Does anyone know how I can extract the first 8 and the last 10 lines of a text file and output them into another text file, or two seperate text files, either way will do.

The first 8 is easy. The last 10 requires you to store the 10 lines until you write them out.

In DOS?

M2

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

Hi Matthew
This I think would be difficult to write in Standalone DOS. WinXP/2000 extended commands would help.
I could write a C.Exe program for you to if nobody comes up with a batch file solution.

C would have the fastest execution time (by fractions of a second) but Perl would be the shortest and easiest to read/write.

#!perl -w
@lines = <>;
print $lines[$_] for (0..7);
print splice(@lines, -10);

The line that uses splice could also be written as
print $lines[$_] for (-10..-1);

For another option, you can install cygwin and use the head and tail utilities.

e.g.

head -n 8 file.txt
tail -n 10 file.txt

In short, my suggestion is to use the power of the unix utilities which have been ported to Windows.

Hi FishMonger,

Neat stuff.

I don't know any of these languages, but the perl must ba as short and sweet as it gets.

I have a few unix-like utils which are indispensable.

As ever, it's challending to do stuff like this in DOS.

A DOS script for this would probably look like a family reuniun size bowl of spaghetti.

M2

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

i'm running win2k server but the command needs to be in a batch file...

can anyone give me some code to try out?
i tried using the code posted on another question (http://computing.net/dos/wwwboard/forum/13852.html) which takes the last n lines. it worked fine.

I need something like this but to take the first and last n lines. anyone please?

I'll work on it.

M2

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

Hi IVO and everybody.

The bat at the link in #9 looks pretty slick. I'm still trying to figure out the syntax.

M2

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

Hi
Thats Win2k syntax. If thats what you want and not stanalone dos. Try this.

@echo off
rem put something in the quotes of the find
rem that would not be found in your file.

type YourFile.txt | find /v /n "~@~" >~tmp.txt
for /f "tokens=1,* delims=[]" %%a in ('type ~tmp.txt') do set NoLines=%%a
set Count=0
set /a NoLines=%NoLines%-11
if exist ~NewFile.txt del ~NewFile.txt
for /f "tokens=*" %%a in (YourFile.txt) do call :GetLines %%a
del ~tmp.txt
exit /b
:GetLines %*
if %Count% LSS 8 echo %* >> ~NewFile.txt
if %Count% GTR %NoLines% echo %* >> ~NewFile.txt
set /a Count+=1
exit /b

Mechanix2Go,

It's clear (from this thread & others) that
Perl is my favorite language. For this task
the Perl script can be rewritten and
executed as a command line instead of a
seperate script. If required, the batch
file could be used as a wrapper to the Perl
command/script. There are only 2 reasons I
can see for Mathtew not to use Perl or head
and tail. 1) This is a homework assignment.
2) Matthew doesn't have admin rights to
install the proper tools.

Have fun everyone!

%00

Hi dtech10,

That's DOS?

Back on the subject of the FIRST several lines, try this:

::** outputs first x lines
@echo off > quit.bat

if %2'==' echo needs two parameters: filename number_of_lines & goto :eof
set /a #=%2
for /f "tokens=*" %%L in (%~f1) do call :output %%L
goto :eof

:output

if %#% LEQ 0 goto :eof
set /a #=%#%-1
echo %*
::**

M2

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

Hi Mechanix
Thanks for the code, We all learn from the way other programmers code, I myself have learned a lot from other peoples code on this site.

Fishmonger maybe right about admin rights.
Perl would be much neater solution. if required nobody in their right would do it with a MsDos Batch file if they did'nt need to.
I and others do it because it a challenge to do it the hard way.
No disrespect intended to anyone.

Hi FishMonger,

Can you show the 'wrapped' perl?

TIA

M2

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

The Perl script would be called from within
the batch file just like any other command,
e.g.,

@echo off
cls

call perlscript.pl file.txt > out.txt
"out of commission" for awhile, I'm leaving
for the hospital for heart surgury so, I'll
check back-in in a coulple weeks.

%00

Hi dTech10... your code seems to work fine. but it gives different results with different lengths of text files....and on the large text file i have it does not return the end lines, just the first few.

M2Go... havent managed to work your code. though that would just give me the first lines? not the end ones too? could you give me an example of the code taking 5 lines from top, 8 lines from bottom of example.txt and putting into output.txt

please :)

Hi Matthew,

In #9 you said the code worked for 'last lines' so is that not useable?

Or is it simply a matter of integrating the two scripts together to get fist# and last# in one go?

I have not studied dtech10's code, but while testing mine, I happened to use it on a file which had email addresses like this:

<joe@domain.org>

[all with leading and trailing 'angle brackets' or whatever you call them]

and the script put out NO lines.

Whithout extensive testing, my instinct is that the < & > trip up the BAT because they are 'redirect' symbols.

With luck, this may only be an issue with those symbols at the beginning or end of a line.

But if your files contain these chars, I would do some testing.

***
I just trie dtech10's and with a file of twenty lines, it got ALL BUT lines 9 and ten. ??

M2

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

Hi
The code as written assume's that the your txt file is greater than the 18 lines required. This could be modified.
Why it's not working on longer files, I'm not sure, Rem out the line "del ~tmp.txt"
and look at it, you will see that it's line-numbered. Is the lines in your file numbered correctly.
How long is the txt.files you are trying to read, maybe the find /n as a limit for it's line numbers.

Hi dtech10,

I REMd out the del and the ~tmp has all 20 lines.

The original yourfile.txt has 20 lines. Short ones; one word each. Size: 152 bytes.

M2

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

M2Go... Yes is it possible to join the two scripts into one to get the first AND last n lines of a text file?

can you give me an example of the script using the values i gave u in my previous post please?

Hi Matthew,

Well, we're getting close.

Try this:

::**
@echo off

if %4'==' echo 4 params #top #bottom infile outfile & goto :eof
If not exist %~f3 Echo File %~f3 not found & Goto :EOF

type nul > %4
set outfile=%4

:main

call :first# %1 %3
set outfile=
set #=
call :last# %2 %3 %4
goto :eof

:first#

set /a #=%1
for /f "tokens=*" %%L in (%2) do call :output %%L
goto :eof

:output

if %#% LEQ 0 goto :eof
set /a #=%#%-1
echo %* >> %outfile%
goto :eof

:last#

Echo Extracting last %1 lines from %~f2 into %~f3

For /F "tokens=3 delims=:" %%A in ('Find /V /C "#~#" %~f2') Do Set RN=%%A
Set /A SN=%RN%-%1
If %SN% leq 0 Echo Too many lines [File contains %RN% lines] & GoTo :EOF
Echo Skipping %SN% lines from %RN%
For /F "tokens=* skip=%SN% delims=" %%A in (%~f2) Do Echo %%A>> %~f3
Set RN=
Set SN=

goto :eof

::**

M2

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

As a person who can't keep track of the variables in a batch, I tried this in BASIC.
I made a possible solution. Worked from 6 to 210,000 lines (5 seconds). If you are interested, email me. Present version runs in DOS prompt, 8.3 file names. EXE Compiled PowerBasic.

Hi wizard-fred,

Yeah, join the club on losing track of cariables.

I'm studying a QBasic tutorial now.

;)

M2

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

Hi Mechanix
Neat bit of programming, very impressive.
Never throught of using the skip.

Hi dtech10,

That part was contributed by IVO.

M2

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

Hi M2Go... thanks for the code but i'm having a bit of trouble interpreting it.... which are the variables i need to change?

where do i put the inputfile.txt?
where do i put the outputfile.txt?
where do i put the number of lines from top?
where do i put the number of lines from bottom?

excuse my ignorance :)

Hi Matthew,

No prob.

If you type the name of the BAT with fewer than 4 parameters, you'll get a syuntax hint.

The first param is the number of lines off the top.

The second is the number of lines off the bottom.

Third is the input file.

Fourth is the output file.

So for your example in #18:

BATname 5 8 example.txt output.txt

M2

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

Great! that worked very well. I have one more request though... is there any way of not ommitting tab indentation when extracting the lines of text?

basically the lines i'm extracting are in a sort of columnar format, when i extract them the headings of the columns are not where they are meant to be.

help?!

Hi Matthew,

Not at all obvious how to keep from losing tabs.

Maybe IVO knows.

I think the 'token' facility sees each string [some chars together with no 'white sapce'] separated by 'white space' which can be a 'space' or a 'tab'.

Hence the tabs get lost in the shuffle.

***
Are the disappearing tabs which cause your columns to get messed up, ONLY ion the first line?

M2

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

Nope they're in the end lines actually. For example:

----------------
Total Copied Skipped
Dirs : 86 0
Files : 402 26
Bytes : 104.1 m 22.1 m 82.0

Ended : Tue Apr 19 20:04:44 2005

Hi Matthew,

I took this block:

Total Copied Skipped
Dirs : 86 0
Files : 402 26
Bytes : 104.1 m 22.1 m 82.0
***
and replaced the spaces with tabs. Now it looks like this:

Total Copied Skipped
Dirs : 86 0
Files : 402 26
Bytes : 104.1 m 22.1 m 82.0
***

I tsurely looks different, but I'm not so sure it looks better.

Notice that the line beginning with 'Bytes' would 'look better' if the was no space [subsequently changed to tab] between the number of bytes and the 'm'.

***
I'm willing to work more on this but we need to be very clear about what we're dealing with.

Are you ALWAYS taking the first 5 and last 8 lines?

Can you zip ut the output.txt and email it? BUT name it output.pzi because I have zips blocked.

mailto:

15622

at

golden-triangle.com

M2

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

Well,

nice try.

It dowsn't look any different posted here. But in a file viewer it dos.

M2

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

Sorry about that mate.... seems that your batch file was working correctly and keeping tabs. it was the mailing program i was using that removed the tabs!!

Thanks for all your help!
Much appreciated.

Hi Matthew,

Yeah, team!

M2

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

Hi folks,

Just for drill, here's an EXE version:

http://www.Golden-Triangle.com/FrstLast.zip

M2

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