Projects / sed / Releases / 4.1.4

Version 4.1.4 of sed

Release Notes: In this version, \B correctly means "not on a word boundary" rather than "inside a word", matching sed 3.x and Perl. Regular expression addresses no longer use leftmost-longest matching: e.g. /.\+/ only looks for a single character, and does not try to find as many of them as possible like it used to do. Documentation about the new interpretation of `s|abc\|def||' in NEWS, and about localization issues, was added. A --disable-nls build problems on Solaris was fixed, as was `make check' in non-English locales. Speedups (up to 30-40%) to regular expression matching were made.

Other releases

  •  28 Jun 2009 06:06

Release Notes: sed -i now preserves SELinux security contexts, and temporary files for sed -i are kept readable only by the user until they are complete. A problem with regular expressions including three or more consecutive brackets has been fixed.

Release Notes: This release improved portability and increased speed in UTF-8 locales. A period in regular expressions now accepts NUL bytes. ACLs correctly preserved in in-place mode. A new --follow-symlinks option is available when editing a file in-place. Hold-space is reset between different files in -i and -s modes. A new 'z' extension clears pattern space even in the presence of invalid multi-byte sequences. Bugs in multi-byte processing were fixed. More GNU extensions are now turned off by the --posix options, including all GNU extensions to regular expressions.

  •  16 May 2006 05:14

Release Notes: Parsing of a negative character class that includes a closed bracket (like [^]] or [^]a-z]) was fixed. Parsing of "[" inside a "y" command (like y/[/A/) was fixed. The result of the "a", "r", and "R" commands is output when a "q" command is found.

  •  28 Jan 2005 01:42

Release Notes: In this version, \B correctly means "not on a word boundary" rather than "inside a word", matching sed 3.x and Perl. Regular expression addresses no longer use leftmost-longest matching: e.g. /.\+/ only looks for a single character, and does not try to find as many of them as possible like it used to do. Documentation about the new interpretation of `s|abc\|def||' in NEWS, and about localization issues, was added. A --disable-nls build problems on Solaris was fixed, as was `make check' in non-English locales. Speedups (up to 30-40%) to regular expression matching were made.

  •  10 Sep 2004 10:57

Release Notes: This release has several important bugfixes, mostly related to localization and support of multi-byte character sets. This release will be in Debian Sarge.

Screenshot

Project Spotlight

openModeller

An ecological niche modeling framework.

Screenshot

Project Spotlight

Barman for PostgreSQL

A backup and recovery manager for PostgreSQL.