MultiFile constructor, and only do the posstack bookkeeping when it is
true.

to pypaper.sty.

Added the use of the "avant" package to use AvantGarde for the sans-serif
font instead of the helvetica font loaded by the "times" package.

like any other header.

.dvi files are properly dependent on it.

for presence unread, not getattr()!

    ^ and the pos argument to re.match().
Also, fixed a typo in libregex.tex.

doesn't work on modern Windows systems, and Control-Z works
everywhere.

goal to use isinstance(x, y) instead of comparing type(x) to y, it
doesn't make sense to change this in the example code for the type()
builtin...

conversions.  Formerly, for example, int('-') would return 0 instead
of raising ValueError, and int(' 0') would raise ValueError
(complaining about a null byte!) instead of 0...

L.append(a,b,c,d) with the correct L.append((a,b,c,d)).

the empty prompt.

The int() function (aliased to getint()) doesn't handle that, so we must
use self.tk.getint() again...

PythonCmd_Error() but failed to return.  The error wasn't very likely
(only when we run out of memory) but since the check is there we might
as well return the error.  (I think that Barry introduced this buglet
when he added error checks everywhere.)

locals().

keyword argument can be used to direct it somewhere else.  Also
documented all the other functions in this module, and even added a
little example.

# Haven't tested the latex for correctness -- all latex installations
# appear broken.

allow negative numbers to specify read until EOF (like for a regular
file's read() method).

eval(raw_input(s)).  The statement about breaking a long expression
over multiple lines is no longer true.

+ Took the "list" argument out of the other functions that no longer need
it.  This speeds things up a little more.

+ Small comment changes in accord with that.

+ Exploited the now-safe ability to cache values in the partitioning loop.
Makes no timing difference on my flavor of Pentium, but this machine ran out
of registers 12 iterations ago.  It should yield a small speedup on a RISC
machine, and not hurt in any case.

instead of testing whether the list changed size after each
comparison, temporarily set the type of the list to an immutable list
type.  This should allow continued use of the list for legitimate
purposes but disallows all operations that can change it in any way.
(Changes to the internals of list items are not caught, of cause;
that's not possible to detect, and it's not necessary to protect the
sort code, either.)

credit for complaining about this and for testing these changes.

From: Dan Pierson <dan@remote.control.com>
To: "Mark Hammond (E-mail)" <MHammond@skippinet.com.au>,
    "Guido van Rossum (E-mail)" <guido@cnri.reston.va.us>
Date: Mon, 8 Jun 1998 17:25:07 -0400

RegistryQueryValue requires that its fourth argument be initialized to
the length of the buffer being passed in, this wasn't being done.  I
also split the call and status test into two lines with a local variable
so that I could look at the status in the debugger.

Added raw strings, imaginary literals, assert and exec (!) keywords, a
table about Resererved classes of identifiers, and more.

# checkin email because my PC doesn't have the "Mail" command.

Add threading (now that it works).  Also some small adaptations to
Unix again.

# from my PC at home, but it can't send email :-(

Add a clarifying comment about the new ENTER_OVERLAP and
LEAVE_OVERLAP_TCL macros; get rid of all the bogus tests for deleted
interpreters (Tcl already tests for this; they were left over from an
earlier misguided attempt to fix the threading).

doesn't depend on the value of os.sep.  (I.e. ported to Windows :-)