Replace match with Stringtheory

[CarlosGtrz]

I need to replace the string matching a regex with another string. The new string is created by a function, which I only want to call it if there is a match. This is my current code, I wonder if there is a simpler or more efficient way to do the replace:

  st.FindMatchPosition(regex,starts,ends,Match:Regular,TRUE)
  IF starts AND ends 
    st.SetValue( st.Slice(1,starts-1) & createnewvalue() & st.Slice(ends+1) )
  END

@vitesse @Bruce Would it be too much work to add methods .ReplaceMatch(...) and .ReplacePosition(...)? If there isn’t something similar already that I missed. Thanks in advance for your comments.

[RchdR]

@CarlosGtrz

ReplaceMatch(...) and .ReplacePosition(...)? If there isn’t something similar already that I missed

Before my system and backups got wiped a few weeks ago I had code that did this and it worked in the templates as well.

I seem to remember posting some code on here which would help you get the start and finish positions using strpos. I think its written in clarion but it might also be a template language code example.

Fwiw

[vitesse]

Hi Carlos

well that’s a coincidence - I wrote code to do this over the weekend and just yesterday sent it to Bruce for consideration for inclusion in a future version of ST along with a bunch of other stuff.

I called the new method ReplaceSlice and the comment in the code says:

!----------------------------------------------------------------------------------
! Sets the value of the specified region within the string to the passed string.
! Unlike setSlice() which either truncates or space fills the new value to fit the
! existing size of the slice, replaceSlice() will expand or contract the slice
! within the value string to exactly fit the newValue.

if you scroll down the in code at

WkHtmlToPdf DLL prototype

yes provided you have access to the file “…/…/logo.jpg” then read it into ST and do more or less as Koen showed above. Maybe simplify it a bit and just replace the reference to logo.jpg with the base64 encoded text. Off the top of my head, something like: st stringTheory img stringTheory fname stringTheory ! filename of image lStart long,auto lEnd long,auto srcStart long,auto srcEnd long,auto replaces long,auto code if ~st.loadFile('myHTML.htm')…

you will see some code of the form:

 st.replace(st.slice(srcStart,srcEnd),'replacementString',1,srcStart,srcEnd)

actually the final parameter is redundant and it could have just been written:

 st.replace(st.slice(srcStart,srcEnd),'replacementString',1,srcStart)

with the new method replaceSlice this would instead look like:

 st.replaceSlice(srcStart,srcEnd,'replacementString')

and would be more efficient as it is not doing a “dummy search” before the replace.

as regards your current code

  st.FindMatchPosition(regex,starts,ends,Match:Regular,TRUE)
  IF starts AND ends 
    st.SetValue( st.Slice(1,starts-1) & createnewvalue() & st.Slice(ends+1) )
  END

I suggest changing the test so instead have

  st.FindMatchPosition(regex,starts,ends,Match:Regular,TRUE)
  IF starts AND ends >= starts
    st.SetValue( st.Slice(1,starts-1) & createnewvalue() & st.Slice(ends+1) )
  END

or use the replace code with a dummy search which might be a little faster:

IF starts AND ends >= starts 
  st.replace(st.slice(starts,ends),createnewvalue(),1,starts)
END

then if/when replaceSlice becomes available, instead use:

IF starts AND ends >= starts 
 st.replaceSlice(starts,ends,createnewvalue())
END

if you are keen to try replaceSlice now then (if Bruce is agreeable) I could send you my current version by PM.

one thing to note is that your st.FindMatchPosition will find the SMALLEST minimal match.

hopefully that is what you are after (and not the maximal “leftmost longest” match).

if you are after that “leftmost longest” match, have a look at the code here which includes StrPosAndLen which returns position and minimum and maximum length of string matched against regex There are a lot of “gotchas” with strPos and this works around them.

StrPos() String Clipping - How to Match a Trailing Space?

Following on from Carl’s post I have found that adding a null (‘<0>’) on the end of a string works on both the data and the regex. You do have to be careful with the nulls as when used with strpos they will terminate the string so they are no good anywhere other than at the very end of a string. What I have done is written a wrapper for strPos called strPosUnclipped: StrPosUnclipped PROCEDURE (*string pText,*string pRegex) !,LONG CODE return strpos(choose(size(pText) >0 and pText […

hth and cheers

Geoff R

[vitesse]

Hi Richard

@RchdR

some code on here which would help you get the start and finish positions using strpos.

I think this may be related to the StrPosAndLen code I mentioned in my previous message - I recall we had several conversations about this over a couple of years and it went through a few iterations before that version.

cheers

Geoff R

[CarlosGtrz]

Hi Geoff

@Vitesse

well that’s a coincidence - I wrote code to do this over the weekend

Excellent!

@vitesse

ReplaceSlice

That’s a better name.

@vitesse

IF starts AND ends >= starts 
  st.replace(st.slice(starts,ends),createnewvalue(),1,starts)
END

Thanks! That’ll work for now, and hopefully, in a future version of ST, I can switch to ReplaceSlice.

@vitesse

your st.FindMatchPosition will find the SMALLEST minimal match.
hopefully that is what you are after (and not the maximal “leftmost longest” match)

Yes, in this case, but I’ll look into the other code you posted since I might need “greedy” matches in the future.

Thanks again your comments and insights.

Regards,

Carlos

[vitesse]

thanks Carlos

if you were after a tiny bit more speed now while waiting for ReplaceSlice, you could use a little “inside knowledge” and alter

@vitesse

IF starts AND ends >= starts 
  st.replace(st.slice(starts,ends),createnewvalue(),1,starts)
END

to instead be

IF starts AND ends >= starts AND ends <= st._DataEnd
  st.replace(st.valuePtr[starts : ends],createnewvalue(),1,starts)
END

“purists” probably would not approve as here we are not using the official public interfaces.

cheers again

Geoff R

[RchdR]

@CarlosGtrz

I might need “greedy” matches in the future.

I had started mirroring the extra regex functionality microsoft provides in dotnet regex so theres no reason why capesoft or softvelocity cant mirror the dotnet regex functionality either.

[CarlosGtrz]

@vitesse

st.valuePtr[starts : ends]

Nice, thank you! I didn’t know valuePtr was available and public. It’s actually documented in the Upgrading section, so the “purists” should be happy .