Remove doubles from string

[freggel99]

Hi there,

I have a macro and everything what is found is stored in a string, is it possible to filter out the exact doubles/triples etc so if 99 is found twice (or more) it is not stored (and processed for later use) twice (or more times)

It would help me a lot if you could help me with this.

TIA

[Gintaras]

Unique items in a list

[freggel99]

Well this should work i think, but how can i implement this in a macro hock:

I mean in that topic you say 'create a member function' which i did but how can i tell my macro to process the data through this 'member function'?

I know i'm stupid

Thanks for your patience

[Gintaras]

See EXAMPLE in the function. For example, if the text is stored in variable s, the code will be

s.RemoveDuplicateLines

[freggel99]

When i try this and i try to compile i get a errorcode 4 unknown member?

Is there a certain place to store these 'member functions' is it possible to give an example of code with a member function which is used and called?

TIA

[ken gray]

did you name the member function "RemoveDuplicateLines"

[Gintaras]

Its name must be str.RemoveDuplicateLines
It can be anywhere.
The variable where the text is stored must be of type str.

str s
...
s.RemoveDuplicateLines

[freggel99]

I Called it RemoveDuplicateLines and the code i copy/pasted:

function [flags] ;;flags: 1 case insensitive

;EXAMPLE
;str s=
;;apple
;;apple
;;orange
;;apple
;;grape
;;orange
;
;s.RemoveDuplicateLines


if(!this.end("[]")) this+"[]"

str s ss
int i j
for i 0 1000000000
,j=s.getl(this i)
,if(j<0) break
,ss.from("[]" s "[]")
,this.findreplace(ss "[]" 8|(flags&1) "" j+s.len)

Something wrong with the procedure?

I placed the member function in the folder in which my macro also is.....

[ken gray]

in your macro, when you type your string variable (eg TestVar) and hit the "." can you see it at the bottom of the list that popsup?

[freggel99]

No it is not visible

[ken gray]

that means you didn't make it a member function correctly
did you put "str." in front of the "RemoveDuplicateLines" part?

[freggel99]

At first i did, later i changed the name several times. Now i made it again and created a new member function named RemoveDuplicateLines without str. in front of it.

[ken gray]

it has to have the "str." in front.

[freggel99]

I don't get it,

I created a new one simular to the first one and it works now...

KEN thanks for your help, i'm very greatfull

[ken gray]

the name of the new member function has to be "str.RemoveDuplicateLines"
that first part tells qm what type of var can use the function ie strings not integers.

[freggel99]

After a few days i had the time to try it but it didn't work, i think it is my stupidity or something what i simply don't understand but i do a findrx over a piece of text with something like this:

findrx(s "(10|20|25)(?:\.\d{1,3}){3}" 0 4 a)

the result is 10.10.10.1010.10.10.1020.20.20.2010.10.10.1025.25.25.25

And with this i do a test for each ip so if there are doubles/triples it is inefficient..

And doesn't give something the member function can split, my questions are:
- is it possible to exclude doubles from the result or
- filter the doubles from this result in a decent way?

Thanks for your time in advance

Freggel

[ken gray]

it would be easier if you could put each of the ips into their own row.
can you do that by adding a "[]" to the string after you put the ip into it?


EDIT
04/03/2007 12:53:09 PM
also, you can put this into an ARRAY to deal with it as well.

[freggel99]

Don't know what you mean, i would like to if it helps but if i have a simple list like

10.10.10.10
20.20.20.20
25.25.25.25

and i search it with the findrx i see in the output field of qm

10.10.10.1020.20.20.2025.25.25.25

so if its possible to separate them by adding a [] then the member function will be able to search and clear i think...

[ken gray]

oh, if your ip are already in list form (one on each line) then just use the string function "RemoveDuplicateLines".

Code:

Copy      Help

_s=
;apple
;apple
;orange
;apple
;grapei    
;orange

_s.RemoveDuplicateLines
out _s

[freggel99]

That was just an example, most of the time i extract it from a piece of text and it is not like the example....

So if there is a possibility to store it like this and then the RemoveDuplicateLines it would be fine to me

TIA

[ken gray]

is there anyway you can do a search/replace to put in a "[]" before you do the rx find? then you can do a rx dump into an Array and then dump the Array into a string.

[freggel99]

I think if i do that it wil destroy my endresult, when i use a header for instance, i clean it up for further usage... So if i place [] in it, it will destroy the layout of the result....

[ken gray]

you could put it back into the "oneline" mode afterwards with a quick "findreplace".

[freggel99]

OK, I'll give it a try in a couple of hours..

Thanks for your time ! :!:

[freggel99]

Hi there,

I'm trying out now to replace ip with ip+[] but can't get it done. I seem to do something wrong:

s.replacerx(".+?(\d{1,3}){4}" ".+?(\d{1,3}){3}[]" 4)

First one seems to be replaced but rest is left alone

Any ideas?

Grz

Freggel99

[ken gray]

4 is "single replacement"
take that flag off and it should work.

[freggel99]

Thx Ken for your help, i've managed it finaly!

Your pathience was wonderfull!