RE: [Flashcoders] E4X filtering strange behaviour
That was it, it works now, thanks. Funny, I didn't notice that, I had it in my code before, but somewhere along the line of messing with things, it just switched back over to .toString() and I hadn't noticed. Thanks everyone! One more question if I may - how can I exclude certain things like blank spaces? I mean, I could check the textfield, but it would seem there is an easier way. If I put a blank space (or nothing for that matter) in the input text field and search for that, it returns all nodes. I could check the input text field, but I'd rather do it with a regular expression (both for elegance and to learn). The help docs are skimpy, so excuse so many questions here. It's also such a huge subject, it's very hard to Google exactly what you need. Jason Merrill Bank of America Enterprise Technology Global Risk LLD Instructional Technology Media Join the Bank of America Flash Platform Developer Community Are you a Bank of America associate interested in innovative learning ideas and technologies? Check out our internal GTO Innovative Learning Blog subscribe. ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
Re: [Flashcoders] E4X filtering strange behaviour
I could be wrong but that (user input validation) should be done prior to RegExp construction. Kenneth Kawamoto http://www.materiaprima.co.uk/ Merrill, Jason wrote: That was it, it works now, thanks. Funny, I didn't notice that, I had it in my code before, but somewhere along the line of messing with things, it just switched back over to .toString() and I hadn't noticed. Thanks everyone! One more question if I may - how can I exclude certain things like blank spaces? I mean, I could check the textfield, but it would seem there is an easier way. If I put a blank space (or nothing for that matter) in the input text field and search for that, it returns all nodes. I could check the input text field, but I'd rather do it with a regular expression (both for elegance and to learn). The help docs are skimpy, so excuse so many questions here. It's also such a huge subject, it's very hard to Google exactly what you need. ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
RE: [Flashcoders] E4X filtering strange behaviour
Ah, I know where the confusion came from, you said the code was based on my example, but that was actually Wagner's example, where he put var test:XML = data...etc. (it was based on my example, but he used test as an instance of the XML when requoting my original question) The use of test as the XML name threw me, because it was mixed in with the regex method, re.test - I never realized you meant test as the XML not test as some regex method I wasn't aware of. So when you wrote var xmlListSearch:XMLList = test..*.(re[test](attribute(*)) || re[test](child(*))); You meant var xmlListSearch:XMLList = theXML..*.(re[test](attribute(*)) || re[test](child(*))); Got it, I'll give it a shot, thanks. Jason Merrill Bank of America Enterprise Technology Global Risk LLD Instructional Technology Media Join the Bank of America Flash Platform Developer Community Are you a Bank of America associate interested in innovative learning ideas and technologies? Check out our internal GTO Innovative Learning Blog subscribe. ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
RE: [Flashcoders] E4X filtering strange behaviour
Also this is shorter: var xmlListSearch:XMLList = test..topic.(re.test(@title)); If we don't know anything about the XML we could do something like: var xmlListSearch:XMLList = test..*.(re[test](attribute(*)) || re[test](child(*))); Kenneth - no, what I mean is, your examples aren't pointing to any XML instance at all (that I could see anyway, maybe you could point it out), so how in the world can the AVM know what you're referring to in order to search though and make an XMLList from? Jason Merrill Bank of America Enterprise Technology Global Risk LLD Instructional Technology Media Join the Bank of America Flash Platform Developer Community Are you a Bank of America associate interested in innovative learning ideas and technologies? Check out our internal GTO Innovative Learning Blog subscribe. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Kenneth Kawamoto Sent: Saturday, July 19, 2008 6:19 AM To: Flash Coders List Subject: Re: [Flashcoders] E4X filtering strange behaviour The code was based on your example and therefore we know the XML schema, i.e. we know the node name and they all have the attribute title. If we don't know anything about the XML we could do something like: var xmlListSearch:XMLList = test..*.(re[test](attribute(*)) || re[test](child(*))); Kenneth Kawamoto http://www.materiaprima.co.uk/ Merrill, Jason wrote: Thanks. However, I have been told using @title is not good in most situations because it will return an error if a node doesn't have the attribute, wheras attribute(title) will not. Also this is shorter: var xmlListSearch:XMLList = test..topic.(re.test(@title)); But you still need to specify what XML to search through, no? How would the above know what XML object topic is within using the code above? Jason Merrill ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
Re: [Flashcoders] E4X filtering strange behaviour
Yes that's that :) Anyway in your case you know the incoming XML schema so you shouldn't use wildcard * - I think wildcards have serious performance hits. i.e. theXML.topics.topic. should be used rather than theXML..*. Kenneth Kawamoto http://www.materiaprima.co.uk/ Merrill, Jason wrote: Ah, I know where the confusion came from, you said the code was based on my example, but that was actually Wagner's example, where he put var test:XML = data...etc. (it was based on my example, but he used test as an instance of the XML when requoting my original question) The use of test as the XML name threw me, because it was mixed in with the regex method, re.test - I never realized you meant test as the XML not test as some regex method I wasn't aware of. So when you wrote var xmlListSearch:XMLList = test..*.(re[test](attribute(*)) || re[test](child(*))); You meant var xmlListSearch:XMLList = theXML..*.(re[test](attribute(*)) || re[test](child(*))); Got it, I'll give it a shot, thanks. ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
RE: [Flashcoders] E4X filtering strange behaviour
OK good to know, thanks! Jason Merrill Bank of America Enterprise Technology Global Risk LLD Instructional Technology Media Join the Bank of America Flash Platform Developer Community Are you a Bank of America associate interested in innovative learning ideas and technologies? Check out our internal GTO Innovative Learning Blog subscribe. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Kenneth Kawamoto Sent: Monday, July 21, 2008 12:52 PM To: Flash Coders List Subject: Re: [Flashcoders] E4X filtering strange behaviour Yes that's that :) Anyway in your case you know the incoming XML schema so you shouldn't use wildcard * - I think wildcards have serious performance hits. i.e. theXML.topics.topic. should be used rather than theXML..*. Kenneth Kawamoto http://www.materiaprima.co.uk/ Merrill, Jason wrote: Ah, I know where the confusion came from, you said the code was based on my example, but that was actually Wagner's example, where he put var test:XML = data...etc. (it was based on my example, but he used test as an instance of the XML when requoting my original question) The use of test as the XML name threw me, because it was mixed in with the regex method, re.test - I never realized you meant test as the XML not test as some regex method I wasn't aware of. So when you wrote var xmlListSearch:XMLList = test..*.(re[test](attribute(*)) || re[test](child(*))); You meant var xmlListSearch:XMLList = theXML..*.(re[test](attribute(*)) || re[test](child(*))); Got it, I'll give it a shot, thanks. ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
RE: [Flashcoders] E4X filtering strange behaviour
OK, so this is almost there, but not quite. If I do as Kenneth suggested (and this should be able to be reproduced), var testXML:XML = data topics topic title=Coldplays New Album / topic title=The Dark Knight / topic title=Arrested Development: The Movie / /topics /data; var re:RegExp = new RegExp(^+searchInput.text); var xmlListSearch:XMLList = testXML.topics.topic.(re[test](@title)); trace(result: +xmlListSearch.toXMLString()); (Assuming there is a TextInput text field on the stage and a function run when a button is clicked to run the regex search), if I enter the search term, The, it only returns the second node containing The Dark Knight, not the third topic node, which has a The in it's title attribute as well. If I put Dark, it doesn't return anything. So it seems it's only searching nodes where the search term is the first item. How can I modify the regular expression above to find all nodes containing an element of the search term? Thanks, Jason Merrill Bank of America Enterprise Technology Global Risk LLD Instructional Technology Media Join the Bank of America Flash Platform Developer Community Are you a Bank of America associate interested in innovative learning ideas and technologies? Check out our internal GTO Innovative Learning Blog subscribe. ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
RE: [Flashcoders] E4X filtering strange behaviour
Hi Jason, This works for me: var testXML:XML = data topics topic title=Coldplays New Album / topic title=The Dark Knight / topic title=Arrested Development: The Movie / /topics /data; var re:RegExp = new RegExp(searchInput.text, g); var xmlListSearch:XMLList = testXML.topics.topic.(re[test](@title)); trace(result: +xmlListSearch.toXMLString()); Hth, Bob -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Merrill, Jason Sent: Monday, July 21, 2008 12:27 PM To: Flash Coders List Subject: RE: [Flashcoders] E4X filtering strange behaviour OK, so this is almost there, but not quite. If I do as Kenneth suggested (and this should be able to be reproduced), var testXML:XML = data topics topic title=Coldplays New Album / topic title=The Dark Knight / topic title=Arrested Development: The Movie / /topics /data; var re:RegExp = new RegExp(^+searchInput.text); var xmlListSearch:XMLList = testXML.topics.topic.(re[test](@title)); trace(result: +xmlListSearch.toXMLString()); (Assuming there is a TextInput text field on the stage and a function run when a button is clicked to run the regex search), if I enter the search term, The, it only returns the second node containing The Dark Knight, not the third topic node, which has a The in it's title attribute as well. If I put Dark, it doesn't return anything. So it seems it's only searching nodes where the search term is the first item. How can I modify the regular expression above to find all nodes containing an element of the search term? Thanks, Jason Merrill Bank of America Enterprise Technology Global Risk LLD Instructional Technology Media Join the Bank of America Flash Platform Developer Community Are you a Bank of America associate interested in innovative learning ideas and technologies? Check out our internal GTO Innovative Learning Blog subscribe. ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
RE: [Flashcoders] E4X filtering strange behaviour
var re:RegExp = new RegExp(searchInput.text, g) Thanks - so it looks like I need the global flag, but how do I also pass the i (ignore case) flag if the RegExp constructor only takes the two arguments? Jason Merrill Bank of America Enterprise Technology Global Risk LLD Instructional Technology Media Join the Bank of America Flash Platform Developer Community Are you a Bank of America associate interested in innovative learning ideas and technologies? Check out our internal GTO Innovative Learning Blog subscribe. ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
RE: [Flashcoders] E4X filtering strange behaviour
Wait, think I got it: var re:RegExp = new RegExp(searchInput.text, g,i); var xmlListSearch:XMLList = testXML..*.( re[test]( attribute(title))); trace(result: +xmlListSearch.toString()); seems to work for me. Thanks eveyrone! Jason Merrill Bank of America Enterprise Technology Global Risk LLD Instructional Technology Media Join the Bank of America Flash Platform Developer Community Are you a Bank of America associate interested in innovative learning ideas and technologies? Check out our internal GTO Innovative Learning Blog subscribe. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Robert Leisle Sent: Monday, July 21, 2008 3:54 PM To: 'Flash Coders List' Subject: RE: [Flashcoders] E4X filtering strange behaviour Hi Jason, This works for me: var testXML:XML = data topics topic title=Coldplays New Album / topic title=The Dark Knight / topic title=Arrested Development: The Movie / /topics /data; var re:RegExp = new RegExp(searchInput.text, g); var xmlListSearch:XMLList = testXML.topics.topic.(re[test](@title)); trace(result: +xmlListSearch.toXMLString()); Hth, Bob -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Merrill, Jason Sent: Monday, July 21, 2008 12:27 PM To: Flash Coders List Subject: RE: [Flashcoders] E4X filtering strange behaviour OK, so this is almost there, but not quite. If I do as Kenneth suggested (and this should be able to be reproduced), var testXML:XML = data topics topic title=Coldplays New Album / topic title=The Dark Knight / topic title=Arrested Development: The Movie / /topics /data; var re:RegExp = new RegExp(^+searchInput.text); var xmlListSearch:XMLList = testXML.topics.topic.(re[test](@title)); trace(result: +xmlListSearch.toXMLString()); (Assuming there is a TextInput text field on the stage and a function run when a button is clicked to run the regex search), if I enter the search term, The, it only returns the second node containing The Dark Knight, not the third topic node, which has a The in it's title attribute as well. If I put Dark, it doesn't return anything. So it seems it's only searching nodes where the search term is the first item. How can I modify the regular expression above to find all nodes containing an element of the search term? Thanks, Jason Merrill Bank of America Enterprise Technology Global Risk LLD Instructional Technology Media Join the Bank of America Flash Platform Developer Community Are you a Bank of America associate interested in innovative learning ideas and technologies? Check out our internal GTO Innovative Learning Blog subscribe. ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
RE: [Flashcoders] E4X filtering strange behaviour
OK, I'm like 95% there, but there is still some kind of bug - here is how to reproduce: var testXML:XML = data topics topic title=Coldplay's New Album, Viva La Vida Or Death And All His Friends / topic title=The Dark Knight / topic title=Arrested Development, The Movie / topic title=Band Of Horses: Everything All The Time / topic title=Master And Commander: The Far Side Of The World / topic title=The Quick Brown Fox Jumped Over The Lazy Dog / topic title=Violet Hill / /topics /data; var re:RegExp = new RegExp(The, g,i); var xmlListSearch:XMLList = testXML..*.topic.( re[test]( attribute(title))); trace(result: +xmlListSearch.toString()); If you search for the word, the, as I do when I construct the regular expression, you should get 5 nodes returned into the XMLList, but I only get these four: topic title=The Dark Knight/ topic title=Arrested Development, The Movie/ topic title=Band Of Horses: Everything All The Time/ topic title=Master And Commander: The Far Side Of The World/ I don't get this node: topic title=The Quick Brown Fox Jumped Over The Lazy Dog / ? Any idea why? Does it fail if there is more than one the in the string? Jason Merrill Bank of America Enterprise Technology Global Risk LLD Instructional Technology Media Join the Bank of America Flash Platform Developer Community Are you a Bank of America associate interested in innovative learning ideas and technologies? Check out our internal GTO Innovative Learning Blog subscribe. ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
Re: [Flashcoders] E4X filtering strange behaviour
I don't think you need g (Global flag) for this. Kenneth Kawamoto http://www.materiaprima.co.uk/ Merrill, Jason wrote: Wait, think I got it: var re:RegExp = new RegExp(searchInput.text, g,i); var xmlListSearch:XMLList = testXML..*.( re[test]( attribute(title))); trace(result: +xmlListSearch.toString()); seems to work for me. Thanks eveyrone! Jason Merrill Bank of America Enterprise Technology Global Risk LLD Instructional Technology Media Join the Bank of America Flash Platform Developer Community Are you a Bank of America associate interested in innovative learning ideas and technologies? Check out our internal GTO Innovative Learning Blog subscribe. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Robert Leisle Sent: Monday, July 21, 2008 3:54 PM To: 'Flash Coders List' Subject: RE: [Flashcoders] E4X filtering strange behaviour Hi Jason, This works for me: var testXML:XML = data topics topic title=Coldplays New Album / topic title=The Dark Knight / topic title=Arrested Development: The Movie / /topics /data; var re:RegExp = new RegExp(searchInput.text, g); var xmlListSearch:XMLList = testXML.topics.topic.(re[test](@title)); trace(result: +xmlListSearch.toXMLString()); Hth, Bob -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Merrill, Jason Sent: Monday, July 21, 2008 12:27 PM To: Flash Coders List Subject: RE: [Flashcoders] E4X filtering strange behaviour OK, so this is almost there, but not quite. If I do as Kenneth suggested (and this should be able to be reproduced), var testXML:XML = data topics topic title=Coldplays New Album / topic title=The Dark Knight / topic title=Arrested Development: The Movie / /topics /data; var re:RegExp = new RegExp(^+searchInput.text); var xmlListSearch:XMLList = testXML.topics.topic.(re[test](@title)); trace(result: +xmlListSearch.toXMLString()); (Assuming there is a TextInput text field on the stage and a function run when a button is clicked to run the regex search), if I enter the search term, The, it only returns the second node containing The Dark Knight, not the third topic node, which has a The in it's title attribute as well. If I put Dark, it doesn't return anything. So it seems it's only searching nodes where the search term is the first item. How can I modify the regular expression above to find all nodes containing an element of the search term? ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
Re: [Flashcoders] E4X filtering strange behaviour
Indeed, the g seems to make it not work. Not a big problem, since it's useless in this case - the g flag in a test() call makes the search start at a specified index (lastIndex property) What may have confused you, Jason, is the example Robert posted fixed your problem (by removing the ^ character from the regexp - the ^ when the first character on a regexp means match at the start of the line), and also added the g flag, which is not quite what you wanted. So just drop it and it will work ;) Side note: I really hate the re[test](...) kind of syntax, so here goes my suggestion: (re as RegExp).test(...) On Mon, Jul 21, 2008 at 5:40 PM, Kenneth Kawamoto [EMAIL PROTECTED] wrote: I don't think you need g (Global flag) for this. Kenneth Kawamoto http://www.materiaprima.co.uk/ Merrill, Jason wrote: Wait, think I got it: var re:RegExp = new RegExp(searchInput.text, g,i); var xmlListSearch:XMLList = testXML..*.( re[test]( attribute(title))); trace(result: +xmlListSearch.toString()); seems to work for me. Thanks eveyrone! Jason Merrill Bank of America Enterprise Technology Global Risk LLD Instructional Technology Media Join the Bank of America Flash Platform Developer Community Are you a Bank of America associate interested in innovative learning ideas and technologies? Check out our internal GTO Innovative Learning Blog subscribe. -Original Message- From: [EMAIL PROTECTED] [mailto: [EMAIL PROTECTED] On Behalf Of Robert Leisle Sent: Monday, July 21, 2008 3:54 PM To: 'Flash Coders List' Subject: RE: [Flashcoders] E4X filtering strange behaviour Hi Jason, This works for me: var testXML:XML = data topics topic title=Coldplays New Album / topic title=The Dark Knight / topic title=Arrested Development: The Movie / /topics /data; var re:RegExp = new RegExp(searchInput.text, g); var xmlListSearch:XMLList = testXML.topics.topic.(re[test](@title)); trace(result: +xmlListSearch.toXMLString()); Hth, Bob -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Merrill, Jason Sent: Monday, July 21, 2008 12:27 PM To: Flash Coders List Subject: RE: [Flashcoders] E4X filtering strange behaviour OK, so this is almost there, but not quite. If I do as Kenneth suggested (and this should be able to be reproduced), var testXML:XML = data topics topic title=Coldplays New Album / topic title=The Dark Knight / topic title=Arrested Development: The Movie / /topics /data; var re:RegExp = new RegExp(^+searchInput.text); var xmlListSearch:XMLList = testXML.topics.topic.(re[test](@title)); trace(result: +xmlListSearch.toXMLString()); (Assuming there is a TextInput text field on the stage and a function run when a button is clicked to run the regex search), if I enter the search term, The, it only returns the second node containing The Dark Knight, not the third topic node, which has a The in it's title attribute as well. If I put Dark, it doesn't return anything. So it seems it's only searching nodes where the search term is the first item. How can I modify the regular expression above to find all nodes containing an element of the search term? ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
Re: [Flashcoders] E4X filtering strange behaviour
If I may repeat myself I don't think you need g (Global flag) for
this. :)
The reason is, I quote from the doc, If the g (global) flag is set for
the regular expression, then the search starts at the index position
specified by the lastIndex property of the regular expression. If the
search matches a substring, the lastIndex property changes to match the
position of the end of the match.
To prove this point, if I do something like:
//
var re:RegExp;
re = new RegExp(The, gi); // with global flag
for each(var prop1:XML in testXML.topics.topic){
trace(re.lastIndex, - , [EMAIL PROTECTED], - ,
re[test]([EMAIL PROTECTED]));
}
trace(***);
re = new RegExp(The, i); // without global flag
for each(var prop2:XML in testXML.topics.topic){
trace(re.lastIndex, - , [EMAIL PROTECTED], - ,
re[test]([EMAIL PROTECTED]));
}
//
I'd get:
0 - Coldplay's New Album, Viva La Vida Or Death And All His Friends -
false
0 - The Dark Knight - true
3 - Arrested Development, The Movie - true
25 - Band Of Horses: Everything All The Time - true
34 - Master And Commander: The Far Side Of The World - true
41 - The Quick Brown Fox Jumped Over The Lazy Dog - false
0 - Violet Hill - false
***
0 - Coldplay's New Album, Viva La Vida Or Death And All His Friends -
false
0 - The Dark Knight - true
0 - Arrested Development, The Movie - true
0 - Band Of Horses: Everything All The Time - true
0 - Master And Commander: The Far Side Of The World - true
0 - The Quick Brown Fox Jumped Over The Lazy Dog - true
0 - Violet Hill - false
With global flag RegExp will not find the in The Quick Brown Fox
Jumped Over The Lazy Dog because it looks for the from the character
41 onwards, namely Dog, which has no the.
Kenneth Kawamoto
http://www.materiaprima.co.uk/
Merrill, Jason wrote:
OK, I'm like 95% there, but there is still some kind of bug - here is
how to reproduce:
var testXML:XML = data
topics
topic title=Coldplay's New Album, Viva La Vida Or Death And All His
Friends /
topic title=The Dark Knight /
topic title=Arrested Development, The Movie /
topic title=Band Of Horses: Everything All The Time /
topic title=Master And Commander: The Far Side Of The World /
topic title=The Quick Brown Fox Jumped Over The Lazy Dog /
topic title=Violet Hill /
/topics
/data;
var re:RegExp = new RegExp(The, g,i);
var xmlListSearch:XMLList = testXML..*.topic.( re[test](
attribute(title)));
trace(result: +xmlListSearch.toString());
If you search for the word, the, as I do when I construct the regular
expression, you should get 5 nodes returned into the XMLList, but I only
get these four:
topic title=The Dark Knight/
topic title=Arrested Development, The Movie/
topic title=Band Of Horses: Everything All The Time/
topic title=Master And Commander: The Far Side Of The World/
I don't get this node:
topic title=The Quick Brown Fox Jumped Over The Lazy Dog /
? Any idea why? Does it fail if there is more than one the in the
string?
___
Flashcoders mailing list
Flashcoders@chattyfig.figleaf.com
http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
Re: [Flashcoders] E4X filtering strange behaviour
I just run a quick and dirty test, and I think I can confirm that. The following test for 10,000 times each: testXML..*.( re[test](attribute(title))); testXML..topic.( re[test](attribute(title))); testXML.topics.topic.( re[test](attribute(title))); testXML.topics.topic.( re[test](@title)); testXML.topics.topic.((re as RegExp).test(@title)); The result: 1114ms 1047ms 993ms 868ms 1087ms So the wildcard * is indeed slower, which is logical. But also attribute() is slower than @, moreover (re as RegExp).test() is slower than re[test]() - well they are all logical too, you may say :) Kenneth Kawamoto http://www.materiaprima.co.uk/ Merrill, Jason wrote: OK good to know, thanks! Jason Merrill Bank of America Enterprise Technology Global Risk LLD Instructional Technology Media Join the Bank of America Flash Platform Developer Community Are you a Bank of America associate interested in innovative learning ideas and technologies? Check out our internal GTO Innovative Learning Blog subscribe. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Kenneth Kawamoto Sent: Monday, July 21, 2008 12:52 PM To: Flash Coders List Subject: Re: [Flashcoders] E4X filtering strange behaviour Yes that's that :) Anyway in your case you know the incoming XML schema so you shouldn't use wildcard * - I think wildcards have serious performance hits. i.e. theXML.topics.topic. should be used rather than theXML..*. Kenneth Kawamoto http://www.materiaprima.co.uk/ Merrill, Jason wrote: Ah, I know where the confusion came from, you said the code was based on my example, but that was actually Wagner's example, where he put var test:XML = data...etc. (it was based on my example, but he used test as an instance of the XML when requoting my original question) The use of test as the XML name threw me, because it was mixed in with the regex method, re.test - I never realized you meant test as the XML not test as some regex method I wasn't aware of. So when you wrote var xmlListSearch:XMLList = test..*.(re[test](attribute(*)) || re[test](child(*))); You meant var xmlListSearch:XMLList = theXML..*.(re[test](attribute(*)) || re[test](child(*))); Got it, I'll give it a shot, thanks. ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
Re: [Flashcoders] E4X filtering strange behaviour
Thanks for showing these results. I'm not missing the attribute method that much...and after this I don't feel any guilt for it. :) -- Keith H -- http://keith-hair.com Kenneth Kawamoto wrote: I just run a quick and dirty test, and I think I can confirm that. The following test for 10,000 times each: testXML..*.( re[test](attribute(title))); testXML..topic.( re[test](attribute(title))); testXML.topics.topic.( re[test](attribute(title))); testXML.topics.topic.( re[test](@title)); testXML.topics.topic.((re as RegExp).test(@title)); The result: 1114ms 1047ms 993ms 868ms 1087ms So the wildcard * is indeed slower, which is logical. But also attribute() is slower than @, moreover (re as RegExp).test() is slower than re[test]() - well they are all logical too, you may say :) Kenneth Kawamoto http://www.materiaprima.co.uk/ ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
RE: [Flashcoders] E4X filtering strange behaviour
I was told on Flexcoders that using @string instead of attribute(string) can be problematic because an error will result if the attribute is missing in the node of the XML. Using attribute() won't throw any errors if the attribute is missing in the XML node, and in most cases (depending on what you're doing) is prefereable to @. Jason Merrill Bank of America Enterprise Technology Global Risk LLD Instructional Technology Media Join the Bank of America Flash Platform Developer Community Are you a Bank of America associate interested in innovative learning ideas and technologies? Check out our internal GTO Innovative Learning Blog subscribe. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Keith Sent: Monday, July 21, 2008 8:19 PM To: Flash Coders List Subject: Re: [Flashcoders] E4X filtering strange behaviour Thanks for showing these results. I'm not missing the attribute method that much...and after this I don't feel any guilt for it. :) -- Keith H -- http://keith-hair.com Kenneth Kawamoto wrote: I just run a quick and dirty test, and I think I can confirm that. The following test for 10,000 times each: testXML..*.( re[test](attribute(title))); testXML..topic.( re[test](attribute(title))); testXML.topics.topic.( re[test](attribute(title))); testXML.topics.topic.( re[test](@title)); testXML.topics.topic.((re as RegExp).test(@title)); The result: 1114ms 1047ms 993ms 868ms 1087ms So the wildcard * is indeed slower, which is logical. But also attribute() is slower than @, moreover (re as RegExp).test() is slower than re[test]() - well they are all logical too, you may say :) Kenneth Kawamoto http://www.materiaprima.co.uk/ ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
RE: [Flashcoders] E4X filtering strange behaviour
So in the following code, why doesn't it find the node containing The Dark Knight? Yet, if I switch out Dark in the RegExp argument for The, it indeed finds all the nodes that have The in it. If I again, switch out the word, Violet, then nothing gets returned. If I put Of, then it returns the two nodes that contain Of. The only logic I have been able to find is that this only works if MORE THAN ONE node contains the search term. Something to do with XMLList or the regex.test method. Ideas? var testXML:XML = data topics topic title=Coldplay's New Album, Viva La Vida Or Death And All His Friends / topic title=The Dark Knight / topic title=Arrested Development, The Movie / topic title=Band Of Horses: Everything All The Time / topic title=Master And Commander: The Far Side Of The World / topic title=The Quick Brown Fox Jumped Over The Lazy Dog / topic title=Violet Hill / /topics /data; var re:RegExp = new RegExp(Dark, i); var xmlListSearch:XMLList = testXML.topics.topic.( re[test](attribute(title))); trace(result: +xmlListSearch.toString()); Jason Merrill Bank of America Enterprise Technology Global Risk LLD Instructional Technology Media Join the Bank of America Flash Platform Developer Community Are you a Bank of America associate interested in innovative learning ideas and technologies? Check out our internal GTO Innovative Learning Blog subscribe. ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
Re: [Flashcoders] E4X filtering strange behaviour
Why not use toXMLString(): trace(result: +xmlListSearch.toXMLString()); When I add more than one node with Dark in the title attribute it finds it with toString(). I've always used toXMLString() so I was unaware of this. -- Keith H -- http://keith-hair.com Merrill, Jason wrote: So in the following code, why doesn't it find the node containing The Dark Knight? Yet, if I switch out Dark in the RegExp argument for The, it indeed finds all the nodes that have The in it. If I again, switch out the word, Violet, then nothing gets returned. If I put Of, then it returns the two nodes that contain Of. The only logic I have been able to find is that this only works if MORE THAN ONE node contains the search term. Something to do with XMLList or the regex.test method. Ideas? var testXML:XML = data topics topic title=Coldplay's New Album, Viva La Vida Or Death And All His Friends / topic title=The Dark Knight / topic title=Arrested Development, The Movie / topic title=Band Of Horses: Everything All The Time / topic title=Master And Commander: The Far Side Of The World / topic title=The Quick Brown Fox Jumped Over The Lazy Dog / topic title=Violet Hill / /topics /data; var re:RegExp = new RegExp(Dark, i); var xmlListSearch:XMLList = testXML.topics.topic.( re[test](attribute(title))); trace(result: +xmlListSearch.toString()); Jason Merrill Bank of America Enterprise Technology Global Risk LLD Instructional Technology Media Join the Bank of America Flash Platform Developer Community Are you a Bank of America associate interested in innovative learning ideas and technologies? Check out our internal GTO Innovative Learning Blog subscribe. ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
Re: [Flashcoders] E4X filtering strange behaviour
The code was based on your example and therefore we know the XML schema, i.e. we know the node name and they all have the attribute title. If we don't know anything about the XML we could do something like: var xmlListSearch:XMLList = test..*.(re[test](attribute(*)) || re[test](child(*))); Kenneth Kawamoto http://www.materiaprima.co.uk/ Merrill, Jason wrote: Thanks. However, I have been told using @title is not good in most situations because it will return an error if a node doesn't have the attribute, wheras attribute(title) will not. Also this is shorter: var xmlListSearch:XMLList = test..topic.(re.test(@title)); But you still need to specify what XML to search through, no? How would the above know what XML object topic is within using the code above? Jason Merrill ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
[Flashcoders] E4X filtering strange behaviour
I was doing some tests here using code from Jason Merril (see previous
thread today), and something has brought my attention.
Here's the xml:
var test:XML = data
topics
topic title=Coldplays New Album /
topic title=The Dark Knight /
topic title=Arrested Development: The Movie /
/topics
/data;
I tried two cases:
var re:RegExp = new RegExp(^Cold);
var xmlListSearch:XMLList = test..*.topic.(
re.test(attribute('title').toString()) );
and:
var re:RegExp = new RegExp(^Cold);
var xmlListSearch:XMLList = test..*.topic.(
!re.test(attribute('title').toString()) );
// this one just negates the RegExp.test() inside topic.() so I should get
the nodes NOT matching
Well, when I run the first case, I get an empty XMLList, so I was thinking
well, that doesn't work.
However, when I run the second test, I get the XMLList with all elements
EXCEPT the one filtered out, so it actually works!
But what is going on that when the regexp returns true, my XMLList is empty?
This seems a bit strange.
Note: after writing all this, I tried something else:
// this does not work
var xmlListSearch:XMLList = test..*.topic.( attribute('title').toString() ==
The Dark Knight );
// this DO work
var xmlListSearch:XMLList = test..*.topic.( attribute('title').toString() !=
The Dark Knight );
Now, THAT's bizarre!
(BTW, I'm quite new to this E4X thing, I'm more of a XPath guy, so bear with
me if I did something completely wrong ;)
___
Flashcoders mailing list
Flashcoders@chattyfig.figleaf.com
http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
Re: [Flashcoders] E4X filtering strange behaviour
I get the correct node (although I also get Warning 3594). If you are
tracing, you must do trace(xmlListSearch.toXMLString());
Also this is shorter:
var xmlListSearch:XMLList = test..topic.(re.test(@title));
Kenneth Kawamoto
http://www.materiaprima.co.uk/
Wagner Amaral wrote:
I was doing some tests here using code from Jason Merril (see previous
thread today), and something has brought my attention.
Here's the xml:
var test:XML = data
topics
topic title=Coldplays New Album /
topic title=The Dark Knight /
topic title=Arrested Development: The Movie /
/topics
/data;
I tried two cases:
var re:RegExp = new RegExp(^Cold);
var xmlListSearch:XMLList = test..*.topic.(
re.test(attribute('title').toString()) );
and:
var re:RegExp = new RegExp(^Cold);
var xmlListSearch:XMLList = test..*.topic.(
!re.test(attribute('title').toString()) );
// this one just negates the RegExp.test() inside topic.() so I should get
the nodes NOT matching
Well, when I run the first case, I get an empty XMLList, so I was thinking
well, that doesn't work.
However, when I run the second test, I get the XMLList with all elements
EXCEPT the one filtered out, so it actually works!
But what is going on that when the regexp returns true, my XMLList is empty?
This seems a bit strange.
Note: after writing all this, I tried something else:
// this does not work
var xmlListSearch:XMLList = test..*.topic.( attribute('title').toString() ==
The Dark Knight );
// this DO work
var xmlListSearch:XMLList = test..*.topic.( attribute('title').toString() !=
The Dark Knight );
Now, THAT's bizarre!
(BTW, I'm quite new to this E4X thing, I'm more of a XPath guy, so bear with
me if I did something completely wrong ;)
___
Flashcoders mailing list
Flashcoders@chattyfig.figleaf.com
http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
RE: [Flashcoders] E4X filtering strange behaviour
Thanks. However, I have been told using @title is not good in most situations because it will return an error if a node doesn't have the attribute, wheras attribute(title) will not. Also this is shorter: var xmlListSearch:XMLList = test..topic.(re.test(@title)); But you still need to specify what XML to search through, no? How would the above know what XML object topic is within using the code above? Jason Merrill Bank of America Enterprise Technology Global Risk LLD Instructional Technology Media Join the Bank of America Flash Platform Developer Community Are you a Bank of America associate interested in innovative learning ideas and technologies? Check out our internal GTO Innovative Learning Blog subscribe. ___ Flashcoders mailing list Flashcoders@chattyfig.figleaf.com http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
