Use python-requests if available. Unit tests not tested without requests library

Use url_params instead of requests session. Removed option of prepared requests to simplify. test case updated to handle absence of requests library

updated comments and exception

Minor fix to test case

Author: Sky NSS

doc/source/whatsnew/v0.21.0.txt

@@ -79,6 +79,62 @@ Other Enhancements
 - :func:`date_range` now accepts 'YS' in addition to 'AS' as an alias for start of year (:issue:`9313`)
 - :func:`date_range` now accepts 'Y' in addition to 'A' as an alias for end of year (:issue:`9313`)
 
+.. _whatsnew_0210.enhancements.read_csv:
+
+``read_csv`` use `python-requests` (if installed) to support basic auth and much more
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+If `python-requests` library is installed try to use it first. If not, continue using urllib
+The :meth:`DataFrame.read_csv`, :meth:`DataFrame.read_html`, :meth:`DataFrame.read_json`,
+:meth:`DataFrame.read_excel` now allow optional param of ``req_session`` to pass in requests.Session()
+
+
+.. ipython:: python
+  import pandas as pd
+  from requests import Session
+
+  # req_session is optional parameter 
+  df = pd.read_csv('https://uname:[email protected]/bb.csv') # now url can contain username and pwd
+
+  # custom auth can be implemented
+  s = Session()
+  s.auth = MyAuthProvider('secret-key') # custom auth provider supported by requests
+  df = pd.read_csv(url, req_session=s)
+
+  # optional advanced scenarios: basic auth, timeout, disable ssl certificate verification, proxy, etc
+  s = Session()
+  s.auth = ('darth', 'l0rd')  # if user wants to perform basic auth Skip if url itself contains username and pwd
+  s.timeout = (3.05, 27)                           # if user wants to modify timeout
+  s.verify = False                                      # if user wants to disable ssl cert verification
+  s.headers.update( {'User-Agent': 'Custom user agent'} )  # extensible to set any custom header needed
+  s.proxies = { 'http': 'http://a.com:100'}  # if user has proxies 
+  s.cert = '/path/client.cert'                     # if custom cert is needed
+  df = pd.read_csv( 'https://aa.com/bbb.csv', req_session=s)
+
+  # support verbs other than 'GET' such as 'POST' using requests.PreparedRequest
+  r = Request('POST', 'http://joker:pwd@nlp_service.api/email_sentiment_extract?out=json')
+  prepped = req.prepare()
+  prepped.body = 'from: [email protected]\nto: [email protected]\nsubject:Complaint letter\n\nbody: I am feeling :(' # multiple lines
+  df = pd.read_json( prepped) # minor update pandas code to detect type(Request) and submit it using requests session in lieu of URL.
+  """
+  [{
+    'from': '[email protected]',
+    'to': '[email protected]',
+    'email_type': 'complaint',
+    'sentiment': 'unhappy',
+  }]
+  """
+  
+  # Event hooks callback (eg log http status codes or other callback related functions)
+  def print_http_status(r, *args, **kwargs):
+      print(r.status_code)
+      print(r.headers['Content-Length'])
+  s = Session()
+  s.hooks = dict(response=print_http_status)
+  df = pd.read_csv( 'https://aa.com/bbb.csv', req_session=s)
+
+
+
 .. _whatsnew_0210.api_breaking:
 
 Backwards incompatible API changes

pandas/io/common.py

@@ -47,6 +47,12 @@
 except:
     _PY_PATH_INSTALLED = False
 
+try:
+    import requests
+    _PY_REQUESTS_INSTALLED = True
+except ImportError:
+    _PY_REQUESTS_INSTALLED = False
+
 
 if compat.PY3:
     from urllib.request import urlopen, pathname2url
@@ -93,6 +99,11 @@ def __next__(self):
     BaseIterator.next = lambda self: self.__next__()
 
 
+def is_requests_pkg_avail():
+    """Checks if 'python-requests' package is already installed."""
+    return _PY_REQUESTS_INSTALLED
+
+
 def _is_url(url):
     """Check to see if a URL has a valid protocol.
 
@@ -176,8 +187,85 @@ def _stringify_path(filepath_or_buffer):
     return filepath_or_buffer
 
 
+def _is_handled_by_requests(o):
+    return _is_url(o) and parse_url(o).scheme in ['http', 'https']
+
+
+def gen_session(url_params):
+    """
+    Generate python-requests session from url_params dict
+    """
+    s = None
+    if url_params and type(url_params) is requests.sessions.Session:
+        s = url_params
+    else:
+        s = requests.Session()
+        s.stream = True
+        # Setting accept-encoding to None for backwards compatibility with
+        # urlopen. ideally we want to allow gzip download
+        # urlopen doesnt decompress automatically, requests does.
+        s.headers.update({'Accept-Encoding': None})
+    if url_params and type(url_params) is dict:
+        if url_params.get('auth', None) and not s.auth:
+            s.auth = url_params.get('auth')
+        if url_params.get('verify', True) is False and s.verify is not False:
+            s.verify = url_params.get('verify')
+    return s
+
+
+def fetch_url(url, url_params=None, skip_requests=False):
+    """
+    If url is url, first try python-requests else try urllib.
+    Note if requests library is used, auto gunzip is
+    disabled for backwards compatibility of code with urlopen
+
+    Parameters
+    ----------
+    url : str
+        Could be:
+            'http://cnn.com'
+            'file:///home/sky/aaa.csv'
+
+    url_params : dict or requests.Session(), default None
+        A python dict containing:
+            'auth': tuple (str, str) eg (unae, pwd)
+            'auth': Any other auth object accepted by requests
+            'verify': boolean, Default True
+                 If False, allow self signed and invalid SSL cert for https
+        or
+        A python requests.Session object if http(s) path to enable basic auth
+        and many other scenarios that requests allows
+
+        .. versionadded:: 0.21.0
+
+   skip_requests : boolean, default False
+       for testing - disable `requests` library Internal use only
+
+        .. versionadded:: 0.21.0
+    Raises
+    ------
+    ValueError if url_params specified without installed python-requests pkg
+    """
+    if is_requests_pkg_avail() and \
+            _is_handled_by_requests(url) and \
+            (not skip_requests):
+        s = gen_session(url_params)
+        resp = s.get(url)
+        resp.raise_for_status()
+        content_bytes = resp.content
+    else:
+        if url_params and (skip_requests or not is_requests_pkg_avail()):
+            msg = 'To utilize url_params, python-requests library is ' + \
+                  'required but not detected'
+            raise ValueError(msg)
+        resp = _urlopen(url)
+        content_bytes = resp.read()
+    return resp, content_bytes
+
+
 def get_filepath_or_buffer(filepath_or_buffer, encoding=None,
-                           compression=None):
+                           compression=None, url_params=None,
+                           skip_requests=False):
     """
     If the filepath_or_buffer is a url, translate and return the buffer.
     Otherwise passthrough.
@@ -188,19 +276,45 @@ def get_filepath_or_buffer(filepath_or_buffer, encoding=None,
                          or buffer
     encoding : the encoding to use to decode py3 bytes, default is 'utf-8'
 
+    compression : str, default None
+        indicate the compression such as 'gzip'.
+
+    url_params : dict or requests.Session(), default None
+        A python dict containing:
+            'auth': tuple (str, str) eg (unae, pwd)
+            'auth': Any other auth object accepted by requests
+            'verify': boolean, Default True
+                 If False, allow self signed and invalid SSL cert for https
+        or
+        A python requests.Session object if http(s) path to enable basic auth
+        and many other scenarios that requests allows
+
+        .. versionadded:: 0.21.0
+
+   skip_requests : boolean, default False
+       for testing - disable `requests` library Internal use only
+
+        .. versionadded:: 0.21.0
+
     Returns
     -------
     a filepath_or_buffer, the encoding, the compression
+
+    Raises
+    ------
+    ValueError if url_params specified without installed python-requests pkg
     """
     filepath_or_buffer = _stringify_path(filepath_or_buffer)
 
     if _is_url(filepath_or_buffer):
-        req = _urlopen(filepath_or_buffer)
+        req, content_bytes = fetch_url(filepath_or_buffer,
+                                       url_params,
+                                       skip_requests)
+        reader = BytesIO(content_bytes)
         content_encoding = req.headers.get('Content-Encoding', None)
         if content_encoding == 'gzip':
             # Override compression based on Content-Encoding header
             compression = 'gzip'
-        reader = BytesIO(req.read())
         return reader, encoding, compression
 
     if _is_s3_url(filepath_or_buffer):

pandas/io/excel.py

@@ -15,12 +15,13 @@
     is_integer, is_float,
     is_bool, is_list_like)
 
+from pandas.compat import BytesIO
 from pandas.core.frame import DataFrame
 from pandas.io.parsers import TextParser
 from pandas.errors import EmptyDataError
-from pandas.io.common import (_is_url, _urlopen, _validate_header_arg,
-                              get_filepath_or_buffer, _NA_VALUES,
-                              _stringify_path)
+from pandas.io.common import (_is_url, fetch_url,
+                              _validate_header_arg, get_filepath_or_buffer,
+                              _NA_VALUES, _stringify_path)
 from pandas.core.indexes.period import Period
 import pandas._libs.json as json
 from pandas.compat import (map, zip, reduce, range, lrange, u, add_metaclass,
@@ -211,7 +212,8 @@ def read_excel(io, sheet_name=0, header=0, skiprows=None, skip_footer=0,
                         "Use just `sheet_name`")
 
     if not isinstance(io, ExcelFile):
-        io = ExcelFile(io, engine=engine)
+        io = ExcelFile(io, engine=engine,
+                       url_params=kwds.get('url_params', None))
 
     return io._parse_excel(
         sheetname=sheet_name, header=header, skiprows=skiprows, names=names,
@@ -258,8 +260,10 @@ def __init__(self, io, **kwds):
 
         # If io is a url, want to keep the data as bytes so can't pass
         # to get_filepath_or_buffer()
-        if _is_url(self._io):
-            io = _urlopen(self._io)
+        if _is_url(self.io):
+            rs = kwds.get('url_params', None)
+            req, content = fetch_url(self.io, url_params=rs)
+            io = BytesIO(content)
         elif not isinstance(self.io, (ExcelFile, xlrd.Book)):
             io, _, _ = get_filepath_or_buffer(self._io)
 

pandas/io/html.py

@@ -14,7 +14,7 @@
 
 from pandas.core.dtypes.common import is_list_like
 from pandas.errors import EmptyDataError
-from pandas.io.common import (_is_url, urlopen,
+from pandas.io.common import (_is_url, fetch_url,
                               parse_url, _validate_header_arg)
 from pandas.io.parsers import TextParser
 from pandas.compat import (lrange, lmap, u, string_types, iteritems,
@@ -116,20 +116,31 @@ def _get_skiprows(skiprows):
                     type(skiprows).__name__)
 
 
-def _read(obj):
+def _read(obj, url_params=None):
     """Try to read from a url, file or string.
 
     Parameters
     ----------
     obj : str, unicode, or file-like
 
+    url_params : dict or requests.Session(), default None
+        A python dict containing:
+            'auth': tuple (str, str) eg (unae, pwd)
+            'auth': Any other auth object accepted by requests
+            'verify': boolean, Default True
+                 If False, allow self signed and invalid SSL certs for https
+        or
+        A python requests.Session object if http(s) path to enable basic auth
+        and many other scenarios that requests allows
+
+        .. versionadded:: 0.21.0
+
     Returns
     -------
     raw_text : str
     """
     if _is_url(obj):
-        with urlopen(obj) as url:
-            text = url.read()
+        req, text = fetch_url(obj, url_params)
     elif hasattr(obj, 'read'):
         text = obj.read()
     elif isinstance(obj, char_types):
@@ -187,11 +198,12 @@ class _HtmlFrameParser(object):
     functionality.
     """
 
-    def __init__(self, io, match, attrs, encoding):
+    def __init__(self, io, match, attrs, encoding, url_params=None):
         self.io = io
         self.match = match
         self.attrs = attrs
         self.encoding = encoding
+        self.url_params = url_params
 
     def parse_tables(self):
         tables = self._parse_tables(self._build_doc(), self.match, self.attrs)
@@ -444,7 +456,7 @@ def _parse_tables(self, doc, match, attrs):
         return result
 
     def _setup_build_doc(self):
-        raw_text = _read(self.io)
+        raw_text = _read(self.io, self.url_params)
         if not raw_text:
             raise ValueError('No text parsed from document: %s' % self.io)
         return raw_text
@@ -731,7 +743,8 @@ def _parse(flavor, io, match, attrs, encoding, **kwargs):
     retained = None
     for flav in flavor:
         parser = _parser_dispatch(flav)
-        p = parser(io, compiled_match, attrs, encoding)
+        p = parser(io, compiled_match, attrs, encoding,
+                   url_params=kwargs.get('url_params', None))
 
         try:
             tables = p.parse_tables()
@@ -755,7 +768,7 @@ def read_html(io, match='.+', flavor=None, header=None, index_col=None,
               skiprows=None, attrs=None, parse_dates=False,
               tupleize_cols=False, thousands=',', encoding=None,
               decimal='.', converters=None, na_values=None,
-              keep_default_na=True):
+              keep_default_na=True, url_params=None):
     r"""Read HTML tables into a ``list`` of ``DataFrame`` objects.
 
     Parameters
@@ -856,6 +869,12 @@ def read_html(io, match='.+', flavor=None, header=None, index_col=None,
 
         .. versionadded:: 0.19.0
 
+    url_params : requests.Session(), default None
+        A python requests.Session object if http(s) path to enable basic auth
+        and many other scenarios that requests allows
+
+        .. versionadded:: 0.21.0
+
     Returns
     -------
     dfs : list of DataFrames
@@ -903,4 +922,4 @@ def read_html(io, match='.+', flavor=None, header=None, index_col=None,
                   parse_dates=parse_dates, tupleize_cols=tupleize_cols,
                   thousands=thousands, attrs=attrs, encoding=encoding,
                   decimal=decimal, converters=converters, na_values=na_values,
-                  keep_default_na=keep_default_na)
+                  keep_default_na=keep_default_na, url_params=url_params)