Author: Armin Rigo <ar...@tunes.org>
Branch: intern-not-immortal
Changeset: r74612:e95ebc7224cd
Date: 2014-11-20 13:52 +0100
http://bitbucket.org/pypy/pypy/changeset/e95ebc7224cd/

Log:    Intern the identifier-like strings that appear in the compiled code.

diff --git a/pypy/interpreter/astcompiler/assemble.py 
b/pypy/interpreter/astcompiler/assemble.py
--- a/pypy/interpreter/astcompiler/assemble.py
+++ b/pypy/interpreter/astcompiler/assemble.py
@@ -2,7 +2,7 @@
 Python control flow graph generation and bytecode assembly.
 """
 
-from pypy.interpreter.astcompiler import ast, symtable
+from pypy.interpreter.astcompiler import ast, symtable, misc
 from pypy.interpreter import pycode
 from pypy.tool import stdlib_opcode as ops
 
@@ -365,7 +365,9 @@
                     raise
                 break
             w_index = space.getitem(w_consts, w_key)
-            consts_w[space.int_w(w_index)] = space.getitem(w_key, first)
+            w_constant = space.getitem(w_key, first)
+            w_constant = misc.intern_if_common_string(space, w_constant)
+            consts_w[space.int_w(w_index)] = w_constant
         return consts_w
 
     def _get_code_flags(self):
diff --git a/pypy/interpreter/astcompiler/misc.py 
b/pypy/interpreter/astcompiler/misc.py
--- a/pypy/interpreter/astcompiler/misc.py
+++ b/pypy/interpreter/astcompiler/misc.py
@@ -106,3 +106,13 @@
     except IndexError:
         return name
     return "_%s%s" % (klass[i:], name)
+
+
+def intern_if_common_string(space, w_const):
+    # only intern identifier-like strings
+    if not space.is_w(space.type(w_const), space.w_str):
+        return w_const
+    for c in space.str_w(w_const):
+        if not (c.isalnum() or c == '_'):
+            return w_const
+    return space.new_interned_w_str(w_const)
diff --git a/pypy/interpreter/astcompiler/optimize.py 
b/pypy/interpreter/astcompiler/optimize.py
--- a/pypy/interpreter/astcompiler/optimize.py
+++ b/pypy/interpreter/astcompiler/optimize.py
@@ -272,6 +272,11 @@
                 if w_const is None:
                     return tup
                 consts_w[i] = w_const
+            # intern the string constants packed into the tuple here,
+            # because assemble.py will see the result as just a tuple constant
+            for i in range(len(consts_w)):
+                consts_w[i] = misc.intern_if_common_string(
+                    self.space, consts_w[i])
         else:
             consts_w = []
         w_consts = self.space.newtuple(consts_w)
diff --git a/pypy/interpreter/test/test_compiler.py 
b/pypy/interpreter/test/test_compiler.py
--- a/pypy/interpreter/test/test_compiler.py
+++ b/pypy/interpreter/test/test_compiler.py
@@ -970,7 +970,12 @@
             sys.stdout = out
         output = s.getvalue()
         assert "CALL_METHOD" in output
-            
+
+    def test_interned_strings(self):
+        source = """x = ('foo_bar42', 5); y = 'foo_bar42'; z = x[0]"""
+        exec source
+        assert y is z
+
 
 class AppTestExceptions:
     def test_indentation_error(self):
_______________________________________________
pypy-commit mailing list
pypy-commit@python.org
https://mail.python.org/mailman/listinfo/pypy-commit

Reply via email to