The program's help gives a comprehensive description of its available options, the wiki gives further details, and the TL;DR section below has a list of quick and easy example usages.
> ned --help
Usage: ned [OPTION...] [-p] PATTERN [FILE...]
ned [OPTION...] [FILE...] -p PATTERN
For regular expression power users, ned is like grep, but with
powerful replace capabilities, and unlike sed, as it
isn't restricted to line oriented editing.
FILEs are ASCII or UTF-8 text files. For regex syntax see:
https://docs.rs/regex/1.4.6/regex/#syntax
Options:
-p, --pattern PATTERN
Specify a pattern. If the option isn't used, the
pattern must precede the files. The option allows the
pattern to be put after the files for more convenient
editing.
-r, --replace REPLACEMENT
Replace matches. Replacements may include numbered and
named groups. Replaces always operate on whole files.
-w, --whole-files Operate on whole files. Otherwise matches are line
oriented.
-n, --number N Match/replace N occurrences.
-k, --skip N Skip N occurrences before matching/replacing.
-b, --backwards Make -n/--number and -k/--skip options count
backwards.
-i, --ignore-case Ignore case.
-s, --single '.' matches newlines, ^ and $ match the beginning and
end of each file. Use with --whole-files.
-m, --multiline Multiline, ^ and $ match the beginning and end of each
line. Use with --whole-files.
-x, --extended Ignore whitespace and # comments.
--case-replacements
Enable \U - uppercase, \L - lowercase, \I - initial
uppercase (title case), \F - first uppercase (sentence
case) replacements. \E marks the end of a case
replacement.
-o, --matches-only Show only matches.
-g, --group GROUP Show the match group, specified by number or name.
-v, --no-match Show only non-matching.
-f, --filenames-only
Show only filenames containing matches. Use with
-v/--no-match to show filenames without matches.
-F, --no-filenames Don't show filenames.
-l, --line-numbers-only
Show only line numbers containing matches. Use with
-v/--no-match to show line numbers without matches.
Use without -w/--whole-files.
-L, --no-line-numbers
Don't show line numbers. Use without -w/--whole-files.
-C, --context LINES Show LINES lines around each matching line. Is the
same as specifying both -B/--before and -A/--after
with the same LINES. Use without -w/--whole-files.
-B, --before LINES Show LINES lines before each matching line. Use
without -w/--whole-files.
-A, --after LINES Show LINES lines after each matching line. Use without
-w/--whole-files.
-R, --recursive Recurse.
-l, --follow Follow symlinks. (Ignored on Windows.)
--include GLOB Match only files that match GLOB.
--exclude GLOB Skip files matching GLOB.
--exclude-dir GLOB
Skip directories matching GLOB.
-u, --ignore-non-utf8
Quietly ignore files that cannot be parsed as UTF-8
(or ASCII). Because this requires reading the file,
the --exclude option should be preferred.
-a, --all Do not ignore files and directories starting with '.'.
-c Show filenames, line numbers, and matches in color. Is
the same as --colors=always.
--colors [WHEN] 'auto' shows filenames, line numbers, and matches in
color when stdout is a terminal, not when it is a
pipe, 'always' shows color even when stdout is a pipe,
and 'never' never shows colors.
--color [WHEN] Synonym for --colors.
--stdout Output to stdout.
-q, --quiet Suppress all normal output. When matching terminate as
soon as a match is found.
-V, --version Output version information and exit.
-h, --help Print this help and exit.
Environment:
NED_DEFAULTS ned options added to the program's arguments. Is
a space delimited list of options and is not first
interpreted by a shell, so quotes are not required
around arguments. For example:
NED_DEFAULTS="-u -R --exclude *.bk --exclude-dir .git"
Exit codes:
0 matches found/replaced
1 no matches
Quiet:
When -q/--quiet is specified, ned tests for matches and returns an exit
code of 0, if a match is found in ANY file. Quiet matches will only read
as many files as needed to find a match. Even without this shortcutting
behaviour, quiet matches are more performant than non-quiet matches.
ned 1.3.1 Copyright (C) 2016-2021 Nev Delap - https://github.com/nevdelap/ned
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>.
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.
Why isn't \U working? (or \L, \I, \F)
Because case replacing is off by default to not waste cycles when you're not doing it, since that is most of the time, generally. See the help: (as of v1.2.0)
--case-replacements
Enable \U - uppercase, \L - lowercase, \I - initial
uppercase (title case), \F - first uppercase (sentence
case) replacements. \E marks the end of a case
replacement.
Why do I get errors like ned: /path/file invalid utf-8 sequence of 1 bytes from index 25?
Because, by default, ned reads everything unless you tell it not to read it. If you want it to always
ignore non-ASCII, non-UTF-8 files, you can put this in NED_DEFAULTS. See the help:
-u, --ignore-non-utf8
Quietly ignore files that cannot be parsed as UTF-8
(or ASCII). Because this requires reading the file,
the --exclude option should be preferred.
Why don't the tests pass in Git Bash?
Git Bash does not support colored output using ansi_term. Run the tests in cmd.exe on Windows.
- Install Rust as per: https://www.rust-lang.org/en-US/install.html
- (Windows) Install Visual Studio Build Tools 2017 as per: https://www.visualstudio.com/downloads/
Last tested on Manjaro up-to-date at 2021/12/12 with Rust 1.57.0, on Windows 10.0.17134.523 and on OS X High Sierra with Rust 1.32.0.
cd ned
cargo build --release
cargo test
...
test result: ok. 137 passed; 0 failed; 0 ignored; 0 measured; 0 filtered outLast tested on Manjaro up-to-date at 2021/12/12 with Rust 1.57.0.
cd ned
rustup target add x86_64-unknown-linux-musl
cargo build --release --target x86_64-unknown-linux-musl
cargo test --target x86_64-unknown-linux-musl
...
test result: ok. 137 passed; 0 failed; 0 ignored; 0 measured; 0 filtered outyaourt -S nedDownload the deb or rpm file from the latest release: https://github.com/nevdelap/ned/releases. They package a single musl binary of ned and its man page with no dependencies. Install with dpkg -i or rpm -i.
Download the appropriate binary and put it on your path. In Windows install the dependencies as described in the latest release: https://github.com/nevdelap/ned/releases
On macOS, ned can also be installed using MacPorts.
IMPORTANT NOTE: The search capabilities of ned are not so interesting, you can do them all with grep or ripgrep, see https://github.com/BurntSushi/ripgrep. It is the replace that is interesting, which grep and ripgrep cannot do - ripgrep can replace as it writes to stdout, it doesn't modify files, and sed can only do in a line oriented way. Examples of searching are shown first, followed by examples of replacing. Replacing with ned is a very powerful way of doing bulk editing from the terminal. Hence ned = Nev's Editor. Stage or commit your files before doing bulk edits, so that you can undo changes until you get your regex right.
These examples use short options and search for 'dog' and replace with 'cat' wherever the example doesn't need a regular expression to demonstrate what it is doing.
Search non-hidden files in the current directory.
ned dog .Search txt files in the current directory.
ned dog *.txtSearch including hidden files.
ned -a dog .Search recursively.
ned -R dog .Search case insensitively.
ned -i dog .Search always showing colored output.
ned -c dog .
ned --colors=always dog .Search never showing colored output.
ned --colors=never dog .Search showing colored output when outputting to a terminal, but don't send colored output, if piped.
ned --colors=auto dog .Set default arguments in your terminal environment.
export NED_DEFAULTS='-i --colors=always'Search showing colored output through less.
ned -c dog . | less -RSearch showing no output, to just use the exit code in a script, if something is found or not found.
This is more efficient when you don't need the output since it shortcuts when it finds the first match.
ned -q dog .; echo $?
0 # Found.
ned -q dinosaur .; echo $?
1 # Not found.Search specifying the pattern at the end of the command - for convenience of editing when you have a lot of options.
ned . -p dogSearch not showing line numbers.
ned -L dog .Search only showing numbers of matched lines.
ned -l dog .Search not showing file names
ned -F dog .Search only showing file names of matched files.
ned -f dog .Search showing only matches.
ned -o dog .Search really showing only matches.
ned -oFL dog .Search matching first 3 occurences per line.
ned -n 3 dog .Search matching first 3 occurences per file.
ned -w -n 3 dog .Search backwards, matching last 3 occurences per line.
ned -b -n 3 dog .Search backwards, matching last 3 occurences per file.
ned -b -w -n 3 dog .Search skipping 3 occurrences and finding 2 occurences.
Note: -k is the short form of --skip. (-s is the short form of the --single option.)
ned -k 3 -n 2 dog .Search backwards, skipping 3 occurrences and finding 2 occurences.
ned -b -k 3 -n 2 dog .Search recursively only including certain files.
ned -R --include '*.txt' dog .Search ignoring certain files.
ned -R --exclude '*.htm' dog .Search ignoring all non-utf8 files.
Quietly ignore files that cannot be parsed as UTF-8 (or ASCII). Because this requires reading the file the --exclude option should be preferred. E.g. --exclude '*.png'
ned -u dog .Search ignoring certain directories.
ned -R --exclude-dir '.git' dog .Search showing context of 5 lines around each match.
ned -C 5 dog .Search showing context of 5 lines before each match.
ned -B 5 dog .Search showing context of 5 lines after match.
ned -A 5 dog .Search matching the beginnings of lines.
ned '^dog' .Search matching the ends of lines.
ned 'dog$' .Search matching the beginnings of files.
ned -w '^dog' .Search matching the ends of files.
ned -w 'dog$' .Search spanning lines.
Search for any and all three consecutive lines containing the word dog.
ned -w 'dog.*\n.*dog.*\n.*dog' .Replace.
ned dog -r cat .Replace using numbered group references.
'the big dog and the smelly dog' replaced with 'the smelly dog and the big dog'.
ned 'the ([a-z]+) dog and the ([a-z]+) dog' -r 'the $2 dog and the $1 dog' .Replace using named group references.
'the big dog and the smelly dog' replaced with 'the smelly dog and the big dog'.
ned 'the (?P<first>[a-z]+) dog and the (?P<second>[a-z]+) dog' -r 'the $second dog and the $first dog' .Replace spanning lines.
Delete any and all three consecutive lines containing the word dog.
ned -w '\n.*dog.*\n.*dog.*\n.*dog.*\n' -r '\n'Replace changing case.
'big dog' and 'smelly dog' replaced with 'BIG! dog' and 'SMELLY! dog'. Available case replacements: \U - uppercase, \L - lowercase, \I - initial uppercase (title case), \F - first uppercase (sentence case).
ned ' ([a-z]+) dog' --case-replacements -r '\U$1\E! dog' --stdout .Replace and see the results in the terminal without updating the target files.
ned dog -r cat --stdout .Replace and treat no replacements as success.
ned trggde -r cat . || true; echo $?
0 # Maybe not found, but if so, not an error.Strip blank lines from files.
ned -w '(\s*\n)+' -r '\n' .Strip blank lines from the ends of files.
ned -w '(\s*\n?)*$' -r '' .Unident tables and lists in XHTML files, ignoring the .git directory.
ned -R --include '*.htm' --exclude-dir '.git' (</?(table|col|tbody|tr|th|td|ol|ul|li)[^>]*>)' -r '$1'