数据样本
电子邮件地址的左列(e)相当长10,000+,没有特别的顺序,它是相应的国家(f)。
电子邮件地址(H)的右列是大小的一小部分,但是它包含许多重复值。
我需要将 e 中的所有值与 h 中的所有值进行比较,如果有匹配副本 f 中的值 i 中的相应细胞。
我已经研究了Vlookup,VBA脚本以及公式。
Sample of the data

The left column (E) of email addresses is quite long 10,000+ and is in no particular order, with it's corresponding country (F).
The right column of email addresses (H) is a fraction of the size however it contains many duplicate values.
I need to compare all of the values in E with all of the values in H, if there is a match copy the value in F to the corresponding cell in I.
I've looked into Vlookup, VBA scripts and if formulas.
发布评论
评论(1)
请尝试下一个代码。它使用数组,并且主要在内存中工作应该非常快:
是否有可能将字符串比较以使其看起来相同,但是有些字母在上层的情况下与较低的情况相比?如果是这样,我也可以调整代码来处理这种情况。现在,它比较了身份字符串。我的意思是“ href =“/cdn-cgi/l/电子邮件保护” class =“ __ cf_email__” data-cfemail =“ baf0d5d2d2494ded5dffac3dffac3dbd25594d94d94d9d9d557d7dd7”>
Please, try the next code. It uses arrays and working mostly in memory should be very fast:
Is there a possibility that the strings to be compared to look the same but some letters to be in Upper case against Lower? If so, I can adapt the code to handle such a situation, too. Now, it compares the identic strings. I mean "[email protected]" is not the same with "[email protected]"...