Home Code Other Software

Table of Contents


paragrep is a “paragraph grep” utility. It searches for a series of regular expressions in a text file (or several text files) and prints out the paragraphs containing those expressions. Normally paragrep displays a paragraph if it contains any of the expressions; this behavior can be modified by using the -a option.

By default, a paragraph is defined as a block of text delimited by an empty or blank line; this behavior can be altered with the -p option.

If no files are specified on the command line, paragrep searches standard input.


The easiest way to install paragrep is via pip:

$ pip install paragrep

WARNING: As of version 3.2.0, paragrep no longer supports Python 2. If you need to run it under Python 2, use an older version. e.g.:

$ pip install paragrep==3.1.3


paragrep [-aiotv] [-p regexp] [-e regexp] … [-f exprfile] … [file] …

paragrep [-itv] [-p eop_regexp] regexp [file] …

paragrep -h –help

Options, in brief

Options, in detail


-a is the and option. It only displays paragraphs that contain all the regular expressions specified on the command line. The default is to display paragraphs that contain any of the regular expressions. (See -o.)


Adds a regular expression to the set of expressions to use when matching paragraphs. More than one -e argument may be specified. If there’s only one expression, the -e may be omitted for brevity. (Think sed.)


-f expfile specifies a file containing regular expressions, one expression per line. Each expression in the file is added to the set of expression against which paragraphs are to be matched. More than one -f argument is permitted. Also, -f and -e may be specified together.

-h (or --help)

Print full usage information and exit.


Specifies case-blind pattern matching. The default is case-sensitive pattern matching.


-o is the or option. It displays a paragraph if it contains any the regular expressions specified. Since this option is the default, it is rarely specified on the command line. It exists primarily to negate the effect of a previous -a option. (e.g., If you’ve defined an alias for paragrep that specifies the -a option, -o would be necessary to force the or behavior.)


Specifies a regular expression to be used match paragraph delimiters. Any line that matches this regular expression is assumed to delimit paragraphs without actually being part of a paragraph (i.e., lines matching this expression are never printed). If this option is not specified, it defaults to:

^[ \t]*$

which matches blank or empty lines. (\t represents the horizontal tab character. If you need to specify a horizontal tab, you’ll need to type the actual character; paragrep doesn’t recognize C-style metacharacters.)


Displays the lines that mark the end of each paragraph, instead of separating matching paragraphs with a newline (which is the default).


Displays all lines that do not match specified expressions. The negation logic works on De Morgan’s laws. Normally, if -a is specified, paragrep uses the following logic to match the paragraph:

match = contains(expr1) AND contains(expr2) ...

Specifying -v along with -a changes this logic to:

match = lacks(expr1) OR lacks(expr2) ...

Likewise, without -a or -v (i.e., using -o, which is the default), the matching logic is:

match = contains(expr1) OR contains(expr2) ...

Negating that logic with -v causes paragrep to match paragraphs with:

match = lacks(expr1) AND lacks(expr2) ...


This is the third implementation of paragrep. The first implementation, in 1989, was in C. The second implementation, in 2003, was in Perl. This is the latest and greatest.


Brian M. Clapper

Copyright © 1998-2019 Brian M. Clapper


BSD license.

Change log

You can find the change log here.


I gladly accept patches from their original authors. Feel free to email patches to me or to fork the GitHub repository and send me a pull request. Along with any patch you send: