[Ecls-list] Bug: STRING= dies on symbol as string designator in compiled code

Prut Flut prutflut at m-net.arbornet.org
Wed Oct 26 00:10:56 UTC 2005


Hello,

ECL 0.9g isn't accepting string-designators for STRING= in compiled code:
I downloaded ecl-0.9g-patch-2.tgz and built it (I can dig out the
configuration options if needed).

When trying to (asdf:operate 'asdf:load-op 'cl-pdf) that itself includes
code from the iterate package, the compiler failed with an error (not
exact wording)

Error: call argument 2 to STRING= is a symbol, not a string.

Indeed: the following test function

(defun test-string= ()
  "Tests whether COMMON-LISP:STRING= accepts symbols as arguments.
This function should return T
for (eq (readtable-case *readtable*) :upcase)"
  (string= "A" '#:A))

can be entered at the prompt and executed, but (compile-file "test.lisp")
(assuming this function is defined in test.lisp) chokes with the same
error.  I didn't try if I could (compile 'test-string=) or if the other
STRING... functions shared that behaviour.

HTH,
-- 
PhC




More information about the ecl-devel mailing list