Seems to work:
@ECHO OFF
REM Show usage and quit if no file name was given
IF [%1]==[] GOTO USAGE
REM Show entire file if no range was given
IF [%2]==[] TYPE %1 & GOTO :EOF
SETLOCAL ENABLEEXTENSIONS ENABLEDELAYEDEXPANSION
SET FILENAME=%1
SET LASTLINE=0
REM Build the array of lines to display
SHIFT
:RANGEREADLOOP
CALL :BUILDARRAY %1
SHIFT
IF NOT [%1]==[] GOTO RANGEREADLOOP
REM Loop through the file and keep track of the lines to display
SET CURRENTLINE=0
SET INDEX=1
FOR /F "delims=" %%l in (%FILENAME%) DO (
SET LINE=%%l
CALL :PRINTLINE
)
GOTO :EOF
REM Adds the lines from the specified range to the array of lines to display
:BUILDARRAY
REM Find out whether we have a single line or a range
SET TEST=%1
SET /A TEST1=%TEST%
SET /A TEST2=%TEST:-=%
IF %TEST1%==%TEST2% (
REM Single line
SET /A LASTLINE+=1
SET LINES[!LASTLINE!]=%1
) ELSE (
REM Range
FOR /F "tokens=1,2 delims=-" %%x IN ("%1") DO (SET RANGESTART=%%x&SET RANGEEND=%%y)
REM Some sanity checking
IF !RANGESTART! GTR !RANGEEND! (
ECHO.Problem with range !RANGESTART!-!RANGEEND!:
ECHO. Ranges must have a start value smaller than the end value.
EXIT /B 1
) ELSE (
FOR /L %%i IN (!RANGESTART!,1,!RANGEEND!) DO (
SET /A LASTLINE+=1
SET LINES[!LASTLINE!]=%%i
)
)
)
GOTO :EOF
REM Prints the specified line if the current line should be printed
:PRINTLINE
SET /A CURRENTLINE+=1
IF %CURRENTLINE%==!LINES[%INDEX%]! (
REM We have a line to print, so do this
ECHO !LINE!
SET /A INDEX+=1
)
GOTO :EOF
REM Prints usage and exits the batch file
:USAGE
ECHO.%~n0 - Displays selected lines from a text file.
ECHO.
ECHO.Usage:
ECHO. %~n0 ^<filename^> ^<range^> ...
ECHO.
ECHO. ^<filename^> - the text file from which to read
ECHO. ^<range^> - the line range(s) to display.
ECHO.
ECHO.Example:
ECHO. %~n0 foo.txt 1-4 13 15 18-20
ECHO.
ECHO. will display the first four lines from the file "foo.txt",
ECHO. the 13th and 15th as well as the lines 18 to 20.
ECHO.
ECHO.Line ranges are separated by comma, semicolon or space. If no range is given,
ECHO.the entire file is displayed.
EXIT /B 1
GOTO :EOF
The whole script could use some better error checking, examples of what not to do or where error checking is a bit wonky:
dl foo.txt 1-2-4
(prints lines 1-2 but no error message)
dl foo.txt -1
(error message that the range 1- isn't correct)
Other limitations:
- Ranges must be sorted. With the current implementation there is no way to do something like
dl foo.txt 7-8,1-2
.
- No line may be selected twice. Something like
dl foo.txt 1,2,2-8,11-15
will stop after the second line.
- No support for UNIX-style line endings (U+000A)
EDIT: Fixed an issue with files that contain parentheses.