convert a php regex to coldfusion regex

[phipps_73]

Hi,

I am afraid that my regex skills are weak and I am in the final stage
of converting a php script to a coldfusion one. I am building a string
which can be of the form:

Every first second third fourth Sunday at 11:15

What the php regex does is changes the above to read:

Every first second third and fourth Sunday at 11:15

It adds the " and".

The php regex is as follows:

$wkstr = preg_replace('/(\s)(\w+)$/i',' and $2',$wkstr);

and I need to convert this to a ReReplace function. I have looked at
the php docs to try and decipher the above regex but I don't seem to
be getting anywhere. I also don't know enough about cf regex to know
what I should convert it to.

Can anyone help, please?

Many thanks,

Dave

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date
Get the Free Trial
http://ad.doubleclick.net/clk;207172674;29440083;f

Archive: http://www.houseoffusion.com/groups/regex/message.cfm/messageid:1197
Subscription: http://www.houseoffusion.com/groups/regex/subscribe.cfm
Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=17837.14401.21

[Peter Boughton]

I'm not certain that regex is doing exactly what you say it is, but
simply converting the PHP directly to CFML isas follows:

$wkstr = reReplaceNoCase('(\s)(\w+)$',' and \2',$wkstr)

Although the $ at the start of the variable names are not required in
CF, and the expression doesn't need to be that complex.

This should do the same thing:

wkstr = reReplace( '\w+$' , 'and \0' ,  wkstr )


For any of these expressions to work though, they can't be ran against
the entire "Every first second third fourth Sunday at 11:15" string,
but only on the "first second third fourth" string, otherwise they
will not work as desired.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date
Get the Free Trial
http://ad.doubleclick.net/clk;207172674;29440083;f

Archive: http://www.houseoffusion.com/groups/regex/message.cfm/messageid:1198
Subscription: http://www.houseoffusion.com/groups/regex/subscribe.cfm
Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=17837.14401.21

[Peter Boughton]

Sorry, brain-dead for a moment. :(

The two expressions above are incorrect - they should be

$wkstr = reReplaceNoCase($wkstr,'(\s)(\w+)$',' and \2')

wkstr = reReplace( wkstr , '\w+$','and \0' )

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date
Get the Free Trial
http://ad.doubleclick.net/clk;207172674;29440083;f

Archive: http://www.houseoffusion.com/groups/regex/message.cfm/messageid:1199
Subscription: http://www.houseoffusion.com/groups/regex/subscribe.cfm
Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=17837.14401.21

[phipps_73]

Many thanks Peter,

The " Sunday" was appended after the regex was applied so your
solution works perfectly.

If you have time could you explain what each part of the regex is
doing, so that I can learn for future?

Cheers,

Dave

2008/11/26 Peter Boughton <boughtonp@...>:
> Sorry, brain-dead for a moment. :(
>
> The two expressions above are incorrect - they should be
>
> $wkstr = reReplaceNoCase($wkstr,'(\s)(\w+)$',' and \2')
>
> wkstr = reReplace( wkstr , '\w+$','and \0' )
>
>

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date
Get the Free Trial
http://ad.doubleclick.net/clk;207172674;29440083;f

Archive: http://www.houseoffusion.com/groups/regex/message.cfm/messageid:1200
Subscription: http://www.houseoffusion.com/groups/regex/subscribe.cfm
Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=17837.14401.21

[Peter Boughton]

> If you have time could you explain what each part of the regex is
> doing, so that I can learn for future?

Sure. :)


Starting with the original PHP one:
/(\s)(\w+)$/i

the /.../i means "a regular expression, using the i flag"
the i flag is "case-Insensitive"

the (\s) is "a group consisting of a single whiteSpace character
(space,tab,newline)"

the (\w+) is "a group consisting of one or more Word characters"
the + part of that means "one or more"

the $ at the end means "end of line" or "end of expression", depending
on if the expression is multi-line or not. It matches the position,
not a character (it is a zero-width match).


The slashes are used for PHP when you want to specify a flag - if you
had no flags, the slashes are not necessary.
Some languages (JavaScript) can work with non-quoted slashes, whilst
others (CFML) do not use the slash convention at all.

In all of these, there is an alternative way to specify flags, using
the construct (?i) - note that anything (?...) is a special group,
that acts differently to other groups. (note: this construct can be
used for a lot more than just flags - primarily lookarounds, but other
stuff too)

In CFML, you can use reReplaceNoCase instead of the i flag, which is
more readable for people that don't know regex.


In the replace string of " and $2" or " and \2" in the PHP/CFML ones,
the $2 or \2 is a backreference referring to group 2 - i.e. the second
pair of parentheses, containing the \w+ ("one or more word
characters")


In the simplified version, I removed the flag, removed the groups,
removed the space, and changed the replace string, so we ended up
with:

reReplace( wkstr , '\w+$','and \0' )

The flag was unnecessary - the \w means "any word character" and is
not case sensitive.

The \s was unnecessary because the \w+ will capture all of the
characters upto the space, and we were just putting the space back in,
so I took both the \s out and the space before the "and" out also.

With the \s gone, there was no need to group the characters and use \2
(or \1 as it would have become) - instead I used \0 in the replace
string, which means "the entire content of the match", rather than one
of the groups within it.


So, I think that covers everything - hopefully it all makes sense, and
wasn't information overload.
Let me know if you'd like any part clarified/re-worded. :)

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date
Get the Free Trial
http://ad.doubleclick.net/clk;207172674;29440083;f

Archive: http://www.houseoffusion.com/groups/regex/message.cfm/messageid:1201
Subscription: http://www.houseoffusion.com/groups/regex/subscribe.cfm
Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=17837.14401.21

[phipps_73]

Brilliant, thanks Peter. One day I will "click" with regex!

Cheers,

Dave

2008/11/26 Peter Boughton <boughtonp@...>:

>> If you have time could you explain what each part of the regex is
>> doing, so that I can learn for future?
>
> Sure. :)
>
>
> Starting with the original PHP one:
> /(\s)(\w+)$/i
>
> the /.../i means "a regular expression, using the i flag"
> the i flag is "case-Insensitive"
>
> the (\s) is "a group consisting of a single whiteSpace character
> (space,tab,newline)"
>
> the (\w+) is "a group consisting of one or more Word characters"
> the + part of that means "one or more"
>
> the $ at the end means "end of line" or "end of expression", depending
> on if the expression is multi-line or not. It matches the position,
> not a character (it is a zero-width match).
>
>
> The slashes are used for PHP when you want to specify a flag - if you
> had no flags, the slashes are not necessary.
> Some languages (JavaScript) can work with non-quoted slashes, whilst
> others (CFML) do not use the slash convention at all.
>
> In all of these, there is an alternative way to specify flags, using
> the construct (?i) - note that anything (?...) is a special group,
> that acts differently to other groups. (note: this construct can be
> used for a lot more than just flags - primarily lookarounds, but other
> stuff too)
>
> In CFML, you can use reReplaceNoCase instead of the i flag, which is
> more readable for people that don't know regex.
>
>
> In the replace string of " and $2" or " and \2" in the PHP/CFML ones,
> the $2 or \2 is a backreference referring to group 2 - i.e. the second
> pair of parentheses, containing the \w+ ("one or more word
> characters")
>
>
> In the simplified version, I removed the flag, removed the groups,
> removed the space, and changed the replace string, so we ended up
> with:
>
> reReplace( wkstr , '\w+$','and \0' )
>
> The flag was unnecessary - the \w means "any word character" and is
> not case sensitive.
>
> The \s was unnecessary because the \w+ will capture all of the
> characters upto the space, and we were just putting the space back in,
> so I took both the \s out and the space before the "and" out also.
>
> With the \s gone, there was no need to group the characters and use \2
> (or \1 as it would have become) - instead I used \0 in the replace
> string, which means "the entire content of the match", rather than one
> of the groups within it.
>
>
> So, I think that covers everything - hopefully it all makes sense, and
> wasn't information overload.
> Let me know if you'd like any part clarified/re-worded. :)
>
>

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date
Get the Free Trial
http://ad.doubleclick.net/clk;207172674;29440083;f

Archive: http://www.houseoffusion.com/groups/regex/message.cfm/messageid:1202
Subscription: http://www.houseoffusion.com/groups/regex/subscribe.cfm
Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=17837.14401.21