character replacemnt in a file

[sheevm]

Hi,

Can someone help me in how to replace globally certain characeters in a file?

Ex: I have several lines looks like

abcd <1?
abcd <20>
abcd <5>
abcd <100>

I want to make abcd <5000> for all the occurance of abcd, the value in <> is different.

Thanks for your help.
Rajim

[siva0123]

Hi ,

Open the file in vi editor
Go to command mode and
Then use the following the syntax
:$s/str1/str2/g

str1 ==>the string pattern to be replaced
str2 ==>The string which has to be there


Please make sure there is no space in the command

[sheevm]

hI,

But str2 has different value on each occurence as you see in my example. Is there a way to use the wild card

abcd <*> for abcd <1> and abcd <20> etc...?

Thanks

[Frank de Vries]

hello

Then I would do this:

in vi mode type :
1,$s/abcd <.*>/abcd <5000>/g

[Frank de Vries · Accepted solution]

sorry, I pressed submit to quickly,
I wanted to explain what it means.

:1,$ is from line 1 to last line

the .* between < > means any character

good luck

[Frank de Vries]

sorry, I pressed submit to quickly,
I wanted to explain what it means.

:1,$ is from line 1 to last line

the .* between < > means any character and any number of characters.

So that should fit <1> or <100> or <23000>
whatever , will be changed to <5000>

good luck

[Frank de Vries]

sorry, I pressed submit to quickly,
I wanted to explain what it means.

:1,$ is from line 1 to last line

the .* between < > means any character and any number of characters.

So that should fit <1> or <100> or <23000>
whatever , will be changed to <5000>

You do not have to repeat the abcd, I
did that for clarity:
just the matching pattern which I hope
in your case is nicely between the < >


good luck

[Sandman!]

# sed 's/^abcd <[0-9]*>/abcd <5000>/g' file

[sheevm]

Frank,

The command you gave did not work. Here is the sample file.

I want to repalce everything in ()


Column1 varchar2 (1),
Column2 varchar2 (14),
Column3 varchar2 (6),
Column4 varchar2 (4000),
Column5 varchar2 (4000),
Column6 varchar2 (10),
Column7 varchar2 (4000),
Column8 varchar2 (16),
Column9 varchar2 (6),
Column10 varchar2 (28),
Column11 varchar2 (4),
Column12 varchar2 (9),
Column13 varchar2 (1),
Column14 varchar2 (4000),
Column15 varchar2 (3),
Column16 varchar2 (4),

Thanks

[Sandman!]

Based on the sample input provided here's a sed(1) version that works:

# sed 's/^\(Column[0-9]*\) \(varchar2\) ([0-9]*)/\1 \2 (5000)/p' file

~hope it helps

[Frank de Vries]

Hi Rajim
Sorry I was away for a long weekend.
I suspect you have your solution already,
but just to complete my answer to what
I would have done.

Yes, naturally I base my command on <>,
when replacing () it becomes a different
ballgame:

If you want to use it with round brackets,
do this:

1,$s!char2 (.*)!char2 (5000)!g

Just to illustrate the different effects:
If you just did

1,$s!(.*)!(5000)!g

It would replace the whole line with (5000).
The brackets are a bit special.