[pypy-svn] r30004 - pypy/extradoc/talk/dls2006
mwh at codespeak.net
mwh at codespeak.net
Thu Jul 13 11:41:25 CEST 2006
Author: mwh
Date: Thu Jul 13 11:41:23 2006
New Revision: 30004
Modified:
pypy/extradoc/talk/dls2006/paper.tex
Log:
a few sentence refactorings and a couple of XXX comments
Modified: pypy/extradoc/talk/dls2006/paper.tex
==============================================================================
--- pypy/extradoc/talk/dls2006/paper.tex (original)
+++ pypy/extradoc/talk/dls2006/paper.tex Thu Jul 13 11:41:23 2006
@@ -131,10 +131,10 @@
\texttt{isinstance} and type queries, but no runtime
reflection. Bindings in class and global namespaces are assumed
constant. RPython code can be run on a Python interpreter without
-severe semantics mismatches, figure \ref{fig_create_frame} shows some
-RPython code from the \textit{Standard Interpreter}, it is still quite
-idiomatic Python code using the built-in dictionary type and
-first-class class values instantiation.
+severe semantics mismatches. Figure \ref{fig_create_frame} shows some
+RPython code from the \textit{Standard Interpreter} and it can be seen
+that it is still quite idiomatic Python code using the built-in
+dictionary type and first-class class values instantiation.
\begin{figure}
\begin{verbatim}
@@ -281,6 +281,10 @@
C (currently based on a global flag to signal the presence of an
exception, which is set and checked around selected function calls).
+%%% XXX can we be more optimistic about the OOTyper? It's pretty
+%%% likely that pypy.net will work in some form by OOPSLA -- hi
+%%% Antonio :)
+
Finally, currently under development is a variant of the very first
transformation step, for use when targeting higher-level,
object-oriented (OO) environments. It is currently being designed
@@ -445,6 +449,9 @@
Python in term of struct- and array-like objects, just by running it.
The same holds for the GC.
+%%% XXX if we're looking to save space, i think this reason can go,
+%%% the others seem more important (to me, at least)
+
Secondly, and anecdotally, as the type inference process (section
\ref{typeinference}) is based on abstract interpretation, we can use
the following trick: the resulting type of most low-level operations
@@ -516,7 +523,7 @@
implement conditional code in a natural way. For example, the code in
figure \ref{fig_insertclean}, which is then specialized for the
possible types of dictionary \texttt{d}, shows this kind of
-compile-time computation and introspection on low-level types,
+compile-time computation and introspection on low-level types --
\texttt{ENTRY} is a C-like structure type and the
\texttt{hasattr}-ibute checks introspect it at translation time. This
kind of expressiveness increased our ability to reuse code even in
@@ -550,6 +557,8 @@
control flow graphs which contain direct references to the prebuilt data
objects, and then handles and transforms these graphs.
+%%% XXX the first sentence seems to come a bit out of the blue.
+
We found in-process debugging sufficient and did not implement dumping
of any intermediate step to disk. Figure \ref{fig_screenshot}
shows the control flow graph
More information about the Pypy-commit
mailing list