- 21 Oct, 2013 1 commit
-
-
Francois-Rene Rideau authored
-
- 15 Oct, 2013 7 commits
-
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
Make some run-program tests Unix-only that fail on Windows for semi-good reasons. Sleep 2 seconds instead of one in test-stamp-propagation, in case that makes things more stable.
-
- 14 Oct, 2013 3 commits
-
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
-
- 13 Oct, 2013 1 commit
-
-
Francois-Rene Rideau authored
Include stamp propagation test, which provides evidence for the Great Defsystem Bug.
-
- 12 Oct, 2013 3 commits
-
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
-
- 11 Oct, 2013 1 commit
-
-
Francois-Rene Rideau authored
Add a test case. (Issue: %active-io-specifier-p takes a non-normalized argument.)
-
- 02 Oct, 2013 1 commit
-
-
Francois-Rene Rideau authored
Fix missing quote for stderr on allegro.
-
- 01 Oct, 2013 1 commit
-
-
Francois-Rene Rideau authored
-
- 30 Sep, 2013 1 commit
-
-
Francois-Rene Rideau authored
Fix corner case on implementations that return a :bidir-stream but no :input-stream. Also test and fix support for :error-output :output.
-
- 28 Sep, 2013 1 commit
-
-
Francois-Rene Rideau authored
* it handles INPUT and ERROR-OUTPUT, not just OUTPUT. * it accepts PATHNAMEs as I/O designators * it works on ECL under SLIME (fixes lp#1232225) * SLIGHT API CHANGE! At the suggestion of stassats, we're always returning an exit code, and in a predictable return value position. However, both for backward compatibility and usability, the OUTPUT slurping results remain the primary return value, followed by the ERROR-OUTPUT slurping results, followed by the exit code. So far, no one was using multiple values in SLURP-INPUT-STREAM methods, so no one will regret that, except people who might have somehow expected the old weird behavior of a return code when OUTPUT is NIL or :INTERACTIVE. * As another API tweak and improvement, keyword options are now specified separately for each of your I/O designators, by putting them your designator in a list. A list also distinguishes a string as input from a namestring to designate a pathname, i.e. :input '("bar") is the string literal, whereas :input "bar" is the file named bar. Several fixes and extensions to stream support. Only tested so far on Linux x64 with ccl clisp sbcl ecl ecl_bytecodes cmucl abcl allegro allegromodern xcl
-
- 16 Sep, 2013 1 commit
-
-
Francois-Rene Rideau authored
-
- 06 Sep, 2013 2 commits
-
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
-
- 05 Sep, 2013 1 commit
-
-
Francois-Rene Rideau authored
-
- 03 Jun, 2013 1 commit
-
-
Robert P. Goldman authored
-
- 21 May, 2013 1 commit
-
-
Robert P. Goldman authored
On ACL, the :INPUT and :OUTPUT arguments to EXCL:RUN-SHELL-COMMAND take NIL for the behavior of inheriting input and output handles from the Lisp process. These were encoded as T in UIOP:RUN-PROGRAM, instead of NIL, causing ACL to get errors trying to open T as a file designator. Added a couple of test cases to test-run-program.script to illustrate the problem. Note that this fix does not provide the prescribed behavior for :OUTPUT NIL (which should take the program's output and throw it on the floor). There is a TODO in uiop:run-program explaining what's missing. In brief, ACL's EXCL:RUN-SHELL-COMMAND does not support this behavior out of the box, so one would have to capture the program's output, and then throw it on the floor.
-
- 30 Apr, 2013 1 commit
-
-
Francois-Rene Rideau authored
-
- 17 Apr, 2013 1 commit
-
-
Francois-Rene Rideau authored
And thus in run-program and run-shell-command. Add a test for capturing output with run-shell-command -- yuck.
-
- 25 Feb, 2013 1 commit
-
-
Francois-Rene Rideau authored
Make test-utilities work on lower-case lisps. Don't assume ln is in /bin. Don't assume SHELL is /bin/sh.
-
- 01 Feb, 2013 1 commit
-
-
Francois-Rene Rideau authored
Thanks to Stelian for finding this issue. Also, make run-program test work where select binaries are not in /bin. Better handle incrementing ASDF versions.
-
- 22 Jan, 2013 1 commit
-
-
Francois-Rene Rideau authored
-
- 21 Jan, 2013 3 commits
-
-
Francois-Rene Rideau authored
Some doc & test frobbing.
-
Francois-Rene Rideau authored
Rename RUN-PROGRAM/ to RUN-PROGRAM. Can't export it from ASDF because it causes a clash in SB-GROVEL. Rename SUB-OBJECT and :PATH to ACCESS-AT and :AT. Get rid of TRANSLATE-OUTPUT-P. Instead, check whether the component's BUILD-OPERATION is the same as the current operation.
-
Francois-Rene Rideau authored
Also, make script re-loadable in a debug session.
-
- 20 Jan, 2013 2 commits
-
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
Don't key upgrade attempt on source-registry, since things can also change in the central-registry, etc. Also rename required-files to direct-dependency-files Improve the new convenience function pattern to work with operation initargs. Implement getcwd on all platforms. Actually catch sb-grovel messages. Tests: make u l=ecl instead of make test-upgrade lisp=ecl Tests: put the setup around the script, not inside; Tests: start using globals, much easier for debugging
-
- 13 Jan, 2013 2 commits
-
-
Francois-Rene Rideau authored
Package tweaks to make CLISP complain less. run-program/ and run-shell-command fixes and test updates. wildening fix to issue found by CLISP. ! is used by EXT on CLISP, so use !p for defparameter. More pathname madness for allegro. Better upgrade.
-
Francois-Rene Rideau authored
Tests: SBCL passes test-lisp and test-upgrade Also: * big refactoring of test infrastructure * contrib/debug.lisp and (asdf/driver:asdf-debug) for debugging. * remove aif and it, use the cleaner if-bind. * backtrace support, with improvements from trivial-backtrace. * integration of asdf-condition-control (originally from xcvb-driver) * simplify the upgrade heuristic: always tries to upgrade once. The previous heuristic was too complex with too many failure cases. You are thus assumed to want a given version asdf if and only if it is in your source-registry.
-
- 21 Oct, 2011 1 commit
-
-
Robert P. Goldman authored
-
- 18 Oct, 2011 1 commit
-
-
Francois-Rene Rideau authored
-