一个老题，有没有好点的思路

后一天有相同规格的,当天排产时排最后可少换一次.

比之前的好点，还是有点麻烦
=LET(a,{"ABC","ACB","BAC","BCA","CAB","CBA"},f,LAMBDA(x,UNIQUE(IFERROR(REGEXREPLACE(a,"["&CONCAT(IF(x="√","",A2:A4))&"]",),a),1)),z,REDUCE(A2:A4,B1:O1,LAMBDA(s,n,LET(p,f(INDEX(B2:O4,,n)),TOCOL(s&","&IF(LEFT(p)=RIGHT(s),"",RIGHT(s))&p)))),DROP(TEXTSPLIT(@SORTBY(z,LEN(z)),","),,1))
=SUM(LEN(B8#)-1)+1

=LET(t,REPT(A2:A4,B2:O4=B2),n,B1:O1,DROP(REDUCE("",n,LAMBDA(w,z,HSTACK(w,CONCAT(UNIQUE(VSTACK(RIGHT(TAKE(w,,-1)),RIGHT(SORT(INDEX(IFERROR(MAP(t,n+P1:P3,LAMBDA(x,y,CONCAT(REPT(CHAR(DROP(n,,y)+64)&x,(x=DROP(t,,y))*(x>"")))&x)),t),,z),,-1)))))))),,1))

****，又是LET，下一个下一个

=DROP(REDUCE("",B1:O1,LAMBDA(x,y,HSTACK(x,CONCAT(UNIQUE(RIGHT(VSTACK(TAKE(x,,-1),SORT(BYROW(REPT(TEXT(DROP(y:P1,,1),0&0),LEFT(INDEX(B2:P4,,y)&0)=DROP(B2:P4,,y)),CONCAT)&REPT(A2:A4,INDEX(B2:P4,,y)>""),,-1)))))))),,1)

=SCAN("",B1:O1,LAMBDA(x,y,CONCAT(UNIQUE(RIGHT(VSTACK(x,SORT(BYROW(REPT(TEXT(DROP(y:P1,,1),0&0),LEFT(INDEX(B2:P4,,y)&0)=DROP(B2:P4,,y)),CONCAT)&REPT(A2:A4,INDEX(B2:P4,,y)>""),,-1)))))))

公式还是很长，但是可以保证多一个模具都不卡
=TEXTSPLIT(MID(@REDUCE(TOROW(A2:A5),B1:O1,LAMBDA(s,n,LET(p,TOCOL(IF(INDEX(B2:O5,,n)="",坤,A2:A5),3),q,REDUCE("",p,LAMBDA(y,x,TOCOL(REPLACE(y,SEQUENCE(,MATCH(x,p,)),,x)))),t,TOROW(s&","&SUBSTITUTE(RIGHT(s),LEFT(q),)&q),FILTER(t,LEN(t)=MIN(LEN(t)))))),3,99),",")

=SCAN("",B1:O1,LAMBDA(x,y,CONCAT(UNIQUE(RIGHT(VSTACK(x,SORT(BYROW(REPT(TEXT(DROP(y:P1,,1),0&0),LEFT(INDEX(B2:P4,,y)&0)=DROP(B2:P4,,y)),CONCAT)&REPT("09",INDEX(B2:P4,,y)>"")&REPT(A2:A4,INDEX(B2:P4,,y)>""),,-1)))))))

加了后边介，你测试一下。=SCAN("",B1:O1,LAMBDA(x,y,CONCAT(UNIQUE(RIGHT(VSTACK(x,SORT(BYROW(REPT(TEXT(DROP(y:P1,,1),0&0),LEFT(INDEX(B2:P4,,y)&0)=DROP(B2:P4,,y)),CONCAT)&REPT("09",INDEX(B2:P4,,y)>"")&REPT(A2:A4,INDEX(B2:P4,,y)>""),,-1)))))))

=SCAN("",B1:O1,LAMBDA(x,y,CONCAT(UNIQUE(RIGHT(VSTACK(x,SORT(BYROW(REPT(TEXT(DROP(y:P1,,1),0&0),LEFT(INDEX(B2:P4,,y)&0)=DROP(B2:P4,,y)),CONCAT)&REPT("09"&A2:A4,INDEX(B2:P4,,y)>""),,-1)))))))

这道题和我昨天发的那道有点像。其实只需故后无需理前。

下面的是你的