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>

.build/Dockerfile

@@ -14,7 +14,8 @@ COPY wheel/* /wheel/
 
 RUN mkdir -p ${REPORT_PATH}
 
-RUN python ./setup.py build bdist_wheel -d /wheel && \
+RUN (cd talon-core && python setup.py build bdist_wheel -d /wheel) && \
+    (cd talon && python setup.py build bdist_wheel -d /wheel) && \
     pip install --no-deps /wheel/*
 
 ENTRYPOINT ["/bin/sh", "/app/run_tests.sh"]

README.rst

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

run_tests.sh

@@ -1,4 +1,4 @@
 #!/usr/bin/env bash
 set -ex
 REPORT_PATH="${REPORT_PATH:-./}"
-nosetests --with-xunit --with-coverage --cover-xml --cover-xml-file $REPORT_PATH/coverage.xml --xunit-file=$REPORT_PATH/nosetests.xml --cover-package=talon .
+nosetests --with-xunit --with-coverage --cover-xml --cover-xml-file $REPORT_PATH/coverage.xml --xunit-file=$REPORT_PATH/nosetests.xml --cover-package=talon --cover-package=talon-core talon talon-core

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 +1,32 @@
+from __future__ import absolute_import
+from setuptools import setup, find_packages
+
+
+setup(name='talon-core',
+      version='1.6.0',
+      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",
+          "regex",
+          "chardet",
+          "cchardet",
+          "cssselect",
+          "six",
+          "html5lib",
+          "joblib",
+          ],
+      tests_require=[
+          "mock",
+          "nose",
+          "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

@@ -14,9 +14,9 @@
 from lxml import etree, html
 from six.moves import range
 
-from talon import html_quotations
-from talon.utils import (get_delimiter, html_document_fromstring,
-                         html_tree_to_text)
+from talon_core import html_quotations
+from talon_core.utils import (get_delimiter, html_document_fromstring,
+                              html_tree_to_text)
 
 log = logging.getLogger(__name__)
 

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

@@ -8,7 +8,7 @@
 from lxml.etree import _Element
 from lxml.html import html5parser
 
-from talon.constants import RE_DELIMITER
+from talon_core.constants import RE_DELIMITER
 
 
 def get_delimiter(msg_body: str) -> str:

talon-core/tests/__init__.py

@@ -0,0 +1,8 @@
+from __future__ import absolute_import
+from nose.tools import *
+from mock import *
+
+import talon_core
+
+
+talon_core.init()

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

@@ -12,7 +12,7 @@
                             OLK_SRC_BODY_SECTION,
                             REPLY_QUOTATIONS_SHARE_BLOCK,
                             REPLY_SEPARATED_BY_HR)
-from talon import quotations, utils as u
+from talon_core import quotations, utils as u
 
 RE_WHITESPACE = re.compile(r"\s")
 RE_DOUBLE_WHITESPACE = re.compile(r"\s")

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

@@ -2,7 +2,7 @@
 
 from __future__ import absolute_import
 
-from talon import utils as u
+from talon_core import utils as u
 from . import *