<br><div class="gmail_quote">On Wed, Dec 1, 2010 at 4:13 PM, Hendrik Boom <span dir="ltr"><<a href="mailto:hendrik@topoi.pooq.com" target="_blank">hendrik@topoi.pooq.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div><div></div><div><br>
</div></div>I'm not sure I understand the question, though I suspect I may know the<br>
answer. Could you be a bit more specific?<br></blockquote><div><br></div><div>Yes - definitely. See below. </div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
I'm not sure what you mean by "optimized away so they won't show up on a<br>
call to eval". And what, spoecifically, are "lexical variables"?<br>
Variables that have scope determined lexically?<br></blockquote><div><br></div><div>Yes I mean variables with scopes determined lexically - I thought that's what they are called - is there another term for them? lexically-scoped variable/identifier, local variable? </div>
<div><br></div><div>So what I am trying to do is to build an interpreter, hopefully a metacircular one so I can leverage racket as much as possible, but without using the built-in eval, i.e. I need to construct my own lambda, etc. </div>
<div><br></div><div>In this interpreter, identifiers are resolved against a custom environment made of an immutable hash table. This works well for toplevel calls, but when coming to lexically-scoped variables the choice isn't as certain. </div>
<div><br></div><div>The simplest approach I can think of is to put it into the same environment. This also work nicely w.r.t. shadowing since they effective overwrite the local copy of the environment (it's an immutable hash table so other references are not affected). </div>
<div><br></div><div>But what I've read about lexical scoping is that the variables do not really exist since they are substituted away - they do not leave trace on the namespace: </div><div><br></div></div><blockquote class="webkit-indent-blockquote" style="margin: 0 0 0 40px; border: none; padding: 0px;">
<div class="gmail_quote"><div><div><font class="Apple-style-span" face="'courier new', monospace">(let ((a 2)) </font></div></div></div><div class="gmail_quote"><div><div><font class="Apple-style-span" face="'courier new', monospace"> (eval 'a)) </font></div>
</div></div><div class="gmail_quote"><div><font class="Apple-style-span" face="'courier new', monospace">;; => Error: reference to undefined identifier: a</font></div></div></blockquote><div class="gmail_quote">
<div><br></div><div>So I am trying to figure out what is a better representation. </div><div><br></div><div>What I can think of so far all involve some sort of activation frame for holding the values. That basically means I will need to maintain my own stack, and I was trying to see if there are simpler ways to approach the problem.</div>
<div><br></div><div>Please let me know if this is still being unclear. Thanks. </div><div>yc</div><div><br></div><div><br></div></div>