@Mathew: Woah... no it doesn't double in size. I've done other quick test using varying sizes of data and node names with follow results
myXML = "<nodelongname>data with lots of data</nodelongname>"; Application.before_myXML = myXML; // 144 bytes Application.after_myXML = XMLParse(myXML); // 216 bytes Even with a slightly more complex structure this seems to hold true myXML = "<node><subnode>data</subnode><subnode>data</subnode></node>"; Application.before_myXML = myXML; // 160 bytes Application.after_myXML = XMLParse(myXML); // 232 bytes Like I said before the XML object is larger than the raw XML by about 80 bytes. Just though I'd clarify that incase people start running away from using XMLParse thinking it has double the memory foot print of the original data. @Simon: Yeah I could do more tests... but its not too hard to do these test using CF8 developer version. So if someone is concerned about the size of a certain object this is a way to do it as Charlie noted. On 22/01/2009, at 12:33 PM, Matthew wrote: > > @Ross: great, thanks for running the test. As you say it could vary > based on a number of varients (such as Ross's comment). If your quick > test is a good guide than it would make a big difference if the parsed > version is double the size. > > Thanks everyone. > Cheers > Matthew > > On Jan 22, 9:31 am, Simon Haddon <simon.had...@gmail.com> wrote: >> It could very well be true for unicode text which would align with >> your >> discovery but I would say that if there was numeric data it would >> be stored >> more efficiently as binary than as text which would occur when you >> parsed >> the xml. 4 bytes for a large float where as the text equivilent >> could be 8 >> or more (in fact 16 since it would probbly use 2 bytes per >> character). >> >> Might be worth running a few more tests. >> >> 2009/1/22 Ross Phillips <r...@fingersdancing.com> >> >> >> >> >> >>> I did a quick test using the CF8 monitor tool to figure out if XML >>> is >>> larger when stored as a String or a XML object. As suspected the >>> XML >>> object is larger but only by about 80 bytes. I was only using a >>> simple XML structure and I would expect this to change when the XML >>> gets more complex. >> >>> myXML = "<node>data</node>"; >>> Application.before_myXML = myXML; // 72 bytes >>> Application.after_myXML = XMLParse(myXML); // 152 bytes >> >>> Cheers, >>> Ross >>> On 22/01/2009, at 10:13 AM, Matthew wrote: >> >>>> @charlie: sorry, I think I lead you a little astray with my >>>> response. >>>> You are right I did want to know both if the output of parsed XML >>>> and >>>> the amount of memory it takes up is large than un-parsed. Your >>>> comment >>>> about all the extra CSS / HTML etc makes sense. >> >>>> @everyone: I ommited one point from my original post. The reason >>>> I'm >>>> interested in how much bigger parsed XML is is because the error >>>> emails I was getting were huge (some were 13Mb!!!). I assume that >>>> if >>>> the are big when being outputted in email than perhaps they are >>>> large >>>> when held in memory. The answer may be that the are only large when >>>> outputted due to all the HTML/CSS etc where as in memory they are >>>> still quite small because it's just struct/arrays etc. >> >>>> @barry: don't worry I'm not stuppid enough to do what your >>>> predecesor >>>> did. My data is customer specific! >> >>>> I'm on CF7 by the way so I can't use the new CF8 feature to work >>>> out >>>> the size of a vaiable. >> >>>> Cheers >>>> Matthew >> >> -- >> Cheers >> Simon Haddon >> >> Woman loves feeling danger and speed. That is why woman wants man. >> They get >> a speed rush that is the most dangerous of all. > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "cfaussie" group. To post to this group, send email to cfaussie@googlegroups.com To unsubscribe from this group, send email to cfaussie+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/cfaussie?hl=en -~----------~----~----~----~------~----~------~--~---
