Colleagues,

I have a dataset with a land-use description codified with a alphanumeric string. Some observations are the followings:

var1 value
R(60ECL(60VAP_20EDFea_10ZAU_10OCT)_40UEN(45VAP_20Z AU_20OCT_15EDFea)) 14
R(60IAS(70EDFnv_20ZAU_10VAP)_40UEN(60VAP_40EDFea)) 12
R(70ERG(60ZAU_15EDFea_15VAP_10OCT)_30EDU(40EDFea_2 5OCT_20VAP_15ZAU)) 46.5
R(90UCS(75EDFea_25VAP)_10ERG(45VAP_35EDFem_20ZAU)) 2


it contains different type of land-use described by letters and its percentages by number. in addition, a land can have two or more description and subdescriptions.

I want to calculate the total value of "ZAU" in each observation but i don´t know how to do that. the complete table has more than 500 observations so it could take me hours to calculate this value for each one. Does anyone have any idea how to do it?