Split non-ML part into a new talon-core package

The nonstandard `setup.py install --no-ml` option didn’t work
correctly with pip.  We could move the ML dependencies into an extra,
but that approach was previously rejected (mailgun#121): extras can’t be
enabled by default, so it could have been disruptive to existing
users.  Instead, we split off a new talon-core package with no ML
dependencies, and have talon re-export everything from it.

Fixes mailgun#130; fixes mailgun#131.

Signed-off-by: Anders Kaseorg <andersk@mit.edu>

README.rst

@@ -0,0 +1 @@
+talon/README.rst

talon-core/MANIFEST.in

@@ -0,0 +1,5 @@
+recursive-exclude tests *.pyc *~
+recursive-exclude talon *.pyc *~
+include LICENSE
+include MANIFEST.in
+include README.rst

talon-core/README.rst

@@ -0,0 +1,6 @@
+talon-core
+==========
+
+This is the part of talon that does not depend on NumPy, SciPy, and
+scikit-learn, and does not include any machine learning functionality.
+See the main talon package for documentation.

talon-core/setup.py

@@ -0,0 +1,31 @@
+from __future__ import absolute_import
+from setuptools import setup, find_packages
+
+
+setup(name='talon-core',
+      version='1.4.8',
+      description=("Mailgun library "
+                   "to extract message quotations and signatures."),
+      long_description=open("README.rst").read(),
+      author='Mailgun Inc.',
+      author_email='admin@mailgunhq.com',
+      url='https://github.com/mailgun/talon',
+      license='APACHE2',
+      packages=find_packages(exclude=['tests', 'tests.*']),
+      include_package_data=True,
+      zip_safe=True,
+      install_requires=[
+          "lxml>=2.3.3",
+          "regex>=1",
+          'chardet>=1.0.1',
+          'cchardet>=0.3.5',
+          'cssselect',
+          'six>=1.10.0',
+          'html5lib'
+          ],
+      tests_require=[
+          "mock",
+          "nose>=1.2.1",
+          "coverage"
+          ]
+      )

talon-core/talon_core/__init__.py

@@ -0,0 +1,6 @@
+from __future__ import absolute_import
+from talon_core.quotations import register_xpath_extensions
+
+
+def init():
+    register_xpath_extensions()

talon/html_quotations.py → talon-core/talon_core/html_quotations.py

@@ -6,7 +6,7 @@
 from __future__ import absolute_import
 import regex as re
 
-from talon.utils import cssselect 
+from talon_core.utils import cssselect 
 
 CHECKPOINT_PREFIX = '#!%!'
 CHECKPOINT_SUFFIX = '!%!#'

talon/quotations.py → talon-core/talon_core/quotations.py

@@ -12,9 +12,9 @@
 
 from lxml import html, etree
 
-from talon.utils import (get_delimiter, html_tree_to_text,
+from talon_core.utils import (get_delimiter, html_tree_to_text,
                          html_document_fromstring)
-from talon import html_quotations
+from talon_core import html_quotations
 from six.moves import range
 import six
 

talon/signature/bruteforce.py → talon-core/talon_core/signature/bruteforce.py

@@ -4,9 +4,9 @@
 
 import regex as re
 
-from talon.signature.constants import (SIGNATURE_MAX_LINES,
+from talon_core.signature.constants import (SIGNATURE_MAX_LINES,
                                        TOO_LONG_SIGNATURE_LINE)
-from talon.utils import get_delimiter
+from talon_core.utils import get_delimiter
 
 log = logging.getLogger(__name__)
 

talon/utils.py → talon-core/talon_core/utils.py

@@ -12,7 +12,7 @@
 from lxml.cssselect import CSSSelector
 from lxml.html import html5parser
 
-from talon.constants import RE_DELIMITER
+from talon_core.constants import RE_DELIMITER
 
 
 def safe_format(format_string, *args, **kwargs):

talon-core/tests/__init__.py

@@ -0,0 +1,11 @@
+from __future__ import absolute_import
+from nose.tools import *
+from mock import *
+
+import talon_core
+
+
+EML_MSG_FILENAME = "tests/fixtures/standard_replies/yahoo.eml"
+
+
+talon_core.init()

tests/html_quotations_test.py → talon-core/tests/html_quotations_test.py

@@ -5,7 +5,7 @@
 # noinspection PyUnresolvedReferences
 import re
 
-from talon import quotations, utils as u
+from talon_core import quotations, utils as u
 from . import *
 from .fixtures import *
 from lxml import html

tests/quotations_test.py → talon-core/tests/quotations_test.py

@@ -4,7 +4,7 @@
 from . import *
 from . fixtures import *
 
-from talon import quotations
+from talon_core import quotations
 
 
 @patch.object(quotations, 'extract_from_html')

tests/signature/bruteforce_test.py → talon-core/tests/signature/bruteforce_test.py

@@ -3,7 +3,7 @@
 from __future__ import absolute_import
 from .. import *
 
-from talon.signature import bruteforce
+from talon_core.signature import bruteforce
 
 
 def test_empty_body():

tests/text_quotations_test.py → talon-core/tests/text_quotations_test.py

@@ -7,7 +7,7 @@
 import os
 
 import email.iterators
-from talon import quotations
+from talon_core import quotations
 import six
 from six.moves import range
 from six import StringIO

tests/utils_test.py → talon-core/tests/utils_test.py

@@ -5,7 +5,7 @@
 import cchardet
 import six
 
-from talon import utils as u
+from talon_core import utils as u
 from . import *