[Python-checkins] r52292 - sandbox/trunk/import_in_py/importer.py

Thu Oct 12 01:56:57 CEST 2006

Author: brett.cannon
Date: Thu Oct 12 01:56:56 2006
New Revision: 52292
Modified:
 sandbox/trunk/import_in_py/importer.py
Log:
Update some ideas.
Modified: sandbox/trunk/import_in_py/importer.py
==============================================================================

--- sandbox/trunk/import_in_py/importer.py	(original)
+++ sandbox/trunk/import_in_py/importer.py	Thu Oct 12 01:56:56 2006
@@ -34,14 +34,18 @@
 Possible Py3K improvements:
 * Have __import__ check for sys.modules entry to alleviate need for every
 loader to do so.
- * Have __import__ pass in module to initialize for imported module so to
- alleviate loader from having to pull from sys.modules.
+ * Have __import__ pass into the loader the module to be initialized so as
+ remove that boilerplate (also keeps sys.modules manipulation within
+ __import__ when the previous suggestion is used).
 * Put importer objects directly into sys.path to remove need for
 sys.path_importer_cache. Could leave string entries on sys.path that do
- not have an importer that can handle them so they can be retried at the
- next needed import (that was not found in sys.modules).
+ not have an importer so they can be re-checked the next time a new import
+ is attempted.
+ + If __import__ handles sys.modules then the impact from having to
+ recheck sys.path entries that lack an importer is minimized as it is
+ only on imports that have not been handled before.
 
-XXX PTL use-case:
+PTL use-case:
 * Tweaked source files that need to be pre-processed before they are imported.
 * Should be able to write out bytecode files easily.
 * Should let them use as much base infrastructure from the source and bytecode
@@ -58,7 +62,7 @@
 - Set 'handles' to bytecode extension
 * 'ptlc'
 
-XXX zipimport use-case:
+zipimport use-case:
 * New path_hooks function.
 + If sys.path entry is an existing zip file, return an importer for it.
 + Initialize the path_hooks function with handlers that are to be considered
@@ -72,7 +76,8 @@
 - Have a 'name' attribute.
 * PySourceHandler() and PyBytecodeHandler() would need to switch to accepting a
 file-like object that has a read() method along with a name attribute (for use
- in compile() for error reporting).
+ in compile() for error reporting). Also do not pass in a source provider for
+ the bytecode handler so as to suppress that writing of new bytecode.
 
 """
 from __future__ import with_statement
@@ -230,6 +235,12 @@
 class PySourceHandler(object):
 
 """Handler for importing Python source modules."""
+ 
+ # XXX Need to have a way to generate bytecode when none was found in the
+ # first place. If there ends up being too much overlap with the bytecode
+ # handler in terms of bytecode generation then consider
+ # following the current convention and have a unified source/bytecode
+ # handler that does it all.
 
 handles = 'py'
 
