- From: <brian.j.mahoney@sybase.com>
- Date: Thu, 15 Apr 2004 11:40:01 -0700
- To: public-cwm-bugs@w3.org
My CVS checkout of 2000/10/swap on Thursday, April 15 gave me only one
failure when trying
retest.py -n -f regression.n3
(in Cygwin on Windows 2000)
File "include/n3ExprFor.n3" has all lines ending with CR CR LF (not just CR
LF) which can happen as a file is saved/written
between Unix and Windows.
These CRs are stripped out when downloading individual CVS files with a
browser, but I did a commandline checkout
of the entire tree with
"checkout 2000/10/swap/"
( an "--lf checkout 2000/10/swap/" prevents this problem, but it isn't in
the cvs example and who remembers to use it?)
Removing the CRs fixes the error. I don't think this will be the last time
that someone ends up with CR CR LF in an
n3 file and perhaps a general purpose catch should be coded (or already in
use somewhere else).
The runtime error is not very enlightening:
66/72 regression.n3#t1053 Parsing strings with n3ExprFor
Files differ, result= 256
# If this is OK, cp ,temp/regression.n3_t1053.out ref/n3ExprFor-out.n3
######### Differences from reference output:
--- ref/n3ExprFor-out.n3 2002-03-30 14:08:19.000000000 -0800
+++ ,temp/regression.n3_t1053.out 2004-03-16 10:07:07.954528200 -0800
@@ -1,14 +1 @@
- @prefix : <#> .
- @prefix log: <http://www.w3.org/2000/10/swap/log#> .
-
- this log:forAll :x .
-
- :Blargh_b :semantics {<http://example.com/#x> <
http://example.com/#y> <http://example.com/#z> .
- } .
- {
- """@prefix : <http://example.com/#>.
- :x :y :z .""" log:n3ExprFor :x .
-
- } log:implies {:Blargh_b :semantics :x .
- } .
-
+# next char: u'\r'
(end)
Received on Thursday, 15 April 2004 14:47:26 UTC