May 2008 - cxml-devel - mailman3.common-lisp.net

[cxml-devel] Problem compiling current git version
by Ralf Mattes 20 May '08

20 May '08

Hello list, this error occurs after a fresh git-clone: -+ Errors (1) `-- SB-INT:SIMPLE-READER-PACKAGE-ERROR at 5492 (line 123, column 30) on #<SB-SYS:FD-STREAM for "file /LISP/cxml/xml/unparse.lisp" {C538FC9}>: Symbol "FIND-OUTPUT-ENCODING" not found in the RUNES package. This is with sbcl-1.0.15 / Ubuntu w. CLC and (CVS) slime. Cheers, RalfD

2 3

[cxml-devel] cxml dependency in closure-html
by Nathan Bird 19 May '08

19 May '08

closure-html/src/parse/unparse.lisp line 33 has a reference to cxml::find-output-encoding. The closure-html.asd doesn't specify cxml as a dependency. From what I understand of the desired layout we don't want it to, so maybe that function should move up to closure common? Hope that helps, Nathan

2 1

[cxml-devel] Error in ystream.lisp while loading closure-common
by Chaitanya Gupta 17 May '08

17 May '08

Hi, I just installed the latest git snapshot of closure-common. I ran into the following error while loading it - -- x -- (LAMBDA (RUNES:ROD RUNES::INDEX) (BLOCK RUNES:RUNE (CHAR RUNES:ROD RUNES::INDEX))) is not a valid function spec [Condition of type TYPE-ERROR] Restarts: 0: [NIL] retry the compilation of /Users/chaitanya/lisp/libs/site/closure-common/ystream.lisp 1: [NIL] continue compiling /Users/chaitanya/lisp/libs/site/closure-common/ystream.lisp but generate no output file 2: [RETRY] Retry performing #<ASDF:COMPILE-OP NIL @ #x113a195a> on #<CLOSURE-COMMON-SYSTEM::CLOSURE-SOURCE-FILE "ystream" @ #x103f0822>. 3: [ACCEPT] Continue, treating #<ASDF:COMPILE-OP NIL @ #x113a195a> on #<CLOSURE-COMMON-SYSTEM::CLOSURE-SOURCE-FILE "ystream" @ #x103f0822> as having been successful. 4: [ABORT-REQUEST] Abort handling SLIME request. 5: [ABORT] Abort entirely from this (lisp) process. Backtrace: 0: (SWANK::DEBUG-IN-EMACS #<TYPE-ERROR @ #x114e80b2>) 1: (SWANK:SWANK-DEBUGGER-HOOK #<TYPE-ERROR @ #x114e80b2> #<Function SWANK-DEBUGGER-HOOK>) 2: (ERROR TYPE-ERROR :DATUM #1=(LAMBDA (RUNES:ROD RUNES::INDEX) (BLOCK RUNES:RUNE (CHAR RUNES:ROD RUNES::INDEX))) :EXPECTED-TYPE EXCL::FUNCTION-NAME :FORMAT-CONTROL "~s is not a valid function spec" :FORMAT-ARGUMENTS (#1#)) 3: (FBOUNDP (LAMBDA (RUNES:ROD RUNES::INDEX) (BLOCK RUNES:RUNE (CHAR RUNES:ROD RUNES::INDEX)))) 4: (SYSTEM:FUNCTION-INFORMATION (LAMBDA (RUNES:ROD RUNES::INDEX) (BLOCK RUNES:RUNE (CHAR RUNES:ROD RUNES::INDEX))) #<Augmentable COMPILER environment 1 1 1 1 1 15>) 5: (GET-SETF-EXPANSION ((LAMBDA (RUNES:ROD RUNES::INDEX) (BLOCK RUNES:RUNE #)) RUNES::TMP 0) #<Augmentable COMPILER environment 1 1 1 1 1 15>) 6: (SETF (SETF (RUNES:RUNE RUNES::TMP 0) RUNES:RUNE) #<Augmentable COMPILER environment 1 1 1 1 1 15>) 7: (FUNCALL #<Function SETF @ #x100dce22> (SETF (RUNES:RUNE RUNES::TMP 0) RUNES:RUNE) #<Augmentable COMPILER environment 1 1 1 1 1 15>) 8: (COMPILER::PA-COMPILE (SETF (RUNES:RUNE RUNES::TMP 0) RUNES:RUNE) NIL) -- x -- The error occurred while Lisp was compiling YSTREAM-WRITE-ESCAPABLE-RUNE in ystream.lisp. To me the problem seems that RUNE is defined as a compiler-macro and its also being used as a SETF form. I am not sure whether its a problem with my Lisp or closure-common. I am using ACL 8.0 on Intel Mac OS X. It loads fine on SBCL, but DEFINLINE doesn't define any compiler macros in SBCL. Maybe that's why it loads fine here and breaks on ACL? Chaitanya

2 2

[cxml-devel] Cxml-rng and allegro issue
by Prashanth 15 May '08

15 May '08

Hi David Lichteblau, I am using cxml-rng in allegro 8.0 and I am unable to load the package, can you help me in fixing the issue? when i try to load this is the error message i get "`NIL' is not of the expected type `CHARACTER'". -- regards, Prashanth http://munichlinux.blogspot.com

2 1

[cxml-devel] ystream-unicode-p errors when serializing to character-stream
by Nathan Bird 09 May '08

09 May '08

With git head I get an error with the following: (with-output-to-string (stream) (let ((doc (dom:create-document 'rune-dom:implementation nil nil (dom:create-document-type 'rune-dom:implementation "html" "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd")))) (dom:map-document (chtml:make-character-stream-sink stream) doc :include-xmlns-attributes nil :include-doctype :full-internal-subset ))) The error as far as I've been able to track it down, is that the ystream created has a encoding slot, that is never filled. Then #'runes::ystream-unicode-p chokes when trying to inspect the encoding. We solved the error by testing the target stream when the encoding slot is null: (defun ystream-unicode-p (ystream) (let ((enc (ystream-encoding ystream))) (if enc (or (eq enc :utf-8) (eq (babel-encodings:enc-name enc) :utf-16)) #+rune-is-character (eq 'character (stream-element-type (ystream-target-stream ystream)))))) But I'm not sure that it is sound, and that function looks kind of strange anyways: what if encoding was the symbol :utf-16 or a babel-encoding object with the name :utf-8? It looks like the shift to babel-encodings is still somewhat a work-in-progress, does this help? We found this today when we switched from asdf-installed version to git HEAD because there was a problem with & getting escaped that looks like it is already fixed in git.

2 1