1. typecasts again
- Posted by gertie at visionsix.com
Jan 10, 2003
This line:
while sequence(TheWebPage[junk][1]) and match(" ",TheWebPage[junk][1])
do
errors out with :
attempt to subscript an atom
(reading from it)
Since TheWebPage[junk] is a sequence,
TheWebPage[junk][1] is a sequence, where is it an atom??
Kat
2. Re: typecasts again
Kat wrote:
> This line:
>
> while sequence(TheWebPage[junk][1]) and match(" ",TheWebPage[junk][1])
> do
>
> errors out with :
>
> attempt to subscript an atom
> (reading from it)
>
> Since TheWebPage[junk] is a sequence,
> TheWebPage[junk][1] is a sequence, where is it an atom??
If by some error, the element at TheWebPage[junk] is an atom and not a
sequence, attempting to further index it with the [1] would cause the error
you're getting.
Try:
while
sequence(TheWebPage[junk]) and
sequence(TheWebPage[junk][1]) and
match(" ",TheWebPage[junk][1])
do
-- etc.
end while
Then again, I'm not sure whether 'while' short-circuits like 'if' does. I
know that return expressions in types and functions don't (In Eu2.3). So,
this would still fail if there's no short-circuiting on the 'while'
condition.
A messy, guaranteed to short-circuit and work as expected workaround is:
while 1 do if
sequence(TheWebPage[junk]) and
sequence(TheWebPage[junk][1]) and
match(" ",TheWebPage[junk][1])
then else exit end if
-- etc.
end while
HTH,
Carl
--
[ Carl R White == aka () = The Domain of Cyrek = ]
[ Cyrek the Illogical /\ www.cyreksoft.yorks.com ]
3. Re: typecasts again
> > Kat wrote:
> > > This line:
> > >
> > > while sequence(TheWebPage[junk][1]) and match(" ",TheWebPage[junk][1])
> > > do
> > >
> > > errors out with :
> > >
> > > attempt to subscript an atom
> > > (reading from it)
> > >
> > > Since TheWebPage[junk] is a sequence,
> > > TheWebPage[junk][1] is a sequence, where is it an atom??
> > >
> > > Kat
> >
Kat,
what do you get if you code it this way instead...
while 1 do
if not sequence(TheWebPage[junk][1]) then
puts(1, "Not a sequence\n")
exit
end if
if not match(" ",TheWebPage[junk][1]) then
exit
end if
. . .
end while
However, I take it that 'TheWebPage[junk]' is supposed to contain a line of
text. If that is so, then I would expect that 'TheWebPage[junk][1]' would
contain an atom - namely the first character in the line.
----------------
cheers,
Derek Parnell
4. Re: typecasts again
Kat:
Don't forget EOF ( end of file ) is an atom.
Bernie
5. Re: typecasts again
- Posted by gertie at visionsix.com
Jan 11, 2003
On 11 Jan 2003, at 8:13, Derek Parnell wrote:
>
> > > Kat wrote:
> > > > This line:
> > > >
> > > > while sequence(TheWebPage[junk][1]) and match(" ",TheWebPage[junk][1])
> > > > do
> > > >
> > > > errors out with :
> > > >
> > > > attempt to subscript an atom
> > > > (reading from it)
> > > >
> > > > Since TheWebPage[junk] is a sequence,
> > > > TheWebPage[junk][1] is a sequence, where is it an atom??
> > > >
> > > > Kat
> > >
>
> Kat,
> what do you get if you code it this way instead...
I recoded the whole block already.
> while 1 do
> if not sequence(TheWebPage[junk][1]) then
> puts(1, "Not a sequence\n")
> exit
> end if
> if not match(" ",TheWebPage[junk][1]) then
> exit
> end if
> . . .
> end while
>
> However, I take it that 'TheWebPage[junk]' is supposed to contain a line of
> text. If that is so, then I would expect that 'TheWebPage[junk][1]' would
> contain an atom - namely the first character in the line.
That stands to reason, however, when i "if sequence()" on it (and it passes),
then "length()" on it (and it passes), and then it crashes when i attempt to
use match() on it, and returns that match needs a sequence, or i am
attempting to read from the atom, it's wasting my time. I want to know if
there is a " " or ' ' or 32 anywhere there, regardless of it's type. I know i
assigned a sequence to it, because the only atoms i declared are used as
integers.
Kat
6. Re: typecasts again
- Posted by gertie at visionsix.com
Jan 11, 2003
On 10 Jan 2003, at 18:21, Bernie Ryan wrote:
>
>
> Kat:
>
> Don't forget EOF ( end of file ) is an atom.
I didn't forget it, i am not reading a disk file. It's not so much that -1 is
the
EOF, it's an error from trying to read past the end of the file, which *could*
be
a ctrl-z, but may not be. And, this is occuring, afaik, in the middle of the
data, not the end.
Kat
