Trying to find ";"

[Patrick Santora]

Hey everyone,

I am trying to find ";" ONLY at the end of a string and not within (like within single or double quotes). I believe I will need to use look aheads and look behinds which I have access too by using the underlying java regex engine in CF. Any ideas on what the expression should look like?

Here is the strings to search against:
sScoped = "&quotedString=value";
sScoped = '&quotedString=value';
sScoped = "&" & ";quotedString=value";

NOTE: This is for varScoper, so if you help me fix this i'm sure we can throw your name out on the next release.

Thanks,
Pat Santora
http://patweb99.avatu.com

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date
Get the Free Trial
http://ad.doubleclick.net/clk;203748912;27390454;j

Archive: http://www.houseoffusion.com/groups/RegEx/message.cfm/messageid:1149
Subscription: http://www.houseoffusion.com/groups/RegEx/subscribe.cfm
Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user837.14401.21

[Rob Wilkerson]

On Thu, Jul 10, 2008 at 10:08 AM, Patrick Santora <patweb99@...> wrote:

> Hey everyone,
>
> I am trying to find ";" ONLY at the end of a string and not within (like within single or double quotes). I believe I will need to use look aheads and look behinds which I have access too by using the underlying java regex engine in CF. Any ideas on what the expression should look like?
>
> Here is the strings to search against:
> sScoped = "&quotedString=value";
> sScoped = '&quotedString=value';
> sScoped = "&" & ";quotedString=value";
>
> NOTE: This is for varScoper, so if you help me fix this i'm sure we can throw your name out on the next release.

At the risk of oversimplifying, could this be as easy as ";$"?  In
other words, match any semicolon that appears as the last character in
the line?

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date
Get the Free Trial
http://ad.doubleclick.net/clk;203748912;27390454;j

Archive: http://www.houseoffusion.com/groups/RegEx/message.cfm/messageid:1150
Subscription: http://www.houseoffusion.com/groups/RegEx/subscribe.cfm
Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user837.14401.21

[Patrick Santora]

>On Thu, Jul 10, 2008 at 10:08 AM, Patrick Santora <patweb99@...> wrote:
>
>At the risk of oversimplifying, could this be as easy as ";$"?  In
>other words, match any semicolon that appears as the last character in
>the line?

I don't think so as we also want to catch multiple variables being set to one line (should have put this in the original post):
var1 = 1; var2 = 2;


~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date
Get the Free Trial
http://ad.doubleclick.net/clk;203748912;27390454;j

Archive: http://www.houseoffusion.com/groups/RegEx/message.cfm/messageid:1151
Subscription: http://www.houseoffusion.com/groups/RegEx/subscribe.cfm
Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user837.14401.21

[Peter Boughton]

> I am trying to find ";" ONLY at the end of a string and not within (like within single or
> double quotes).

> I don't think so as we also want to catch multiple variables being set to one line (should have put this in the original post):
> var1 = 1; var2 = 2;

So end of string is irrelevant?
What you actually want to find is all semi-colons excluding those within quotes?

In which case, just replace/remove quoted values first and then search
for any semi-colons.

Text = REReplace ( Text , '"([^"]|"")+"' , '{quotedstring}' , 'all')

(and same again with single/double inverted)

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date
Get the Free Trial
http://ad.doubleclick.net/clk;203748912;27390454;j

Archive: http://www.houseoffusion.com/groups/RegEx/message.cfm/messageid:1152
Subscription: http://www.houseoffusion.com/groups/RegEx/subscribe.cfm
Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user837.14401.21

[Patrick Santora]

hmm. I'm not sure if that will work for me as I don't want to try to alter
the text. I just want to ignore ";" within quotes hence why doing a look
ahead / look behind may be an answer.

I will keep your approach in mind though as it could work if I have to go
down that road.

On Thu, Jul 10, 2008 at 8:29 AM, Peter Boughton <boughtonp@...> wrote:

> > I am trying to find ";" ONLY at the end of a string and not within (like
> within single or
> > double quotes).
>
> > I don't think so as we also want to catch multiple variables being set to
> one line (should have put this in the original post):
> > var1 = 1; var2 = 2;
>
> So end of string is irrelevant?
> What you actually want to find is all semi-colons excluding those within
> quotes?
>
> In which case, just replace/remove quoted values first and then search
> for any semi-colons.
>
> Text = REReplace ( Text , '"([^"]|"")+"' , '{quotedstring}' , 'all')
>
> (and same again with single/double inverted)
>
>

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date
Get the Free Trial
http://ad.doubleclick.net/clk;203748912;27390454;j

Archive: http://www.houseoffusion.com/groups/RegEx/message.cfm/messageid:1153
Subscription: http://www.houseoffusion.com/groups/RegEx/subscribe.cfm
Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user837.14401.21

[Peter Boughton]

> hmm. I'm not sure if that will work for me as I don't want to try to alter
> the text.

Why not?

If you need the original text (e.g. for showing context of errors),
then you just create a copy of the text and work on that, then refer
back to the original version once you know which line numbers the
errors are on.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date
Get the Free Trial
http://ad.doubleclick.net/clk;203748912;27390454;j

Archive: http://www.houseoffusion.com/groups/RegEx/message.cfm/messageid:1154
Subscription: http://www.houseoffusion.com/groups/RegEx/subscribe.cfm
Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user837.14401.21

[Patrick Santora]

Let me rephrase what I stated before. Your approach should work fine, but I
would like to avoid removing characters if possible. I will definatly keep
your idea in mind if I can't do with a look ahead/behind.

The bottom line is I am just trying my best not to have to alter the code
being searched.


On Thu, Jul 10, 2008 at 9:01 AM, Peter Boughton <boughtonp@...> wrote:

> > hmm. I'm not sure if that will work for me as I don't want to try to
> alter
> > the text.
>
> Why not?
>
> If you need the original text (e.g. for showing context of errors),
> then you just create a copy of the text and work on that, then refer
> back to the original version once you know which line numbers the
> errors are on.
>
>

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date
Get the Free Trial
http://ad.doubleclick.net/clk;203748912;27390454;j

Archive: http://www.houseoffusion.com/groups/RegEx/message.cfm/messageid:1155
Subscription: http://www.houseoffusion.com/groups/RegEx/subscribe.cfm
Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user837.14401.21

[Patrick Santora]

Peter,

I just may be overthinking this. heh. I am going to toss your idea in and
see how it floats. It should work fine. I just may need to make some
additional modifictions elsewhere in varScoper.

Thanks for the quick assist.

-Pat
On Thu, Jul 10, 2008 at 9:01 AM, Peter Boughton <boughtonp@...> wrote:

> > hmm. I'm not sure if that will work for me as I don't want to try to
> alter
> > the text.
>
> Why not?
>
> If you need the original text (e.g. for showing context of errors),
> then you just create a copy of the text and work on that, then refer
> back to the original version once you know which line numbers the
> errors are on.
>
>

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date
Get the Free Trial
http://ad.doubleclick.net/clk;203748912;27390454;j

Archive: http://www.houseoffusion.com/groups/RegEx/message.cfm/messageid:1156
Subscription: http://www.houseoffusion.com/groups/RegEx/subscribe.cfm
Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user837.14401.21

[Peter Boughton]

Yes, it's very easy to get carried away with overthinking regex - I've
done it plenty of times. :)


There may be situations where working with only the original text is
required, but I couldn't think of any - once linebreaks are preserved
it should be fine - and I think actually cutting out unwanted parts
way is a simpler method.


However, I have thought up a situation where the regex I supplied would fail...

Consider this:
varY = "ab#LCase("C;DEF;G")#hi";

The nested quotes inside the hashes are valid syntax, but confuse the
regex, and the semicolons would be left when they shouldn't be.

One way of solving that is to first strip all hashed variables - I
think that makes sense, but haven't fully thought it through.

Here's the quick test code I've thrown together for that:

<cfsavecontent variable="reHashed">#([^#]|##)+#</cfsavecontent>
<cfsavecontent variable="reDouble">"([^"]|"")+"</cfsavecontent>
<cfsavecontent variable="reSingle">'([^']|'')+'</cfsavecontent>

<cfoutput>
<pre>
#SourceText#<hr/>
<cfset SourceCopy = jre.replace( SourceCopy , reHashed ,'{hashed}' , 'all' )/>
#SourceCopy#<hr/>
<cfset SourceCopy = jre.replace( SourceCopy , reDouble ,'{dquoted}' , 'all' )/>
#SourceCopy#<hr/>
<cfset SourceCopy = jre.replace( SourceCopy , reSingle  ,'{squoted}' , 'all' )/>
#SourceCopy#<hr/>
</pre>
</cfoutput>

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date
Get the Free Trial
http://ad.doubleclick.net/clk;203748912;27390454;j

Archive: http://www.houseoffusion.com/groups/RegEx/message.cfm/messageid:1157
Subscription: http://www.houseoffusion.com/groups/RegEx/subscribe.cfm
Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user837.14401.21

[Patrick Santora]

Ahh I see, yeah that could pose and issue. I will check out what you sent
and let you know. I probably will not get to it until later today.

Thanks Peter.

On Thu, Jul 10, 2008 at 9:58 AM, Peter Boughton <boughtonp@...> wrote:

> Yes, it's very easy to get carried away with overthinking regex - I've
> done it plenty of times. :)
>
>
> There may be situations where working with only the original text is
> required, but I couldn't think of any - once linebreaks are preserved
> it should be fine - and I think actually cutting out unwanted parts
> way is a simpler method.
>
>
> However, I have thought up a situation where the regex I supplied would
> fail...
>
> Consider this:
> varY = "ab#LCase("C;DEF;G")#hi";
>
> The nested quotes inside the hashes are valid syntax, but confuse the
> regex, and the semicolons would be left when they shouldn't be.
>
> One way of solving that is to first strip all hashed variables - I
> think that makes sense, but haven't fully thought it through.
>
> Here's the quick test code I've thrown together for that:
>
> <cfsavecontent variable="reHashed">#([^#]|##)+#</cfsavecontent>
> <cfsavecontent variable="reDouble">"([^"]|"")+"</cfsavecontent>
> <cfsavecontent variable="reSingle">'([^']|'')+'</cfsavecontent>
>
> <cfoutput>
> <pre>
> #SourceText#<hr/>
> <cfset SourceCopy = jre.replace( SourceCopy , reHashed ,'{hashed}' , 'all'
> )/>
> #SourceCopy#<hr/>
> <cfset SourceCopy = jre.replace( SourceCopy , reDouble ,'{dquoted}' , 'all'
> )/>
> #SourceCopy#<hr/>
> <cfset SourceCopy = jre.replace( SourceCopy , reSingle  ,'{squoted}' ,
> 'all' )/>
> #SourceCopy#<hr/>
> </pre>
> </cfoutput>
>
>

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date
Get the Free Trial
http://ad.doubleclick.net/clk;203748912;27390454;j

Archive: http://www.houseoffusion.com/groups/RegEx/message.cfm/messageid:1158
Subscription: http://www.houseoffusion.com/groups/RegEx/subscribe.cfm
Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user837.14401.21