switch text

Rushlan Khan

Hello Friends

Lets say I have a document below with the following entries

Name: rushlan
account : 1234
user: abcd
Name: Tom
account: 5678
user: efgh

What i would like to do is replace account with user and vice versa. So my correct output will be :

Name: rushlan
account : abcd
user: 1234
Name: Tom
account: efgh
user: 5678

Thank you
Rush

PeterJones

@Rushlan-Khan ,

Making the assumptions that there is no indentation, and that the extra space between account and : is really there, and that you have windows CRLF newlines:

FIND = (?-s)account : (.*)\r\nuser: (.*)$
REPLACE = account : $2\r\nuser: $1
Search Mode = Regular expression
Replace All

The first set of parens says to make sure . doesn’t match newline; the second the old account name and puts it into capture-group1, the third matches the old user name and stores it into capture-group2. Then the replacement puts capture-group2 with account : prefix and capture-group1 with user: prefix.

----

Useful References

Terry R

@Rushlan-Khan said in switch text:

What i would like to do is replace account with user and vice versa. So my correct output will be :

Since you did not include the examples inside of a “code box” (see the pinned post “Please Read This Before Posting” and the subsequent “Formatting Forum Posts”) your real data may not be exactly as you have portrayed in the example. Nonetheless I have made an assumption that the space between account and the : was in error. This is a regular expression so search mode must be regular expression.

Find What:(?-s)(account:\x20*)(.+\R)(user:\x20*)(.+\R)
Replace With:${1}${4}${3}${2}

Terry

PS I see @PeterJones got there just before me. You will see we are both very similar, that the beauty of regular expressions. They can achieve the same results with different methods.

PeterJones

@Rushlan-Khan ,

until Terry’s reply, I didn’t notice that one account had a space before the colon and the other didn’t. If I change my assumption to account followed by 0 or 1 space followed by colon, and the replacement always needs to have 0 spaces between account and colon, then it becomes:

Find What = (?-s)account\h?: (.*)\r\nuser: (.*)$
Replace with = account: $2\r\nuser: $1

If you added extra groups to mine, like Terry used four groups, then you can preserve whether or not there’s a space in each account :

Find What = (?-s)(account\h?: )(.*)\r\n(user: )(.*)$
Replace with = $1$4\r\n$3$2

Please learn from this that giving an accurate representation of the data is critical when asking such questions, because small variations can change the answer. The more effort you put into asking a clean question, the better the answers will be.

Alan Kilborn

This post is deleted!

guy038

Hello, @rushlan-khan, @peterjones, @terry-R and All,

Here is a new regex variant which does not care about anything before the ending values to invert :

SEARCH (?xi) ^ ( account \W+ ) ( \d+ \R ) ( user \W+ ) ( \w+ \R )

REPLACE \1\4\3\2

Best Regards,

guy038

guy038

Hi, @rushlan-khan, @peterjones, @terry-R and All,

I did not do it on purpose but my regex S/R allows splitting text, with a line-break, between the names of the fields and their corresponding values !

For instance, with this INPUT text :

Name: rushlan
account :
 1234
user
: abcd
Name: Tom
account: 5678
user: efgh

and the regex S/R :

SEARCH (?xi) ^ ( account \W+ ) ( \d+ \R ) ( user \W+ ) ( \w+ \R )

REPLACE \1\4\3\2

it would correctly return this OUTPUT text :

Name: rushlan
account :
 abcd
user
: 1234
Name: Tom
account: efgh
user: 5678

Why : just because the regex syntax \W, equivalent to [^\w], means any character which is not a word char ! So this char may also be a line-break character ( \r or \n )

Best Regards,

guy038

Rushlan Khan

@guy038 WOW!! thank you so much . Let me absorb this, I will get a back to you if I have any questions.

Rushlan Khan

@Terry-R thanks for your help !! Sorry this is my first time here, I will read the rules for sure now.

Rushlan Khan

@PeterJones Thanks for your help !