Services & ResourcesWolfram ForumsMathGroup Archive

[Date Index] [Thread Index] [Author Index]

Bug with Sequence and Assignment by Part...

• To: mathgroup at smc.vnet.net
• Subject: [mg121615] Bug with Sequence and Assignment by Part...
• From: BernieTheJet <berniethejet at gmail.com>
• Date: Thu, 22 Sep 2011 07:26:18 -0400 (EDT)
• Delivered-to: l-mathgroup@mail-archive0.wolfram.com

If one patches a list of items into a list using Sequence[], it
appears that Mathemaitca doesn't properly update the data structure:

temp = Range[3];
temp[[2]] = Sequence @@ {a, a, a};
temp

as expected :
    -> {1, a, a, a, 3}

FullForm[temp]
  also looks right:
   -> List[1,a,a,a,3]

As does:
Length[temp]
   -> 5

Ok, so far all is fine.  But subsequent additions to the same place do
not increase the length of temp, as one might expect:
temp[[2]] = Sequence @@ {b, b, b};
temp
FullForm[temp]

       -> {1, b, b, b, 3}
and  -> List[1,b,b,b,3]

And, despite the Length of temp being '5', in fact there is no way to
put anything into position '5' of temp:

Length[temp]
   -> 5
temp[[5]] = Sequence @@ {c, c, c}
   -> Set::partw: Part 5 of {1,Sequence[b,b,b],3} does not exist. >>

Here we finally see what is going on, somewhere temp is being stored
as {1,Sequence[b,b,b],3}, but all the functions that go to request
info from it first perform the evaluation of Sequence hence returning
the expected results.

One can force Mathematica to sort things out with the otherwise meaningless:
temp=temp

Surely I am not the first to encounter this, and surely it constitutes
at least one bug, (and surely it is shouldn't be in Mathematica 8!).  I mean,
it seems to me that FullForm and Length should also detect the
presence of Sequence and act differently than they do here.

Bernard

• Follow-Ups:
  • Re: Bug with Sequence and Assignment by Part...
    • From: Leonid Shifrin <lshifr@gmail.com>