- cffi-devel - mailman3.common-lisp.net

[cffi-devel] uffi and CUSTOM:*FOREIGN-ENCODING* from clisp
by Yaroslav Kavenchuk 14 Nov '05

14 Nov '05

Pascal Bourguignon make patch http://sourceforge.net/mailarchive/forum.php?thread_id=8889229&forum_id=6767 I test it - it is work! :) Please, append it. Thanks! File strings.diff ========================================================= --- /src/strings.lisp Fri Oct 7 03:32:50 2005 +++ /src/strings.lisp Fri Nov 4 23:56:57 2005 @@ -33,6 +33,7 @@ ;;; and vice versa. Currently this is blithely ignorant of encoding ;;; and assumes characters can fit in 8 bits. +#-clisp (defun lisp-string-to-foreign (string ptr size) "Copy at most SIZE-1 characters from a Lisp STRING to PTR. The foreign string will be null-terminated." @@ -42,6 +43,7 @@ do (setf (%mem-ref ptr :unsigned-char (post-incf i)) (char-code char)) finally (setf (%mem-ref ptr :unsigned-char i) 0))) +#-clisp (defun foreign-string-to-lisp (ptr &optional (size most-positive-fixnum) (null-terminated-p t)) "Copy at most SIZE characters from PTR into a Lisp string. @@ -52,6 +54,41 @@ for code = (mem-ref ptr :unsigned-char i) until (and null-terminated-p (zerop code)) do (write-char (code-char code) s))))) + +#+clisp +(defun lisp-string-to-foreign (string ptr size) + "Copy at most SIZE-1 characters from a Lisp STRING to PTR. +The foreign string will be null-terminated." + (decf size) + (loop + :with bytes = (ext:convert-string-to-bytes string custom:*foreign-encoding*) + :with i = 0 + :for byte :across bytes + :while (< i size) + :do (setf (%mem-ref ptr :unsigned-char (post-incf i)) byte) + :finally (setf (%mem-ref ptr :unsigned-char i) 0))) + +#+clisp +(defun clength (ptr size) + (loop + :for i :from 0 :below size + :do (when (zerop (mem-ref ptr :unsigned-char i)) + (return-from clength i)) + :finally (return-from clength size))) + +#+clisp +(defun foreign-string-to-lisp (ptr &optional (size most-positive-fixnum) + (null-terminated-p t)) + "Copy at most SIZE characters from PTR into a Lisp string. +If PTR is a null pointer, returns nil." + (unless (null-ptr-p ptr) + (loop + :with clen = (if null-terminated-p (clength ptr size) size) + :with bytes = (make-array clen :element-type '(unsigned-byte 8)) + :for i fixnum from 0 below clen + :do (setf (aref bytes i) (mem-ref ptr :unsigned-char i)) + :finally (return (ext:convert-string-from-bytes bytes + custom:*foreign-encoding*))))) ;;;# Using Foreign Strings

2 1

[cffi-devel] uffi and CUSTOM:*FOREIGN-ENCODING* from clisp
by Yaroslav Kavenchuk 04 Nov '05

04 Nov '05

odbc return error message in UTF-8 encoding (may be) I get: *** - A database error occurred: NIL / HY000 *** - Character #\u00C4 cannot be represented in the character set CHARSET:CP866 And under emacs+slime: A database error occurred: NIL / HY000 [Microsoft][Äðàéâåð ODBC Microsoft Access] Óäàëåíèå èíäåêñà íåâîçìîæíî. Îí ÿâëÿåòñÿ òåêóùèì èíäåêñîì èëè èñïîëüçóåòñÿ äëÿ óñòàíîâëåíèÿ ñâÿçè. [Condition of type CLSQL-SYS:SQL-DATABASE-ERROR] (setf CUSTOM:*FOREIGN-ENCODING* CHARSET:UTF-8) not helped How correct it? Thanks! -- WBR, Yaroslav Kavenchuk.

2 2

[cffi-devel] uffi tests
by Yaroslav Kavenchuk 04 Nov '05

04 Nov '05

I have tried banish the test from the uffi-1.5.4 under win2k (most likely in it the reason of failure of several tests): 30 out of 65 total tests failed: :GETENV.1, :GETENV.2, :TIMEOFDAY.1, :TIME.3, :C-TEST.1, :C-TEST.2, :C-TEST.3, :C-TEST.4, :C-TEST.5, :C-TEST.6, :C-TEST.7, :C-TEST.8, :CAST.1, :CAST.2, :FVAR.1, :FVAR.2, :FVAR.3, :FVAR.4, :FVAR.5, :FVAR.6, :FVAR.7, :FVAR.8, :FVARST.1, :FVARST.2, :FVARST.3, :FVARST.4, :FVARST.5, UFFI-TESTS::FVARST.6, UFFI-TESTS::FVARST.7, UFFI-TESTS::FVARST.8. *** - test-op failed Thanks! -- WBR, Yaroslav Kavenchuk.

2 1

[cffi-devel] def-foreign-type and with-foreign-object
by Yaroslav Kavenchuk 04 Nov '05

04 Nov '05

odbc-ff-interface.lisp: (in-package #:odbc) (def-foreign-type sql-handle :pointer-void) ... odbc-api.lisp: ... (defun %new-db-connection-handle (henv) (with-foreign-object (phdbc 'sql-handle) (setf (deref-pointer phdbc sql-handle) +null-handle-ptr+) (with-error-handling (:henv henv) (SQLAllocHandle $SQL_HANDLE_DBC henv phdbc) (deref-pointer phdbc 'sql-handle)))) result: *** - ODBC::%NEW-DB-CONNECTION-HANDLE-1: symbol ODBC::SQL-HANDLE has no value [4]> (describe 'ODBC::SQL-HANDLE) ODBC::SQL-HANDLE is the symbol ODBC::SQL-HANDLE, lies in #<PACKAGE ODBC>, is accessible in 1 package ODBC. #<PACKAGE ODBC> is the package named ODBC. It imports the external symbols of 2 packages COMMON-LISP, CFFI-UFFI-COMPAT and exports 43 symbols to 1 package ODBC-DBI. [5]> (apropos "SQL-HANDLE" :ODBC) ODBC::SQL-HANDLE ODBC::SQL-HANDLE-PTR Thanks! -- WBR, Yaroslav Kavenchuk.

2 1

[cffi-devel] uffi:make-pointer
by Yaroslav Kavenchuk 04 Nov '05

04 Nov '05

Possible add make-pointer to uffi-compat? This need for clsql-odbc. Thanks! -- WBR, Yaroslav Kavenchuk.

2 7

[cffi-devel] uffi-compat and clsql
by Yaroslav Kavenchuk 03 Nov '05

03 Nov '05

clisp 2.35 / mingw, cffi-luis-051102-2042, clsql-3.3.3 [1]> (asdf:operate 'asdf:load-op :cffi-uffi-compat) ... 0 errors, 0 warnings NIL [2]> (asdf:operate 'asdf:load-op :clsql-odbc) ... ;; Loaded file C:\CL\lib\clsql\clsql.asd *** - component CLSQL-UFFI-SYSTEM::UFFI not found, required by #<SYSTEM "clsql-uffi" #x1A2B4949> The following restarts are available: ABORT :R1 ABORT Hmm... If copy cffi-uffi-compat.asd to uffi.asd and replace in it "cffi-uffi-compat" to "uffi": [4]> (asdf:operate 'asdf:load-op :clsql-odbc) ... ;; Loaded file C:\CL\lib\clsql\clsql.asd ; loading system definition from C:\CL\lib\cffi\uffi.asd into #<PACKAGE ASDF4287> ... ;; Compiling file C:\CL\lib\clsql\uffi\clsql-uffi-loader.lisp ... *** - READ from #<INPUT BUFFERED FILE-STREAM CHARACTER #P"C:\\CL\\lib\\clsql\\uffi\\clsql-uffi-loader.lisp" @29> : #<PACKAGE CFFI-UFFI-COMPAT> has no external symbol with name "FOREIGN-LIBRARY-TYPES" What is FOREIGN-LIBRARY-TYPES? Thanks!

2 3

[cffi-devel] uffi-compat and clsql
by Yaroslav Kavenchuk 03 Nov '05

03 Nov '05

clisp 2.35 / mingw, cffi-luis-051102-2042, clsql-3.3.3 [1]> (asdf:operate 'asdf:load-op :cffi-uffi-compat) ... 0 errors, 0 warnings NIL [2]> (asdf:operate 'asdf:load-op :clsql-odbc) ... ;; Loaded file C:\CL\lib\clsql\clsql.asd *** - component CLSQL-UFFI-SYSTEM::UFFI not found, required by #<SYSTEM "clsql-uffi" #x1A2B4949> The following restarts are available: ABORT :R1 ABORT Hmm... If copy cffi-uffi-compat.asd to uffi.asd and replace in it "cffi-uffi-compat" to "uffi": [4]> (asdf:operate 'asdf:load-op :clsql-odbc) ... ;; Loaded file C:\CL\lib\clsql\clsql.asd ; loading system definition from C:\CL\lib\cffi\uffi.asd into #<PACKAGE ASDF4287> ... ;; Compiling file C:\CL\lib\clsql\uffi\clsql-uffi-loader.lisp ... *** - READ from #<INPUT BUFFERED FILE-STREAM CHARACTER #P"C:\\CL\\lib\\clsql\\uffi\\clsql-uffi-loader.lisp" @29> : #<PACKAGE CFFI-UFFI-COMPAT> has no external symbol with name "FOREIGN-LIBRARY-TYPES" What is FOREIGN-LIBRARY-TYPES? Thanks!

1 1

[cffi-devel] Recent changes in cffi-luis: long-long and docs
by Luis Oliveira 02 Nov '05

02 Nov '05

Hello, I pushed support for :(unsigned-)long-long but unfortunately it seems to be supported only by SBCL, OpenMCL and CMUCL (and the last two have bugs which have been reported). Also, I added a new section to the manual with known issues. Changelog follows. Wed Nov 2 15:32:20 WET 2005 Luis Oliveira <loliveira(a)common-lisp.net> * New chapter: Implementation support and known issues. Wed Nov 2 15:34:16 WET 2005 Luis Oliveira <loliveira(a)common-lisp.net> * Light reformating of the ECL changes to fit 80 columns. Wed Nov 2 20:36:13 WET 2005 Luis Oliveira <loliveira(a)common-lisp.net> * Support for the (unsigned) long long type. - New CFFI primitive types :unsigned-long-long and :long-long. - Implemented in the following backends: OpenMCL, SBCL and CMUCL. - Implementations that don't support long long push :cffi/no-long-long into *features*: Allegro, CLISP, Corman, ECL and Lispworks. - New tests: FOREIGN-GLOBALS.REF.LONG-LONG, FOREIGN-GLOBALS.REF.UNSIGNED-LONG-LONG, FUNCALL.LONG-LONG, CALLBACKS.LONG-LONG, CALLBACKS.UNSIGNED-LONG-LONG and DEFCFUN.LONG-LONG. - libtest.c is now compiled with -std=c99 instead of -ansi. - Add information about the new type in the manual and Mention known issues in the various implementations. -- Luís Oliveira http://student.dei.uc.pt/~lmoliv/ Equipa Portuguesa do Translation Project http://www.iro.umontreal.ca/translation/registry.cgi?team=pt

1 0

[cffi-devel] Re: [fetter-devel] wrong code generated
by Kenny Tilton 28 Oct '05

28 Oct '05

Rayiner Hashem wrote: >On 27 Oct 2005 16:06:56 +0200, Immanuel Litzroth <immanuell(a)enfocus.be> wrote: > > >>A header containing only: >>typedef long long unsigned int tp; >>generates: >>(defpackage "TEST" (:use #:CFFI) (:nicknames) (:export "TP")) >> >>(in-package "TEST") >>(asdf:operate 'asdf:load-op 'verrazano-support) >>(cffi:define-foreign-type tp () 'long long unsigned int) >>(cl:progn) >> >>which is nonsense to cffi. >>This is sbcl-0.9.6, fetter+cffi from cvs, linux. >> >> > >Now, this is an interesting problem. There are a number of issues with >"long long" that keep cropping up. To be pedantic, "long long" isn't a >valid C++ type. This is not in and of itself a problem, because >GCC-XML accepts it anyway and the only thing Vzn needs to do is have >it in its type translation table (apparently, I've missed a few >permutations of it, though). The bigger problem is that there is >nothing sensible to map "long long" too, at least on a 32-bit system. >On such a system, long long maps to a 64-bit word (a pair of 32-bit >registers), and there is no way to express that in C-FFI. Anybody have >any ideas? > Sounds like CFFI needs an extension. What are other FFIs such as AllegroCL and CLisp doing? Will such a creature actually only work on a 64-bit system? If not, what do they become on 32-bit systems? Can they be treated benignly as 32-bit longs on 32-bit systems, failing only where runtime values actually exceed 32 bits? I am just hunting around for some imperfect solution that fails no more often than would using the library on a 32-but system even from a C client application. -- Kenny Why Lisp? http://wiki.alu.org/RtL_Highlight_Film "I've wrestled with reality for 35 years, Doctor, and I'm happy to state I finally won out over it." Elwood P. Dowd, "Harvey", 1950

3 2

[cffi-devel] Port to ECL
by Juan Jose Garcia Ripoll 27 Oct '05

27 Oct '05

Hi, Michael Goffioul, one of the developers of ECL, has passed some patches which port cffi-luis to the CVS version of ECL. The changes include support for callbacks and also for dynamic loading of libraries, calling of foreign functions which are defined in these libraries and also runtime definition of C variables. Please feel free to incorporate this into your source tree. All tests are passed except for one that has a memory leak. Juanjo

2 2