CL-USER 1 > (ql:quickload "clpython")
...
**++++ Error in (DEFCLASS CLPYTHON:PY-TYPE): 
  #<STANDARD-CLASS DICT-MIXIN 40200E136B> is an invalid superclass of #<PY-META-TYPE PY-TYPE 402013771B>.

The following file cannot be loaded with clpython:run on recent versions of SBCL:

import CL
CL.PRINT("FOO")

Tested as failing on 1.4.13 and likely starting with the ASDF update in 1.4.2. Here is the backtrace:

ImportError: Loading of module `CL' was aborted. Binary: /tmp/user/1000/clpython-cl-fasl-p6tnmpa2Imported by: /home/ntd/git/tmkit/test.py

1: ((FLET CLPYTHON::LOG-ABORT :IN CLPYTHON::PY-IMPORT) T)
2: ((FLET "CLEANUP-FUN-54" :IN CLPYTHON::PY-IMPORT)) [cleanup]
3: (CLPYTHON::PY-IMPORT (CLPYTHON.USER::CL) :WITHIN-MOD-PATH #P"/home/ntd/git/tmkit/test.py" :WITHIN-MOD-NAME "__main__")
4: (CLPYTHON::|python-module __main__|)
5: (CLPYTHON.USER::|{__main__[0] "import CL" #4138581656636548740}| #<HASH-TABLE :TEST EQ :COUNT 2 {100343ACE3}>)
6: ((FLET CLPYTHON::RUN-TOP-LEVEL-FORMS :IN CLPYTHON::MODULE-INIT) #<HASH-TABLE :TEST EQ :COUNT 2 {100343ACE3}>)
7: ((FLET CLPYTHON:RUN :IN CLPYTHON:RUN-PYTHON-AST))

The issue appears to arise in CLPYTHON::COMPILED-FILE-NAME with the call to (make-instance 'asdf:compile-op). On newer SBCL, this throws the following error:

OPERATION instances must only be created through MAKE-OPERATION.

Replacing cl:make-instance with asdf:make-operation does not throw the error in COMPILED-FILE-NAME and the above CL import works on SBCL 1.4.13 and 1.3.14

cl-pythong source code uses BREAK several in several places.

It's inconvenient. For example, when I do (ql:quickload :clpython) on ABCL when testing libraries with cl-test-grid, clpython calls BREAK with message "A CHECKING-READER-CONDITIONALS expression returned zero forms: (CLPYTHON.UTIL:CHECKING-READER-CONDITIONALS)".

The ABCL process remains hanging, I can't capture this situation programmatically.

It it was ERROR instead of BREAK, test-grig-agent could intercept it using either HANDLER-BIND or DEBUGGER-HOOK and record error.

When I do (ql:quickload "clpython") fails with the following:

Error: Symbol "OPEN-TEMPORARY" not found at all in the CL-FAD package. 1 (continue) Make a new symbol. 2 Try reading the next form. 3 Try compiling /Users/deepaksurti/quicklisp/dists/quicklisp/software/cl-python-20140211-git/runtime/import.lisp again. 4 Skip compiling /Users/deepaksurti/quicklisp/dists/quicklisp/software/cl-python-20140211-git/runtime/import.lisp. 5 Retry compiling #<ASDF:CL-SOURCE-FILE "clpython.runtime" "runtime" "import">. 6 Continue, treating compiling #<ASDF:CL-SOURCE-FILE "clpython.runtime" "runtime" "import"> as having been successful. 7 (abort) Give up on "clpython" 8 Return to level 0. 9 Return to top-level loop. 10 Return from multiprocessing.

Is it related to https://github.com/metawilm/cl-python/issues/9?

After ql:quickload "clpython", the packages were downloaded and they began to be processed:

... [package clpython.module]......................... [package clpython]................................ [package clpython.app.repl]...... Error: No translations for logical pathname host "TEMPORARY-FILES". 1 (continue) Try loading /home/caldwell/.cache/common-lisp/lw-6.1.1-linux-x64/home/caldwell/quicklisp/dists/quicklisp/software/temporary-file-20120909-git/temporary-file.64ufasl again. ...

Roughly:

• (CLPYTHON:RUN #p"/usr/lib/python2.6/test/pystone.py" :compile nil :args '("pystone.py" "1000000")) Pystone(1.1) time for 1000000 passes = 28.630005
  This machine benchmarks at 34928.39 pystones/second
• (CLPYTHON:RUN #p"/usr/lib/python2.6/test/pystone.py" :compile t :args '("pystone.py" "1000000")) Pystone(1.1) time for 1000000 passes = 29.440002
  This machine benchmarks at 33967.387 pystones/second

As a contrast, this machine marks at roughly 90000 pystones with cpython 2.6.

Upcoming ECL release will have this problem solved (and clpython does not load with this fix present, because class is not available at compile time).

(defun |tee| (iterable &optional (n 2))
  "Create N independent generators"
  (check-type n integer)
  (cond ((zerop n)
         (make-tuple-from-list ()))
        (t
         (let ((iter-func (clpython::get-py-iterate-fun iterable))
               (deques (loop repeat n collect (clpython.module._collections:|deque|))))
           (flet ((generator (deque)
                    (clpython::make-iterator-from-function
                     :name :tee-iterator
                     :func (lambda () 
                             (if (clpython::py-val->lisp-bool deque)
                                 (clpython.module._collections::deque.popleft deque)
                               (let ((new-val (funcall iter-func)))
                                 ;; also continue if NEW-VAL is NIL
                                 (dolist (d deques)
                                   (unless (eq d deque)
                                     (clpython.module._collections::deque.append d new-val)))
                                 new-val))))))
             (make-tuple-from-list (mapcar 'generator deques)))))))

Probably the MAPCAR should use #'generator and not 'generator

In reference to issue https://github.com/metawilm/cl-python/issues/1, I was still unable to load python files with mixed case on SBCL. CLPYTHON:PARSE works fine on the mixed-case pathname and just calls SLURP-FILE on the provided pathname. I can then call CLPYTHON:RUN-PYTHON-AST on the parsed AST.

However, giving the pathname to CLPYTHON:RUN would result in an additional call to SLURP-FILE during macroexpansion of [MODULE-STMT]. In [MODULE-STMT], the pathname is reinterpreted through DERIVE-PATHNAME, which on SBCL alters the provided case. Removing the call to DERIVE-PATHNAME in [MODULE-STMT] appears to workaround this issue.

Please consider adding :description, :author and :license information to your ASDF system(s). This will greatly help Quicklisp users and make it easier for them to report bugs.

More information:

http://blog.quicklisp.org/2015/05/looking-for-more-metadata.html https://www.quicklisp.org

Hey,

I have been playing around with the grammar for cl-python trying to get to more compatable with the python 2.7.5 grammar. This is an extension of the existing import grammar that will allow trailing commas.

from sys import (path, argv,)

I have been having some problems getting the test cases to re-compile as expected. I seem to have to quickload the whole clpython.test to actually get correct results after i have had one failed run. Otherewise it seems to report the same failure over and over.

Thanks, Russell

When attempting to load the library bdb, it fails with the following error

There is no applicable method for the generic function
  #<STANDARD-GENERIC-FUNCTION CLPYTHON::PY-FUNCTION-CODE (1)>
when called with arguments
  (#<FUNCTION # {10100EF89B}>).
   [Condition of type SIMPLE-ERROR]

After loading the latest HEAD from github with ql:quickload (by placing the repo in ~/quicklisp/local-projects), I run the following to produce this error

CL-USER> (clpython.app.repl:repl)
Welcome to CLPython, an implementation of Python in Common Lisp.
Running on: SBCL 1.1.2.35-2b29a7c
REPL shortcuts: `:q' = quit, `:h' = help.

;; Consider customizing variable *CLPYTHON-MODULE-SEARCH-PATHS*
;; to be a list of paths that are tried when locating a module in order to import it
;; (in addition to the current directory and `sys.path'). Typically it should at least
;; contain the path to the Python (2.5) standard libraries, as those are not
;; distributed with CLPython.
>>> import sys
#<module SYS (package "CLPYTHON.MODULE.SYS" with 41 external symbols) {10044C1113}>
>>> print sys.path
[.]
>>> sys.path = ['/usr/lib/python2.7/']

[/usr/lib/python2.7/]
>>> import bdb
;; Parsing #P"/usr/lib/python2.7/bdb.py" into #P"/tmp/clpython-bdb-lisp-58ac5use"
;; Parsing #P"/usr/lib/python2.7/fnmatch.py" into #P"/tmp/clpython-fnmatch-lisp-gfao9rlf"
;; Parsing #P"/usr/lib/python2.7/os.py" into #P"/tmp/clpython-os-lisp-bcztmntz"
;; Parsing #P"/usr/lib/python2.7/posixpath.py" into #P"/tmp/clpython-posixpath-lisp-v3d8ocj1"
;; Parsing #P"/usr/lib/python2.7/stat.py" into #P"/tmp/clpython-stat-lisp-hm7j90p1"
;; Parsing #P"/usr/lib/python2.7/genericpath.py" into #P"/tmp/clpython-genericpath-lisp-uiuj8gke"
;; Parsing #P"/usr/lib/python2.7/warnings.py" into #P"/tmp/clpython-warnings-lisp-u1dmflid"
;; Parsing #P"/usr/lib/python2.7/linecache.py" into #P"/tmp/clpython-linecache-lisp-qjgoy79x"
;; Parsing #P"/usr/lib/python2.7/types.py" into #P"/tmp/clpython-types-lisp-2mfjlxbg"

Let me know if there's any other information I can provide to help with this.