So, I've looked around and tried to solve this on my own. This isn't an absolutely crucial question currently, I just want to know if it could be done.
So let's say I've got a list with some data that looks like
Date Location
01/24/14 H-12
01/25/14 BB-44
01/30/14 G-12
01/29/14 7A-55
01/28/14 NN-15
01/24/14 GG-47
What I want is to be able to sort the data by Location, but I don't want it to be the general way, otherwise I'll end up with 7A-55, BB-44, G-12, H-12, NN-15. I want the data to be sorted so that double letters and single letters are sorted together. E.G. it should be G-12, H-12, BB-44, NN-15, 7A-55 once everything has been sorted.
I've tried creating a custom list sort, but it doesn't work. the way intended. The custom list I tried was A-Z, AA-ZZ, 7A (items were listed out, but for saving space I wrote them like that).
Like I said, this isn't a particularly huge deal if it can't be done, it just would have made it a little easier.
Edit 1 Here is what I would like to be the output
Date Location
01/30/12 G-12
01/24/14 H-12
01/25/14 BB-44
01/24/14 GG-47
01/28/14 NN-15
01/29/14 7A-55
Edit
All of these worked in the regards i wanted to, although if I had to choose a favorite it would be the base 36 number conversion one. That was some real out-of-the-box thinking and the math geek in me appreciated it. Thanks everyone!
Well it works, but is a bit complex, so rather just for fun:
This UDF returns a value that can be used as sort key. It transforms the code into a four-digit base 36-number, i.e. using A-Z and 0-9 as symbols (like hex uses 0-9 and A-F). To get at your desired output, I literally put the symbols in this order, letters first (so "A" = 0 and "0" = 26).
(The missing 'digits' are filled up with zeros, which are in this case "A"s)
It works ;)
Public Function Base36Transform(r As Range) As Long
Dim s As String, c As String
Dim v
Dim i As Integer
Dim rv As Long
v = Split(r.Text, "-")
s1 = v(0)
s2 = v(1)
s = Right("A" & s1, 2) & Right("A" & s2, 2)
rv = 0
For i = 1 To Len(s)
c = Mid(s, Len(s) - i + 1, 1)
If c Like "#" Then
rv = rv + (Val(c) + 26) * (36 ^ (i - 1))
Else
' c is like "[A-Z]"
rv = rv + (Asc(c) - Asc("A")) * (36 ^ (i - 1))
End If
Next
Base36Transform = rv
End Function
Sorting is often a very creative process. VBA can ease up the process, but a little extension of the data will work just as well.
See my results:
The way I did it is by getting the length of each string, just to be safe. This is gotten by simply going =LEN(B2), dragged down.
Then I check if it starts with 7. If it does, assign 1, otherwise keep at 0. I used this formula: =(LEFT(B2,1)="7")*1, dragged down.
Now, my custom sort is this:
Now I might have gotten some things wrong here, or I might even have done overkill by going the Length column. However, the logic is pretty much what you're aiming for.
Hope this helps in a way! Let us know. :)
I am a little lazy here and assuming your data sits in Column A,B. You mightneed to adjust your range or the starting point of your list. But here's the code:
Sub sortttttt()
Dim rng As Range
Dim i As Integer
Range("B2").Activate
Do While Not IsEmpty(ActiveCell)
ActiveCell.Value = Len(ActiveCell.Value) & ActiveCell.Value
ActiveCell.Offset(1, 0).Activate
Loop
Set rng = Range("A1:B6")
rng.Sort Key1:=Range("B2"), Order1:=xlAscending, Header:=xlYes
Range("B2").Activate
Do While Not IsEmpty(ActiveCell)
ActiveCell.Value = Right(ActiveCell.Value, Len(ActiveCell.Value) - 1)
ActiveCell.Offset(1, 0).Activate
Loop
End Sub
Assuming your data is in columns B:C with labels in Row1 and no intervening blank rows, add a column with:
=IF(ISNUMBER(VALUE(LEFT(C2))),3,IF(FIND("-",C2)>2,2,1))
in D1 copied down to suit and sort ascending Location within sort ascending of the added column.
Related
I need to sort a column containing cells with the following format : "TITLE text". I know the list of possible titles, but not the texts, so what I would like to do is sort the title in a custom order (for example : PLA, ARG, FHI, BRT) that is not alphabetical. The problem is that the title and the text are in the same cell.
So, for example, here is a screen of datas I might want to work on :
How can I sort this if the cells doesn't perfectly match the list members ?
And, if possible, how to do that using a macro and not manually ?
It's not very difficult. I will try to explain how this is done.
First of all, we need to figure out a way to transfer the range of cells to be sorted to the macro. There are different ways - write the address directly in the macro code, pass it as a parameter to the UDF, get it from the current selection. We use the third method - it is not the easiest to code, but it will work with any data sets.
The main difficulty when using the current selection is that the selection can be one single cell (nothing to sort), a range of cells (and may be several columns - how to sort this?) or several ranges of cells (this is if you hold down the CTRL key and select several unconnected ranges).
A good macro should handle each of these situations. But now we are not writing a good macro, we are getting acquainted with the principle of solving such problems (Since StackOfflow is a resource for programmers, the answers here help you write code yourself, and not get ready-made programs for free). Therefore, we will ignore a single cell and
multiple ranges - we will just stop execution of macro. Moreover, if there is more than one column in the selected range, then we will not do anything either.
Also, in case a full column is selected, we restrict the range to be sorted to the used area. This will sort the real data, but not the million empty cells.
The code that does all this looks like this:
Sub SortByTitles()
Dim oCurrentSelection As Variant
Dim oSortRange As Variant
Dim oSheet As Variant
Dim oCursor As Variant
Dim oDataArray As Variant
Dim sList As String
sList = "PLA,ARG,FHI,BRT"
oCurrentSelection = ThisComponent.getCurrentSelection()
Rem Is it one singl cell?
If oCurrentSelection.supportsService("com.sun.star.sheet.SheetCell") Then Exit Sub
Rem Is it several ranges of cells?
If oCurrentSelection.supportsService("com.sun.star.sheet.SheetCellRanges") Then Exit Sub
Rem Is this one range of cells? (It can be a graphic item or a control.
Rem Or it may not even be a Calc spreadsheet at all)
If Not oCurrentSelection.supportsService("com.sun.star.sheet.SheetCellRange") Then Exit Sub
Rem Is there only one column selected?
If oCurrentSelection.getColumns().getCount() <> 1 Then Exit Sub
Rem Is the current selection outside of the used area?
oSheet = oCurrentSelection.getSpreadsheet()
oCursor = oSheet.createCursor()
oCursor.gotoEndOfUsedArea(True)
oSortRange = oCursor.queryIntersection(oCurrentSelection.getRangeAddress())
If oSortRange.getCount() <> 1 Then Exit Sub
Rem Redim oSortRange as single range (not any ranges)
oSortRange = oSortRange.getByIndex(0)
Rem Get data from oSortRange
oDataArray = oSortRange.getDataArray()
Rem Paste sorted data to the same place:
oSortRange.setDataArray(getSorted(oDataArray, Split(sList,",")))
End Sub
The getSorted() function, which is mentioned in the last line of the procedure, must take two arrays as parameters — the values of the cells to be sorted and the sort list — and return one array of sorted values.
One aspect of working with data from ranges of cells should be mentioned here. If in Excel after receiving data from the range we get a two-dimensional array, then in OpenOffice/LibreOffice we get a one-dimensional "array of arrays", each element of which is a one-dimensional array of cell values of one row. Writing to a range is done from exactly the same structure, from an "array of arrays". The first parameter of the getSorted() function is oDataArray - just such an array of arrays, this will need to be taken into account when processing data.
What will getSorted() function do? It will build a "tree" sorted by Headers from the oDataArray values. In fact, this is not a tree - it is an ascending sorted array of all Headers and all values with these Headers. The values are also a sorted array. Then the function will select from the tree those Headings that are listed in the List and remove them from the tree. If, after all the actions, some elements still remain in the sorted tree, they will be displayed at the very end.
The function will accumulate the result in a separate array of the same size as the original one. In other words, the algorithm will use three times more memory than the original sorted range - source data, a tree and result array. The function will accumulate the result in a separate array of the same size as the original one. In other words, the algorithm will use three times more memory than the original sorted range - source data, a tree and result array.
You can try to save resources and write the results directly to the original array. But I strongly advise against doing this.
The fact is that an array cell may contain not a value, but a reference to a value, and in the case of inaccurate coding, you will not get a large sorted array, but a large array of the same value (the last cell).
I deliberately do not comment on all the following code - if you can read and understand this without comment, then you will understand how actions are programmed to process data from ranges:
Function getSorted(aData As Variant, aList As Variant) As Variant
Dim aRes As Variant
Dim i As Long, pos As Long, j As Long, k As Long, m As Long, uB As Long
Dim aTemp As Variant
aTemp = Array()
ReDim aRes(LBound(aData) To UBound(aData))
For i = LBound(aData) To UBound(aData)
pos = InStr(aData(i)(0), " ")
If pos > 0 Then
AddToArray(Left(aData(i)(0),pos-1), aData(i)(0), aTemp)
Else
AddToArray(aData(i)(0), aData(i)(0), aTemp)
EndIf
Next i
m = LBound(aData) - 1
For i = LBound(aList) To UBound(aList)
k = getIndex(aList(i), aTemp)
If k > -1 Then
uB = UBound(aTemp) - 1
For j = LBound(aTemp(k)(1)) To UBound(aTemp(k)(1))
m = m + 1
aRes(m) = Array(aTemp(k)(1)(j))
Next j
For j = k To uB
aTemp(j) = aTemp(j+1)
Next j
ReDim Preserve aTemp(uB)
EndIf
Next i
For k = LBound(aTemp) To UBound(aTemp)
For j = LBound(aTemp(k)(1)) To UBound(aTemp(k)(1))
m = m + 1
aRes(m) = Array(aTemp(k)(1)(j))
Next j
Next k
getSorted = aRes
End Function
To build a sorted tree, two subroutines are used - AddToArray() and InsertToArray(). They are very similar - the first eight lines are a normal binary search, and the remaining 10-12 lines are actions when an element is not found at the end of the array, when it is found and when it is not found in the middle of the array:
Sub AddToArray(key As Variant, value As Variant, aData As Variant)
Dim l&, r&, m&, N&, i&
l=LBound(aData)
r=UBound(aData)+1
N=r
While (l<r)
m=l+Int((r-l)/2)
If aData(m)(0)<key Then
l=m+1
Else
r=m
EndIf
Wend
If r=N Then
ReDim Preserve aData(0 To N)
aData(N) = Array(key, Array(value))
ElseIf aData(r)(0)=key Then
InsertToArray(value, aData(r)(1))
Else
ReDim Preserve aData(0 To N)
For i = N-1 To r Step -1
aData(i+1)=aData(i)
Next i
aData(r) = Array(key, Array(value))
EndIf
End Sub
Sub InsertToArray(key As Variant, aData As Variant)
Dim l&, r&, m&, N&, i&
l=LBound(aData)
r=UBound(aData)+1
N=r
While (l<r)
m=l+Int((r-l)/2)
If aData(m)<key Then
l=m+1
Else
r=m
EndIf
Wend
If r=N Then
ReDim Preserve aData(0 To N)
aData(N) = key
Else
ReDim Preserve aData(0 To N)
For i = N-1 To r Step -1
aData(i+1)=aData(i)
Next i
aData(r) = key
EndIf
End Sub
The getIndex() function uses the same binary search. It will return the index of the element in the array if it can find it, or -1 otherwise:
Function getIndex(key As Variant, aData As Variant) As Long
Dim l&, r&, m&, N&
l=LBound(aData)
r=UBound(aData)+1
N=r
While (l<r)
m=l+Int((r-l)/2)
If aData(m)(0)<key Then
l=m+1
Else
r=m
EndIf
Wend
If r=N Then
getIndex = -1
ElseIf aData(r)(0)=key Then
getIndex = r
Else
getIndex = -1
EndIf
End Function
And that's all that is needed to solve the task:
Demo file with code - SortByTitle.ods
I have a fairly complex look of code where I am looking through multiple control variables.
I am getting an error 'Invalid 'for' loop control variable
the line in questions is
for w(1) = 32 to 127
I am more familiar with VBA where I would have zero problem with this statement.
I'm guessing it has something to do with the fact that i will be looping through w(1),w(2),w(3) etc. in the same tree. I initialize the variable as dim x(10) but have also tried dim w() , dim w() redim w(10)
Any thoughts? its a fairly critical aspect of the script; as such I am unwilling to swap out all my w 1,2... for individual variables
Thoughts?
EDIT:
As per comments I should clarify a Few things:
Essentially there is a alpha numeric association with an ID in a system that I am working with which I was not handed down the key too. So I have a multi-dimensional array of rates that are used for multiplying out costs.
What I am doing is working backwards through invoices and matching a material with very subtle differences that have different pricings.
For simplicity sake, say theres a 2 dimensional material where AA, AB, ... A9 are all priced through several multiplication factors in what would just be a 2x2 grid. So maintaining a pivot point based on the position in string is very important. For this code you could take tier to mean how many characters in the string (aka how complex the composition of the material):
dim x(), w()
for tier = 1 to 2
for w(1) = 32 to 127
x(1)= chr(w(1))
If tier = 2 then
for w(2)= 32 to 127
X(2)=chr(w(2))
next
end if
str = ""
for y = 1 to (tier)
str = trim(str & x(y))
next
'''msgbox str 'debug
next
end if
str = ""
for y = 1 to (tier)
str = trim(str & x(y))
next
'' msgbox str ' debug
next 'tier
This is just an excerpt i pulled to get a basic idea of the structure w/o any calculations. this is in essence what is not working
The error is quite clear, you cannot use an Array as the control variable. The definition in For...Next Statement is even clearer;
Numeric variable used as a loop counter. The variable cannot be an array element or an element of a user-defined type.
This is one of the key differences between VBA and VBScript.
You won't loop through x(1),x(2)...on what you write it's going like this 32(1),33(1)....what type it's your w(1) and how you define him?
I have the following VBA code:
Sub test2()
Dim w1 As Worksheet
Dim w2 As Worksheet
Dim k As Long
Dim c As Range
Dim d As Range
Dim strFA As String
Set w1 = Sheets("Sheet1")
Set w2 = Sheets("Sheet2")
w2.Cells.Clear
k = 1
With w1.Range("A:A")
Set c = .Cells.Find("FirstThing", After:=.Cells(.Cells.Count), lookat:=xlWhole)
strFA = ""
While Not c Is Nothing And strFA <> c.Address
If strFA = "" Then strFA = c.Address
If IsError(Application.Match(c.Offset(0, 1).value, w2.Range("A:A"), False)) Then
Set d = .Cells.Find("SecondThing", c, , xlWhole)
w2.Range("A" & k).value = c.Offset(1, 0).value
w2.Range("B" & k).value = d.Offset(0, 1).value
k = k + 1
End If
Set c = .Cells.Find("FirstThing", After:=c, lookat:=xlWhole)
Wend
End With
End Sub
The code works essentially like this:
Look through Sheet1 for a certain phrase.
Once the phrase is found, place the value from the cell one row over in Sheet2
Search for a second phrase.
Place the value from the cell one row over in the cell beside the other value in Sheet2
Repeat
Now. I have the same data that, don't ask me why, is in .doc files. I'd like to create something similar to this code that will go through and look for the first phrase, and place the next n characters in an Excel sheet, and then look for the second phrase and place the next m characters in the row beside the cell housing the previous n characters.
I'm not sure whether it's better to do this with a bash script or whether it's possible to do this with VBA, so I've attached both as tags.
Your question seems to be: "I'm not sure whether it's better to do this with a bash script or whether it's possible to do this with VBA"
The answer to that is: You'd need VBA, especially since this is a *.doc file - docx would be a different matter.
In order to figure out what that is, start by trying to do the task manually in Word. More specifically, how to use Word's "Find" functionality. When you get that figured out, record those actions in a macro to get the starting point for your syntax. The code on the Excel side for writing the data across will essentially stay the same.
You'll also need to decide where the code should reside: in Word or in Excel. That will mean researching how to run the other application from within the one you choose - lots of examples here on SO and on the Internet...
Here's part of my code.
<%
Dim lineData,fso,filea,fileb,filec
s=request.querystring("query")
set fso = Server.CreateObject("Scripting.FileSystemObject")
a(0,0)=0
a(1,0)=" - Entries in File A"
set filea = fso.OpenTextFile(Server.MapPath("FileA.txt"), 1, true)
do until lone.AtEndOfStream
lineData = lcase(filea.ReadLine())
if instr(lineData,s)>0 then
a(0,0)=a(0,0)+1
end if
Loop
a(0,1)=0
a(1,1)=" - Entries in File B"
set fileb = fso.OpenTextFile(Server.MapPath("FileB.txt"), 1, true)
do until mile.AtEndOfStream
lineData = lcase(fileb.ReadLine())
if instr(lineData,s)>0 then
a(0,1)=a(0,1)+1
end if
Loop
a(0,2)=0
a(1,2)=" - Entries in File C"
set filec = fso.OpenTextFile(Server.MapPath("FileC.txt"), 1, true)
do until payne.AtEndOfStream
lineData = lcase(filec.ReadLine())
if instr(lineData,s)>0 then
a(0,2)=a(0,2)+1
end if
Loop
%>
The code essentially looks for the number of entries in a text file. What I need is it to be sorted such that the file with the most number of entries comes first.
Suppose there are 10 entries in FileA, 12 in FileB and 7 in FileC. I'd like the output to be displayed like this:
12 - Entries in File B
10 - Entries in File A
7 - Entries in File C
I'm guessing it won't be too complicated since response.write(a(0,i)&a(1,i)) will work. I just need help with the loop or any sorting method if there is one.
Any help I can get in here will be much appreciated.
This will be a "neo-answer" that should help you get to where you want to go, both in the short- and long-term.
1) First, a suggestion for further reading to help you address this sort of problem in a more general way -- and to help you develop your "chops" as you go. You can Google the term "bubble sort" and get a whole host of interesting and mostly helpful input, but here's a link you probably will find most directly helpful, from a brief series of articles on sorting from the 4 Guys from Rolla site, which back in the day was THE place for quality writing on ASP:
https://web.archive.org/web/20211020153403/https://www.4guysfromrolla.com/webtech/011601-1.shtml
You will see that there is a link to an introductory article at the top of this one that covers one-dimensional array sorting, and I recommend it as well. For one, it introduces another sort method, QuickSort, and having multiple tools in your toolbox is almost never a bad idea. (As you will discover, bubble sorting is often the easiest to envision and implement, but because its performance is essentially linear based on the number of items being sorted, can become a performance problem on larger datasets.) Go ahead, check it out; I'll wait 'til you get back...
2) OK, to give you a more concrete approach to address your specific situation here, if the number of files you're reviewing isn't going to be too large, you can do a sort of "final pass" sort to present your results in the desired order.
First, you'll want to introduce a simple global counting variable up toward the top of your code:
dim intMaxEntries
intMaxEntries = 0
Then, at the end of each of your file-parsing runs, you'll want to check the number of entries against intMaxEntries and update intMaxEntries if the number of entries just read in is greater.
if a(0, 1) > intMaxEntries then
intMaxEntries = a(0, 1)
end if
You'll do right after each file reading loop, so the comparison in the above snippet would be done for a(0, 1), a(1, 1), and a(2, 1). More on that repetitive logic at the end.
After you've done all the file reads, intMaxEntries will have the maximum number of entries you've found in one of the files. Then, you can just step down from that value and print out entry counts in the correct order when they match your countdown:
dim i, j
for i = intMaxEntries to 0 step -1
for j = 0 to ubound(a) 'By default gives the upper bound of the 1st dimen.
if a(j, 1) = i then
Response.Write i & a(j, 2) & "<BR>"
end if
next j
next i
This is more than a bit of a hack, and I would encourage you to opt instead for doing a proper sort of your array so that you have something more generalizably useful, but it will work to get you where you want to go, especially if the number of files -- or the maximum number of entries -- isn't too large. You could also clean up my example by introducing the possibility of breaking out of the loops when all the files are accounted for, but I'll let you figure out if that's necessary.
3) You may have just simplified the codebase to get the concept across more cleanly (for which I applaud you if true), but just in case, I would encourage you to look at ways to modularize your work by building your file reading functionality as a function that is simply called with the file and string comparison information needed. (Also, probably an artifact of your snipping, but the "lone", "mile" and "Payne" references in there don't make sense; assuming those are the FSOs you are instantiating and have just forgotten to change them to fileA, fileB and fileC.)
Hope that helps a bit,
Bret
#bret
Someone else came through.
Here's a code that worked perfectly.
Would this be an example of "bubble sort"?
for k=23 to 0 Step-1
for j=0 to k
if (a(0,j)<a(0,j+1)) then
t1=a(0,j+1)
t2=a(1,j+1)
a(0,j+1)=a(0,j)
a(1,j+1)=a(1,j)
a(0,j)=t1
a(1,j)=t2
end If
next
next
for i=0 to 24
if a(0,i)>0 then
response.write (a(0,i)&a(1,i)&"<br>")
end if
next
set objFSO = Server.CreateObject("Scripting.FileSystemObject")
set objFolder = objFSO.GetFolder(server.mappath("Files"))
set objfiles = objFolder.Files
Function filesearch(name)
set searchname = objFSO.OpenTextFile(server.mappath(filename),1, true)
do until searchname.AtEndOfStream
lineData = lcase(searchname.ReadLine())
if instr(lineData,s)>0 then
instances = instances + 1
end if
Loop
End Function
For Each objFile in objFolder.Files
filesearch(objFile)
Response.Write filename & "<br>" & instances & "<br>" & "<br>"
Next
Set objFolder = Nothing
Set objFSO = Nothing
There are a few rough edges but what really bothers me now is the sorting. Where do I keep the bubble sort code?
EDIT:
I've got it work perfect with the following code.
For Each objFile in objFolder.Files
filesearch(objFile)
i = i + 1
a(0,i) = instances
a(1,i) = filename
Next
I was also wondering if there's anyway I could also write the total number of instances. I was able to do it before with:
for i=0 to 43
entries=entries+a(0,i)
next
I cant seem to make it work now.
EDIT:
Works now with:
for i = 0 to n
entries = entries + a(0,i)
next
I'm tying to make something in VBA that will basically list all the files in one or more directories starting from a root folder. Long story short, I'm using filesystemobject to run through all of the folders and then getting all the files in those folders. Moving to the next folder, etc.
The problem I'm running into is that I need to spit out my data (onto a sheet) in the same folder sort order as one might find in Windows. I know this isn't a fixed concept per say, so here's a quick example, as it's displayed in Windows(for me):
Windows Sort Order:
FolderTest\000
FolderTest\0
FolderTest\0001
Not too surprisingly, when using FSO it returns the sub folders in a different (perhaps more logical) order:
FolderTest\0
FolderTest\000
FolderTest\0001
I was hoping someone might have an idea of what one could do to get this to be resorted as it's displaying in Windows. This is just an example obviously, the files could be named anything, but it certainly seems to behave a lot better with alpha characters in the name. I'm not necessarily married to using FSO, but I don't even know where else to look for an alternative. I know I could potentially resort these in an array, but I'm not sure what kind of wizardry would be required to make it sort in the "proper" order. For all I know, there's some method or something that makes this all better. Thanks in advance for any help!
To whoever it may end up helping, the following code looks like it's giving me the results I was looking for, converting a list of subfolders into the same sort orders you (probably) find in Windows Explorer. Feeding in Subfolders from a Filesystem object, it spits the results out in an array (fnames). The code... it's not pretty. I'll be the first to admit it. Don't judge me too harshly. Big thanks #Paddy (see above) for pointing me towards StrCmpLogicalW (http://msdn.microsoft.com/en-us/library/windows/desktop/bb759947(v=vs.85).aspx)
Private Declare PtrSafe Function StrCmpLogicalW Lib "shlwapi" _
(ByVal s1 As String, ByVal s2 As String) As Integer
Sub filefoldersortWindows()
Dim folder As String
Dim fnames() As String, buffer As String, content As String
folder = "Your Path"
Set fsol = CreateObject("Scripting.fileSystemObject")
Set fold = fsol.GetFolder(folder)
FoldCount = fold.SubFolders.Count
ReDim fnames(FoldCount)
cFcount = 0
For Each fld In fold.SubFolders
cFcount = cFcount + 1
Namer$ = fld.Name
fnames(cFcount) = StrConv(Namer, vbUnicode)
Next
For AName = 1 To FoldCount
For BName = (AName + 1) To FoldCount
If StrCmpLogicalW(fnames(AName), fnames(BName)) = 1 Then
buffer = fnames(BName)
fnames(BName) = fnames(AName)
fnames(AName) = buffer
End If
Next
Next
For i = 1 To FoldCount
fnames(i) = StrConv(fnames(i), vbFromUnicode)
If i > 1 Then
content = content & "," & fnames(i)
Else
content = fnames(i)
End If
Next
End Sub
Ahh, I see now. I made a bunch of directories with numeric names to see what's going on. Windows explorer does an integer conversion on the value. The sort rule is like this:
numeric value : ascending
padding length : descending
So, if have 01 and 001, both evaluate to the integer 1, but 001 will appear first because it is longer (has more zero-padding). The 'length' in this case only refers to the numeric part (ie the padding), and is not affected by any characters that appear after (they only matter if the numeric value and the padding length are the same - then normal ordering applies):