Split, escaping certain splits - excel

I have a cell that contains multiple questions and answers and is organised like a CSV. So to get all these questions and answers separated a simple split using the comma as the delimiter should separate this easily.
Unfortunately, there are some values that use the comma as the decimal separator. Is there a way to escape the split for those occurrences?
Fortunately, my data can be split using ", " as separator, but if this wouldn't be the case, would there still be a solution besides manually replacing the decimal delimiter from a comma to a dot?
Example:
"Price: 0,09,Quantity: 12,Sold: Yes"
Using Split("Price: 0,09,Quantity: 12,Sold: Yes",",") would yield:
Price: 0
09
Quantity: 12
Sold: Yes

One possibility, given this test data, is to loop through the array after splitting, and whenever there's no : in the string, add this entry to the previous one.
The function that does this might look like this:
Public Function CleanUpSeparator(celldata As String) As String()
Dim ret() As String
Dim tmp() As String
Dim i As Integer, j As Integer
tmp = Split(celldata, ",")
For i = 0 To UBound(tmp)
If InStr(1, tmp(i), ":") < 1 Then
' Put this value on the previous line, and restore the comma
tmp(i - 1) = tmp(i - 1) & "," & tmp(i)
tmp(i) = ""
End If
Next i
j = 0
ReDim ret(j)
For i = 0 To UBound(tmp)
If tmp(i) <> "" Then
ret(j) = tmp(i)
j = j + 1
ReDim Preserve ret(j)
End If
Next i
ReDim Preserve ret(j - 1)
CleanUpSeparator = ret
End Function
Note that there's room for improvement by making the separator caharacters : and , into parameters, for instance.

I spent the last 24 hours or so puzzling over what I THINK is a completely analogous problem, so I'll share my solution here. Forgive me if I'm wrong about the applicability of my solution to this question. :-)
My Problem: I have a SharePoint list in which teachers (I'm an elementary school technology specialist) enter end-of-year award certificates for me to print. Teachers can enter multiple students' names for a given award, separating each name using a comma. I have a VBA macro in Access that turns each name into a separate record for mail merging. Okay, I lied. That was more of a story. HERE'S the problem: How can teachers add a student name like Hank Williams, Jr. (note the comma) without having the comma cause "Jr." to be interpreted as a separate student in my macro?
The full contents of the (SharePoint exported to Excel) field "Students" are stored within the macro in a variable called strStudentsBeforeSplit, and this string is eventually split with this statement:
strStudents = Split(strStudentsBeforeSplit, ",", -1, vbTextCompare)
So there's the problem, really. The Split function is using a comma as a separator, but poor student Hank Williams, Jr. has a comma in his name. What to do?
I spent a long time trying to figure out how to escape the comma. If this is possible, I never figured it out.
Lots of forum posts suggested using a different character as the separator. That's okay, I guess, but here's the solution I came up with:
Replace only the special commas preceding "Jr" with a different, uncommon character BEFORE the Split function runs.
Swap back to the commas after Split runs.
That's really the end of my post, but here are the lines from my macro that accomplish step 1. This may or may not be of interest because it really just deals with the minutiae of making the swap. Note that the code handles several different (mostly wrong) ways my teachers might type the "Jr" part of the name.
'Dealing with the comma before Jr. This will handle ", Jr." and ", Jr" and " Jr." and " Jr".
'Replaces the comma with ~ because commas are used to separate fields in Split function below.
'Will swap ~ back to comma later in UpdateQ_Comma_for_Jr query.
strStudentsBeforeSplit = Replace(strStudentsBeforeSplit, "Jr", "~ Jr.") 'Every Jr gets this treatment regardless of what else is around it.
'Note that because of previous Replace functions a few lines prior, the space between the comma and Jr will have been removed. This adds it back.
strStudentsBeforeSplit = Replace(strStudentsBeforeSplit, ",~ Jr", "~ Jr") 'If teacher had added a comma, strip it.
strStudentsBeforeSplit = Replace(strStudentsBeforeSplit, " ~ Jr", "~ Jr") 'In cases when teacher added Jr but no comma, remove the (now extra)...
'...space that was before Jr.

Related

How to match accented characters but not tab

I'm trying to match the company name in this string delimited with tabs.
Below table does not have tabs when you copy it, but I have replaced tabs with two spaces, which I assume will work fine for testing.
1025164 HERBEX IBERIA, S.L.U. KY01 4600292091
1016379 DRISCOLL´S OF EUROPE B.V. KY01 4600322589
1008809 LANDGARD NORD OBST & GEMÜSE GM KY01 4600347315
1008835 C.A.S.I. : COOPERATIVA PROVINC KY01 4600348112
1019258 SYDGRÖNT EKONOMISK FÖRENING KY02 4600343422
(The second column of the above, between 7 digit number and KY0 above)
In real life the columns are not always in the same order since it's a user preference.
I just took a few examples but names could also include /éèáà()´, pretty much anything (sadly).
I found another question here Concrete Javascript Regex for Accented Characters (Diacritics)
When I use the regex patterns in that thread, example: "\t([A-zÀ-ÿ0-9\s\.\,\_\-\'\&]+)\t" (I know some characters are still missing) to match between two tabs it becomes greedy and matches the whole line.
Is there any pattern that could match any character in a company name between tabs (or two spaces as the example above)?
Instead of returning a matched part, I matched everything and replaced it with the 1st capture group. Hope it helps.
Sub Test()
Dim str As String: str = "1025164" & vbTab & "HERBEX IBERIA, S.L.U." & vbTab & "KY01" & vbTab & "4600292091"
With CreateObject("vbscript.regexp")
.Global = True
.Pattern = "(?:^|\t)(?:\d+|KY\d+|([^\t]+))(?=\t|$)"
Debug.Print .Replace(str, "$1")
End With
End Sub
Have a look at this online demo to test the pattern:
(?:^|\t) - Match either start line anchor or a tab. Unfortunately the VBA-regex object does not support lookbehinds.
(?: - Open a non-capture group to start matching all parts you don't want to capture first:
\d+ - match 1+ digits;
| - Or:
KY\d+ - Match "KY" followed by 1+ digits;
| - Or:
([^\t]+) - nest a capture group to capture 1+ non-tabs.
) - Close non-capture group.
(?=\t|$) - Positive lookahead to assert captured text is followed by either a tab or end-line anchor.
I would use a different attempt using the split-command. The following code assumes that you have Tabs as separator and that the company name is found if the column is not numeric (only digits) and does not start with 'KY'.
Function getCompanyName(line As String) As String
Const separator = vbTab ' Replace with " " if you need that.
Dim tokens() As String, i As Integer
tokens = Split(line, separator)
For i = 0 To UBound(tokens)
If Not IsNumeric(tokens(i)) And Left(tokens(i) <> "KY") Then
getCompanyName = tokens(i)
Exit Function
End If
Next
End Function

Multiple space delimiter split string

I have a string like
string = "computer prog <5spaces> data mining <5spaces> oops concept"
As we can see clearly computer prog, data mining etc., are one continuous string and the delimiter is 5 spaces between the strings " ".
I need to split based on this in vb.net - so far I tried regex.split which works but results in giving 2 empty strings additionally and it's tedious to remove those additional strings.
I also tried using the string.split method but again it's taking even single white space also delimiters.
Below are the tried options:
regex.split
string.split
None give me the required result. I am not sure what I need to use. I even tried the option of stringsplitoption.removesapceentry (something like that) to get the desired result inside the split method, but none worked.
Dim array_keyskills As String() = res.Split(" ".ToCharArray,StringSplitOptions.RemoveEmptyEntries)
system.Windows.MessageBox.Show(array_keyskills(2) & array_keyskills.Length & " key skills") 'Display
The following short program:
Module Module1
Sub Main()
Dim s = "computer prog data mining oops concept"
Dim parts = s.Split({" "}, StringSplitOptions.None)
For Each p In parts
Console.WriteLine(p)
Next
Console.ReadLine()
End Sub
End Module
outputs:
computer prog
data mining
oops concept
If your data does not work that way then you should examine it to find which whitespace characters are in it which appear to be spaces but are not.
This did the trick:
array_keyskills = System.Text.RegularExpressions.Regex.Split(res," ").Where(
Function(s) Not String.IsNullOrWhitespace(s)
).ToArray()

Why does Excel treat double spaces as a comma?

I wrote an export to CSV file in my vb.net application, and I then exported it into Outlook.
The issue I've got, is that when the CSV file is being written, my code is checking for a comma in the current field, but while doing this, it also mistakes a double space for a comma, or space followed by 'Enter' key being pressed (for multiline textboxes)
An example would be if in the notes section of the customer, there is 4 lines of text, and one ends in a space - The user has then pressed enter to go to the next line, however the program is taking the next line of text and creating a new record for it, as it thinks it's a comma...
What is the reason for this? This means that data has to be super validated (ie checking for no double spaces etc) before it can be exported, which is far too time consuming.
Hopefully this makes sense!
This is the code:
Dim result As Boolean = True
Try
Dim sb As New StringBuilder()
Dim separator As String = ","
Dim group As String = """"
Dim newLine As String = Environment.NewLine
For Each column As DataColumn In dtable.Columns
sb.Append(wrapValue(column.ColumnName, group, separator) & separator)
Next
sb.Append(newLine)
For Each row As DataRow In dtable.Rows
For Each col As DataColumn In dtable.Columns
sb.Append(wrapValue(row(col).ToString(), group, separator) & separator)
Next
sb.Append(newLine)
Next
The code for wrapValue
Function wrapValue(value As String, group As String, separator As String) As String
If value.Contains(separator) Then
If value.Contains(group) Then
value = value.Replace(group, group + group)
End If
value = group & value & group
End If
Return value
End Function
Based on the fact that it's shortening it by 430 lines, I'd suggest it's something to do with the fact you're adding a load of "" before and after the value variable.
If it's removing a value at the start, then it will be removing a " before the first column header. As to why it's importing one record as you mentioned in the comments, I'm not entirely sure, however, I would suggest the issue lies in your wrapValue code.
Can you try changing
value = group & value & group
to
value = value
and see if that changes anything?

Automatically convert fixed length strings into delimited strings

I have many files containing such lines :
HUIHOJ OPKKA LK
ASOIJS AISJJ PL
AOSKSI ASIJD YA
I want to convert theses lines into something like this :
HUI;HOJ ;OPKKA ;L;K
ASO;IJS ;AISJJ ;P;L
AOS;KSI ;ASIJD ;Y;A
So the first field would be 3 characters, second would be 4, third 6, fourth 1 and fifth 1 character.
I know that it's possible to do it manually with excel, but I need to have automatically conversion solution, because I have many files with exactly the same structure.
VBA has a simple command to format text, so you can do this fairly easily, and with just a single line of code
Function SpFmt(S As String) As String
SpFmt = Format(S, "###\;####\;######\;#\;#")
End Function
If you want to use a worksheet function, you can do this with a nested replace formula on the worksheet:
=REPLACE(REPLACE(REPLACE(REPLACE(A1,4,0,";"),9,0,";"),16,0,";"),18,0,";")
Use the VBA Join Function after splitting (Split function) on a space or simply replace (Replace functoin) all of the spaces with a space & semi-colon.
dim str as string
str = range("A1").value2 'HUIHOJ OPKKA LK
range("A1") = Join(Split(str, char(32)), chr(32) & chr(59)) 'HUIHOJ ;OPKKA ;LK
str = range("A1").value2 'HUIHOJ OPKKA LK
range("A1") = Replace(str, chr(32), chr(32) & chr(59)) 'HUIHOJ ;OPKKA ;LK
If you are not bounded to Excel, you can use unix tools (also available for Windows) to do this very efficiently with just one command:
cut --output-delimiter=";" -c 1-3,4-7,8-13,14,15 fixed.txt > delimited.csv
The same command in a loop:
for f in *.txt ; do
cut --output-delimiter=";" -c 1-3,4-7,8-13,14,15 "${f}" > "${f}.csv"
done
Edit : the output delimiter option does not seem to work on every platform.
Alternatively, you can use sed :
sed "s/^\(.\{3\}\)\(.\{4\}\)\(.\{6\}\)\(.\)\(.\)/\1;\2;\3;\4;\5/" fixed.txt > delimited.csv

Separate words with commas in Excel 2010

I'm trying to use a formula in Excel to separate a bunch of words in a cell with a comma. If there are more than 5 words in the cell, I just want to get the first 5 words. To get the first five words in a cell and separate them by a comma I use this:
=SUBSTITUTE(LEFT(A1,FIND("^",SUBSTITUTE(A1," ","^",5))-1), " ", ", ")
This works fine. But the problem with this, because of the number 5 here, if I a cell contains less than 5 words, I get an error. I tried to substitute the 5 with this:
LEN(TRIM(A1))-LEN(SUBSTITUTE(A1," ",""))+1
So my function becomes this:
=SUBSTITUTE(LEFT(A1,FIND("^",SUBSTITUTE(A1," ","^",LEN(TRIM(A1))-LEN(SUBSTITUTE(A1," ",""))+1))-1), " ", ", ")
But this doesn't work, it gives me an error. Any idea how I can do this please?
Also I would like to ignore the first word if its first character is "-" (without the quotes) and just start from the second word. So in other words, I want something like this:
I love my life very much should return I, love, my, life, very
- I love my life very much should return I, love, my, life, very (the "-" is ignored")
I love my should return I, love, my
Thanks in advance for any help
Here's a somewhat different approach. Aside from the "less than 5" issue, it also deals with the "5 words with no space at the end" issue:
=LEFT(A1,FIND("^",SUBSTITUTE(A1 & "^"," ","^",5))-1)
EDIT 1: I just noticed the part about the leading "- ". My addition isn't very elegant, but it deals with it, and also TRIMS any trailing spaces:
=TRIM(LEFT(IF(LEFT(A1,2)="- ",MID(A1,3,999),A1),FIND("^",SUBSTITUTE(IF(LEFT(A1,2)="- ",MID(A1,3,999),A1) & "^"," ","^",5))-1))
EDIT 2: Oh yeah, commas:
=SUBSTITUTE(TRIM(LEFT(IF(LEFT(A1,2)="- ",MID(A1,3,999),A1),FIND("^",SUBSTITUTE(IF(LEFT(A1,2)="- ",MID(A1,3,999),A1) & "^"," ","^",5))-1))," ",",")
Try this:
=TRIM(LEFT(SUBSTITUTE(SUBSTITUTE(TRIM(SUBSTITUTE(A1,"-"," "))," ",","),",",REPT(" ",99),5),99))
This will work even if there is not a space after the dash or if there are extra spaces in the text. Often I find that input is not very clean.
=SUBSTITUTE(LEFT(SUBSTITUTE(TRIM(SUBSTITUTE(A1,"-","",1)),
" ","*",5),IFERROR(FIND("*",SUBSTITUTE(TRIM(SUBSTITUTE(A1,"-","",1)),
" ","*",5))-1,999))," ",",")
Edit: After commenting on István's, I made mine flawless too.
=SUBSTITUTE(LEFT(SUBSTITUTE(TRIM(SUBSTITUTE(LEFT(TRIM(A1),1),"-"," ",1)
&MID(TRIM(A1),2,999))," ","*",5),IFERROR(FIND("*",SUBSTITUTE(
TRIM(SUBSTITUTE(LEFT(TRIM(A1),1),"-","",1)&MID(TRIM(A1),2,999))," ","*",5))-1,999))," ",",")
But I think his is more elegant.
Try this:
=SUBSTITUTE(LEFT(SUBSTITUTE(SUBSTITUTE(TRIM(SUBSTITUTE(A1,"- ","",1))&" "," ",", "),", ","|",MIN(LEN(SUBSTITUTE(TRIM(SUBSTITUTE(A1,"- ","",1))&" "," ",", "))-LEN(SUBSTITUTE(SUBSTITUTE(TRIM(SUBSTITUTE(A1,"- ","",1))&" "," ",", ")," ","")),5)),FIND("|",SUBSTITUTE(SUBSTITUTE(TRIM(SUBSTITUTE(A1,"- ","",1))&" "," ",", "),", ","|",MIN(LEN(SUBSTITUTE(TRIM(SUBSTITUTE(A1,"- ","",1))&" "," ",", "))-LEN(SUBSTITUTE(SUBSTITUTE(TRIM(SUBSTITUTE(A1,"- ","",1))&" "," ",", ")," ","")),5)))-1),",,",",")
The formula works by taking the following steps:
Remove any leading dash-space
Trim any leading or trailing spaces
Insert comma-spaces in place of spaces and add a trailing comma-space
Calculate the lesser of 5 and the number of words in the string
Put in "|" in place of either the fifth comma-space or the trailing comma-space if the string is less than five words
Determine the position of the "|"
Strip off the "|" and all characters to the right of it
Remove any doubled commas due to any single embedded commas in the initial string
If you are willing to consider a VBA solution, this complex expression can be replaced by a user-defined function:
Function words5(InputString As String) As String
Dim wordArray As Variant
wordArray = Split(Trim(Replace(InputString, _ 'remove "-", put words into array
"-", "", , 1)), " ")
ReDim Preserve wordArray(LBound(wordArray) To _ 'drop all but the first 5 words
WorksheetFunction.Min(UBound(wordArray), 5 - 1))
words5 = Replace(Join(wordArray, ", "), ",,", ",") 'rejoin the words with ", "
End Function 'separator
On the plus side of using this code is its maintainability compared to the worksheet formula, which impossible to understand or safely alter without access to the original building blocks that were combined into the single expression.
The code would have to be installed in the workbook in which it is used or in either the standard Personal.xlsb workbook or an addin workbook.
To use the function, copy and paste it into a standard module, which can be inserted into a workbook via the VBA editor. You can open the editor with the Visual Basic button on the `Developer tab of the ribbon.
Figured I'd throw my hat in the ring also. I think this formula should cover the bases:
=SUBSTITUTE(TRIM(LEFT(SUBSTITUTE(TRIM(SUBSTITUTE(A1&" ","- ",""))," ",REPT(" ",99)),99*5))," ",",")

Resources