Could ABCL set up a #P"SRC:", or #P"SYS:SRC:" logical pathname to the source directory that was used for the build?
-T.
"Tobias C. Rittweiler" writes:
Could ABCL set up a #P"SRC:", or #P"SYS:SRC:" logical pathname to the source directory that was used for the build?
Tentatively related might be the *LISP-HOME* stuff in compile-system.lisp. In particular, I want the SRC logical pathname to easily get at the TAGS file which GROVEL-JAVA-DEFINITIONS also tries to do.
-T.
On Mar 3, 2010, at 11:02 AM, Tobias C. Rittweiler wrote:
"Tobias C. Rittweiler" writes:
Could ABCL set up a #P"SRC:", or #P"SYS:SRC:" logical pathname to the source directory that was used for the build?
Tentatively related might be the *LISP-HOME* stuff in compile-system.lisp. In particular, I want the SRC logical pathname to easily get at the TAGS file which GROVEL-JAVA-DEFINITIONS also tries to do.
*LISP-HOME* is badly named at this point, as it refers to the location of "boot.lisp" which is where the FASLs are located when ABCL system source is compiled (which it is in about every "normal" deployment situation).
The way I thinking at this point is that COMPILE-SYSTEM compiles and dumps a form to be picked up at runtime.
-- "A screaming comes across the sky. It has happened before, but there is nothing to compare to it now."
On Mar 3, 2010, at 9:49 AM, Tobias C. Rittweiler wrote:
Could ABCL set up a #P"SRC:", or #P"SYS:SRC:" logical pathname to the source directory that was used for the build?
I've committed an implementation of logical pathname translations for SYS:SRC and SYS:JAVA [in r12514][1]. SYS:SRC translates to the root of the Lisp files, while SYS:JAVA translates to the root of the Java tree.
[1]: http://code.google.com/p/abcl-dynamic-install/source/detail?r=c5003c8270fb30...
-- "A screaming comes across the sky. It has happened before, but there is nothing to compare to it now."
Mark Evenson evenson@panix.com writes:
On Mar 3, 2010, at 9:49 AM, Tobias C. Rittweiler wrote:
Could ABCL set up a #P"SRC:", or #P"SYS:SRC:" logical pathname to the source directory that was used for the build?
I've committed an implementation of logical pathname translations for SYS:SRC and SYS:JAVA [in r12514][1]. SYS:SRC translates to the root of the Lisp files, while SYS:JAVA translates to the root of the Java tree.
CL-USER(1): (probe-file #P"SYS:SRC") Debugger invoked on condition of type FILE-ERROR: No translation for ~S Restarts: 0: TOP-LEVEL Return to top level.
On Mar 3, 2010, at 4:51 PM, Tobias C. Rittweiler wrote:
Mark Evenson evenson@panix.com writes:
On Mar 3, 2010, at 9:49 AM, Tobias C. Rittweiler wrote:
Could ABCL set up a #P"SRC:", or #P"SYS:SRC:" logical pathname to the source directory that was used for the build?
I've committed an implementation of logical pathname translations for SYS:SRC and SYS:JAVA [in r12514][1]. SYS:SRC translates to the root of the Lisp files, while SYS:JAVA translates to the root of the Java tree.
CL-USER(1): (probe-file #P"SYS:SRC") Debugger invoked on condition of type FILE-ERROR: No translation for ~S Restarts: 0: TOP-LEVEL Return to top level.
Yeah, but
(translate-logical-pathname "SYS:SRC;boot.lisp")
works, right?
As far as I understand logical pathnames (admittedly not a huge distance), your form shouldn't work, as (probe-file #p"SYS:SRC") fails on SBCL.
-- "A screaming comes across the sky. It has happened before, but there is nothing to compare to it now."
armedbear-devel@common-lisp.net
-
Mark Evenson -
Tobias C. Rittweiler