Subject: Problem with READ-BYTE-SEQUENCE on Windows
Newsgroups: gmane.lisp.clisp.general
Date: 2004-01-15 12:26:21 GMT (4 years, 24 weeks, 4 days, 10 hours and 35 minutes ago)
The test code below works fine with binary downloads like, say,
(foo "weitz.de" "/files/cl-interpol.tgz")
if I invoke it from Linux. But if I use it from Windows most of the
contents of the generated "out" file (towards the end) are nulls. Am I
doing something wrong or is this a bug? (Looks to me as if it acted
like the :NO-HANG argument were T.)
I had the same problem with the standard functions
READ-SEQUENCE/WRITE-SEQUENCE.
Both are CLISP 2.32. On Linux (SuSE 9.0) it was built from source, on
Windows (XP pro) I used the "clisp-2.32-win32.zip" binary from
Sourceforge.
Thanks,
Edi.
(defun foo (host url)
(let ((stream (socket:socket-connect
80 host
:external-format
(ext:make-encoding :charset 'charset:iso-8859-1
:line-terminator :unix))))
(format stream "GET ~A HTTP/1.0~C~CHost: ~A~C~C~C~C"
url #\Return #\Linefeed
host #\Return #\Linefeed #\Return #\Linefeed)
(force-output stream)
(let (length)
(loop for line = (read-line stream nil nil)
until (or (null line)
(zerop (length line))
(eql (elt line 0) (code-char 13)))
when (= (mismatch line "Content-length: "
:test #'char-equal)
16)
do (setq length
(parse-integer line
:start 15
:junk-allowed t)))
(unless length
(error "No Content-Length header found"))
(setf (stream-element-type stream) '(unsigned-byte 8))
(with-open-file (out "out" :direction :output
:element-type '(unsigned-byte 8)
:if-exists :supersede)
(let ((buf (make-array length
:element-type
(stream-element-type stream))))
(ext:read-byte-sequence buf stream :no-hang nil)
(ext:write-byte-sequence buf out)))
(values))))
-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software.
Perforce is the Fast Software Configuration Management System offering
advanced branching capabilities and atomic changes on 50+ platforms.
Free Eval! http://www.perforce.com/perforce/loadprog.html