using XSLT on large xml files (~25meg). The problem is
that all the XSLT engines I know load the entire
document into memory before performing the
transformation. At first I cut up the large files by
hand into two or three pieces so they could be
processed by XSLT, but that was just too much of a
pain. I found that by first processing the file
line-by-line using sed, I could remove a lot of the
unneeded attributes. The resultant xml could then be
run through my XSLT engine without having to break it
apart. Hope this helps.
--- Dick Applebaum <[EMAIL PROTECTED]> wrote:
> On Jun 14, 2004, at 7:17 AM, Dave Watts wrote:
>
> > That's a lot of data to output to a browser.
> Where exactly is it
> > failing? On
> > the CFFILE, the XmlTransform function, or the
> CFOUTPUT?
> >
>
> That's the size of the entire file (3400 songs & 30
> playlists) -- all
> of it is not outputted to the browser -- only 10
> playlists and selected
> info for each song in the playlist about 150 songs
> total.
>
> There is one playlist that contains every song in
> the library -- i want
> to exclude it and any of the 3400 songs not in the
> target playlists
>
> What I really wanted to do was read the song file
> into a cf prog,
> extract selected playlists. and the songs for the
> lists, then output
> the whole thing as follows:
>
> Playlist
> song title---artist---duration
> song title---artist---duration
> song title---artist---duration
> *
> *
> Playlist
> song title---artist---duration
> song title---artist---duration
> song title---artist---duration
> *
> *
>
> The song title would be a link to the actual song
> file
>
> So the max output would be about 150 rows of 3
> columns (and I could
> page/cache that, if needed)
>
> Rob put together a xls style sheet & i figured how
> use it to do a
> transform in the cf prog.
>
> But, I wonder if it wouldn't be more efficient to
> parse the xml file
> myself, and extract the playlists & songs I need.
>
> I would think that a compiled, byte code, cf app
> would be more
> efficient than an interpretive transform of a text
> xsl file against a
> text xml file.
>
> Because the xml file is so large and contains so
> much extraneous (to
> this app) information and verbose tags, I even
> thought of parsing the
> file myself with regexp.
>
> Dick
>
>
> "Seek simplicity, and distrust it."
> - Alfred North Whitehead -
>
>
>
[Todays Threads] [This Message] [Subscription] [Fast Unsubscribe] [User Settings]
