Anyone good with text files?

[IceShock]

Howdy, im in need of some easy way to extract things from text files for example, i have 8000 lines which consist of things like:
{ 1, 47855, "", "-", "=ds=#s10#, #a1#"};
{ 2, 47857, "", "=q4", "=ds=#s3#, #a2#"};

etc, the part i need is the 5 didget number from each row.
Is there a way to extract this part from each row?

thanks

 

[HecFam]

First thought is that you could probably open it as a CSV in excel. Then the B column will contain the 5 digit numbers as both sides of it is separated by a comma. Whether or not that's how you would like it formatted, but it should isolate the just the numbers.

 

[IceShock]

ah i dont have excel, but that's a good idea

 

[adolf hamster]

First thought is that you could probably open it as a CSV in excel. Then the B column will contain the 5 digit numbers as both sides of it is separated by a comma. Whether or not that's how you would like it formatted, but it should isolate the just the numbers.

you can also manually select the "width" of each column doing this, so you could extract your data that way assuming the 5 digit numbers all appear in the same place on each row.

 

[ErNciLator]

Paste to excel. Text to columns button. Separate by comma. Copy column back out to notepad or wherever.

 

[arty]

Any decent text editor + regex? Something like /^.+, (\d+?),.+$/\1/g should do, I think (un-tested).

 

[GSXRMovistar]

Can't check at the mo but would suggest Excel, paste data into column A then in column B, "=MID(A1,6,5)", fill down for all rows.

 

[IceShock]

Just downloading excel now, will try these suggestions and report back, ty

 

[russell664]

ah i dont have excel, but that's a good idea

You could probably bang it into Google sheets to see what happens.

 

[h4rm0ny]

Howdy, im in need of some easy way to extract things from text files for example, i have 8000 lines which consist of things like:
{ 1, 47855, "", "-", "=ds=#s10#, #a1#"};
{ 2, 47857, "", "=q4", "=ds=#s3#, #a2#"};

etc, the part i need is the 5 didget number from each row.
Is there a way to extract this part from each row?

thanks

Here is an easy way to get what you want:

Install Notepad++. (this will take only a moment).

Open the text file in it.

Hold down the LEFT ALT-KEY and drag the mouse cursor in a column over only the five digit numbers. When done, hit Copy (Ctrl-C) and then paste the copied numbers into a new text file in a new tab in Notepad++.

Note, I expect you will need to do this four times as your initial number presumably increments beyond single digits and thus the position of the five digit group will shift one to the right when you transition from units, to tens, to hundreds to thousands.

If that's no good, it's a fairly easy regular expression.

 

[Armageus]

ah i dont have excel, but that's a good idea

LibreOffice is free and Libreoffice Calc will do 95% of what Excel can do - certainly fine for your use case

 

[h4rm0ny]

Can't check at the mo but would suggest Excel, paste data into column A then in column B, "=MID(A1,6,5)", fill down for all rows.

Just downloading excel now, will try these suggestions and report back, ty

This is unlikely to work as written because the moment your initial number grows beyond 9, the formula is no longer correct. Seriously, try Notepad++ and just highlight the numbers you want.

EDIT:

 

[Clark Nova]

If your file always positions the 5 digit number in the same place, use awk to grab it and then sed to strip the comma:

[simon@clarknova ~]$ cat testfile | awk {'print $3'} | sed 's/,//g'
47855
47857

 

[IceShock]

tyvm guys, excels working brilliantly, i dont know why i didnt download this years ago haha. ty

 

[h4rm0ny]

If your file always positions the 5 digit number in the same place, use awk to grab it and then sed to strip the comma:

[simon@clarknova ~]$ cat testfile | awk {'print $3'} | sed 's/,//g'
47855
47857

Ah, sed and awk. GNU's answer to the question: "how do I combine different tools on an OS too old to have object pipelining and inexcusably uses unnamed parameters." ;P

 

[dl8860]

Agree with h4rm0ny's approach, notepad++ is great, useful for tonnes of other things too

 

[Pho]

As mentioned Notepad++ and Regex will do it. It has a JSON plug you could probably use instead of Regex to extract what you need, but anyway:

Search > Find

Find what:{ ?(\d+), ?(\d+), ?\"(.*?)\", ?\"(.*?)\". ?\"(.*?)\"\};
Replace with: \1 for first column (1, 2)
Replace with: \2 for second column (47855, 47857)
Replace with: \3 for third column
Replace with: \4 for fourth column
Replace with: \4 for fifth column

Make sure you have search mode set to regular expression and click replace all.

You'll want to use \2 to extract the ID.

That's a crude Regex but it works

 

[Clark Nova]

Ah, sed and awk. GNU's answer to the question: "how do I combine different tools on an OS too old to have object pipelining and inexcusably uses unnamed parameters." ;P

It's all there in the man pages!

 

[h4rm0ny]

As mentioned Notepad++ and Regex will do it. It has a JSON plug you could probably use instead of Regex to extract what you need, but anyway:

Search > Find

Find what:{ ?(\d+), ?(\d+), ?\"(.*?)\", ?\"(.*?)\". ?\"(.*?)\"\};
Replace with: \1 for first column (1, 2)
Replace with: \2 for second column (47855, 47857)
Replace with: \3 for third column
Replace with: \4 for fourth column
Replace with: \4 for fifth column

Make sure you have search mode set to regular expression and click replace all.

You'll want to use \2 to extract the ID.

That's a crude Regex but it works

Nice that you gave a regex for every column and explained how to pick out a column of choice. I'm concerned by the OP's comment that they need to do this for "files". I hope the OP isn't planning to regularly do this. If so, I think something has gone wrong as the sample file is clearly meant to be read programmatically so the OP is working with an incomplete solution, I suspect.

 

[h4rm0ny]

It's all there in the man pages!

So named because everybody refuses to read the instructions.