4956: Add maximum request size checking to http_request patch in api.py.

[arvados.git] / sdk / python / arvados / api.py

diff --git a/sdk/python/arvados/api.py b/sdk/python/arvados/api.py
index c618fc3c6623ef18e8663609340b55bcce219f78..949ead36443eae97f34fdb6be8badfb7b3ba8248 100644 (file)
--- a/sdk/python/arvados/api.py
+++ b/sdk/python/arvados/api.py
@@ -13,7 +13,6 @@ import errors
 import util
 
 _logger = logging.getLogger('arvados.api')
-conncache = {}
 
 class CredentialsFromToken(object):
     def __init__(self, api_token):
@@ -22,6 +21,12 @@ class CredentialsFromToken(object):
     @staticmethod
     def http_request(self, uri, **kwargs):
         from httplib import BadStatusLine
+
+        if (self.max_request_size and
+            kwargs.get('body') and
+            self.max_request_size < len(kwargs['body'])):
+            raise apiclient_errors.MediaUploadSizeError("Request size %i bytes exceeds published limit of %i bytes" % (len(kwargs['body']), self.max_request_size))
+
         if 'headers' not in kwargs:
             kwargs['headers'] = {}
 
@@ -39,10 +44,12 @@ class CredentialsFromToken(object):
             # previous call did not succeed, so this is slightly
             # risky.
             return self.orig_http_request(uri, **kwargs)
+
     def authorize(self, http):
         http.arvados_api_token = self.api_token
         http.orig_http_request = http.request
         http.request = types.MethodType(self.http_request, http)
+        http.max_request_size = 0
         return http
 
 # Monkey patch discovery._cast() so objects and arrays get serialized
@@ -77,17 +84,22 @@ def http_cache(data_type):
 def api(version=None, cache=True, host=None, token=None, insecure=False, **kwargs):
     """Return an apiclient Resources object for an Arvados instance.
 
-    Arguments:
-    * version: A string naming the version of the Arvados API to use (for
+    :version:
+      A string naming the version of the Arvados API to use (for
       example, 'v1').
-    * cache: If True (default), return an existing Resources object if
-      one already exists with the same endpoint and credentials. If
-      False, create a new one, and do not keep it in the cache (i.e.,
-      do not return it from subsequent api(cache=True) calls with
-      matching endpoint and credentials).
-    * host: The Arvados API server host (and optional :port) to connect to.
-    * token: The authentication token to send with each API call.
-    * insecure: If True, ignore SSL certificate validation errors.
+
+    :cache:
+      Use a cache (~/.cache/arvados/discovery) for the discovery
+      document.
+
+    :host:
+      The Arvados API server host (and optional :port) to connect to.
+
+    :token:
+      The authentication token to send with each API call.
+
+    :insecure:
+      If True, ignore SSL certificate validation errors.
 
     Additional keyword arguments will be passed directly to
     `apiclient_discovery.build` if a new Resource object is created.
@@ -113,13 +125,7 @@ def api(version=None, cache=True, host=None, token=None, insecure=False, **kwarg
     elif host and token:
         pass
     elif not host and not token:
-        # Load from user configuration or environment
-        for x in ['ARVADOS_API_HOST', 'ARVADOS_API_TOKEN']:
-            if x not in config.settings():
-                raise ValueError("%s is not set. Aborting." % x)
-        host = config.get('ARVADOS_API_HOST')
-        token = config.get('ARVADOS_API_TOKEN')
-        insecure = config.flag_is_true('ARVADOS_API_HOST_INSECURE')
+        return api_from_config(version=version, cache=cache, **kwargs)
     else:
         # Caller provided one but not the other
         if not host:
@@ -132,12 +138,6 @@ def api(version=None, cache=True, host=None, token=None, insecure=False, **kwarg
         kwargs['discoveryServiceUrl'] = (
             'https://%s/discovery/v1/apis/{api}/{apiVersion}/rest' % (host,))
 
-    if cache:
-        connprofile = (version, host, token, insecure)
-        svc = conncache.get(connprofile)
-        if svc:
-            return svc
-
     if 'http' not in kwargs:
         http_kwargs = {}
         # Prefer system's CA certificates (if available) over httplib2's.
@@ -155,7 +155,37 @@ def api(version=None, cache=True, host=None, token=None, insecure=False, **kwarg
 
     svc = apiclient_discovery.build('arvados', version, **kwargs)
     svc.api_token = token
+    kwargs['http'].max_request_size = svc._rootDesc.get('maxRequestSize', 0)
     kwargs['http'].cache = None
-    if cache:
-        conncache[connprofile] = svc
     return svc
+
+def api_from_config(version=None, apiconfig=None, **kwargs):
+    """Return an apiclient Resources object enabling access to an Arvados server
+    instance.
+
+    :version:
+      A string naming the version of the Arvados REST API to use (for
+      example, 'v1').
+
+    :apiconfig:
+      If provided, this should be a dict-like object (must support the get()
+      method) with entries for ARVADOS_API_HOST, ARVADOS_API_TOKEN, and
+      optionally ARVADOS_API_HOST_INSECURE.  If not provided, use
+      arvados.config (which gets these parameters from the environment by
+      default.)
+
+    Other keyword arguments such as `cache` will be passed along `api()`
+
+    """
+    # Load from user configuration or environment
+    if apiconfig is None:
+        apiconfig = config.settings()
+
+    for x in ['ARVADOS_API_HOST', 'ARVADOS_API_TOKEN']:
+        if x not in apiconfig:
+            raise ValueError("%s is not set. Aborting." % x)
+    host = apiconfig.get('ARVADOS_API_HOST')
+    token = apiconfig.get('ARVADOS_API_TOKEN')
+    insecure = config.flag_is_true('ARVADOS_API_HOST_INSECURE', apiconfig)
+
+    return api(version=version, host=host, token=token, insecure=insecure, **kwargs)