Merge changes from upstream

docs/conf.py

@@ -14,20 +14,14 @@
 import sphinx.environment
 from docutils.utils import get_source_line
 
-
-def _warn_node(self, msg, node):
-    """Do not warn on external images."""
-    if not msg.startswith('nonlocal image URI found:'):
-        self._warnfunc(msg, '{0}:{1}'.format(get_source_line(node)))
-
-sphinx.environment.BuildEnvironment.warn_node = _warn_node
-
-
 # -- General configuration ------------------------------------------------
 
 # If your documentation needs a minimal Sphinx version, state it here.
 #needs_sphinx = '1.0'
 
+# Do not warn on external images.
+suppress_warnings = ['image.nonlocal_uri']
+
 # Add any Sphinx extension module names here, as strings. They can be
 # extensions coming with Sphinx (named 'sphinx.ext.*') or your custom
 # ones.

setup.py

@@ -34,7 +34,7 @@
         'ConfigParser>=3.3.0r2',
     ],
     'docs': [
-        'Sphinx>=1.3',
+        'Sphinx>=1.4.2',
     ],
     'uwsgi': [
         'uWSGI>=2.0.3',
@@ -98,12 +98,11 @@
             'nara = timegate.examples.nara:NaraHandler',
             'orain = timegate.examples.orain:OrainHandler',
             'pastpages = timegate.examples.pastpages:PastpagesHandler',
-            'po = timegate.examples.po:PoHandler',
             'sg = timegate.examples.sg:SgHandler',
             'si = timegate.examples.si:SiHandler',
             'simple = timegate.examples.simple:ExampleHandler',
             'w3c = timegate.examples.w3c:W3cHandler',
-            'webcite = timegate.examples.webcite:WebciteHandler',
+            'webcite = timegate.examples.webcite:WebCiteHandler',
             'wikia = timegate.examples.wikia:WikiaHandler',
             'wikipedia = timegate.examples.wikipedia:WikipediaHandler',
         ],

timegate/application.py

@@ -12,17 +12,15 @@
 
 from __future__ import absolute_import, print_function
 
-import glob
-import importlib
-import inspect
 import json
 import logging
 import os
 from datetime import datetime
 
+from pkg_resources import iter_entry_points
+
 from dateutil.tz import tzutc
 from link_header import Link, LinkHeader
-from pkg_resources import iter_entry_points
 from werkzeug.exceptions import HTTPException, abort
 from werkzeug.http import http_date, parse_date
 from werkzeug.local import Local, LocalManager
@@ -31,10 +29,8 @@
 from werkzeug.wrappers import Request, Response
 
 from . import constants
-from ._compat import quote, unquote
 from .cache import Cache
 from .config import Config
-from .errors import TimegateError, URIRequestError
 from .handler import Handler, parsed_request
 from .utils import best
 
@@ -47,6 +43,9 @@
 request = local('request')
 """Proxy to request object."""
 
+# logging.getLogger(__name__)
+# logging.basicConfig(level=logging.DEBUG)
+
 
 def url_for(*args, **kwargs):
     """Proxy to URL Map adapter builder."""
@@ -106,6 +105,7 @@ def __init__(self, config=None, cache=None):
         self.config = Config(None)
         self.config.from_object(constants)
         self.config.update(config or {})
+        self.cache = None
         if cache:
             self.cache = cache
         elif self.config['CACHE_USE']:

timegate/conf/config.ini

@@ -14,11 +14,15 @@ strict_datetime = true
 # Timeout for any API request in seconds
 api_time_out = 6
 
+# user-agent
+# Provide a user-agent to be added to the requests made by the timegate server
+user_agent = Memento TimeGate
+
 [handler]
 # handler_class
 # Optional path to handler class. If not provided the program will
 # search core extensions for a possible handler.
-handler_class = timegate.examples.simple:ExampleHandler
+handler_class = timegate.examples.es.EsHandler
 
 # use_timemap
 # Optional boolean to define wether the program can handle timemap requests.
@@ -37,7 +41,7 @@ is_vcs = true
 # (Optional) String that will be prepended to requested URI if it is not already present
 # For example, if the server runs at `http://timegate.example.com` and all original resources begin with `http://example.com/res/{resource ID}`,
 # then setting `base_uri = http://example.com/res/` will allow short requests such `http://timegate.example.com/{resource ID}`
-base_uri = http://www.example.com/
+base_uri =
 
 [cache]
 

timegate/conf/timegate.ini

@@ -1,15 +1,16 @@
 # uWSGI launch configuration file
 [uwsgi]
-home = /data/venv/timegate/
-socket = /data/var/run/timegate/w3c/w3c.sock
-chdir = /data/web/timegate/w3c
+home = /Users/harihar/venv/timegate/
+#socket = uwsgi.sock
+http = :9000
+#chdir = /data/web/timegate/w3c
 
-daemonize = /data/var/logs/timegate/w3c.log
+#daemonize = /data/var/logs/timegate/w3c.log
 module = timegate.application
 callable = application
 master = true
-pidfile = /data/var/run/timegate/w3c/w3c.pid
-harakiri = 120
+#pidfile = /data/var/run/timegate/w3c/w3c.pid
+#harakiri = 120
 
 memory-report
 processes = 4

timegate/config.py

@@ -19,8 +19,22 @@
 class Config(dict):
     """Implement custom loaders to populate dict."""
 
+    _instance = None
+
+    def __new__(cls, root_path, defaults=None):
+        """
+        Converting this into a singleton for cached access.
+        :param root_path:
+        :param defaults:
+        :return:
+        """
+        if not cls._instance:
+            cls._instance = super(Config, cls).__new__(cls)
+        return cls._instance
+
     def __init__(self, root_path, defaults=None):
-        """Build an empty config wrapper.
+        """
+        Build an empty config wrapper.
 
         :param root_path: Path to which files are read relative from.
         :param defaults: An optional dictionary of default values.
@@ -36,6 +50,7 @@ def from_inifile(self, filename, silent=True):
 
         # Server configuration
         self['HOST'] = conf.get('server', 'host').rstrip('/')
+        self['USER_AGENT'] = conf.get('server', 'user_agent')
         self['STRICT_TIME'] = conf.getboolean('server', 'strict_datetime')
         if conf.has_option('server', 'api_time_out'):
             self['API_TIME_OUT'] = conf.getfloat('server', 'api_time_out')

timegate/examples/arxiv.py

@@ -28,8 +28,7 @@ def __init__(self):
         Handler.__init__(self)
 
         # Resources
-
-        # Ignores all that trails the identifier (? params, vX version,...)
+        # Ignores all that trails the identifier (? params, vX version,...info)
         self.rex = re.compile(
             r'(http://arxiv.org)/((?:pdf)|(?:abs))/(\d+\.\d+)(.*)')
         self.api_base = 'http://export.arxiv.org/oai2'

timegate/handler.py

@@ -19,6 +19,7 @@
 
 from . import utils as timegate_utils
 from ._compat import quote
+from .config import Config
 from .constants import API_TIME_OUT, TM_MAX_SIZE
 from .errors import HandlerError
 
@@ -43,6 +44,11 @@ def request(self, resource, timeout=API_TIME_OUT, **kwargs):
         :raises HandlerError: if the requests fails to access the API.
         """
         uri = resource
+        config = Config(None)
+        user_agent = config.get("USER_AGENT")
+        headers = {}
+        if user_agent:
+            headers["User-Agent"] = user_agent
 
         # Request logging with params
         try:
@@ -55,7 +61,7 @@ def request(self, resource, timeout=API_TIME_OUT, **kwargs):
             logging.info('Sending request for %s' % uri)
 
         try:
-            req = requests.get(uri, timeout=timeout, **kwargs)
+            req = requests.get(uri, timeout=timeout, headers=headers, **kwargs)
         except Exception as e:
             logging.error('Cannot request server (%s): %s' % (uri, e))
             raise HandlerError('Cannot request version server.', 502)