Move python source understanding into python.py

7 files changed, 142 insertions, 137 deletions

diff --git a/coverage/codeunit.py b/coverage/codeunit.py
index e9efa396..998aa098 100644
--- a/coverage/codeunit.py
+++ b/coverage/codeunit.py
@@ -1,12 +1,9 @@
 """Code unit (module) handling for Coverage."""
 
 import os
-import sys
 
 from coverage.backward import unicode_class
-from coverage.files import get_python_source, FileLocator
-from coverage.parser import PythonParser
-from coverage.phystokens import source_token_lines, source_encoding
+from coverage.files import FileLocator
 
 
 class CodeUnit(object):
@@ -114,54 +111,3 @@ class CodeUnit(object):
 
     def get_parser(self, exclude=None):
         raise NotImplementedError
-
-
-class PythonCodeUnit(CodeUnit):
-    """Represents a Python file."""
-
-    def __init__(self, morf, file_locator=None):
-        super(PythonCodeUnit, self).__init__(morf, file_locator)
-        self._source = None
-
-    def _adjust_filename(self, fname):
-        # .pyc files should always refer to a .py instead.
-        if fname.endswith(('.pyc', '.pyo')):
-            fname = fname[:-1]
-        elif fname.endswith('$py.class'):   # Jython
-            fname = fname[:-9] + ".py"
-        return fname
-
-    def source(self):
-        if self._source is None:
-            self._source = get_python_source(self.filename)
-            if sys.version_info < (3, 0):
-                encoding = source_encoding(self._source)
-                self._source = self._source.decode(encoding, "replace")
-            assert isinstance(self._source, unicode_class)
-        return self._source
-
-    def get_parser(self, exclude=None):
-        return PythonParser(filename=self.filename, exclude=exclude)
-
-    def should_be_python(self):
-        """Does it seem like this file should contain Python?
-
-        This is used to decide if a file reported as part of the execution of
-        a program was really likely to have contained Python in the first
-        place.
-
-        """
-        # Get the file extension.
-        _, ext = os.path.splitext(self.filename)
-
-        # Anything named *.py* should be Python.
-        if ext.startswith('.py'):
-            return True
-        # A file with no extension should be Python.
-        if not ext:
-            return True
-        # Everything else is probably not Python.
-        return False
-
-    def source_token_lines(self):
-        return source_token_lines(self.source())
diff --git a/coverage/control.py b/coverage/control.py
index 0ca1e95c..4aaf1af3 100644
--- a/coverage/control.py
+++ b/coverage/control.py
@@ -10,7 +10,7 @@ import sys
 
 from coverage.annotate import AnnotateReporter
 from coverage.backward import string_class, iitems
-from coverage.codeunit import CodeUnit, PythonCodeUnit
+from coverage.codeunit import CodeUnit
 from coverage.collector import Collector
 from coverage.config import CoverageConfig
 from coverage.data import CoverageData
@@ -22,6 +22,7 @@ from coverage.files import ModuleMatcher
 from coverage.html import HtmlReporter
 from coverage.misc import CoverageException, bool_or_none, join_regex
 from coverage.misc import file_be_gone, overrides
+from coverage.python import PythonCodeUnit
 from coverage.results import Analysis, Numbers
 from coverage.summary import SummaryReporter
 from coverage.xmlreport import XmlReporter
diff --git a/coverage/execfile.py b/coverage/execfile.py
index fc7ad174..2d856897 100644
--- a/coverage/execfile.py
+++ b/coverage/execfile.py
@@ -7,8 +7,8 @@ import types
 
 from coverage.backward import BUILTINS
 from coverage.backward import PYC_MAGIC_NUMBER, imp, importlib_util_find_spec
-from coverage.files import get_python_source
 from coverage.misc import ExceptionDuringRun, NoCode, NoSource
+from coverage.python import get_python_source
 
 
 class DummyLoader(object):
diff --git a/coverage/files.py b/coverage/files.py
index 9f0827fa..15ccabce 100644
--- a/coverage/files.py
+++ b/coverage/files.py
@@ -7,10 +7,8 @@ import os.path
 import posixpath
 import re
 import sys
-import tokenize
 
-from coverage.misc import CoverageException, NoSource, join_regex
-from coverage.phystokens import source_encoding
+from coverage.misc import CoverageException, join_regex
 
 
 class FileLocator(object):
@@ -56,80 +54,6 @@ class FileLocator(object):
         return self.canonical_filename_cache[filename]
 
 
-def read_python_source(filename):
-    """Read the Python source text from `filename`.
-
-    Returns a str: unicode on Python 3, bytes on Python 2.
-
-    """
-    # Python 3.2 provides `tokenize.open`, the best way to open source files.
-    if sys.version_info >= (3, 2):
-        f = tokenize.open(filename)
-    else:
-        f = open(filename, "rU")
-
-    with f:
-        return f.read()
-
-
-def get_python_source(filename):
-    """Return the source code, as a str."""
-    base, ext = os.path.splitext(filename)
-    if ext == ".py" and sys.platform == "win32":
-        exts = [".py", ".pyw"]
-    else:
-        exts = [ext]
-
-    for ext in exts:
-        try_filename = base + ext
-        if os.path.exists(try_filename):
-            # A regular text file: open it.
-            source = read_python_source(try_filename)
-            break
-
-        # Maybe it's in a zip file?
-        source = get_zip_bytes(try_filename)
-        if source is not None:
-            if sys.version_info >= (3, 0):
-                source = source.decode(source_encoding(source))
-            break
-    else:
-        # Couldn't find source.
-        raise NoSource("No source for code: '%s'." % filename)
-
-    # Python code should always end with a line with a newline.
-    if source and source[-1] != '\n':
-        source += '\n'
-
-    return source
-
-
-def get_zip_bytes(filename):
-    """Get data from `filename` if it is a zip file path.
-
-    Returns the bytestring data read from the zip file, or None if no zip file
-    could be found or `filename` isn't in it.  The data returned will be
-    an empty string if the file is empty.
-
-    """
-    import zipimport
-    markers = ['.zip'+os.sep, '.egg'+os.sep]
-    for marker in markers:
-        if marker in filename:
-            parts = filename.split(marker)
-            try:
-                zi = zipimport.zipimporter(parts[0]+marker[:-1])
-            except zipimport.ZipImportError:
-                continue
-            try:
-                data = zi.get_data(parts[1])
-            except IOError:
-                continue
-            assert isinstance(data, bytes)
-            return data
-    return None
-
-
 if sys.platform == 'win32':
 
     def actual_path(path):
diff --git a/coverage/parser.py b/coverage/parser.py
index a2fa1dfe..f488367d 100644
--- a/coverage/parser.py
+++ b/coverage/parser.py
@@ -9,7 +9,6 @@ import tokenize
 from coverage.backward import range    # pylint: disable=redefined-builtin
 from coverage.backward import bytes_to_ints
 from coverage.bytecode import ByteCodes, CodeObjects
-from coverage.files import get_python_source
 from coverage.misc import nice_pair, expensive, join_regex
 from coverage.misc import CoverageException, NoSource, NotPython
 from coverage.phystokens import generate_tokens
@@ -46,6 +45,7 @@ class PythonParser(CodeParser):
         self.filename = filename or "<code>"
         self.text = text
         if not self.text:
+            from coverage.python import get_python_source
             try:
                 self.text = get_python_source(self.filename)
             except IOError as err:
diff --git a/coverage/plugin.py b/coverage/plugin.py
index a0a65e44..362e561a 100644
--- a/coverage/plugin.py
+++ b/coverage/plugin.py
@@ -1,7 +1,5 @@
 """Plugin management for coverage.py"""
 
-import sys
-
 
 class CoveragePlugin(object):
     """Base class for coverage.py plugins."""
diff --git a/coverage/python.py b/coverage/python.py
new file mode 100644
index 00000000..62376922
--- /dev/null
+++ b/coverage/python.py
@@ -0,0 +1,136 @@
+"""Python source expertise for coverage.py"""
+
+import os.path
+import sys
+import tokenize
+import zipimport
+
+from coverage.backward import unicode_class
+from coverage.codeunit import CodeUnit
+from coverage.misc import NoSource
+from coverage.parser import PythonParser
+from coverage.phystokens import source_token_lines, source_encoding
+
+
+def read_python_source(filename):
+    """Read the Python source text from `filename`.
+
+    Returns a str: unicode on Python 3, bytes on Python 2.
+
+    """
+    # Python 3.2 provides `tokenize.open`, the best way to open source files.
+    if sys.version_info >= (3, 2):
+        f = tokenize.open(filename)
+    else:
+        f = open(filename, "rU")
+
+    with f:
+        return f.read()
+
+
+def get_python_source(filename):
+    """Return the source code, as a str."""
+    base, ext = os.path.splitext(filename)
+    if ext == ".py" and sys.platform == "win32":
+        exts = [".py", ".pyw"]
+    else:
+        exts = [ext]
+
+    for ext in exts:
+        try_filename = base + ext
+        if os.path.exists(try_filename):
+            # A regular text file: open it.
+            source = read_python_source(try_filename)
+            break
+
+        # Maybe it's in a zip file?
+        source = get_zip_bytes(try_filename)
+        if source is not None:
+            if sys.version_info >= (3, 0):
+                source = source.decode(source_encoding(source))
+            break
+    else:
+        # Couldn't find source.
+        raise NoSource("No source for code: '%s'." % filename)
+
+    # Python code should always end with a line with a newline.
+    if source and source[-1] != '\n':
+        source += '\n'
+
+    return source
+
+
+def get_zip_bytes(filename):
+    """Get data from `filename` if it is a zip file path.
+
+    Returns the bytestring data read from the zip file, or None if no zip file
+    could be found or `filename` isn't in it.  The data returned will be
+    an empty string if the file is empty.
+
+    """
+    markers = ['.zip'+os.sep, '.egg'+os.sep]
+    for marker in markers:
+        if marker in filename:
+            parts = filename.split(marker)
+            try:
+                zi = zipimport.zipimporter(parts[0]+marker[:-1])
+            except zipimport.ZipImportError:
+                continue
+            try:
+                data = zi.get_data(parts[1])
+            except IOError:
+                continue
+            assert isinstance(data, bytes)
+            return data
+    return None
+
+
+class PythonCodeUnit(CodeUnit):
+    """Represents a Python file."""
+
+    def __init__(self, morf, file_locator=None):
+        super(PythonCodeUnit, self).__init__(morf, file_locator)
+        self._source = None
+
+    def _adjust_filename(self, fname):
+        # .pyc files should always refer to a .py instead.
+        if fname.endswith(('.pyc', '.pyo')):
+            fname = fname[:-1]
+        elif fname.endswith('$py.class'):   # Jython
+            fname = fname[:-9] + ".py"
+        return fname
+
+    def source(self):
+        if self._source is None:
+            self._source = get_python_source(self.filename)
+            if sys.version_info < (3, 0):
+                encoding = source_encoding(self._source)
+                self._source = self._source.decode(encoding, "replace")
+            assert isinstance(self._source, unicode_class)
+        return self._source
+
+    def get_parser(self, exclude=None):
+        return PythonParser(filename=self.filename, exclude=exclude)
+
+    def should_be_python(self):
+        """Does it seem like this file should contain Python?
+
+        This is used to decide if a file reported as part of the execution of
+        a program was really likely to have contained Python in the first
+        place.
+
+        """
+        # Get the file extension.
+        _, ext = os.path.splitext(self.filename)
+
+        # Anything named *.py* should be Python.
+        if ext.startswith('.py'):
+            return True
+        # A file with no extension should be Python.
+        if not ext:
+            return True
+        # Everything else is probably not Python.
+        return False
+
+    def source_token_lines(self):
+        return source_token_lines(self.source())