CI: Update deps, docs

Author: TomAugspurger

ci/requirements-2.7-64.run

@@ -11,7 +11,7 @@ sqlalchemy
 lxml=3.2.1
 scipy
 xlsxwriter
-boto
+s3fs
 bottleneck
 html5lib
 beautiful-soup

ci/requirements-2.7.run

@@ -11,7 +11,7 @@ sqlalchemy=0.9.6
 lxml=3.2.1
 scipy
 xlsxwriter=0.4.6
-boto=2.36.0
+s3fs
 bottleneck
 psycopg2=2.5.2
 patsy

ci/requirements-2.7_SLOW.run

@@ -13,7 +13,7 @@ numexpr
 pytables
 sqlalchemy
 lxml
-boto
+s3fs
 bottleneck
 psycopg2
 pymysql

ci/requirements-3.5.run

@@ -17,7 +17,7 @@ sqlalchemy
 pymysql
 psycopg2
 xarray
-boto
+s3fs
 
 # incompat with conda ATM
 # beautiful-soup

ci/requirements-3.5_OSX.run

@@ -12,7 +12,7 @@ matplotlib
 jinja2
 bottleneck
 xarray
-boto
+s3fs
 
 # incompat with conda ATM
 # beautiful-soup

doc/source/install.rst

@@ -262,7 +262,7 @@ Optional Dependencies
   * `XlsxWriter <https://pypi.python.org/pypi/XlsxWriter>`__: Alternative Excel writer
 
 * `Jinja2 <http://jinja.pocoo.org/>`__: Template engine for conditional HTML formatting.
-* `boto <https://pypi.python.org/pypi/boto>`__: necessary for Amazon S3 access.
+* `s3fs <http://s3fs.readthedocs.io/>`__: necessary for Amazon S3 access.
 * `blosc <https://pypi.python.org/pypi/blosc>`__: for msgpack compression using ``blosc``
 * One of `PyQt4
   <http://www.riverbankcomputing.com/software/pyqt/download>`__, `PySide

doc/source/whatsnew/v0.20.0.txt

@@ -93,25 +93,15 @@ Backwards incompatible API changes
 
 .. _whatsnew_0200.api:
 
-.. _whatsnew_2000.api.s3
-
-S3 File Handling
-^^^^^^^^^^^^^^^^
-
-pandas now uses `s3fs <http://s3fs.readthedocs.io/>`_ for handling S3 connections. This shouldn't break
+- pandas now uses `s3fs <http://s3fs.readthedocs.io/>`_ for handling S3 connections. This shouldn't break
 any code, but since s3fs is not a required dependency, you will need to install it separately (like boto
 in prior versions of pandas).
-
 - ``CParserError`` has been renamed to ``ParserError`` in ``pd.read_csv`` and will be removed in the future (:issue:`12665`)
 - ``SparseArray.cumsum()`` and ``SparseSeries.cumsum()`` will now always return ``SparseArray`` and ``SparseSeries`` respectively (:issue:`12855`)
 
-
-- ``CParserError`` has been renamed to ``ParserError`` in ``pd.read_csv`` and will be removed in the future (:issue:`12665`)
-
 Other API Changes
 ^^^^^^^^^^^^^^^^^
 
-
 .. _whatsnew_0200.deprecations:
 
 Deprecations

pandas/io/s3.py

@@ -2,6 +2,7 @@
 from pandas import compat
 try:
     import s3fs
+    from botocore.exceptions import NoCredentialsError
 except:
     raise ImportError("The s3fs library is required to handle s3 files")
 
@@ -19,15 +20,16 @@ def _strip_schema(url):
 
 def get_filepath_or_buffer(filepath_or_buffer, encoding=None,
                            compression=None):
-
-    # Assuming AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY and AWS_S3_HOST
-    # are environment variables
     fs = s3fs.S3FileSystem(anon=False)
     try:
         filepath_or_buffer = fs.open(_strip_schema(filepath_or_buffer))
-    except OSError:
+    except (OSError, NoCredentialsError):
         # boto3 has troubles when trying to access a public file
         # when credentialed...
+        # An OSError is raised if you have credentials, but they
+        # aren't valid for that bucket.
+        # A NoCredentialsError is raised if you don't have creds
+        # for that bucket.
         fs = s3fs.S3FileSystem(anon=True)
         filepath_or_buffer = fs.open(_strip_schema(filepath_or_buffer))
     return filepath_or_buffer, None, compression

pandas/util/print_versions.py

@@ -94,7 +94,7 @@ def show_versions(as_json=False):
         ("pymysql", lambda mod: mod.__version__),
         ("psycopg2", lambda mod: mod.__version__),
         ("jinja2", lambda mod: mod.__version__),
-        ("boto", lambda mod: mod.__version__),
+        ("s3fs", lambda mod: mod.__version__),
         ("pandas_datareader", lambda mod: mod.__version__)
     ]