text
sed
Stream editor that applies a small program to each line of input. Widely used for search-and-replace and inline file edits.
sed [OPTION]... {script} [FILE]...Common flags
| Flag | Purpose |
|---|---|
| -i | Edit files in place (accepts a backup suffix, e.g. -i.bak) |
| -e | Add a script expression |
| -n | Suppress default output; use with p to print explicitly |
| -E | Use extended regular expressions |
| -r | GNU alias for -E |
Examples
sed 's/foo/bar/g' file.txtReplace every foo with bar and print to stdout
sed -i 's|/old|/new|g' config.ymlIn-place replacement using | as delimiter to avoid escaping slashes
sed -n '10,20p' log.txtPrint lines 10 through 20 only
sed -E 's/([0-9]+)/[\1]/g' data.txtWrap every number in brackets using extended regex
Gotcha
sed -i behavior differs between GNU and BSD (macOS) -- GNU accepts an empty suffix, BSD requires one; scripts targeting both should be explicit.
Related commands
awk
Pattern-action language for processing columnar text. Excellent for reshaping tabular data, aggregating fields, and quick reports.
grep
Searches text for lines matching a regular expression. Central to log analysis and pipeline filtering.
cut
Extracts sections from each input line by byte, character, or delimited field. Handy for pulling columns out of CSV or colon-delimited files.
cat
Concatenates files and prints them to standard output. Also useful as a simple viewer or to feed content into pipelines.
sort
Sorts lines of text. Supports numeric, human-numeric, reverse, unique, and keyed sorting.