I have a directory with .txt files in it. I'm trying to parse out a value in the files and output to a .csv file. The value will always be a letter, V, and 7 digits, example V1234567. The letter will always be the same, the digits will always be different, but always 7 in number.

I've tried some of the examples I've found on the board, but, I am stumped.

When in doubt, format.

Can you paste the first few lines of the file so we can see what format it's in?

Here are the first three lines of a sample. This will be the general format for all the files. I have added the 1.), 2.), and 3.) in front of the lines:

1.)testrun|^~\&|ITDEPT|SAMPLE|ENGINE||||TRVB||2|68|
2.)TID||||||||||||||||||V1234567||
3.)CID||||||||||||||||||||||||||||||||||||||||||||08 13500|

I hope that helps. The V1234567 is the value that I'm trying to abstract.

When in doubt, format.

As an FYI, this is a custom report we run on workstations. TID is terminal id, I have no idea what CID is, some date thing that I don't really need to parse out.

When in doubt, format.

Will the string of interest always be in the line[s] starting with:

TID

?

If so, things get easier.

=====================================
If at first you don't succeed, you're about average.

M2

Also, in addition to M2's question, on the TID line is there always exactly one instance of the letter Z (i.e. the one that's followed by the number)? If so, things get easier still.

Yes, the TID line is the only one of interest, and there is only one instance of the letter V, followed by the numbers.

When in doubt, format.

get GNU awk for Windows from http://gnuwin32.sourceforge.net/pac...
then

c:\> gawk -F "|" "/TID/{print $(NF-2)}" file
V1234567

you can incorporate that command in batch if you wish

I'm downloading a copy of the Gawk for Windows now. I'll install it on a VM server I have and test it out. With the example you gave, c:\> gawk -F "|" "/TID/{print $(NF-2)}" file
V1234567, wouldn't that only abstract the V1234567? The next file may have the value V0987654.

I haven't had a chance to look at the software yet, but how would you handle wildcards?

When in doubt, format.

Well, a whole different thing to learn now. I'm going through the pdf's with the program. Could you explain the command you wrote? I can't find any mention of a printing command NF- etc.

When in doubt, format.

Your original post indicated that you needed to output to a csv file, which presumably means that you need to incorporate this parsed value with other data. Is that correct? If so, the gawk command may not do everything you need.

Here's a Perl command that accomplishes the same thing as the gawk command, but my assumption is that we need to expand the logic to handle the additional csv data.

perl -ne "print $1 if /TID.+(V\d{7})/" file.txt

I'm trying these ideas out on a 2003 Server VMware machine. What would I need to run Perl on that?

When in doubt, format.

You need to install Perl, which is free.

http://activestate.com/Products/act...

ok, I'll give it a shot. At this point, I will try anything.

When in doubt, format.

@ID123

NF means number of fields in gawk. -F means field delimiter.

gawk -F "|" means set the field delimiter to "|".

If you try this on the command line:
gawk -F "|" '{print NF}' file

it shows you the number of fields you have separated by "|". you can verify by counting.

/TID/ simply means to match lines where it contains TID.

$(NF-2) means to get the VALUE of the field that is 2 places before the final field. That would be your V1234567, and its assumed that VXXXXXXX is always at that field.

$NF means to get the last field VALUE.

NF means the last field number.

to extract that VXXXXXX value to a new file, just pipe it.

gawk -F .... > newfile

If the Vxxxxxxx is in the TID line and the intervening chars are |, this may do it.

::==
@echo off
setLocal EnableDelayedExpansion
for /f "tokens=* delims= " %%a in ('find "TID" ^< parse.txt') do (
set str=%%a
set str=!str:^|=!
set str=!str:TID=!
echo !str!
)

=====================================
If at first you don't succeed, you're about average.

M2

Report Offensive Follow Up For Removal

Why not use the letter V as the field delimiter? Then all you have to do is look at the first seven characters of the second field.

Mech,

I've run the bat against a sample file, but no dice. The file runs but no output, even though the sample file has the format and value mentioned above. Any ideas?

Thank you for the help.

When in doubt, format.

Forgot to mention, the bat runs against the file without any errors, just no output.

When in doubt, format.

Did you change parse.txt to whatever yours is?

Is the file SRTICTLY text?

try this:

type parse.txt

If it doesn't display 'clean' it may have some non-standard line breaks.

=====================================
If at first you don't succeed, you're about average.

M2