New in version 4.7:
- Fixed signal handling to make grepmail easier to debug. Thanks to Ilya
Zakharevich for providing the solution.
- Fixed a possible performance problem in reading of emails (Perl
implementation), and documented the settings in the README.
- Expanded the pattern for matching the start of emails to allow different
types of emails to be parsed.
- Fixed a bug where -R was reported as not recognized. (Thanks to Nicholas
Riley <nriley@sf.net> for the bug report and fix.)
- "anonymize_mailbox" utility included to help people submit bug reports
- If a mailbox can not be found, grepmail now searches the $home/mail,
$home/Mail, $home/Mailbox directories (or the directory specified by the
MAIL environment variable). (Thanks to Michael Friendly <friendly@yorku.ca>
for the feature suggestion and initial patch.)
- Added -X flag to allow the user to specify a pattern for the signature
separator. (Thanks to Michael Friendly <friendly@yorku.ca> for the feature
suggestion.)
- Added -Y flag to search specific headers. (Thanks to Terry Jones
<terry@eatoni.com> for the idea to automatically wrap header lines as
necessary.)
New in version 4.60:
- Removed -B flag and added -S flag. -B is now performed using -bS.
- Added installation flags to suppress interactive installation. (Thanks to
Joey Hess <joey@kitenet.net> for the problem report. He had to patch
Makefile.PL for his Debian packaging.)
- Fixed a slow implementation of searching for signatures that would cause
grepmail to crawl for very large emails. Thanks to Joey Hess
<joey@kitenet.net> for discovering the inefficiency.
- Fixed a short-circuit which should have bypassed the search for signatures
if -B was not specified. Thanks to Joey Hess <joey@kitenet.net> for finding
the bug.
- Implemented a new Perl parser which is 5% to 50% faster depending on how
I/O-bound your system is.
- Restructured the code a bit and improved detection of invalid arguments.
New in version 4.51:
- grepmail now dies gracefully when invalid patterns like 'strcpy(' are
supplied. (It should be 'strcpy\('.)
- Fixed a bug in attachment boundary matching which would cause the boundary
match to fail if the boundary contained special pattern matching characters.
(Thanks to Richard Everson <rme at users.sourceforge.net> for identifying
the bug, and providing a sample email which demonstrates the problem.)
- Added a check for Inline 0.41 or better during "perl Makefile.PL" when
Mail::Folder::FastReader is selected to be installed. (Thanks to Brian L.
Johnson <blj8@blj8.com> for the problem report.)
- Fixed a bug where grepmail would fail to print matching emails which had
signatures, and added a test case for it. This bug was introduced with -B
support in version 4.49. (A *huge* thanks to Moritz Barsnick
<barsnick@gmx.net> for reporting the bug and doing the initial analysis of
the cause.)
- Modified Makefile.PL to ask whether the user wants FastReader regardless of
whether they specified arguments to "make Makefile.PL".
- Modified Makefile.PL to allow the user to interactively specify the
installation path.
- Fixed a typo in debugging output for emails without "Date:" headers.
- Improved error messages.
- Usage message now displays just the flags, --help shows a summary of their
meanings as well.
New in version 4.50:
- Added X-Draft-From to support newer versions of Gnus (Thanks to Nevin Kapur
<nevin@jhu.edu> for the patch).
New in version 4.49:
- Fixed test cases to work around PATH modifications made by Date::Manip
- Added -B to search the body but not the signature. (Thanks to Helmut
Springer <delta@faveve.uni-stuttgart.de> for the feature request.)
- Added LICENSE file. (Thanks to Janet Casey <jcasey@gnu.org> for the
reminder.)
New in version 4.48:
- Mail::Folder::FastReader migrated from XS to Inline.
- -H flag added by Nevin Kapur <nevin@jhu.edu>
- Error messages are localized in test cases. (Thanks to cpan-testers and in
particular Jost Krieger <Jost.Krieger@ruhr-uni-bochum.de> for finding this
bug and diagnosing it.)
- Fixed a problem with timezones in the test cases. (Thanks to Roy Lanek
<lanek@ranahminang.net> for helping me debug this.)
- Added a check in the test cases for determining if the user's timezone is
not recognized by Date::Manip.
New in version 4.47:
- Grepmail now converts Gnus emails into mbox emails before printing them.
(Thanks to Johan Vromans <JVromans@squirrel.nl> for supplying a patch and
explaining the need for it.
- Fixed a couple bugs in -M processing
- -M is now 19% faster. (It's now only about 9% slower than without -M.)
New in version 4.46:
- Fixed a bug in -u message id processing. (Thanks go to an anonymous bug
reporter on SourceForge.)
- Added more workarounds to prevent warnings resulting from a bug in Perl 5.6
(Thanks to Joey Hess <joey@kitenet.net>)
- Added -Z flag to tell grepmail not to use Mail::Folder::FastReader even if
it is installed.
- Fixed a bug introduced in version 4.44 where -m would not work unless used
with -n. (Thanks to Imre Simon <is@ime.usp.br> for catching this.)
- --help anywhere on the command line now causes the help to be printed,
instead of only when used as the first argument.
- Test script now exercises both the Mail::Folder::FastReader and perl mailbox
implementations as appropriate.
- Restructured code to separate out Perl mailbox implementation as
Mail::Folder::SlowReader. (This module is embedded in the grepmail script.)
New in version 4.45:
- Added -n and -V to usage message. (Thanks to Wolfgang Rohdewald
<WRohdewald@dplanet.ch> for catching this.)
- Added workarounds to prevent warnings resulting from a bug in Perl 5.6
- A blank line before the start of an email is not required now. This
allows broken folders created by Netscape to be read. (Thanks to Jeremy
Malcolm <terminus@terminus.net.au> for the bug report.)
- Mail::FastReader is 8% faster.
New in version 4.44:
- execution with -M flag now 35% faster
- Added a Mail::Folder::FastReader module which can cause grepmail to
run 10-40% faster, depending on your system. Since this module is
experimental, the installation script will allow you to not install
the module. A C compiler is required.
- "-----Original Message-----" now recognized as beginning an included
message
- Fixed a bug where errors would not be displayed if compressed data
was piped to grepmail
- Date::Parse is now only required if -d is used. (Date::Manip is
still only required if you use complex date specfications.) (Thanks
to Richard Stallman <rms@gnu.org> for pointing this out.)
- Added -n flag to print line numbers a la grep. (Thanks to Richard
Stallman <rms@gnu.org> for the suggestion)
- Fixed a bug in debug output where the email subject was actually the
sender
- Fixed an undefined value in the printing of flag information
- An ASCII file is now determined to be a mailbox if it has a line
whose prefix is 'From' or 'X-From-Line:', and another line whose
prefix is 'Received ', 'Date:', 'Subject:', 'X-Status:', or
'Status:'.
- Error exits now return 1 (Thanks to Wolfgang Rohdewald
<WRohdewald@dplanet.ch> for the bug report)
- -V flag prints the version (Thanks to Wolfgang Rohdewald
<WRohdewald@dplanet.ch> for the feature request)
- Restructured code: localized reading of the emails from the file,
removed two functions
New in version 4.43:
- Fixed a bug in -r counting when used with -h. (Thanks to Andrew
<andrew@ugh.net.au> for the bug report.)
- Fixed a bug in the handling of included messages. (Thanks to Antoine
Martin <antonus@libertysurf.fr> for the bug report and suggestion
for the fix.)
New in version 4.42:
- Added -a flag to use received date instead of sent date for -d matches.
(Thanks to Michael Liljeblad <liljeblad at users.sourceforge.net> for the
patch.)
- Included emails are now ignored properly (Thanks to an anonymous submittor
for the bug report and part of a patch
https://sourceforge.net/bugs/?func=detailbug&bug_id=112159&group_id=2207)
- If an email has no date, the -d switch now issues a warning and does not
treat the email as a match (Thanks to David Blaine <blained at
users.sourceforge.net> for the bug report.)
- -d "" can be used to find emails without dates
- Mailbox files are now detected as files using
if ($buffer =~ /^(X-From-Line:|From) /im && $buffer =~ /^Subject: /im)
rather than
if ($buffer =~ /^(X-From-Line:|From) /im && $buffer =~ /^Date: /im)
- Improved detection of binary files. (Thanks to Dan Sugalski
<dan@tuatha.sidhe.org> for the sample code.)
- STDERR and STDOUT now checked separately during "make test"
- Headers can now be in the format "Date:<tab>" in addition to "Date: "
(Thanks to Benjamin Korvemaker <korvemaker at users.sourceforge.net> for the
patch and concept.)
New in version 4.41:
- Support for Gnus-style mail files added (Thanks to Werner Bosse
<Werner.Bosse@alcatel.de> for the patch.)
- Test mail files tweaked to make the test cases work better across all time
zones. (Thanks to Andreas Kutschera <Andreas.Kutschera@bln.siemens.de> for
the patch.)
- Added check for unparsable dates in email headers.
- Fixed a Perl warning raised when date parsing fails.
- Added instructions for getting the necessary modules to README.
New in version 4.40:
- Date::Parse and Date::Manip version unified -- Date::Parse is now required,
and Date::Manip (if present) is used to parse complex dates. (Patch by
Seth Cohn <sethcohn@users.sourceforge.net>, modified by David Coppit)
New in version 4.31:
- Distro has missing test cases for -u functionality. Doh!
New in version 4.30:
- Updated code to avoid warnings under Perl 5.6 (Thanks to Andreas Kutschera
<Andreas.Kutschera@bln.siemens.de> for the bug report.)
- Fixed a bug in the test script where bzip2, gzip, and tzip support would not
be tested even though the programs were available. (Thanks to Andreas
Kutschera <Andreas.Kutschera@bln.siemens.de> for the patch.)
- Added standard --help flag (Patch by Seth Cohn <seth@users.sourceforge.net>)
- Added -u ("unique") flag, which ensures that no duplicate messages will be
output. (A BIG thanks to Seth Cohn <seth@users.sourceforge.net>.)
New in version 4.23:
- Updated the test cases to work better in timezones close to +0000 and +2300.
(email if you have problems with tests 1 and 23. Thanks to Harald Krause
<harald.krause@a-city.de> for first finding the bug, and Adam Huffman
<adam.huffman@man.ac.uk> for his help debugging it.)
- Fixed a bug in the "ignore attachments" code
New in version 4.22:
- grepmail now behaves better when tzip, bzip2 or gunzip aren't present on the
system.
- The code has been restructured to compile more easily with perlcc.
New in version 4.21:
- Fixed a bug that would cause grepmail to runaway when a pipe following it
was broken. (Thanks to Gerald Pfeifer <pfeifer@dbai.tuwien.ac.at> for the
bug report)
New in version 4.20:
- grepmail development has been moved to SourceForge, and made public. Visit
http://grepmail.sourceforge.net/
- Added -s flag, which limits matched emails to a given size
- Restructured the code to be more robust with respect to feature interaction.
(At a 5-10% slowdown cost.)
- Fixed an uninitialized variable warning caused by emails without subjects in
debug mode.
New in version 4.11:
- Fixed a bug where an ASCII file would not be recognized as a mailbox when
the first couple emails did not have a "From:" line. (Thanks to Jeff Flowers
<jeff.flowers@npl.co.uk>)
- Added standard Perl testing.
New in version 4.1:
- Stripped auto-perl execution code, since it never works on all platforms.
(Installation instructions modified to require the user to fix the #! line.)
- Minor changes to allow grepmail to run without -w complaints.
New in version 4.0:
- Fixed a bug where shell characters needed to be escaped for compressed
files. (Bug found by Richard Clamp <richardc@tw2.com>
- Added #!/bin/sh as first line to make the rest compatible with csh/tcsh
users. (Bug found by Ed Arnold <era@ucar.edu>)
New in version 3.9:
- Took out specialization engine because there wasn't enough support to
program in that style.
- Offering 2 main versions now -- Date::Manip and Date::Parse
- Added -R option, which causes grepmail to recurse any directories
encountered. (Thanks to Emil Tiller <etiller@asc.corp.mot.com> for the
initial code.)
- Fixed a small bug that would cause some attachments not to be identified.
New in version 3.8:
- Added a prototype engine to allow users to specialize grepmail at
installation time. (See below)
- Fixed buggy mailbox detection algorithm
- Fixed bug in identification of email headers.
- Fixed bug in parsing timezones of emails. (Thanks to Wolfgang Weisselberg
<weisselberg@element5.de>)
- Fixed bug in handling of date specifications like "2 days ago" and "2 weeks
ago"
- Added -M switch, which causes grepmail to ignore non-text attachments.
- Added "quiet mode", -q switch, which supresses warnings about directories,
non-mailbox ASCII files, binary files that can't be decompressed, etc.
- Restructured code a bit. Moved file and STDIN processing out of main. The
whole email is now read before the match is made to the body, instead of
trying to match the pattern while reading the body. (This simplifies the
algorithms and makes -M support a lot easier, at the cost of increasing the
required memory slightly.) Now uses ungetc to put the test characters used
during file type detection back on the stream.
New in version 3.7:
- Added -D for debugging output
- Now ignores ASCII files that don't look like mailboxes. Thanks to oak
<oak@eskimo.com> for pointing this out.
- Uses Date::Parse instead of Date::Manip, which results in faster execution
time at the expense of less flexibility. (e.g. You can't do "12pm January 5
1997" any more)
New in version 3.6:
- No more temporary files! This addresses the security issues that a few
people have sent me email about. The script may use slightly more memory,
depending on the size of the largest email. (Email is now buffered as it is
read, and the whole buffer is printed when a match occurs. This is in
contrast to storing the file pointer and seeking back to the start of the
email.) The script is substantially faster for large amounts of data piped
as STDIN. Many thanks go to Joey Hess (<joey@kitenet.net>), who supplied
insights and contributions toward making this release happen, especially the
buffering code.
New in version 3.5:
- grepmail will not try to decompress piped input that is empty.
- Temporary files are now placed in the user's home directory to help avoid
privacy attacks (or in the directory specified by the TMPDIR environment
variable, if it exists).
- Fixed a bug that would occasionally leave a tempfile around.
New in version 3.4.1:
- Fixed a bug that added an extra line to the start of output. (Thanks to
Moritz Barsnick <barsnick@gmx.net> for helping to find this.)
New in version 3.4:
- Added tzip support. (thanks to Marc Lehmann <pcg@goof.com>)
- Reordered flags to better match grep.
- Changed command line syntax again. (Last time, I hope.)
New in version 3.3:
- Added bzip2 support. (thanks to Josh Plautz <plautz@milner.com>).
- Improved error checking on piped binary input.
- Added debugging code.
New in version 3.2:
- Added TMPFILE environment variable support, and a signal handler. (thanks
to Ulli Horlacher (<framstag@moep.bb.bawue.de>).
- Fixed a bug where the last paragraph of the last email in a mailbox would
not be printed on Linux. (How's that for obscure? Thanks to Eli Criffield
<rexracer@mammoth.org> for discovering it.)
New in version 3.1:
- Modified the decompression to be more compatible with older versions of gzip.
- Improved error checking so that "grepmail -h" prints a usage message.
- Added -m flag, which causes an "X-Mailfolder" line to be added to the
header, thereby showing which folder contained the message. (by Ulli
Horlacher <framstag@moep.bb.bawue.de>).
- Improved error checking on flags.
- Changed "zcat" to "gunzip -c" to help with backwards compatibility with older
versions of gzip (thanks to Eugene Kim <eekim@eekim.com>).
New in version 3.0:
- -h and -b can be used together.
- Rewrote the ProcessMailFile to run 2 to 3 times faster, and use less memory.
- Correctly diagnoses directories as such (by Gerald Pfeifer
<pfeifer@dbai.tuwien.ac.at>).
New in version 2.1:
- Added -l,-r, and -e, as suggested by Reinhard Max <max@suse.de>.
- Now uses about 1/3 the memory, and is a little faster.
New in version 2.0.1:
- Added POD documentation at the end of the script (thanks, Jeffrey
Haemer <jsh@boulder.qms.com>).
- -h for headers only -b for body only
New in version 1.9:
- "Ignore empty files" by Gerald Pfeifer <pfeifer@dbai.tuwien.ac.at>.
- Emails without dates are now automatically output no matter what the
date specification is. (Better safe than sorry!)
New in version 1.7:
- Sped up by Andrew Johnson. It no longer looks for dates unless the email
matches the search string.
New in version 1.6:
- Removed use of Compress::Zlib because it was 30% slower, complicated the
code, and because any user with gzip'd mail has zcat...
New in version 1.5:
- Andrew Johnson <ajohnson@gpu.srv.ualberta.ca> fixed a couple of bugs.
New in version 1.4:
- Incorporated conditional loading of the date module (submitted by Andrew
Johnson <ajohnson@gpu.srv.ualberta.ca> Many thanks!).
- compress::Zlib used instead of shelling out to gunzip (submitted by Andrew
Johnson <ajohnson@gpu.srv.ualberta.ca> Many thanks!).
- Some bug fixes (submitted by Andrew Johnson <ajohnson@gpu.srv.ualberta.ca>
Many thanks!).
- Also restructured the code a bit.
New in version 1.3:
- Made it pipeable so you can do: grepmail <pattern> file | grepmail <pattern>
New in version 1.2:
- Restructured the code a bit.
New in version 1.1:
- Support for dates.
New in version 1.0:
- Initial version, with -v -i, and gzip support