I have an excel sheet with strings in the cells in column C. In the
corresponding cell in column D, I want words of the type "TKCDT" or
"TKR25" or "02587" or "25YHT" ie words having capital letters and/or
numbers to be displayed. It does not matter whether this is
accomplished using a fucnction or vba.

Thanks in advance for all the help.

Rajendra

I think you are going to have to provide a little more detail of what you
want to do for us. What kind of string values do you have in column C and
what link is there between those strings and the "words" you want to place
in column D? For example, what in column C dictates that TKCDT is put on
column D? Same question for TKR25 or 02587 etc.

Rick


"Raj" wrote in message
...
I have an excel sheet with strings in the cells in column C. In the
corresponding cell in column D, I want words of the type "TKCDT" or
"TKR25" or "02587" or "25YHT" ie words having capital letters and/or
numbers to be displayed. It does not matter whether this is
accomplished using a fucnction or vba.

Thanks in advance for all the help.

Rajendra

On Nov 29, 9:25 pm, "Rick Rothstein \(MVP - VB\)"
wrote:
I think you are going to have to provide a little more detail of what you
want to do for us. What kind of string values do you have in column C and
what link is there between those strings and the "words" you want to place
in column D? For example, what in column C dictates that TKCDT is put on
column D? Same question for TKR25 or 02587 etc.

Rick

"Raj" wrote in message

...



I have an excel sheet with strings in the cells in column C. In the
corresponding cell in column D, I want words of the type "TKCDT" or
"TKR25" or "02587" or "25YHT" ie words having capital letters and/or
numbers to be displayed. It does not matter whether this is
accomplished using a fucnction or vba.

Thanks in advance for all the help.

Rajendra- Hide quoted text -

- Show quoted text -

Thanks Rick.

Cell C2 (and other cells in column C) have sentences like "The TKRTC
value in the 765TW field is not the default.". In cell
D2(corresponding cell in column D), I want TKRTC and 765TW to be
extracted ie words made up of capital letters and/or numbers in the
sentence in cell C2. And repeat the same for all cells in column C.
(Maybe I was wrong in using "string" in place of sentence in my
previous post).

Thanks in advance.

I think this macro will do what you want. Right click the sheet tab where
your data is, select View Code from the popup menu that appears and
copy/paste the macro below into the code window that appears. Run it from
either the VBA editor environment or click Alt+F8 from the sheet where your
data is and run it from there.

Rick

Sub GetCapWords()
Dim I As Long
Dim J As Long
Dim Result As String
Dim Words() As String
Const FirstDataRow = 2
Const DataColumn = 3 'Column C
Const CopyColumn = 4 'Column D
For I = FirstDataRow To Cells(Rows.Count, DataColumn).End(xlUp).Row
Words = Split(Cells(I, DataColumn).Value, " ")
For J = 0 To UBound(Words)
If Words(J) = UCase(Words(J)) Then
Result = Result & ", " & Words(J)
End If
Next
If Len(Result) 0 Then
Result = Mid(Result, 3)
End If
Cells(I, CopyColumn).Value = Result
Result = ""
Next
End Sub



"Raj" wrote in message
...
On Nov 29, 9:25 pm, "Rick Rothstein \(MVP - VB\)"
wrote:
I think you are going to have to provide a little more detail of what you
want to do for us. What kind of string values do you have in column C and
what link is there between those strings and the "words" you want to
place
in column D? For example, what in column C dictates that TKCDT is put on
column D? Same question for TKR25 or 02587 etc.

Rick

"Raj" wrote in message

...



I have an excel sheet with strings in the cells in column C. In the
corresponding cell in column D, I want words of the type "TKCDT" or
"TKR25" or "02587" or "25YHT" ie words having capital letters and/or
numbers to be displayed. It does not matter whether this is
accomplished using a fucnction or vba.

Thanks in advance for all the help.

Rajendra- Hide quoted text -

- Show quoted text -

Thanks Rick.

Cell C2 (and other cells in column C) have sentences like "The TKRTC
value in the 765TW field is not the default.". In cell
D2(corresponding cell in column D), I want TKRTC and 765TW to be
extracted ie words made up of capital letters and/or numbers in the
sentence in cell C2. And repeat the same for all cells in column C.
(Maybe I was wrong in using "string" in place of sentence in my
previous post).

Thanks in advance.

On Thu, 29 Nov 2007 07:47:27 -0800 (PST), Raj wrote:

I have an excel sheet with strings in the cells in column C. In the
corresponding cell in column D, I want words of the type "TKCDT" or
"TKR25" or "02587" or "25YHT" ie words having capital letters and/or
numbers to be displayed. It does not matter whether this is
accomplished using a fucnction or vba.

Thanks in advance for all the help.

Rajendra

Here is a UDF that will return the words specified, separated by a <space. If
you want a different separator, or none, the UDF can be modified to allow that.

To enter it, <alt-F11 opens the VB Editor. Ensure your project is highlighted
in the Project Explorer window, then Insert/Module and paste the code below
into the window that opens.

To use this, enter the formula =reSpecial(cell_ref) in the cell in column D
where you want the results. cell_ref is the corresponding cell in column C.

====================================
Option Explicit
Function reSpecial(str As String) As String
Dim re As Object, mc As Object
Dim I As Long
Const sPat As String = "\b[A-Z0-9]+\b"

Set re = CreateObject("vbscript.regexp")
re.Global = True
re.Pattern = sPat
re.ignorecase = False

If re.test(str) = True Then
Set mc = re.Execute(str)
For I = 0 To mc.Count - 1
reSpecial = reSpecial & mc(I) & " " '<space as separator
Next I
reSpecial = Trim(reSpecial)
End If
End Function
=======================================
--ron

"Ron Rosenfeld" wrote in message
...
On Thu, 29 Nov 2007 07:47:27 -0800 (PST), Raj wrote:

I have an excel sheet with strings in the cells in column C. In the
corresponding cell in column D, I want words of the type "TKCDT" or
"TKR25" or "02587" or "25YHT" ie words having capital letters and/or
numbers to be displayed. It does not matter whether this is
accomplished using a fucnction or vba.

Thanks in advance for all the help.

Rajendra

Here is a UDF that will return the words specified, separated by a
<space. If
you want a different separator, or none, the UDF can be modified to allow
that.

To enter it, <alt-F11 opens the VB Editor. Ensure your project is
highlighted
in the Project Explorer window, then Insert/Module and paste the code
below
into the window that opens.

To use this, enter the formula =reSpecial(cell_ref) in the cell in column
D
where you want the results. cell_ref is the corresponding cell in column
C.

====================================
Option Explicit
Function reSpecial(str As String) As String
Dim re As Object, mc As Object
Dim I As Long
Const sPat As String = "\b[A-Z0-9]+\b"

Set re = CreateObject("vbscript.regexp")
re.Global = True
re.Pattern = sPat
re.ignorecase = False

If re.test(str) = True Then
Set mc = re.Execute(str)
For I = 0 To mc.Count - 1
reSpecial = reSpecial & mc(I) & " " '<space as separator
Next I
reSpecial = Trim(reSpecial)
End If
End Function
=======================================
--ron

For those who might be interested in a non-RegEx solution, here is my
offering for a UDF (also using a space delimiter between found words)...

Function GetCapWords(R As Range) As String
Dim X As Long
Dim Words() As String
Words = Split(R, " ")
For X = 0 To UBound(Words)
If Words(X) = UCase(Words(X)) Then
GetCapWords = GetCapWords & " " & Words(X)
End If
Next
If Len(GetCapWords) 0 Then
GetCapWords = Mid(GetCapWords, 3)
End If
End Function

What surprises me a little, Ron, is that it looks shorter than the RegEx
solution... not what I would have expected.

Rick

On Thu, 29 Nov 2007 15:39:15 -0500, "Rick Rothstein \(MVP - VB\)"
wrote:

What surprises me a little, Ron, is that it looks shorter than the RegEx
solution... not what I would have expected.

Well, here's a shorter, possibly better optimized version:

===============================
Function reSpecial(str As String) As String
Dim re As Object, mc As Object, m As Object
Set re = CreateObject("vbscript.regexp")
re.Global = True
re.Pattern = "\b[A-Z0-9]+\b"
Set mc = re.Execute(str)
For Each m In mc
reSpecial = reSpecial & m & " "
Next m
reSpecial = Trim(reSpecial)
End Function
======================================

But I think the big advantage is in development time. If the OP wants to
change a parameter, it's probably just going to involve changing the regex.
--ron

On Thu, 29 Nov 2007 15:39:15 -0500, "Rick Rothstein \(MVP - VB\)"
wrote:

Function GetCapWords(R As Range) As String
Dim X As Long
Dim Words() As String
Words = Split(R, " ")
For X = 0 To UBound(Words)
If Words(X) = UCase(Words(X)) Then
GetCapWords = GetCapWords & " " & Words(X)
End If
Next
If Len(GetCapWords) 0 Then
GetCapWords = Mid(GetCapWords, 3)
End If
End Function

What surprises me a little, Ron, is that it looks shorter than the RegEx
solution... not what I would have expected.

Rick,

Given this string:

The TKRTC value in the 765TW field is not the default

Your routine, on my machine, returns:
KRTC 765TW
(The initial T is missing)

Also, given a multiline string:

The TKRTC
value in the 765TW field is not the default

(with no space after TKRTC), your routine returns:
65TW

missing TKRTC and also again missing the first character.
--ron

As for the missing first letter... that's what happens when you rush. My
brother is due any minute for dinner, so I had to get ready... I made a last
minute change from my original delimiter of a comma-space to just the space
that you used and I forgot to adjust the code for the length difference. I
change my code to include a Delimiter constant (set to equal a space; but
changeable to any combination of characters) and put code in to account for
it. As for the multi-line problem... I never even gave it a thought
originally. Below is a revised function that handles both problems (thanks
for catching them)...

Function GetCapWords(R As Range) As String
Dim X As Long
Dim Words() As String
Const Delimiter = " "
Words = Split(Replace(Replace(R.Value, vbLf, " "), vbCr, " "))
For X = 0 To UBound(Words)
If Words(X) = UCase(Words(X)) Then
GetCapWords = GetCapWords & Delimiter & Words(X)
End If
Next
If Len(GetCapWords) 0 Then
GetCapWords = Mid(GetCapWords, 1 + Len(Delimiter))
End If
End Function

Rick


"Ron Rosenfeld" wrote in message
...
On Thu, 29 Nov 2007 15:39:15 -0500, "Rick Rothstein \(MVP - VB\)"
wrote:

Function GetCapWords(R As Range) As String
Dim X As Long
Dim Words() As String
Words = Split(R, " ")
For X = 0 To UBound(Words)
If Words(X) = UCase(Words(X)) Then
GetCapWords = GetCapWords & " " & Words(X)
End If
Next
If Len(GetCapWords) 0 Then
GetCapWords = Mid(GetCapWords, 3)
End If
End Function

What surprises me a little, Ron, is that it looks shorter than the RegEx
solution... not what I would have expected.

Rick,

Given this string:

The TKRTC value in the 765TW field is not the default

Your routine, on my machine, returns:
KRTC 765TW
(The initial T is missing)

Also, given a multiline string:

The TKRTC
value in the 765TW field is not the default

(with no space after TKRTC), your routine returns:
65TW

missing TKRTC and also again missing the first character.
--ron

On Thu, 29 Nov 2007 17:39:14 -0500, "Rick Rothstein \(MVP - VB\)"
wrote:

As for the missing first letter... that's what happens when you rush. My
brother is due any minute for dinner, so I had to get ready... I made a last
minute change from my original delimiter of a comma-space to just the space
that you used and I forgot to adjust the code for the length difference. I
change my code to include a Delimiter constant (set to equal a space; but
changeable to any combination of characters) and put code in to account for
it. As for the multi-line problem... I never even gave it a thought
originally. Below is a revised function that handles both problems (thanks
for catching them)...

Function GetCapWords(R As Range) As String
Dim X As Long
Dim Words() As String
Const Delimiter = " "
Words = Split(Replace(Replace(R.Value, vbLf, " "), vbCr, " "))
For X = 0 To UBound(Words)
If Words(X) = UCase(Words(X)) Then
GetCapWords = GetCapWords & Delimiter & Words(X)
End If
Next
If Len(GetCapWords) 0 Then
GetCapWords = Mid(GetCapWords, 1 + Len(Delimiter))
End If
End Function

Rick

You're getting there. Still a problem with returning extra spaces -- if there
are multiple spaces after the CapWord, your routine returns them, also.

It seems that what is happening is that if there are multiple, sequential
spaces, the Split function returns a null string ("") for the spaces after the
first. Since "" = Ucase(""), the null string gets concatenated, followed by
your Delimiter.

Enjoy your dinner! We just had a wonderful one. My mother-in-law and niece
are visiting from the Azores, and they go home tomorrow. My wife and niece
made a great "going away" meal, and some of our kids were over, too.
--ron

You're getting there. Still a problem with returning extra spaces -- if
there
are multiple spaces after the CapWord, your routine returns them, also.

It seems that what is happening is that if there are multiple, sequential
spaces, the Split function returns a null string ("") for the spaces after
the
first. Since "" = Ucase(""), the null string gets concatenated, followed
by
your Delimiter.

Yes, the blanks do cause a problem because the test I was doing was too
simple. All I checked for was that the word equaled its capitalization.
Well, as it turns out, an empty string equals the empty string with the
UCase function applied and so it passed the test. Actually, in examining
this problem, another problem was highlighted... if there is stand-alone
non-letters, these too passed my simple test; hence, something like "(*)"
was added to the return value. This part is easy enough to fix and the code
to do that is shown after my signature below. However, there is another
difference between our codes and I am not sure which of our treatments
should be considered the "correct" one. If the string of text is this, for
example...

"One (TWO) Three"

my routine returns (TWO) with the parentheses and your routine returns just
TWO. The OP said in his Subject line as well as the body of his first
posting that the items he wanted to find were "words"... so, should the
surrounding parentheses be retained as I do (because they are attached to
the upper case letters and/or digits) or removed as you do (even though they
are attached)? Or are we both wrong and the "word" should be rejected
altogether because it is not composed of just upper case letters and/or
digits? We will need input from the OP on this issue.

Now, while testing the above, I came across a problem with your function. If
there are any internal non-letters/non-digits, your regular expression
appears to treat them as blanks and separates the word into parts, using a
blank, for each such character. So, for this sentence...

"One TWENTY-ONE Three"

your routine returns TWENTY ONE with a space in the middle. Your code does a
similar thing with something like...

"Model #AB(12)CD"

returning AB CD EF with two internal blanks.


Enjoy your dinner!

Thank you... I did.


We just had a wonderful one. My mother-in-law and niece
are visiting from the Azores, and they go home tomorrow. My wife and
niece
made a great "going away" meal, and some of our kids were over, too.


Yes, that sounds like it was nice too. I glad you had such a nice time.


Rick

'The repaired function
'==============================
Function GetCapWords(R As Range) As String
Dim X As Long
Dim Words() As String
Const Delimiter = " "
Words = Split(Replace(Replace(R.Value, vbLf, " "), vbCr, " "))
For X = 0 To UBound(Words)
If Words(X) = UCase(Words(X)) And Words(X) Like "*[A-Z]*" Then
GetCapWords = GetCapWords & Delimiter & Words(X)
End If
Next
If Len(GetCapWords) 0 Then
GetCapWords = Mid(GetCapWords, 1 + Len(Delimiter))
End If
End Function

On Nov 30, 8:39 am, "Rick Rothstein \(MVP - VB\)"
wrote:
You're getting there. Still a problem with returning extra spaces -- if
there
are multiple spaces after the CapWord, your routine returns them, also.

It seems that what is happening is that if there are multiple, sequential
spaces, the Split function returns a null string ("") for the spaces after
the
first. Since "" = Ucase(""), the null string gets concatenated, followed
by
your Delimiter.

Yes, the blanks do cause a problem because the test I was doing was too
simple. All I checked for was that the word equaled its capitalization.
Well, as it turns out, an empty string equals the empty string with the
UCase function applied and so it passed the test. Actually, in examining
this problem, another problem was highlighted... if there is stand-alone
non-letters, these too passed my simple test; hence, something like "(*)"
was added to the return value. This part is easy enough to fix and the code
to do that is shown after my signature below. However, there is another
difference between our codes and I am not sure which of our treatments
should be considered the "correct" one. If the string of text is this, for
example...

"One (TWO) Three"

my routine returns (TWO) with the parentheses and your routine returns just
TWO. The OP said in his Subject line as well as the body of his first
posting that the items he wanted to find were "words"... so, should the
surrounding parentheses be retained as I do (because they are attached to
the upper case letters and/or digits) or removed as you do (even though they
are attached)? Or are we both wrong and the "word" should be rejected
altogether because it is not composed of just upper case letters and/or
digits? We will need input from the OP on this issue.

Now, while testing the above, I came across a problem with your function. If
there are any internal non-letters/non-digits, your regular expression
appears to treat them as blanks and separates the word into parts, using a
blank, for each such character. So, for this sentence...

"One TWENTY-ONE Three"

your routine returns TWENTY ONE with a space in the middle. Your code does a
similar thing with something like...

"Model #AB(12)CD"

returning AB CD EF with two internal blanks.

Enjoy your dinner!

Thank you... I did.

We just had a wonderful one. My mother-in-law and niece
are visiting from the Azores, and they go home tomorrow. My wife and
niece
made a great "going away" meal, and some of our kids were over, too.

Yes, that sounds like it was nice too. I glad you had such a nice time.

Rick

'The repaired function
'==============================
Function GetCapWords(R As Range) As String
Dim X As Long
Dim Words() As String
Const Delimiter = " "
Words = Split(Replace(Replace(R.Value, vbLf, " "), vbCr, " "))
For X = 0 To UBound(Words)
If Words(X) = UCase(Words(X)) And Words(X) Like "*[A-Z]*" Then
GetCapWords = GetCapWords & Delimiter & Words(X)
End If
Next
If Len(GetCapWords) 0 Then
GetCapWords = Mid(GetCapWords, 1 + Len(Delimiter))
End If
End Function

Thanks Rick and Ron for all the effort.

In column C is the sentence: " His code is CND8599, and pin is 2588."
In column D the function GetCapWords is extracting "CND8599," but not
extracting "2588"
In column E the function reSpecial is extracting "CND8599 2588"
Observations:
reSpecial meets my requirements as it extracts Words Capitals and/or
Numbers. An additional bonus(meeting my unstated requirement) is that
it ignores the punctuation marks ie "," at the end of "CND8599".
I would still love to see Rick's non regexp solution with necessary
changes for displaying words containing only numbers and also ignoring
punctuations at the end of words (if possible)

Thanks once again
Raj

I would still love to see Rick's non regexp solution with necessary
changes for displaying words containing only numbers and also ignoring
punctuations at the end of words (if possible)

I think this will work...

Function GetCapWords(R As Range) As String
Dim X As Long
Dim Words() As String
Const Delimiter = " "
Words = Split(Replace(Replace(R.Value, vbLf, " "), vbCr, " "))
For X = 0 To UBound(Words)
If Words(X) = UCase(Words(X)) And Words(X) Like "*[A-Z0-9]*" Then
GetCapWords = GetCapWords & Delimiter & Words(X)
End If
Next
If Len(GetCapWords) 0 Then
GetCapWords = Mid$(GetCapWords, 1 + Len(Delimiter))
End If
For X = 1 To Len(GetCapWords)
If Mid$(GetCapWords, X, 1) Like "[!A-Z0-9 ]" Then
Mid$(GetCapWords, X, 1) = "`"
End If
Next
GetCapWords = Replace(GetCapWords, "`", "")
End Function


Rick

On Thu, 29 Nov 2007 22:39:42 -0500, "Rick Rothstein \(MVP - VB\)"
wrote:

You're getting there. Still a problem with returning extra spaces -- if
there
are multiple spaces after the CapWord, your routine returns them, also.

It seems that what is happening is that if there are multiple, sequential
spaces, the Split function returns a null string ("") for the spaces after
the
first. Since "" = Ucase(""), the null string gets concatenated, followed
by
your Delimiter.

Yes, the blanks do cause a problem because the test I was doing was too
simple. All I checked for was that the word equaled its capitalization.
Well, as it turns out, an empty string equals the empty string with the
UCase function applied and so it passed the test. Actually, in examining
this problem, another problem was highlighted... if there is stand-alone
non-letters, these too passed my simple test; hence, something like "(*)"
was added to the return value. This part is easy enough to fix and the code
to do that is shown after my signature below. However, there is another
difference between our codes and I am not sure which of our treatments
should be considered the "correct" one. If the string of text is this, for
example...

"One (TWO) Three"

my routine returns (TWO) with the parentheses and your routine returns just
TWO. The OP said in his Subject line as well as the body of his first
posting that the items he wanted to find were "words"... so, should the
surrounding parentheses be retained as I do (because they are attached to
the upper case letters and/or digits) or removed as you do (even though they
are attached)? Or are we both wrong and the "word" should be rejected
altogether because it is not composed of just upper case letters and/or
digits? We will need input from the OP on this issue.

Now, while testing the above, I came across a problem with your function. If
there are any internal non-letters/non-digits, your regular expression
appears to treat them as blanks and separates the word into parts, using a
blank, for each such character. So, for this sentence...

"One TWENTY-ONE Three"

your routine returns TWENTY ONE with a space in the middle. Your code does a
similar thing with something like...

"Model #AB(12)CD"

returning AB CD EF with two internal blanks.



Those results depend on the definition of a "word containing caps and letters"

My interpretation was that special characters are not in the class of caps and
letters.

But if the OP wanted to include them, it would be fairly simple to modify the
regex to do so.

Also, for word boundaries, I just used the regex definition which is the point
between a word character (letter, digit, underscore) and a non-word character.

Again, this is something that can be modified merely by changing the regex, if
required by the OP. The rest of the VBA code does not need to be changed at
all.
--ron

On Thu, 29 Nov 2007 20:34:30 -0800 (PST), Raj wrote:

reSpecial meets my requirements as it extracts Words Capitals and/or
Numbers. An additional bonus(meeting my unstated requirement) is that
it ignores the punctuation marks ie "," at the end of "CND8599".

Raj,

I'm glad it meets your requirements. And I would not consider that omitting
the punctuation was an "unstated" requirement. You did state you wanted
"words" returned and that they were defined as consisting only of Capital
letters and Numbers. So punctuation would not have been in the class.

What you did leave unstated was what kind of delimiter you wanted between the
words. I assumed a <space -- that turned out to be correct.

In any event, I was happy to help, and I'm glad it meets your requirements. IT
also gave me an opportunity to clean up some of the surrounding VBA code.
--ron

Also, for word boundaries, I just used the regex definition which is the
point
between a word character (letter, digit, underscore) and a non-word
character.

I just looked back at your code and realized that the regex "language" that
I knew must have changed over the years or, more likely, having been a UNIX
based implementation, simply differs in construction. I pulled some old
notes I had laying around from my "old days" and the \b which you used (for
a blank space I am guessing) used to me the backspace character to me "back
in the day". We had a \s character which stood for "white space" which were
the blank space (obviously<g), horizontal tabs, vertical tabs (line feeds)
and carriage returns. I think, but am not sure, that other characters stood
for themselves (unless they were a meta-character like the backslash in
which they were "escaped" by preceding them with a backslash) and, again I
think, that the words were separated by white space only. As I said earlier,
I've forgotten almost everything I knew about regular expressions because I
haven't touched them for 20+ years; however, I'm getting the impression they
have evolved over the years and "work" differently than how they did back in
the mid-1980s when I worked with them.

Rick

On Fri, 30 Nov 2007 12:43:40 -0500, "Rick Rothstein \(MVP - VB\)"
wrote:

Also, for word boundaries, I just used the regex definition which is the
point
between a word character (letter, digit, underscore) and a non-word
character.

I just looked back at your code and realized that the regex "language" that
I knew must have changed over the years or, more likely, having been a UNIX
based implementation, simply differs in construction. I pulled some old
notes I had laying around from my "old days" and the \b which you used (for
a blank space I am guessing) used to me the backspace character to me "back
in the day". We had a \s character which stood for "white space" which were
the blank space (obviously<g), horizontal tabs, vertical tabs (line feeds)
and carriage returns. I think, but am not sure, that other characters stood
for themselves (unless they were a meta-character like the backslash in
which they were "escaped" by preceding them with a backslash) and, again I
think, that the words were separated by white space only. As I said earlier,
I've forgotten almost everything I knew about regular expressions because I
haven't touched them for 20+ years; however, I'm getting the impression they
have evolved over the years and "work" differently than how they did back in
the mid-1980s when I worked with them.

Rick

I didn't work with regular expressions back then so can't comment.

But in the usage with which I am familiar, \b only stands for a backspace if it
is within a character class e.g. [\b] would be a backspace.

\b does NOT stand for a blank space but rather for a "word boundary" which is
defined as the junction of \w and \W as well as at the start and/or end of the
string if the first and/or last characters in the string are word characters.

\w stands for any word character, which is further defined as in the class:
[A-Za-z0-9_] (note the underscore)

\W is the negation of \w (any character NOT in the class)

Also, \d stands for any digit, and \D stands for a non-digit.

\s is the same as your recollection. And \S is any character not in the class
of \s.

And also there are a number of different flavors, with subtle (and not so
subtle) rule differences. I restrict myself to VBScript (which is the same as
Javascript).

So the regex I used could also be written:
"\b[A-Z\d]+\b"

and explained as:

"\b" & ' Assert position at a word boundary
"[A-Z\d]" & ' Match a single character present in the list below
' A character in the range between “A” and “Z”
' A single digit 0..9
"+" & ' Between one and unlimited times, as many times as possible,
giving back as needed (greedy)
"\b" ' Assert position at a word boundary

Best,
--ron

"Ron Rosenfeld" wrote in message
...
On Fri, 30 Nov 2007 12:43:40 -0500, "Rick Rothstein \(MVP - VB\)"
wrote:

Also, for word boundaries, I just used the regex definition which is the
point
between a word character (letter, digit, underscore) and a non-word
character.

I just looked back at your code and realized that the regex "language"
that
I knew must have changed over the years or, more likely, having been a
UNIX
based implementation, simply differs in construction. I pulled some old
notes I had laying around from my "old days" and the \b which you used
(for
a blank space I am guessing) used to me the backspace character to me
"back
in the day". We had a \s character which stood for "white space" which
were
the blank space (obviously<g), horizontal tabs, vertical tabs (line
feeds)
and carriage returns. I think, but am not sure, that other characters
stood
for themselves (unless they were a meta-character like the backslash in
which they were "escaped" by preceding them with a backslash) and, again I
think, that the words were separated by white space only. As I said
earlier,
I've forgotten almost everything I knew about regular expressions because
I
haven't touched them for 20+ years; however, I'm getting the impression
they
have evolved over the years and "work" differently than how they did back
in
the mid-1980s when I worked with them.

Rick

I didn't work with regular expressions back then so can't comment.

But in the usage with which I am familiar, \b only stands for a backspace
if it
is within a character class e.g. [\b] would be a backspace.

\b does NOT stand for a blank space but rather for a "word boundary" which
is
defined as the junction of \w and \W as well as at the start and/or end of
the
string if the first and/or last characters in the string are word
characters.

\w stands for any word character, which is further defined as in the
class:
[A-Za-z0-9_] (note the underscore)

\W is the negation of \w (any character NOT in the class)

Also, \d stands for any digit, and \D stands for a non-digit.

\s is the same as your recollection. And \S is any character not in the
class
of \s.

And also there are a number of different flavors, with subtle (and not so
subtle) rule differences. I restrict myself to VBScript (which is the
same as
Javascript).

So the regex I used could also be written:
"\b[A-Z\d]+\b"

and explained as:

"\b" & ' Assert position at a word boundary
"[A-Z\d]" & ' Match a single character present in the list below
' A character in the range between "A" and "Z"
' A single digit 0..9
"+" & ' Between one and unlimited times, as many times as
possible,
giving back as needed (greedy)
"\b" ' Assert position at a word boundary

Best,

Thanks for the mini-lesson. Some of what you posted looks familiar, but I am
afraid I've been away from it for far too long to be sure of anything any
more. I do remember our UNIX clone was named CLIX and a company named
Intergraph (which made our CADD workstations and software) was the vendor. I
also remember doing most of my work in something called the Bourne Shell and
the scripting language I used mostly was called awk (but I am afraid I
remember very little about it either). Too many years have passed, I'm
getting old, brain cells are dying...<g

Rick

On Sat, 1 Dec 2007 00:32:31 -0500, "Rick Rothstein \(MVP - VB\)"
wrote:

Thanks for the mini-lesson. Some of what you posted looks familiar, but I am
afraid I've been away from it for far too long to be sure of anything any
more. I do remember our UNIX clone was named CLIX and a company named
Intergraph (which made our CADD workstations and software) was the vendor. I
also remember doing most of my work in something called the Bourne Shell and
the scripting language I used mostly was called awk (but I am afraid I
remember very little about it either). Too many years have passed, I'm
getting old, brain cells are dying...<g

Well, the ones that remain seem to be well tuned in to writing concise VBA
code!

I'm really a novice with regular expressions. Harlan Grove got me interested
in them within the past year, and I've found them to be a much easier method of
dealing with all kinds of text manipulations, expecially after getting used to
the vagaries of VB Script.
--ron