Add untokenize() function to allow full round-trip tokenization.

Should significantly enhance the utility of the module by supporting
the creation of tools that modify the token stream and writeback the
modified result.

5 files changed, 182 insertions, 7 deletions

diff --git a/Doc/lib/libtokenize.tex b/Doc/lib/libtokenize.tex
index 6cd9348..dc5f8c1 100644
--- a/Doc/lib/libtokenize.tex
+++ b/Doc/lib/libtokenize.tex
@@ -45,6 +45,9 @@ An older entry point is retained for backward compatibility:
   provides the same interface as the \method{readline()} method of
   built-in file objects (see section~\ref{bltin-file-objects}).  Each
   call to the function should return one line of input as a string.
+  Alternately, \var{readline} may be a callable object that signals
+  completion by raising \exception{StopIteration}.
+  \versionchanged[Added StopIteration support]{2.5}
 
   The second parameter, \var{tokeneater}, must also be a callable
   object.  It is called once for each token, with five arguments,
@@ -65,3 +68,52 @@ passed to the \var{tokeneater} function by \function{tokenize()}:
   are generated when a logical line of code is continued over multiple
   physical lines.
 \end{datadesc}
+
+Another function is provided to reverse the tokenization process.
+This is useful for creating tools that tokenize a script, modify
+the token stream, and write back the modified script.
+
+\begin{funcdesc}{untokenize}{iterable}
+  Converts tokens back into Python source code.  The \variable{iterable}
+  must return sequences with at least two elements, the token type and
+  the token string.  Any additional sequence elements are ignored.
+
+  The reconstructed script is returned as a single string.  The
+  result is guaranteed to tokenize back to match the input so that
+  the conversion is lossless and round-trips are assured.  The
+  guarantee applies only to the token type and token string as
+  the spacing between tokens (column positions) may change.
+  \versionadded{2.5}
+\end{funcdesc}
+
+Example of a script re-writer that transforms float literals into
+Decimal objects:
+\begin{verbatim}
+def decistmt(s):
+    """Substitute Decimals for floats in a string of statements.
+
+    >>> from decimal import Decimal
+    >>> s = 'print +21.3e-5*-.1234/81.7'
+    >>> decistmt(s)
+    "print +Decimal ('21.3e-5')*-Decimal ('.1234')/Decimal ('81.7')"
+
+    >>> exec(s)
+    -3.21716034272e-007
+    >>> exec(decistmt(s))
+    -3.217160342717258261933904529E-7
+
+    """
+    result = []
+    g = generate_tokens(StringIO(s).readline)   # tokenize the string
+    for toknum, tokval, _, _, _  in g:
+        if toknum == NUMBER and '.' in tokval:  # replace NUMBER tokens
+            result.extend([
+                (NAME, 'Decimal'),
+                (OP, '('),
+                (STRING, repr(tokval)),
+                (OP, ')')
+            ])
+        else:
+            result.append((toknum, tokval))
+    return untokenize(result)
+\end{verbatim}
diff --git a/Lib/test/regrtest.py b/Lib/test/regrtest.py
index 6160b3d..85e784b 100755
--- a/Lib/test/regrtest.py
+++ b/Lib/test/regrtest.py
@@ -91,7 +91,9 @@ resources to test.  Currently only the following are defined:
 
     compiler -  Test the compiler package by compiling all the source
                 in the standard library and test suite.  This takes
-                a long time.
+                a long time.  Enabling this resource also allows
+                test_tokenize to verify round-trip lexing on every
+                file in the test library.
 
     subprocess  Run all tests for the subprocess module.
 
diff --git a/Lib/test/test_tokenize.py b/Lib/test/test_tokenize.py
index d217404..2ce435f 100644
--- a/Lib/test/test_tokenize.py
+++ b/Lib/test/test_tokenize.py
@@ -1,12 +1,82 @@
-from test.test_support import verbose, findfile
-import tokenize, os, sys
+from test.test_support import verbose, findfile, is_resource_enabled
+import os, glob, random
+from tokenize import (tokenize, generate_tokens, untokenize,
+                      NUMBER, NAME, OP, STRING)
 
 if verbose:
     print 'starting...'
 
 f = file(findfile('tokenize_tests' + os.extsep + 'txt'))
-tokenize.tokenize(f.readline)
+tokenize(f.readline)
 f.close()
 
+
+
+###### Test roundtrip for untokenize ##########################
+
+def test_roundtrip(f):
+    ## print 'Testing:', f
+    f = file(f)
+    try:
+        fulltok = list(generate_tokens(f.readline))
+    finally:
+        f.close()
+
+    t1 = [tok[:2] for tok in fulltok]
+    newtext = untokenize(t1)
+    readline = iter(newtext.splitlines(1)).next
+    t2 = [tok[:2] for tok in generate_tokens(readline)]
+    assert t1 == t2
+
+
+f = findfile('tokenize_tests' + os.extsep + 'txt')
+test_roundtrip(f)
+
+testdir = os.path.dirname(f) or os.curdir
+testfiles = glob.glob(testdir + os.sep + 'test*.py')
+if not is_resource_enabled('compiler'):
+    testfiles = random.sample(testfiles, 10)
+
+for f in testfiles:
+    test_roundtrip(f)
+
+
+
+###### Test example in the docs ###############################
+
+from decimal import Decimal
+from cStringIO import StringIO
+
+def decistmt(s):
+    """Substitute Decimals for floats in a string of statements.
+
+    >>> from decimal import Decimal
+    >>> s = 'print +21.3e-5*-.1234/81.7'
+    >>> decistmt(s)
+    "print +Decimal ('21.3e-5')*-Decimal ('.1234')/Decimal ('81.7')"
+
+    >>> exec(s)
+    -3.21716034272e-007
+    >>> exec(decistmt(s))
+    -3.217160342717258261933904529E-7
+
+    """
+    result = []
+    g = generate_tokens(StringIO(s).readline)   # tokenize the string
+    for toknum, tokval, _, _, _  in g:
+        if toknum == NUMBER and '.' in tokval:  # replace NUMBER tokens
+            result.extend([
+                (NAME, 'Decimal'),
+                (OP, '('),
+                (STRING, repr(tokval)),
+                (OP, ')')
+            ])
+        else:
+            result.append((toknum, tokval))
+    return untokenize(result)
+
+import doctest
+doctest.testmod()
+
 if verbose:
     print 'finished'
diff --git a/Lib/tokenize.py b/Lib/tokenize.py
index 9087e84..b29da6b 100644
--- a/Lib/tokenize.py
+++ b/Lib/tokenize.py
@@ -31,7 +31,7 @@ from token import *
 
 import token
 __all__ = [x for x in dir(token) if x[0] != '_'] + ["COMMENT", "tokenize",
-           "generate_tokens", "NL"]
+           "generate_tokens", "NL", "untokenize"]
 del x
 del token
 
@@ -159,12 +159,55 @@ def tokenize_loop(readline, tokeneater):
     for token_info in generate_tokens(readline):
         tokeneater(*token_info)
 
+
+def untokenize(iterable):
+    """Transform tokens back into Python source code.
+
+    Each element returned by the iterable must be a token sequence
+    with at least two elements, a token number and token value.
+
+    Round-trip invariant:
+        # Output text will tokenize the back to the input
+        t1 = [tok[:2] for tok in generate_tokens(f.readline)]
+        newcode = untokenize(t1)
+        readline = iter(newcode.splitlines(1)).next
+        t2 = [tok[:2] for tokin generate_tokens(readline)]
+        assert t1 == t2
+    """
+
+    startline = False
+    indents = []
+    toks = []
+    toks_append = toks.append
+    for tok in iterable:
+        toknum, tokval = tok[:2]
+
+        if toknum == NAME:
+            tokval += ' '
+
+        if toknum == INDENT:
+            indents.append(tokval)
+            continue
+        elif toknum == DEDENT:
+            indents.pop()
+            continue
+        elif toknum in (NEWLINE, COMMENT, NL):
+            startline = True
+        elif startline and indents:
+            toks_append(indents[-1])
+            startline = False
+        toks_append(tokval)
+    return ''.join(toks)
+
+
 def generate_tokens(readline):
     """
     The generate_tokens() generator requires one argment, readline, which
     must be a callable object which provides the same interface as the
     readline() method of built-in file objects. Each call to the function
-    should return one line of input as a string.
+    should return one line of input as a string.  Alternately, readline
+    can be a callable function terminating with StopIteration:
+        readline = open(myfile).next    # Example of alternate readline
 
     The generator produces 5-tuples with these members: the token type; the
     token string; a 2-tuple (srow, scol) of ints specifying the row and
@@ -180,7 +223,10 @@ def generate_tokens(readline):
     indents = [0]
 
     while 1:                                   # loop over lines in stream
-        line = readline()
+        try:
+            line = readline()
+        except StopIteration:
+            line = ''
         lnum = lnum + 1
         pos, max = 0, len(line)
 
diff --git a/Misc/NEWS b/Misc/NEWS
index f15e873..4763598 100644
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -141,6 +141,11 @@ Extension Modules
 Library
 -------
 
+- The tokenize module has a new untokenize() function to support a full
+  roundtrip from lexed tokens back to Python sourcecode.  In addition,
+  the generate_tokens() function now accepts a callable argument that
+  terminates by raising StopIteration.
+
 - Bug #1196315: fix weakref.WeakValueDictionary constructor.
 
 - Bug #1213894: os.path.realpath didn't resolve symlinks that were the first