1. 07 Jun, 2017 1 commit
    • mevenson@1c010e3e-69d0-11dd-93a8-456734b0d56f's avatar
      Restore compilation on Java 6 · ba0019c9
      JSR-223 is always present so usage in <file:build.xml> is deprecated.
      Fix compilation for SYS:RUN-PROGRAM on Java 6.  Not expected to be
      working very well.  The general strategy is to undeprecate
      SYS:RUN-SHELL-COMMAND is some manner that uses the Java 6 NIO plus
      thread pools to drain IO.
      Conditionalize on runtime platform of parts of our RUN-PROGRAM
      Stub implementation of NIO async processor to drain queue.
      TODO: figure out how why "cat /etc/passwd" doesn't ever give output?
  2. 23 May, 2017 1 commit
  3. 22 May, 2017 1 commit
    • mevenson@1c010e3e-69d0-11dd-93a8-456734b0d56f's avatar
      Rework ABCL-BUILD as a contrib · f8592464
      ABCL-BUILD now more usefully installs necessary build machinery.
      Refactor the ABCL-BUILD Lisp based build to use use Ant, which allows
      us to canonicalize the build instructions in a much more configurable
      manner, restoring the cross-implementation build of ABCL for CL users
      who really do want to know about Java tooling.
      Since we want to use ASDF/UIOP anyways, this code is best expressed as
      an abcl contrib where such behavior can profitably take place.
      TODO: does not pass all tests
      TODO: API compatibility so that build-from-lisp.bash works
      TODO: If an invocable Ant cannot be found, then offer to install one via restart
      TODO: work backwards to the scenario that we are booting in pure
      interepreted mode without access to compilers.
  4. 16 May, 2017 1 commit
  5. 16 Apr, 2017 1 commit
  6. 07 Apr, 2017 1 commit
  7. 30 Mar, 2017 2 commits
  8. 28 Mar, 2017 1 commit
  9. 19 Nov, 2016 1 commit
    • mevenson@1c010e3e-69d0-11dd-93a8-456734b0d56f's avatar
      Re-write the ABCL ASDF description using secondary systems · 0088b052
      Future versions of ASDF will start complaining when multiple DEFSYSTEM
      forms occupy a given file unit, but the systems named therein don't
      use the "PRIMARY[/SECONDARY..]" naming conventions.
      (asdf:test-system :abcl)
      Run the ABCL tests located under <file:test/lisp/abcl/>
      (asdf:test-system :abcl/test/ansi/compiled)
      Run the compiled version of the ANSI tests in <file:../ansi-test/>.
      (asdf:test-system :abcl/test/ansi/interpreted)
      Run the interpreted version of the ANSI tests in <file:../ansi-test/>.
      (asdf:test-system :abcl/test/cl-bench)
      Run the CL-BENCH test suite in <file:../cl-bench/>.
  10. 08 Oct, 2016 2 commits
  11. 04 Sep, 2016 4 commits
  12. 09 Sep, 2015 1 commit
  13. 31 Aug, 2015 1 commit
  14. 14 Apr, 2015 1 commit
  15. 17 Aug, 2014 1 commit
  16. 25 Apr, 2014 1 commit
  17. 09 Mar, 2013 1 commit
  18. 17 Oct, 2012 1 commit
  19. 03 Aug, 2012 1 commit
  20. 04 May, 2012 2 commits
    • mevenson@1c010e3e-69d0-11dd-93a8-456734b0d56f's avatar
      ansi-tests: Fix ANSI-RT load definition. · 50ee397b
      A bit of a circular dance here to be able to use the ANSI-TESTS via
      ASDF.  In order to define functions like
      ABCL.TEST.ANSI:DO-TESTS-MATCHING we need to have loaded the
      REGRESSION-TEST package included with the ANSI-TESTS, so we can no
      longer use the value of the ABCL.TEST.ANSI:*ANSI-TESTS-DIRECTORY*.  We
      now do this via a relative pathname of the form '../ansi-tests/' to
      the location of 'abcl.asd' file, whereas our api previously suggested
      that one could set the contents of
      ABCL.TEST.ANSI:*ANSI-TESTS-DIRECTORY* to an arbitrary pathname.  This
      fixes the location of the ANSI-TESTS on the filesystem to be a sibling
      directory named 'ansi-tests' to the truename of the directory
      containing 'abcl.asd'.
      I tried to add some sort of :before method to either the ASDF:LOAD-OP
      or the ASDF:COMPILE-OP for the ANSI-RT definition, but assuming that
      the source named in an system definition exists locally seems too
      baked into ASDF2 to be cleanly intercepted.
    • mevenson@1c010e3e-69d0-11dd-93a8-456734b0d56f's avatar
      ansi-tests: DO-TESTS-MATCHING will run all case-insensitive matching tests. · b6a3ad17
      An ASDF load of ANSI-INTERPRETED now loads the interpreted tests into
      memory via an :around specialization.
      Changes in ABCL.TEST.ANSI (aka ANSI-TESTS):
      DO-TESTS-MATCHING implemented (cribbed from ABCL-TEST-LISP).
      Refactored DEFPACKAGE forms out of ABCL-ANSI-TESTS code, creating
      abcl-ansi.lisp to hold all other forms.
      Import symbols from REGRESSION-TEST where it makes sense (list could
      probably be larger; why not just use the package?)
  21. 30 Sep, 2011 1 commit
  22. 16 Jun, 2011 1 commit
    • mevenson@1c010e3e-69d0-11dd-93a8-456734b0d56f's avatar
      Create form of SYSTEM:ZIP that uses a hashtable to map files to entries. · 6e60f1a0
      SYSTEM:ZIP PATH HASHTABLE now creates entries in a zipfile at PATH
      whose entries are the contents of for each (KEY VALUE) in HASHTABLE
      for which KEY refers to an object on the filesystem and VALUE is the
      location in the zip archive.
      Introduce Java interfaces in org.armedbear.lisp.protocol to start
      encapsulating behavior of Java system.  By retroactively adding
      markers to the object hierarchy rooted on LispObject we gain the
      ability to have our JVM code optionally work with interfaces but we
      leave the core dispatch functions alone for speed.
  23. 07 Jun, 2011 1 commit
    • mevenson@1c010e3e-69d0-11dd-93a8-456734b0d56f's avatar
      Implementation of hashtables with weak keys and/or values. · bd90e2da
      MAKE-HASH-TABLE now has an optional :WEAKNESS argument that can take
      the values :KEY, :VALUE, :KEY-AND-VALUE, or :KEY-OR-VALUE.  :KEY means
      that the key of an entry must be live to guarantee that the entry is
      preserved. VALUE means that the value of an entry must be live to
      guarantee that the entry is preserved.  :KEY-AND-VALUE means that both
      the key and the value must be live to guarantee that the entry is
      preserved.  :KEY-OR-VALUE means that either the key or the value must
      be live to guarantee that the entry is preserved.
      The tests simply excercise the various types of weak hash tables
      enough that a GC phase should show that the table indeed does decrease
      in size.
      Changed the defition of functions in HashTableFunctions to match
      current docstring/pf_XXX() naming conventions.
      This implementation is only lightly tested in single-threaded use, and
      untested in multiple threading scenarios.
      Addresses ticket:140.
  24. 04 Apr, 2011 1 commit
  25. 03 Apr, 2011 1 commit
  26. 27 Nov, 2010 1 commit
  27. 07 Nov, 2010 2 commits
  28. 24 Sep, 2010 1 commit
  29. 28 Aug, 2010 1 commit
  30. 15 Aug, 2010 1 commit
  31. 07 May, 2010 1 commit
  32. 15 Apr, 2010 2 commits