Windows happens. So do text files. Sometimes, they happen together.
(Put Cygwin down.)
Encodings
Powershell, like the rest of Windows, thinks in UTF-16. One charming consequence of this is that passing an ASCII data file through a Powershell pipeline will double its size on disk. This is often not what you or your sysadmin wants, so instead of redirecting output to a file using >,
> cat data.csv > data_copy.csv
use the out-file cmdlet (> is just synctatic sugar for out-file anyway):
> cat data.csv | out-file data_copy.csv -encoding utf8
You can also use -encoding ascii.
$ $CMD --help
> help $CMD
The command help is an alias for get-help. The “REMARKS” section at the bottom of the help output has good further pointers, like what command to use to see examples and more detailed information.
$ cheat
(If you’re not already using cheat, then let me introduce you to sunshine and rainbows.)
> get-help $CMD -examples
Weirdly, help $CMD -examples is presented slightly differently; try both and see which output you prefer.
$ head
> get-content $FILE -TotalCount 10
> cat $FILE -t 10
$ tail
> get-content $FILE | select-object -last 10
> cat $FILE | select -l 10
$ grep, vanilla
> select-string $PATTERN $FILE
> sls $PATTERN $FILE # sls is a defualt alias only in PowerShell 3.0 and later
select-string accepts perl regular expressions:
> (echo "<foo> <bar>" | select-string "<.*?>").matches | select-object -exp value
<foo>
You can also use the -match command to iterate regexes quickly:
> "foo bar" -match "b|c"
True
$ grep, recursive
> get-childitem -include *.txt -recurse | select-string $PATTERN
> ls -i *.txt -r | sls $PATTERN
$ sed
> get-content $FILE | % {$_ -replace "$PATTERN","$REPLACEMENT"}
You can capture back references with parens (), but to reference them it is best to use single quotes ' around the second argument to -replace, otherwise the back references will be evaluated as Powershell variables:
> $1 = "NOPE"
> "bar" -replace "(bar)","foo $1"
foo NOPE
> "bar" -replace "(bar)",'foo $1'
foo bar
$ cut
> echo 'fst_col,snd_col' | % {$_.split(':')[0]}
fst_col