如何使用EXCEL公式从一列单元格中提取具有特定模式的文本？

Question

Help please, I have a dataset that contains a column of text which contain users mentions and their tweets. I would like to extract all the users mentioned in the tweets in EXCEL. In another words, I would like to do this task: for each row in a column if the cell value of the row contains a text starts with @ and end with a space then put that string value in another column, each cell might contain more than one occurance of the string (more than one user mentioned in a tweet).The question is that possible to do with EXCEL formulas not with coding? If yes, would you please direct me to which formula I should use.. and if not, do you know a good way/method to accomplish this task? Please do not send me links to documentations, if you have a well documented code to do this task or know a software/tool that can do that it will be great. Thanks for your help in advance.

Answer 1

This is only a partial solution. It retrieves the first instance of text bounded by "@" and a single space. With data in A1 , in B1 enter:

=LEFT(MID(A1,FIND("@",A1)+1,9999),FIND(" ",MID(A1,FIND("@",A1)+1,9999)))

For example:

Answer 2

I suggest that you break this task down into multiple columns, to understand how the formulas work together to get to your answer.

Column Headers

Column A = Your Data
Column B = First Start = Find the first occurrence of @
Column C = First End = Find the end of the first occurrence with a space
Column D = Second Start = Find the Second occurrence of @ 
Column E = Second End = Find the end of the second occurrence with a space
Column F = First Twitter Account = MID the First Start/End
Column G = Second Twitter Account = Mid the Second Start/End

Formulas

Column A = "An Example @Tweet with @two mentions"
Column B = Find("@",A2)
Column C = FIND(" ",A2,B2)
Column D = FIND("@",A3,C3)
Column E = FIND(" ",A3,D3)
Column F = MID(A2,B2,C2-B2)
Column G = MID(A2,D2,E2-D2)

You can repeat the above pattern for as many "mentions" as needed. You can learn that by counting how many @ signs are in each string, and making sure to write enough formulas to accommodate that number.

You could mash all of the above into one formula, but it would be a beast to read.

Keep in mind as well, if a "mention" is made at the END of a string, the above formulas will not count it. ie In your question you mention that mentions end in a space, which may not be the case if it happens at the end of the tweet.

Answer 3

Although tagged with [excel-vba] you offer no code and do have formula in your Title, so I suggest:

1. Replace @ with say |@ where the first character is distinctive (working on a copy of your data, select all, HOME > Editing - Find & Select, Replace..., Find what: @ Replace with: |@ , Replace All.)
2. Select the relevant column, DATA, Text to Columns, Delimited, Next, Delimiters Other: (only) | , Finish.

3. In the first completely empty column:

    =IF(LEFT(A1)="@",LEFT(A1,FIND(" ",A1)),"") 

copied across as many columns as previously were occupied and all formulae then copied down to suit.

This should cope with an indeterminate number of @ instances in any one cell and also where the last instance is not followed by a space.