On 3/9/06 1:03 AM, "Alex Tweedly" <[EMAIL PROTECTED]> wrote:
> Jim, you can't do that (at least, not if I interpreted Liam's email > correctly). I'm pretty sure he is saying that the file contains > multiple "entries", each entry consisting of the multiple "fields". If > you had a single entry, then this 'split'-based technique is the one I'd > use - but it can't be used while there are multiple entries .... you'd > finish up with an array element for each field name, containing the data > for the last entry. > > You could keep the outer loop from Devin's code, to create tCurrRecord, > and then split tCurrRecord Yes, Alex, you are correct about the multiple 'records' per file, and I was thinking that in the beginning, but was not sure since the description of the task was not clear. As you say, you would need to put another repeat level to parse the multiple 'records' per file, then do the array stuff. After assuming one-per-file, I offered the attached alternative. I would design a more robust approach by using regEx, assuming some errors, the possibility multiple records per file, and even test for incomplete records, missing values, as well as incomplete html downloads. I have several programs that do very robust html download and error checking for my business projects. One program I use as a lynch pin is AltBrowser by Altuit. Glad you studied this well enough to see the limitations :-) Jim Ault Las Vegas > Jim Ault wrote: > >> --------------- >> Another approach would have an advantage or two. >> ONE - using a web address like http://website.com would not use the colon as >> a delimiter, since the 'split' command only considers the first occurance >> TWO - if any glitches occurred that added spurious lines that did not match >> an existing field, these would be ignored. (eg. tab, space) >> THREE - if your labels have a leading/trailing space this will get cleaned >> off >> FOUR -- filter rawData without empty >> >> NOTE: I would recommend that you not use "/" in your field label (eg. >> Principal/Director) can lead to problems later with scripting. >> >> ------------------------------------------------------------- _______________________________________________ use-revolution mailing list [email protected] Please visit this url to subscribe, unsubscribe and manage your subscription preferences: http://lists.runrev.com/mailman/listinfo/use-revolution
