[loren s]

silly question: with grep how could I delete all lines of text that contain a word like cat ? In apple script ?

[benb]

Why not use perl?

Code:
perl -i.bak -p -e 's/search/replace/g;' *.html

This will work for all files in the directory with the extension html. Change the extension to what file types you need. It will also create backups of the original files with the suffix .bak.

To search multiple directories use

Code:
find /directory/to/search/ -type f | xargs perl -i.bak -p -e 's/search/replace/g;' *.html

[benb]

Funny you should ask this, too. Because not just five minutes ago I did the same thing to batch change some URLs on my website.

[loren s]

because my text looks like this

Code:
005  9022_B29 V     D    010 01:06:03:13 01:06:03:23 01:00:03:28 01:00:04:08
M2   9022_B29       090.5                01:06:03:13
* FROM CLIP NAME:  QUEBRADA 1KM LOADING GEAR.03
* TO CLIP NAME:  LENS FLARE SUN.02
* REPAIR: TO SOURCE TRUE SPEED IS 90.483871 FPS
006  9022_B29 V     C        01:06:04:13 01:06:04:13 01:00:04:08 01:00:04:08

and every entire line I have that has " Repair" in it needs be be removed. If it can be done in perl I will be happy, does not matter as long as it works with applescript.

[benb]

Originally posted by loren s:
because my text looks like this

Code:
005  9022_B29 V     D    010 01:06:03:13 01:06:03:23 01:00:03:28 01:00:04:08
M2   9022_B29       090.5                01:06:03:13
* FROM CLIP NAME:  QUEBRADA 1KM LOADING GEAR.03
* TO CLIP NAME:  LENS FLARE SUN.02
* REPAIR: TO SOURCE TRUE SPEED IS 90.483871 FPS
006  9022_B29 V     C        01:06:04:13 01:06:04:13 01:00:04:08 01:00:04:08

and every entire line I have that has " Repair" in it needs be be removed. If it can be done in perl I will be happy, does not matter as long as it works with applescript.

Alright, I understand now. Use this then:

Code:
perl -ni -e 'print unless m/search/' *.txt

This tells pearl to read lines and search for your text, then delete the lines that contain your search string. You can pipe this through xargs from a find as well.

[benb]

So you would need something like

Code:
find /directory/ type -f | xargs perl -ni -e 'print unless m/REPAIR/' *.txt

[loren s]

like this ?

Code:
[filemaker:~/Public/akm/test] filemake% cd /Users/filemake/Public/akm/testfind /Users/filemake/Public/akm/test/ type -f | okyu perl -ni -e 'print unless m/REPAIR/' *.txt
tcsh: okyu: No match.
tcsh: cd: Too many arguments.

[filemaker:~/Public/akm/test] filemake% 
[filemaker:~/Public/akm/test] filemake%

I was tring this in terminal, but I had a working script in applescript that would edit the open file of BBedit.

like so

tell application "BBEdit"
activate

replace "^TITLE:.*\\r|^FCM:.*\\r|^M2.*\\r|^\\n+\\*.*\\r|^\ \x1A\\x1A\\x1A\\x1A+.*" using "" searching in text 1 of text window 1 options {search mode:grep, starting at top:true, wrap around:false, backwards:false, case sensitive:false, match words:false, extend selection:false}
replace "V\\s+|C\\s+|D\\s+|K\\s+|B\\s+" using "" searching in text 1 of text window 1 options {search mode:grep, starting at top:true, wrap around:false, backwards:false, case sensitive:false, match words:false, extend selection:false}
replace " \\d\\d\\d " using "" searching in text 1 of text window 1 options {search mode:grep, starting at top:true, wrap around:false, backwards:false, case sensitive:false, match words:false, extend selection:false}
replace " \\([A-Z]\\) " using "" searching in text 1 of text window 1 options {search mode:grep, starting at top:true, wrap around:false, backwards:false, case sensitive:false, match words:false, extend selection:false}
replace "\\sO\\s" using "" searching in text 1 of text window 1 options {search mode:grep, starting at top:true, wrap around:false, backwards:false, case sensitive:false, match words:false, extend selection:false}

replace "([:\\d\\d]) ([\\d\\d:])" using "\\1\\t\\2" searching in text 1 of text window 1 options {search mode:grep, starting at top:true, wrap around:false, backwards:false, case sensitive:false, match words:false, extend selection:false}
replace "( +)" using "\\t" searching in text 1 of text window 1 options {search mode:grep, starting at top:true, wrap around:false, backwards:false, case sensitive:false, match words:false, extend selection:false}
end tell

[larkost]

Why do you keep mentioning AppleScript... there are much better tools on MacOS X for this use. My throw into the hat would be sed:

Code:
cat myfile.txt | sed "/Repair/d" > mycleanedFile.txt

where myfile.txt and mycleanedFile.txt have to replaced with the proper names of the files you want.

[benb]

Originally posted by loren s:
like this ?

Code:
[filemaker:~/Public/akm/test] filemake% cd /Users/filemake/Public/akm/testfind /Users/filemake/Public/akm/test/ type -f | okyu perl -ni -e 'print unless m/REPAIR/' *.txt
tcsh: okyu: No match.
tcsh: cd: Too many arguments.

[filemaker:~/Public/akm/test] filemake% 
[filemaker:~/Public/akm/test] filemake%

Well, by using find you don't have to use cd.

Code:
find /Users/filemake/Public/akm/test/ type -f | xargs perl -ni -e 'print unless m/REPAIR/' *.txt

That should do it if that is the directory you want and the file type is .txt

[loren s]

Originally posted by larkost:
Why do you keep mentioning AppleScript... there are much better tools on MacOS X for this use. My throw into the hat would be sed:

Code:
cat myfile.txt | sed "/Repair/d" > mycleanedFile.txt

where myfile.txt and mycleanedFile.txt have to replaced with the proper names of the files you want.

I thought people liked apple script < I like it for the instant visual feedback. But it does not matter as long as it is quick and it works.

But doing this
cat okyu.txt | sed "/Repair/d" > mycleanedFile.txt
created a new file but did not clean up the text file,
* REPAIR: TO SOURCE TRUE SPEEIS 90.483871 FPS
does the : in REPAIR: mater ?

[loren s]

you know I have had a problem with sed before. Can you help me tell if sed is working on my machine ?
10.3.3. Is there a way to test and or repair it ?

[loren s]

Originally posted by benb:
Well, by using find you don't have to use cd.

Code:
find /Users/filemake/Public/akm/test/ type -f | xargs perl -ni -e 'print unless m/REPAIR/' *.txt

That should do it if that is the directory you want and the file type is .txt

Code:
[filemaker:~] filemake% find /Users/filemake/Public/akm/test/ type -f | okyu.txt perl -ni -e 'print unless m/REPAIR/' *.txt
tcsh: okyu.txt: No match.
find: -f: unknown expression primary
[filemaker:~] filemake% find /Users/filemake/Public/akm/test/ type -f | okyu perl -ni -e 'print unless m/REPAIR/' *.txt
tcsh: okyu: No match.
find: -f: unknown expression primary
[filemaker:~] filemake%

[benb]

Originally posted by loren s:

Code:
[filemaker:~] filemake% find /Users/filemake/Public/akm/test/ type -f | okyu.txt perl -ni -e 'print unless m/REPAIR/' *.txt
tcsh: okyu.txt: No match.
find: -f: unknown expression primary
[filemaker:~] filemake% find /Users/filemake/Public/akm/test/ type -f | okyu perl -ni -e 'print unless m/REPAIR/' *.txt
tcsh: okyu: No match.
find: -f: unknown expression primary
[filemaker:~] filemake%

Oops, I made a typo. It should be -type f

So:

Code:
find /Users/filemake/Public/akm/test/ -type f | xargs perl -ni -e 'print unless m/REPAIR/' *.txt

If the file you want to search and replace is okyu.txt use

Code:
perl -ni -e 'print unless m/REPAIR/' okyu.txt

[loren s]

Originally posted by benb:
Oops, I made a typo. It should be -type f

So:

Code:
find /Users/filemake/Public/akm/test/ -type f | xargs perl -ni -e 'print unless m/REPAIR/' *.txt

If the file you want to search and replace is okyu.txt use

Code:
perl -ni -e 'print unless m/REPAIR/' okyu.txt

This deleted all of the text in my file, glad I made a back up

No wait ! YAY tried again this time this worked ! yaY !Thankyou

[loren s]

alright one more please please

for peral or grep or sed,

how can I remove a complete line that begins with a patteren ? I need this one for later.

something like,
* ___ Text cats !

please and thankyou ^v^

[MrBS]

Quick simple solution

grep -v cats oldcrappyfile > newcoolfile

~BS

[MrBS]

Originally posted by loren s:
alright one more please please

for peral or grep or sed,

how can I remove a complete line that begins with a patteren ? I need this one for later.

something like,
* ___ Text cats !

please and thankyou ^v^

grep -v -G "^____ Text cats" oldcrappyfile > newcoolfile

(if you want the * escape it with a \ -G "^\*____") <- parenthesis denoting the end of a parenthetical comment, not syntax


~BS

[loren s]

Originally posted by MrBS:
grep -v -G "^____ Text cats" oldcrappyfile > newcoolfile

(if you want the * escape it with a \ -G "^\*____")


~BS

Sweet!

But how do I use the last line for the * ?
like this ?

grep -v "^\*____") okyu.txt > newcoolfile.txt

or
grep -v \ -G "^\*____") okyu.txt > newcoolfile.txt

with those I get when trying to remove a line with * . But anything else is prrrrrrfect .

[Doritoss-Computer:Jerry Trip/trinkets/test] ranch% grep -v "^\*____") okyu.txt > newcoolfile.txt
tcsh: Too many )'s.

[MrBS]

If you want to drop lines with the string cat at the beginning, just do this:

grep -v -G ^cat old > new

If you want the string to have a space like this kitty cat, do this:

grep -v -G "^kitty cat" old >new

If you are going to use special characters (of which * is one of), you'll need to
escape them by putting a \ in front, so I like *s (at the beginning of the line again) would be:

grep -v -G "^I like \*s" old > new

~BS

[loren s]

sweet happy !
grep -v \* okyu.txt > newcoolfile.txt

Thankyou ^v^ this saved me hours and days of work . !

[loren s]

More help please ....

I found these sites while trying to learn this tool. Or any regrex's for that matter.
http://unix.about.com/library/weekly/aa082000a.htm
http://www.microdot.co.uk/grep/

But I am still to stupid to get it all fixxed.

My most current method is driving me nuts.

the grep -v textstuff makes since now to me, but how can I use it to do somthing like

all selected lines that do not have this word find and replace all run of spaces with tabs.

so I would think it is
grep -v REPAIR \s \t oldfile> newhappyfile

but that will not work amoung the various other I tried.

FCM: DROP FRAME
002 BL V D 015 00:00:00:00 00:00:00:15 01:00:18:05 01:00:18:20
M2 5510 031.5 02:17:08:03
* FROM CLIP NAME: XCARET RECREATION A-NIGHT.05
* REPAIR: FROM SOURCE TRUE SPEED IS 31.489141 FPS
003 BL V C 00:00:00:00 00:00:00:00 01:00:19:29 01:00:19:29

I go cookoo waisting hours still tryinmg this stuff. Also is there enev one GUI program out there for this ? like a simple python something to test really fast if something works or not ?

[loren s]

pretty please with sugar as a flavor center core , yummy yum ^v^

[MrBS]

Grep isn't the right tool for editing a file. For cli I'd recommend vi (takes a bit to learn, but nice) or in the gui bbedit has a really nice find/replace functionality.

To do what you want in vi open the file

% vi filename

Then type this

:%s/old/new/g

and hit enter. "old" would be what you want to replace (regex... don't use / either) and "new" is what you want to replace old with.

Here's a nice regex tutorial.

~BS

[MrBS]

(oh yeah... if you want to save your file then type

: wq

and then hit enter)
~BS

[loren s]

Originally posted by MrBS:
Grep isn't the right tool for editing a file. For cli I'd recommend vi (takes a bit to learn, but nice) or in the gui bbedit has a really nice find/replace functionality.

To do what you want in vi open the file

% vi filename

Then type this

:%s/old/new/g

and hit enter. "old" would be what you want to replace (regex... don't use / either) and "new" is what you want to replace old with.

Here's a nice regex tutorial.

~BS

Aw what even more regex's I did not know of ! Boy coders like to make lots of stuff !

Ok I will try this, thankyou . ^v^