9:07 p.m.
On 2014年01月18日 02:01, Aline Manera wrote:
On 01/14/2014 01:47 PM, lvroyce0210(a)gmail.com wrote:
> From: Royce Lv <lvroyce(a)linux.vnet.ibm.com>
>
> Add support for passing params for GET method,
> This include add param to get_list method,
> and filter the result with given params.
> we will call it like:
> GET /collection?filter_field=value
> With current implementation with a wrapper,
> we can easily support any collection query with param.
>
> Signed-off-by: Royce Lv <lvroyce(a)linux.vnet.ibm.com>
> ---
> src/kimchi/control/base.py | 28 ++++++++++++++++++++--------
> src/kimchi/control/storagepools.py | 4 ++--
> src/kimchi/control/storagevolumes.py | 2 +-
> src/kimchi/control/utils.py | 2 ++
> 4 files changed, 25 insertions(+), 11 deletions(-)
>
> diff --git a/src/kimchi/control/base.py b/src/kimchi/control/base.py
> index 185c8d8..a477abb 100644
> --- a/src/kimchi/control/base.py
> +++ b/src/kimchi/control/base.py
> @@ -28,7 +28,7 @@ import urllib2
>
> import kimchi.template
> from kimchi.control.utils import get_class_name, internal_redirect,
> model_fn
> -from kimchi.control.utils import parse_request, validate_method
> +from kimchi.control.utils import parse_request, parse_param,
> validate_method
> from kimchi.control.utils import validate_params
> from kimchi.exception import InvalidOperation, InvalidParameter
> from kimchi.exception import MissingParameter, NotFoundError,
> OperationFailed
> @@ -212,10 +212,10 @@ class Collection(object):
>
> return res.get()
>
> - def _get_resources(self):
> + def _get_resources(self, filter_params):
> try:
> get_list = getattr(self.model, model_fn(self, 'get_list'))
> - idents = get_list(*self.model_args)
> + idents = get_list(*self.model_args, **filter_params)
If the filter will be done in filter_data() below we don't need to
pass filter_params to get_list
My plan is to make it effective in either of these
two places:
reasons are according to http://libvirt.org/html/libvirt-libvirt.html
1. Many list API takes flags to indicate what kinds of resource they
will query, this act as a filter, is supported by libvirt itself, we
don't need to filter on our own.
2. Some full resource query is quite time consuming and not needed:
such as: user want to query 'nfs' targets on a server, if without
get_list filter, we will do:
1> query all type of targets: nfs+ iscsi+ gluster, etc
2> filter out iscsi + gluster
which is quite waste of effort.
> res_list = []
> for ident in idents:
> # internal text, get_list changes ident to unicode for sorted
> @@ -234,19 +234,31 @@ class Collection(object):
> args = self.resource_args + [ident.decode("utf-8")]
> return self.resource(self.model, *args)
>
> - def get(self):
> - resources = self._get_resources()
> + def filter_data(self, resources, filter_params):
> data = []
> for res in resources:
> - data.append(res.data)
> + valid = True
> + for key, val in filter_params.items():
> + if key in res.data and res.data[key] != val:
> + valid = False
> + break
You should continue the loop even when one resource doesn't have the
filter parameter.
yeah, my condition is "if key in res.data and res.data[key]
!= val" it
means if resource doesn't have filter parameter, it will ignore this key
and iterate next filter key.
For example, some storagepools have "task_id" parameter and
other ones
don't have.
> + if valid:
> + data.append(res.data)
> + return data
> +
> + def get(self, filter_params):
> + resources = self._get_resources(filter_params)
> + data = self.filter_data(resources, filter_params)
> return kimchi.template.render(get_class_name(self), data)
>
> @cherrypy.expose
> - def index(self, *args):
> + def index(self, *args, **kwargs):
> method = validate_method(('GET', 'POST'))
> if method == 'GET':
> try:
> - return self.get()
> + filter_params = parse_param()
> + validate_params(filter_params, self, 'get_list')
> + return self.get(filter_params)
> except InvalidOperation, param:
> error = "Invalid operation: '%s'" % param
> raise cherrypy.HTTPError(400, error)
> diff --git a/src/kimchi/control/storagepools.py
> b/src/kimchi/control/storagepools.py
> index 782f5a6..e3236a7 100644
> --- a/src/kimchi/control/storagepools.py
> +++ b/src/kimchi/control/storagepools.py
> @@ -63,9 +63,9 @@ class StoragePools(Collection):
>
> return resp
>
> - def _get_resources(self):
> + def _get_resources(self, filter_params):
> try:
> - res_list = super(StoragePools, self)._get_resources()
> + res_list = super(StoragePools, self)._get_resources(filter_params)
> # Append reserved pools
> isos = getattr(self, ISO_POOL_NAME)
> isos.lookup()
> diff --git a/src/kimchi/control/storagevolumes.py
> b/src/kimchi/control/storagevolumes.py
> index d541807..cd15bcc 100644
> --- a/src/kimchi/control/storagevolumes.py
> +++ b/src/kimchi/control/storagevolumes.py
> @@ -70,7 +70,7 @@ class IsoVolumes(Collection):
> super(IsoVolumes, self).__init__(model)
> self.pool = pool
>
> - def get(self):
> + def get(self, filter_params):
> res_list = []
> try:
> get_list = getattr(self.model, model_fn(self, 'get_list'))
> diff --git a/src/kimchi/control/utils.py b/src/kimchi/control/utils.py
> index c3c5f8e..4b3c4b0 100644
> --- a/src/kimchi/control/utils.py
> +++ b/src/kimchi/control/utils.py
> @@ -81,6 +81,8 @@ def parse_request():
> raise cherrypy.HTTPError(415, "This API only supports"
> " 'application/json'")
>
> +def parse_param():
> + return cherrypy.request.params
You don't need a function for it.
It is just one line.
Use cherrypy.request.params directly
ACK
>
> def internal_redirect(url):
> raise cherrypy.InternalRedirect(url.encode("utf-8"))