Re: Pointers to "why C behaves like that ?"

Peter Flass <peter_flass@yahoo.com> wrote:
+ Christian Bau wrote:
+> > A typo in the spelling of the variable "CurrentAmunt" in line 2
+> > creates a new variable "CurrentAmunt" that receives the result of the
+> > addition. Since the CurrentAmount variable is stored, the account
+> > receives a wrong value, and there is NO WAY to spot this bug until
+> > runtime!
+
+ Not true. To use a PL/I example again the compiler would simply have
+ to somehow flag usage of an undeclared variable and the programmer
+ could easily spot the typo. Of course this assumes you declare all
+ your variables even when you don't have to...
+ [I don't see a useful difference between a system that requires that you
+ declare everything, vs. one that doesn't but producesundeclared variable
+ warnings that the programmer treats like error messages. -John]


Some of my students are Pythonistas. They insist that the lack of
such assistance from the compiler in detecting typos has never been an
issue. (I wouldn't know.)


In the case at hand, the typo would (in most cases) yield an
unused-variable warning, even if the undeclared variable were accepted
by the language.


IIRC, some implicitly typed languages use the keyword "let" to
introduce a new variable. In which case, the absence of "let" just
before the typo would yield an error.


Tom Payne

Return to the comp.compilers page.
Search the comp.compilers archives again.