, post those. Then someone can look an see exactly how the files were created.

[GerdW]

Hi Vincent,

from your supplied file I would guess the remaining data consists of DBL values.
The DBL values start directly after the 'file information end' (plus 0dh=CR). Also there seems to be 4 DBL values per data entry, as there is a pattern every 20h bytes... Hope this helps!
To check this: read in a chunk of data and plot it in a graph (after splitting into 4 data columns)...

Message Edited by GerdW on 07-10-2007 04:05 PM

[vgathmann]

Dennis,

I'm reluctant to post the VI files for legal reasons. I don't want to infringe on the rights of the author of the code. I've been trying to contact him for help, but I haven't located him. I don't want him to find me by way of his attorney. There is a 'toolkit' with a code in a folder called 'History Utilities". I just don't know how to read the code to determine what happens to write the files out in the format. I'd be willing to send you the toolkit portion if it would help, but I don't want to post to the board. Thanks again for your help.

Vincent Gathmann
MIS/IT Director
Pictorial Offset Corporation
111 Amor Avenue
Carlstadt, NJ 07072
(201) 935-7101 x429
vgathmann@pictorialoffset.com

[vgathmann]

GerdW,

Thanks for your response. I looked at the file with a hex editor and concluded that the records are in 32 byte lengths, as you stated. I will take your infornation and try to chart them out, as you suggested. I cannot imagine where i get header information for this data, however. I'm only going to be able to guess what the values mean. We are going to look for some historical printouts to see if the headers are evident. Again, thanks for you help.

Vincent Gathmann
MIS/IT Director
Pictorial Offset Corporation
111 Amor Avenue
Carlstadt, NJ 07072
(201) 935-7101 x429
vgathmann@pictorialoffset.com

[tst]

It seems to me that you do not have any experience with LV yourself. I assume the history utilities folder includes a VI for reading the data from the file.

Basically, all you need is that VI, because if it will load the data into a format you can use, you can then easily convert the data to a 2D string array and save it as a CSV file. I would personally doubt the original author would object if you upload it, but it really isn't my place to tell. If you have an available version of LV installed anywhere, I suggest you find the VI which reads the file, open it and try to understand how it works. If you want to see examples of working with files, have a look at the example finder (Help>>Find Examples).

[vgathmann]

You are correct. I have no experience in Labview. I am an experienced utility programmer in C, so I understand file I/O -- thus my desire to get the file format.

I will compress the 'history' portion of the toolkit and upload it. Perhaps that will give you enough information to give me an idea of how those programs write out the data. I don't have enough time to learn LV to figure out how to comprehend how the toolkit produces it's output. Thank you for your help.

[GerdW]

Hi,

the attached vis look quite good (nice programming, good comments, etc.) but a lot of other vi are missing.
There have to be folders named RTDB, Utility, Type Defs with lots of subVIs.

The vi writing the data seems to be 'Write Coeff to History File.vi'. And yes, the data is DBL...
From 'Get History Buffer.vi' it seems the data contains one column of time stamps (LabView7.1 internal format, DBL value, seconds since 12am, 1. Jan. 1904 UT) and one column of data (also DBL).