The last part of Guy’s expression:

\d+

means 1 or more digits. If you want a digit, a period, and exactly 3 digits, one way is to use this:

\d\.\d{3}

The first \d is for a single digit. The \. is for a period (the period means any character whatsoever unless it has a \ in front of it). The final \d{3} means exactly 3 digits.

This kind of manipulation is very easily performed, with regular expressions. So :

• Open your XML file in N++

• Move back to its very beginning, if necessary ( Ctrl + Origin )

• Open the Replace dialog ( Ctrl + H )

• Type, in Find what: zone, the regex (?-i)<entityGroup="spiderZombie" prob="\d+"/>

• Let the Replace with: zone EMPTY

• Select the Regular expression search mode

• Click on the Replace All button

Et voilà !

Assuming, for instance, the 7 lines original text, below :

blablablah <entityGroup="spiderZombie" prob="000"/> Other text
blablablah <entityGroup="spiderZombie" prob="123"/> Other text
blablablah <entityGroup="superHero" prob="999"/> Other text
blablablah <entityGroup="spiderZombie" prob="456"/> Other text
blablablah <entityGroup="total_Destroyer" prob="10"/> Other text
blablablah <entityGroup="spiderZombie" prob="789"/> Other text
blablablah <entityGroup="spiderZombie" prob="4589"/> Other text

We would obtain, after global replacement :

blablablah  Other text
blablablah  Other text
blablablah <entityGroup="superHero" prob="999"/> Other text
blablablah  Other text
blablablah <entityGroup="total_Destroyer" prob="10"/> Other text
blablablah  Other text
blablablah  Other text

Notes :

• I supposed that search must be sensitive to case. If not, change the modifier (?-i) by (?i)

• I supposed that the prob number is any integer. If it has to be restricted to three digits, only, change the part \d+ by \d{3}

• On the contrary, if any string entityGroup is concerned by the suppression, use the regex, below :

(?-i)<entityGroup="\w+" prob="\d+"/>

With this regex, any entityGroup name, containing word characters ( letters, digits, or the low line character _ ) would be matched and suppressed, in replacement, along with the prob string. So, it would give the resulting text :

blablablah  Other text
blablablah  Other text
blablablah  Other text
blablablah  Other text
blablablah  Other text
blablablah  Other text
blablablah  Other text

Best Regards,

guy038