Unix administration - Finding all e-mail addresses in files on server.

Author

Finding all e-mail addresses in files on server.

ifoutch@gmail.com

2005-04-03, 8:49 pm

Looking for advice and/or examples on how to audit a server to find all
possible e-mail addresses listed in files. Need to verify and possibly
consolidate addresses for server monitoring and administration. What I
have been doing is greping thru the filesystem and checking for
matches to entries in /etc/mail/aliases, /etc/password and any lines
containing "mail" (case insensitive) or "@" in crontabs, and scripts.

What I would like to do is create an exclude list of certain
directories. Then go through the remaining dirs. and extract mail
addresses and print them to a file with the filename. This would be run
on a production systems during off hours but still needs to be "nice"
to system resources. I want to automate this as much as possible. So
that we can periodically do an audit of all our systems.

Has anybody done anything like this previously? Any pointers or advice
greatly appreciated.

Alan Connor

2005-04-04, 2:48 am

On comp.unix.shell, in
<1112576768.533990.100930@z14g2000cwz.googlegroups.com>,
"ifoutch@gmail.com" wrote:

> Looking for advice and/or examples on how to audit a server
> to find all possible e-mail addresses listed in files. Need
> to verify and possibly consolidate addresses for server
> monitoring and administration. What I have been doing is
> greping thru the filesystem and checking for matches to entries
> in /etc/mail/aliases, /etc/password and any lines containing
> "mail" (case insensitive) or "@" in crontabs, and scripts.
>
> What I would like to do is create an exclude list of certain
> directories. Then go through the remaining dirs. and extract
> mail addresses and print them to a file with the filename. This
> would be run on a production systems during off hours but still
> needs to be "nice" to system resources. I want to automate this
> as much as possible. So that we can periodically do an audit of
> all our systems.
>
> Has anybody done anything like this previously? Any pointers or
> advice greatly appreciated.
>

I ran a search at http://groups.google.com/advanced_group_search

on ifoutch@gmail.com

This is the only post ever made on a Usenet textgroup under that
alias.

If you feel the need to hide, and want to extract email addresses
from a bunch of dirs, then you are probably a spammer. Or a
wannabee spammer.

Here's the script I wrote that does everything you want. All the
numbers and letters have been converted to Os with my handy
vi macro.

I learned some good things putting it together. Thanks.

And XXXX off.

(OOO OOO OOO OOOO)

#!/OOO/OOOO

OOO _OOOO OO $(OOOO / -OOOO "/OOO" -O -OOOO "/OOOO" -O \
-OOOO "/OOOO"\) -OOOOO -O -OOOO O -OOOOO)

do

OOO -O 'O/$.*$$[O-OO-OO-O_.+-]+@[O-OO-OO-O_.+-\
]+\.[O-OO-OO-O]+$$.*$/\O/O' $_OOOO >> OOOOOOOOOO

OOOO

#OO OOOO OO OOO OOO OOOO OOOOO OOO OOO OOOO \
#(OOO,OOOO...) OO OOO OOOO OOOOOO.

OOOO O

AC

ifoutch@gmail.com

2005-04-04, 2:48 am

I have posted as ifoutch@yahoo.com and possibly even as
ifoutch@hotmail.com. I am a newbie SA trying to find my way in my new
job.

The reason for needing to extract mail addresses is that different
"admins" have created scripts that functioned as some sort of monitor
and notification system. The problem is that the notification may not
be going to an address that will be of any use anymore. So I am trying
to track these down and either retire the scripts or change the address
to something appropriate.

-- Ian Foutch
P.S. Nice manners you have when addressing someone you know nothing
about!

Alan Connor

2005-04-04, 2:48 am

On comp.unix.shell, in
<hC14e.2758$EE2.2426@newsread2.news.pas.earthlink.net>, "Alan
Connor" wrote:

> On comp.unix.shell, in
><1112576768.533990.100930@z14g2000cwz.googlegroups.com>,
>"ifoutch@gmail.com" wrote:
>
>
>
>
> I ran a search at
> http://groups.google.com/advanced_group_search
>
> on ifoutch@gmail.com
>
> This is the only post ever made on a Usenet textgroup under
> that alias.
>
> If you feel the need to hide, and want to extract email
> addresses from a bunch of dirs, then you are probably a
> spammer. Or a wannabee spammer.
>
> Here's the script I wrote that does everything you want. All
> the numbers and letters have been converted to Os with my handy
> vi macro.
>
> I learned some good things putting it together. Thanks.
>
> And XXXX off.
>
>
> (OOO OOO OOO OOOO)
>
> #!/OOO/OOOO
>
> OOO _OOOO OO $(OOOO / -OOOO "/OOO" -O -OOOO "/OOOO" -O \
> -OOOO "/OOOO"\) -OOOOO -O -OOOO O -OOOOO)
>
> OO
>
> OOO -O 'O/$.*$$[O-OO-OO-O_.+-]+@[O-OO-OO-O_.+-\
> ]+\.[O-OO-OO-O]+$$.*$/\O/O' $_OOOO >> OOOOOOOOOO
>
> OOOO
>
> #OO OOOO OO OOO OOO OOOO OOOOO OOO OOO OOOO \
> #(OOO,OOOO...) OO OOO OOOO OOOOOO.
>
> OOOO O
>
> AC
>
>

Oops! Made a booboo. Used eregexes where they won't work.

#!/OOO/OOOO

OOO _OOOO OO $(OOOO / -OOOO "/OOO" -O -OOOO "/OOOO" -O \
-OOOO "/OOOO"\) -OOOOO -O -OOOO O -OOOOO)

OO

OOO -O 'O/$.*$$[O-OO-OO-O_.+-]\{O,\}@[O-OO-OO-O_.+-\
]\{O,\}\.[O-OO-OO-O]\{O,\}$$.*$/\O/O' $_OOOO >> OOOOOOOOOO

OOOO

#OO OOOO OO OOO OOO OOOO OOOOO OOO OOO OOOO \
#(OOO,OOOO...) OO OOO OOOO OOOOOO.

OOOO O

There. Perfect! Just fill in the dirs you don't want searched
and run the script and it will produce a lovely list of
email addresses, one per line, no garbage.

Hope this helps,

AC

Alan Connor

2005-04-04, 2:48 am

On comp.unix.shell, in
<1112582256.601080.108550@z14g2000cwz.googlegroups.com>,
"ifoutch@gmail.com" wrote:

> I have posted as ifoutch@yahoo.com and possibly even as
> ifoutch@hotmail.com. I am a newbie SA trying to find my way in
> my new job.
>
> The reason for needing to extract mail addresses is that
> different "admins" have created scripts that functioned as
> some sort of monitor and notification system. The problem is
> that the notification may not be going to an address that will
> be of any use anymore. So I am trying to track these down and
> either retire the scripts or change the address to something
> appropriate.
>
> -- Ian Foutch P.S. Nice manners you have when addressing
> someone you know nothing about!
>

That's the problem, fool: I don't know anything about you.

Except that you want to extract email addresses from a bunch
of files. Which is what spammers do.

Like whether you tell the truth or not on the Usenet.

And I can't search your posting history to find out, and
have no idea whether you have anything to do with any
formal organization at all.

What company do you work for? What's your name so that I
can ask them if you work there?

Are you ashamed of them? Is that why you don't use the name?

What's wrong with the RCN Corporation in Princeton N.J.?

Do they know that you are using their computers or connection?

What I do know about you is that you are hiding by changing
your alias regularly and are evasive about even that.

Like I said: XXXX Off.

AC

Alan Connor

2005-04-04, 2:48 am

On comp.unix.shell, in
<1112576768.533990.100930@z14g2000cwz.googlegroups.com>,
"ifoutch@gmail.com" wrote:

I just sent this whole thread, including all headers, to
noc@rcn.com

You had better be legit.

If you are, I'm not going to apologize because you should have
said what company you worked for and used your real name and
work email address.

I just can't figure out why _I_ would know how to do what you
want, almost off the top of my head, being only an amateur linux
user, while you, a "sysadmin", has to ask someone how to do it.

Makes no sense.

I suspect that you have gained the root password at your
workplace (or someone else's workplace) illicitly, and are doing
some very un-authorized snooping.

You sure ran like a rabbit when you found out I knew where you
were posting from. Or _through_.

Kind of late for someone in NJ to be working. Especially on
a Sunday night....

Are you a windoze cracker that finds himself lost on a
nix system?

A janitor there?

AC

Bit Twister

2005-04-04, 2:48 am

On Mon, 04 Apr 2005 03:48:18 GMT, Alan Connor wrote:
>
> If you are, I'm not going to apologize because you should have
> said what company you worked for and used your real name and
> work email address.

There is no good reason to do so.
Far too many idiots out in Usenet land.

> I just can't figure out why _I_ would know how to do what you
> want, almost off the top of my head, being only an amateur linux
> user, while you, a "sysadmin", has to ask someone how to do it.
>
> Makes no sense.

Unless you are out in the real world. Maybe the senior sysadmin was
layed off and the second team was thrown into the breach. Or
management decided to outsource the IT department and the new people
are having to scramble to understand what they signed up for.

> I suspect that you have gained the root password at your
> workplace (or someone else's workplace) illicitly, and are doing
> some very un-authorized snooping.

Well if he a totally clueless as you indicate and has managed to
gained root access, him searching for email addresses
are the least of their worries.

> Kind of late for someone in NJ to be working. Especially on
> a Sunday night....

See, you realy are totaly clueless.
When do you think sysadmins get to do whatever they like on production
systems. It sure as h*11 is not 9am Monday morning.

Friday is also a good massive layoff day to allow support people to disable
accouts, change ownership of file, copying employee files into
management directories, ...... Depending on the size of the layoff
cleanup work can still be going on come Monday morning.

Chris F.A. Johnson

2005-04-04, 2:48 am

On Mon, 04 Apr 2005 at 02:37 GMT, ifoutch@gmail.com wrote:
> I have posted as ifoutch@yahoo.com and possibly even as
> ifoutch@hotmail.com. I am a newbie SA trying to find my way in my new
> job.
>
> The reason for needing to extract mail addresses is that different
> "admins" have created scripts that functioned as some sort of monitor
> and notification system. The problem is that the notification may not
> be going to an address that will be of any use anymore. So I am trying
> to track these down and either retire the scripts or change the address
> to something appropriate.
>
> -- Ian Foutch
> P.S. Nice manners you have when addressing someone you know nothing
> about!

Don't bother responding to our resident kook. Most people here have
killfiled him.

--
Chris F.A. Johnson http://cfaj.freeshell.org/shell
========================================
===========================
Everything in moderation -- including moderation

ifoutch@gmail.com

2005-04-04, 2:48 am

Thank you BT and CJ... Actually Chris has offered extremely useful
solutions to a few of my previous posts in the past and I've much
appreciated it. I even pre-ordered his book, still waiting

I ended up an SA because I showed some abilties with my shell scripting
while working as a NOC operator. Unfortunately without any previous
experience, schooling or training. The job is not as easy I would like.
Every task I get I have to learn it as I go and much of that ends up
during the evening on my own time.

This task sounds easy enough except that it gets messy really quick and
I dislike messy scripts. I can figure out how to use grep, find, sed,
xargs, awk, cut and whatever else need. What I am looking for is
pointers in the "best" way to do this and any advice on what I might
likely overlook.

I keep telling myself i will join a LUG in my home town (not NJ by the
way!) so that I might find a couple mentors/peers to work with. Looks
like that has become a high priority todo. People have much nicer
manners when they are face to face and they need to be concerned about
the condition of such after an exchange.

-- Ian

Bit Twister

2005-04-04, 2:48 am

On 3 Apr 2005 22:29:18 -0700, ifoutch@gmail.com wrote:
> Thank you BT and CJ... Actually Chris has offered extremely useful
> solutions to a few of my previous posts in the past and I've much
> appreciated it.

Yes Chris does help alot of people.

> I ended up an SA because I showed some abilties with my shell scripting
> while working as a NOC operator. Unfortunately without any previous
> experience, schooling or training. The job is not as easy I would like.

Been there done that.

> Every task I get I have to learn it as I go and much of that ends up
> during the evening on my own time.

Yes, I have take home lots of work or dialed in after getting home.

> This task sounds easy enough except that it gets messy really quick and
> I dislike messy scripts.

I find pushing the acutal work as far down into a subroutine/funmction helps on
messy stuff.

> I can figure out how to use grep, find, sed, xargs, awk, cut and
> whatever else need. What I am looking for is pointers in the "best"
> way to do this and any advice on what I might likely overlook.

From what you have described, I would have created the list of files
to munge. Look at the lines to be munged, and see if a small
function/subroutine can make the changes.

Then use a loop to read the list to process files into a tmp workspace
and check the results. If they are valid, backup the production
scripts into a fallback directory, and install the new ones. If
production throws up, run the fallback script to copy fallback scripts
back into production.

No way would I run a script to mass update a bunch of scripts on a
production box with my fingers crossed.

> People have much nicer manners when they are face to face and they
> need to be concerned about the condition of such after an exchange.

Never let anyone's post bother you. Watch them for awhile to see if
they are worth your time, kill file them if they are of no use.

Anytime I run into a mess/problem like yours, I'll put the variable
element into a config file, source it in a script, and use a variable in the
scripts.

As an application programmer, I put lots of tests in my scripts to
verify nodes I need, database listners, mount points
needed by my app were realy there. When a check failed, it would
create a message file, and call the nofity funcition with who to
contact, subject, filename of problem description.
Example email variable names: $dba, $sysops, $crash, $coder, $mgr,
$sysadmin, $fyi...

Notify function would then add in node, time, script name, application
and call mailx -s "$2" $1 < $3

That reduced the O dark thirty calls to me when my production scripts
aborted because of external failures.
Also allowed me to run the scripts on the development and test
bed systems without bothering the production group by changing names
in my config file.

Laurenz Albe

2005-04-04, 7:49 am

In comp.unix.shell ifoutch@gmail.com <ifoutch@gmail.com> wrote:
> Looking for advice and/or examples on how to audit a server to find all
> possible e-mail addresses listed in files. Need to verify and possibly
> consolidate addresses for server monitoring and administration. What I
> have been doing is greping thru the filesystem and checking for
> matches to entries in /etc/mail/aliases, /etc/password and any lines
> containing "mail" (case insensitive) or "@" in crontabs, and scripts.
>
> What I would like to do is create an exclude list of certain
> directories. Then go through the remaining dirs. and extract mail
> addresses and print them to a file with the filename.

You have received flames and encouragement, but no answers.

Maybe you can use a find command, something like:

find / -path /first/excluded -prune -o -path /second/excluded -prune \
-o -type f -exec egrep -li '([^ @]+@[^ @]+|mail)' {} \;

Yours,
Laurenz Albe

Constantin Wiemer

2005-04-04, 7:49 am

ifoutch@gmail.com wrote:
> Looking for advice and/or examples on how to audit a server to find all
> possible e-mail addresses listed in files. Need to verify and possibly
> consolidate addresses for server monitoring and administration. What I
> have been doing is greping thru the filesystem and checking for
> matches to entries in /etc/mail/aliases, /etc/password and any lines
> containing "mail" (case insensitive) or "@" in crontabs, and scripts.
>
> What I would like to do is create an exclude list of certain
> directories. Then go through the remaining dirs. and extract mail
> addresses and print them to a file with the filename. This would be run
> on a production systems during off hours but still needs to be "nice"
> to system resources. I want to automate this as much as possible. So
> that we can periodically do an audit of all our systems.
>
> Has anybody done anything like this previously? Any pointers or advice
> greatly appreciated.
>

This may not help you in the short term but I find PERL very useful for
tasks like that. I think the time you have to spend to learn PERL is
well spent as a system administrator. For starters you can have a look
at www.perl.org.

Hope this helps,

Constantin Wiemer

c0ldbyte

2005-04-04, 5:58 pm

On 2005-04-04, Alan Connor <zzzzzz@xxx.yyy> wrote:
> On comp.unix.shell, in
><1112576768.533990.100930@z14g2000cwz.googlegroups.com>,
> "ifoutch@gmail.com" wrote:
>
>
>
> I ran a search at http://groups.google.com/advanced_group_search
>
> on ifoutch@gmail.com
>
> This is the only post ever made on a Usenet textgroup under that
> alias.
>
> If you feel the need to hide, and want to extract email addresses
> from a bunch of dirs, then you are probably a spammer. Or a
> wannabee spammer.
>
> Here's the script I wrote that does everything you want. All the
> numbers and letters have been converted to Os with my handy
> vi macro.
>
> I learned some good things putting it together. Thanks.
>
> And XXXX off.
>
>
> (OOO OOO OOO OOOO)
>
> #!/OOO/OOOO
>
> OOO _OOOO OO $(OOOO / -OOOO "/OOO" -O -OOOO "/OOOO" -O \
> -OOOO "/OOOO"\) -OOOOO -O -OOOO O -OOOOO)
>
> do
>
> OOO -O 'O/$.*$$[O-OO-OO-O_.+-]+@[O-OO-OO-O_.+-\
> ]+\.[O-OO-OO-O]+$$.*$/\O/O' $_OOOO >> OOOOOOOOOO
>
> OOOO
>
> #OO OOOO OO OOO OOO OOOO OOOOO OOO OOO OOOO \
> #(OOO,OOOO...) OO OOO OOOO OOOOOO.
>
> OOOO O
>
> AC
>

LOL I like that hahahahaha.

--
( When in doubt, use brute force. -- Ken Thompson 1998 )

c0ldbyte

2005-04-04, 5:58 pm

On 2005-04-04, Bit Twister <BitTwister@mouse-potato.com> wrote:
> On 3 Apr 2005 22:29:18 -0700, ifoutch@gmail.com wrote:
>
> Yes Chris does help alot of people.
>
>
> Been there done that.
>
>
> Yes, I have take home lots of work or dialed in after getting home.
>
>
>
> I find pushing the acutal work as far down into a subroutine/funmction helps on
> messy stuff.
>
>
> From what you have described, I would have created the list of files
> to munge. Look at the lines to be munged, and see if a small
> function/subroutine can make the changes.
>
> Then use a loop to read the list to process files into a tmp workspace
> and check the results. If they are valid, backup the production
> scripts into a fallback directory, and install the new ones. If
> production throws up, run the fallback script to copy fallback scripts
> back into production.
>
> No way would I run a script to mass update a bunch of scripts on a
> production box with my fingers crossed.
>
>
> Never let anyone's post bother you. Watch them for awhile to see if
> they are worth your time, kill file them if they are of no use.
>
> Anytime I run into a mess/problem like yours, I'll put the variable
> element into a config file, source it in a script, and use a variable in the
> scripts.
>
> As an application programmer, I put lots of tests in my scripts to
> verify nodes I need, database listners, mount points
> needed by my app were realy there. When a check failed, it would
> create a message file, and call the nofity funcition with who to
> contact, subject, filename of problem description.
> Example email variable names: $dba, $sysops, $crash, $coder, $mgr,
> $sysadmin, $fyi...
>
> Notify function would then add in node, time, script name, application
> and call mailx -s "$2" $1 < $3
>
> That reduced the O dark thirty calls to me when my production scripts
> aborted because of external failures.
> Also allowed me to run the scripts on the development and test
> bed systems without bothering the production group by changing names
> in my config file.

Sounds to me like you could use a system like GNATS: http://www.gnu.org/software/gnats/ with a nice send-pr.

--
( When in doubt, use brute force. -- Ken Thompson 1998 )

Bit Twister

2005-04-04, 5:58 pm

On Mon, 04 Apr 2005 13:21:38 GMT, c0ldbyte wrote:
>
> Sounds to me like you could use a system like GNATS:
> http://www.gnu.org/software/gnats/ with a nice send-pr.

I hear you talking, but in strict control environment, sysopts would
windup in control on nats and I would not be able to make a change in
rules until the second Sunday of the month, after the change control
board approved my change submitted 7 days in advance of the change
date. :-(

Much easier to submit a change to copy in a no impact config file for
10pm.

Greg Beeker

2005-04-04, 5:58 pm

Chris F.A. Johnson wrote:
> On Mon, 04 Apr 2005 at 02:37 GMT, ifoutch@gmail.com wrote:
new[vbcol=seagreen]
monitor[vbcol=seagreen]
not[vbcol=seagreen]
trying[vbcol=seagreen]
address[vbcol=seagreen]
>
> Don't bother responding to our resident kook. Most people here
have
> killfiled him.

I finally figured that out. Alan Conner = AC = Anonymous Coward.
I learned that one from the FAQ on slashdot.
"Don't feed the trolls"

>
> --
> Chris F.A. Johnson
http://cfaj.freeshell.org/shell
>
========================================
===========================
> Everything in moderation -- including moderation

Kevin Collins

2005-04-04, 5:58 pm

In article <2124e.4754$x4.4607@newsread1.news.pas.earthlink.net>, Alan Connor
wrote:
> On comp.unix.shell, in
><1112582256.601080.108550@z14g2000cwz.googlegroups.com>,
> "ifoutch@gmail.com" wrote:
>
>
>
> That's the problem, fool: I don't know anything about you.
>
> Except that you want to extract email addresses from a bunch
> of files. Which is what spammers do.
>
> Like whether you tell the truth or not on the Usenet.
>
> And I can't search your posting history to find out, and
> have no idea whether you have anything to do with any
> formal organization at all.
>
> What company do you work for? What's your name so that I
> can ask them if you work there?
>
> Are you ashamed of them? Is that why you don't use the name?
>
> What's wrong with the RCN Corporation in Princeton N.J.?
>
> Do they know that you are using their computers or connection?
>
> What I do know about you is that you are hiding by changing
> your alias regularly and are evasive about even that.
>
> Like I said: XXXX Off.
>
> AC

Is it just me, or have you been forgetting your meds more frequently?

Oh, and don't bother, I know - I've been killfiled, right?

Kevin

Kevin Collins

2005-04-04, 5:58 pm

In article <1112576768.533990.100930@z14g2000cwz.googlegroups.com>, ifoutch@gmail.com wrote:
> Looking for advice and/or examples on how to audit a server to find all
> possible e-mail addresses listed in files. Need to verify and possibly
> consolidate addresses for server monitoring and administration. What I
> have been doing is greping thru the filesystem and checking for
> matches to entries in /etc/mail/aliases, /etc/password and any lines
> containing "mail" (case insensitive) or "@" in crontabs, and scripts.
>
> What I would like to do is create an exclude list of certain
> directories. Then go through the remaining dirs. and extract mail
> addresses and print them to a file with the filename. This would be run
> on a production systems during off hours but still needs to be "nice"
> to system resources. I want to automate this as much as possible. So
> that we can periodically do an audit of all our systems.
>
> Has anybody done anything like this previously? Any pointers or advice
> greatly appreciated.
>

Since this topic (recursively searching) comes up a lot, I'm posting a shar
file (shell archive) of one of my favorite tools ever. This script called
'rgrep' uses PERL and is very quick and powerful. It uses Perl's regexp engine,
which is more powerful than a grep, too.

Save the text below as 'rgrep.shar' and then run "sh rgrep.shar". It will
create rgrep. Run 'rgrep -h' for help.

Before I post it, I should mention that searching for email addresses sounds
like a trivial task, but see the O'Reilly book 'Mastering Regular Expressions'
for the regexp that actually matches all valid email addresses per the RFCs -
its just under 6600 characters! It might be simpler to search for all the mail
programs such as sendmail, mail, mailx, pine, elm, etc...

Kevin

--- Cut Here ---
#!/bin/sh
# This is a shell archive (produced by GNU sharutils 4.2.1).
# To extract the files from this archive, save it to some FILE, remove
# everything before the `!/bin/sh' line above, then type `sh FILE'.
#
# Made on 2005-04-04 14:08 PDT by <cokm@cpafisxw>.
# Source directory was `/util/bin'.
#
# Existing files will *not* be overwritten unless `-c' is specified.
#
# This shar contains:
# length mode name
# ------ ---------- ------------------------------------------
# 6119 -rwxrwxr-x rgrep
#
save_IFS="${IFS}"
IFS="${IFS}:"
gettext_dir=FAILED
locale_dir=FAILED
first_param="$1"
for dir in $PATH
do
if test "$gettext_dir" = FAILED && test -f $dir/gettext \
&& ($dir/gettext --version >/dev/null 2>&1)
then
set `$dir/gettext --version 2>&1`
if test "$3" = GNU
then
gettext_dir=$dir
fi
fi
if test "$locale_dir" = FAILED && test -f $dir/shar \
&& ($dir/shar --print-text-domain-dir >/dev/null 2>&1)
then
locale_dir=`$dir/shar --print-text-domain-dir`
fi
done
IFS="$save_IFS"
if test "$locale_dir" = FAILED || test "$gettext_dir" = FAILED
echo=echo
else
TEXTDOMAINDIR=$locale_dir
export TEXTDOMAINDIR
TEXTDOMAIN=sharutils
export TEXTDOMAIN
echo="$gettext_dir/gettext -s"
fi
if touch -am -t 200112312359.59 $$.touch >/dev/null 2>&1 && test ! -f 200112312359.59 -a -f $$.touch; then
shar_touch='touch -am -t $1$2$3$4$5$6.$7 "$8"'
elif touch -am 123123592001.59 $$.touch >/dev/null 2>&1 && test ! -f 123123592001.59 -a ! -f 123123592001.5 -a -f $$.touch; then
shar_touch='touch -am $3$4$5$6$1$2.$7 "$8"'
elif touch -am 1231235901 $$.touch >/dev/null 2>&1 && test ! -f 1231235901 -a -f $$.touch; then
shar_touch='touch -am $3$4$5$6$2 "$8"'
else
shar_touch=:
echo
$echo 'WARNING: not restoring timestamps. Consider getting and'
$echo "installing GNU \`touch', distributed in GNU File Utilities..."
echo
fi
rm -f 200112312359.59 123123592001.59 123123592001.5 1231235901 $$.touch
#
if mkdir _sh15995; then
$echo 'x -' 'creating lock directory'
else
$echo 'failed to create lock directory'
exit 1
fi
# ============= rgrep ==============
if test -f 'rgrep' && test "$first_param" != -c; then
$echo 'x -' SKIPPING 'rgrep' '(file already exists)'
else
$echo 'x -' extracting 'rgrep' '(text)'
sed 's/^X//' << 'SHAR_EOF' > 'rgrep' &&
#!/usr/bin/perl
X
die "Usage: rgrep [-iredblL] regexp filepat ...\n rgrep -h for help\n"
X if $#ARGV < $[;
X
# Written by Piet van Oostrum <piet@cs.ruu.nl>
# This is really free software
X
# added "-n" option for name-only
X
$nextopt = 1;
$igncase = '';
$regpat = 0;
$links = 0;
$error = 0;
$skipbin = 1;
$debug = 0;
$nameonly = 0;
X
do { $regexp = shift (@ARGV); } while &checkopt ($regexp);
$icreg = $igncase;
$igncase = '';
X
eval 'sub grep_file {
X while (<F> ) {
X $ln++;
X if (/$regexp/o' . $icreg .') {
X if (! $nameonly)
X {
X print "$file:$ln:$_";
X print "\n" if substr($_, -1, 1) ne "\n";
X }
X else
X {
X print "$file\n";
X last;
X }
X }
X }
}';
X
for (@ARGV) {
X if (! &checkopt ($_)) {
X if ($igncase || $regpat || /[?*[]/ || ! -e) {
X if ($regpat) {
X s/#/\\#/g;
X $_ = "#$_#";
X } else { # translate File pattern into regexp
X $re = '#($|/)'; $save = $_;
X while (/[[*?+()|.^$#]/) {
X $re .= $`;
X $c = $&;
X $_ = $';
X if ($c eq '*') { $c = '[^/]*'; }
X elsif ($c eq '?') { $c = '[^/]'; }
X elsif ($c eq '[') {
X if (/.\]/) { $c = "[$`$&"; $_ = $'; }
X else {
X $error++;
X printf stderr "Illegal filepattern %s\n", $save;
X }
X } else { $c = "\\$c"; }
X $re .= $c;
X }
X $_ = "$re$_\$#$igncase";
X }
X print "filepat: $_\n" if $debug;
X push (@filepat, $_);
X }
X else { push (@files, $_); print "file: $_\n" if $debug; }
X }
}
X
exit 1 if $errors ;
X
if ($#filepat < $[) {
n_pat {1;}" ;
}
else {
X $subtxt = 'sub in_pat { local ($f) = @_;';
X $or = "";
X for (@filepat) {
X $subtxt .= $or . '$f =~ m' . $_;
X $or = " || ";
X }
X $subtxt .= ';};1';
X
X if (! eval $subtxt) {
X print $@;
X exit 1;
X }
}
X
@files = (".") if $#files < $[;
X
for $file (@files) {
X &do_grep ($file);
}
X
sub do_grep {
X local ($file) = @_;
X local (*F, $ln, $f, $g, @dirfiles);
X if (-f $file) {
X if (open (F, $file)) {
X if (-B $file) { # binary file -- may be compressed/compacted
X if (($cx1 = getc(F)) eq "\377" && (getc(F) eq "\037")) {
X open (F, "uncompact < $file|");
X if ($skipbin && -B $file) { close (F); return; }
X }
X elsif ($cx1 eq "\037" && (getc(F) eq "\235")) {
X open (F, "uncompress < $file|");
X if ($skipbin && -B $file) { close (F); return; }
X }
X elsif ($skipbin) {
X close (F); return;
X }
X }
X print "Reading $file\n" if $debug;
X &grep_file;
X } else {
X print stderr "Cannot open $file\n";
X }
X }
X elsif (-d $file) {
X print "Entering $file\n" if $debug;
X if (opendir (F, $file)) {
X @dirfiles = readdir (F);
X closedir (F);
X for $f (@dirfiles) {
X next if ($f eq '.' || $f eq '..');
X $g = "$file/$f";
X next if (-l $g && ($links < 1 || $links == 1 && -d $g));
X if (-f $g && &in_pat ($g) || -d _) {
X &do_grep ($g);
X }
X }
X } else {
X print stderr "Can't open $file\n";
X }
X }
}
X
sub checkopt {
X local ($_) = $_[0];
X if (/^-/ && $nextopt) {
X $nextopt = 1;
X @opt = split (/-*/,$_); shift (@opt);
X for $opt (@opt) {
X if ($opt eq 'i') { $igncase = 'i'; }
X elsif ($opt eq 'd') { $debug = 1; }
X elsif ($opt eq 'l') { $links = 1; }
X elsif ($opt eq 'L') { $links = 2; }
X elsif ($opt eq 'b') { $skipbin = 0; }
X elsif ($opt eq 'r') { $regpat = 1; }
X elsif ($opt eq 'e') { $nextopt = 0; }
X elsif ($opt eq 'n') { $nameonly = 1; }
X elsif ($opt eq 'h' || $opt eq 'H') { & help; }
X else { $error++; printf stderr "Unknown option -%s\n", $opt; }
X }
X return 1;
X }
X $nextopt = 1;
X return 0;
}
X
sub help {
X print <<'HELP'; exit 0;
Usage: rgrep [-iredblLn] regexp filepat ...
X regexp = PERL regular expression to search
X filepat ... = a list of files and directories to be searched or
X file patterns to match filenames.
X filepat will be interpreted as file or directory name if it exists
X as such, and does not contain the metacharacters [ ] ? or *. After
X the options -i and -r all filepats will be considered patterns.
X rgrep will search all files in any of the directories given (and its
X subdirectories) that match any of the filepats, except binary files.
X Compressed files will be searched in uncompressed form.
X Note: filepats may contain / contrary to find usage.
X -b Don't skip binary files.
X -i Ignore case, either in the regexp or in filename matching (depending
X on the location). Before the regexp only applies to the regexp,
X otherwise to the filepats following it.
X -r The following filepats are treated as real PERL regexps rather than
X shell style filename patterns. In this case / is not a special
X character, i.e. it is matched by . and matching is not anchored (you
X must supply ^ and $ yourself). E.g. a.b matches the file /xa/by/zz.
X -l Do follow symbolic links only for files (default is do not follow).
X -L Do follow symbolic links for files and directories.
X -e Do not interpret following argument as option. Useful if regexp or
X filepat starts with a -.
X -d Debugging: Give a lot of output on what happens.
X -n print only the filename of matching files.
X -h print this message and exit.
Piet van Oostrum <piet@cs.ruu.nl>
HELP
}
SHAR_EOF
(set 20 04 01 21 13 58 13 'rgrep'; eval "$shar_touch") &&
chmod 0775 'rgrep' ||
$echo 'restore of' 'rgrep' 'failed'
if ( md5sum --help 2>&1 | grep 'sage: md5sum \[' ) >/dev/null 2>&1 \
&& ( md5sum --version 2>&1 | grep -v 'textutils 1.12' ) >/dev/null; then
md5sum -c << SHAR_EOF >/dev/null 2>&1 \
|| $echo 'rgrep:' 'MD5 check failed'
1eb760efd19050c0bc7911504e83dc45 rgrep
SHAR_EOF
else
shar_count="`LC_ALL= LC_CTYPE= LANG= wc -c < 'rgrep'`"
test 6119 -eq "$shar_count" ||
$echo 'rgrep:' 'original size' '6119,' 'current size' "$shar_count!"
fi
fi
rm -fr _sh15995
exit 0