[session] allow writable tx_attr and use it for commit_state
This keep clarifying the code.
# copyright 2003-2012 LOGILAB S.A. (Paris, FRANCE), all rights reserved.
# contact http://www.logilab.fr/ -- mailto:contact@logilab.fr
#
# This file is part of CubicWeb.
#
# CubicWeb is free software: you can redistribute it and/or modify it under the
# terms of the GNU Lesser General Public License as published by the Free
# Software Foundation, either version 2.1 of the License, or (at your option)
# any later version.
#
# CubicWeb is distributed in the hope that it will be useful, but WITHOUT
# ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
# FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
# details.
#
# You should have received a copy of the GNU Lesser General Public License along
# with CubicWeb. If not, see <http://www.gnu.org/licenses/>.
"""abstract views and templates classes for CubicWeb web client"""
__docformat__ = "restructuredtext en"
_ = unicode
import types, new
from cStringIO import StringIO
from warnings import warn
from functools import partial
from logilab.common.deprecation import deprecated
from logilab.common.registry import yes
from logilab.mtconverter import xml_escape
from rql import nodes
from cubicweb import NotAnEntity
from cubicweb.predicates import non_final_entity, nonempty_rset, none_rset
from cubicweb.appobject import AppObject
from cubicweb.utils import UStringIO, HTMLStream
from cubicweb.uilib import domid, js
from cubicweb.schema import display_name
# robots control
NOINDEX = u'<meta name="ROBOTS" content="NOINDEX" />'
NOFOLLOW = u'<meta name="ROBOTS" content="NOFOLLOW" />'
CW_XHTML_EXTENSIONS = '''[
<!ATTLIST html xmlns:cubicweb CDATA #FIXED \'http://www.logilab.org/2008/cubicweb\' >
<!ENTITY % coreattrs
"id ID #IMPLIED
class CDATA #IMPLIED
style CDATA #IMPLIED
title CDATA #IMPLIED
cubicweb:accesskey CDATA #IMPLIED
cubicweb:actualrql CDATA #IMPLIED
cubicweb:dataurl CDATA #IMPLIED
cubicweb:facetName CDATA #IMPLIED
cubicweb:facetargs CDATA #IMPLIED
cubicweb:fallbackvid CDATA #IMPLIED
cubicweb:fname CDATA #IMPLIED
cubicweb:initfunc CDATA #IMPLIED
cubicweb:inputid CDATA #IMPLIED
cubicweb:inputname CDATA #IMPLIED
cubicweb:limit CDATA #IMPLIED
cubicweb:loadtype CDATA #IMPLIED
cubicweb:loadurl CDATA #IMPLIED
cubicweb:maxlength CDATA #IMPLIED
cubicweb:required CDATA #IMPLIED
cubicweb:rooteid CDATA #IMPLIED
cubicweb:rql CDATA #IMPLIED
cubicweb:size CDATA #IMPLIED
cubicweb:sortvalue CDATA #IMPLIED
cubicweb:target CDATA #IMPLIED
cubicweb:tindex CDATA #IMPLIED
cubicweb:tlunit CDATA #IMPLIED
cubicweb:type CDATA #IMPLIED
cubicweb:unselimg CDATA #IMPLIED
cubicweb:uselabel CDATA #IMPLIED
cubicweb:value CDATA #IMPLIED
cubicweb:variables CDATA #IMPLIED
cubicweb:vid CDATA #IMPLIED
cubicweb:wdgtype CDATA #IMPLIED
"> ] '''
TRANSITIONAL_DOCTYPE = u'<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd" %s>\n' % CW_XHTML_EXTENSIONS
TRANSITIONAL_DOCTYPE_NOEXT = u'<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">\n'
STRICT_DOCTYPE = u'<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd" %s>\n' % CW_XHTML_EXTENSIONS
STRICT_DOCTYPE_NOEXT = u'<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">\n'
# base view object ############################################################
class View(AppObject):
"""This class is an abstraction of a view class, used as a base class for
every renderable object such as views, templates and other user interface
components.
A `View` is instantiated to render a result set or part of a result
set. `View` subclasses may be parametrized using the following class
attributes:
:py:attr:`templatable` indicates if the view may be embedded in a main
template or if it has to be rendered standalone (i.e. pure XML views must
not be embedded in the main template of HTML pages)
:py:attr:`content_type` if the view is not templatable, it should set the
`content_type` class attribute to the correct MIME type (text/xhtml being
the default)
:py:attr:`category` this attribute may be used in the interface to regroup
related objects (view kinds) together
:py:attr:`paginable`
:py:attr:`binary`
A view writes to its output stream thanks to its attribute `w` (the
append method of an `UStreamIO`, except for binary views).
At instantiation time, the standard `_cw`, and `cw_rset` attributes are
added and the `w` attribute will be set at rendering time to a write
function to use.
"""
__registry__ = 'views'
templatable = True
# content_type = 'application/xhtml+xml' # text/xhtml'
binary = False
add_to_breadcrumbs = True
category = 'view'
paginable = True
def __init__(self, req=None, rset=None, **kwargs):
super(View, self).__init__(req, rset=rset, **kwargs)
self.w = None
@property
def content_type(self):
return self._cw.html_content_type()
def set_stream(self, w=None):
if self.w is not None:
return
if w is None:
if self.binary:
self._stream = stream = StringIO()
else:
self._stream = stream = UStringIO()
w = stream.write
else:
stream = None
self.w = w
return stream
# main view interface #####################################################
def render(self, w=None, **context):
"""called to render a view object for a result set.
This method is a dispatched to an actual method selected
according to optional row and col parameters, which are locating
a particular row or cell in the result set:
* if row is specified, `cell_call` is called
* if none of them is supplied, the view is considered to apply on
the whole result set (which may be None in this case), `call` is
called
"""
# XXX use .cw_row/.cw_col
row = context.get('row')
if row is not None:
context.setdefault('col', 0)
view_func = self.cell_call
else:
view_func = self.call
stream = self.set_stream(w)
try:
view_func(**context)
except Exception:
self.debug('view call %s failed (context=%s)', view_func, context)
raise
# return stream content if we have created it
if stream is not None:
return self._stream.getvalue()
def tal_render(self, template, variables):
"""render a precompiled page template with variables in the given
dictionary as context
"""
from cubicweb.ext.tal import CubicWebContext
context = CubicWebContext()
context.update({'self': self, 'rset': self.cw_rset, '_' : self._cw._,
'req': self._cw, 'user': self._cw.user})
context.update(variables)
output = UStringIO()
template.expand(context, output)
return output.getvalue()
# should default .call() method add a <div classs="section"> around each
# rset item
add_div_section = True
def call(self, **kwargs):
"""the view is called for an entire result set, by default loop
other rows of the result set and call the same view on the
particular row
Views applicable on None result sets have to override this method
"""
rset = self.cw_rset
if rset is None:
raise NotImplementedError("%r an rset is required" % self)
wrap = self.templatable and len(rset) > 1 and self.add_div_section
# avoid re-selection if rset of size 1, we already have the most
# specific view
if rset.rowcount != 1:
kwargs.setdefault('initargs', self.cw_extra_kwargs)
for i in xrange(len(rset)):
if wrap:
self.w(u'<div class="section">')
self.wview(self.__regid__, rset, row=i, **kwargs)
if wrap:
self.w(u"</div>")
else:
if wrap:
self.w(u'<div class="section">')
kwargs.setdefault('col', 0)
self.cell_call(row=0, **kwargs)
if wrap:
self.w(u"</div>")
def cell_call(self, row, col, **kwargs):
"""the view is called for a particular result set cell"""
raise NotImplementedError(repr(self))
def linkable(self):
"""return True if the view may be linked in a menu
by default views without title are not meant to be displayed
"""
if not getattr(self, 'title', None):
return False
return True
def is_primary(self):
return self.cw_extra_kwargs.get('is_primary', self.__regid__ == 'primary')
def url(self):
"""return the url associated with this view. Should not be
necessary for non linkable views, but a default implementation
is provided anyway.
"""
rset = self.cw_rset
if rset is None:
return self._cw.build_url('view', vid=self.__regid__)
coltypes = rset.column_types(0)
if len(coltypes) == 1:
etype = iter(coltypes).next()
if not self._cw.vreg.schema.eschema(etype).final:
if len(rset) == 1:
entity = rset.get_entity(0, 0)
return entity.absolute_url(vid=self.__regid__)
# don't want to generate /<etype> url if there is some restriction
# on something else than the entity type
restr = rset.syntax_tree().children[0].where
# XXX norestriction is not correct here. For instance, in cases like
# "Any P,N WHERE P is Project, P name N" norestriction should equal
# True
norestriction = (isinstance(restr, nodes.Relation) and
restr.is_types_restriction())
if norestriction:
return self._cw.build_url(etype.lower(), vid=self.__regid__)
return self._cw.build_url('view', rql=rset.printable_rql(), vid=self.__regid__)
def set_request_content_type(self):
"""set the content type returned by this view"""
self._cw.set_content_type(self.content_type)
# view utilities ##########################################################
def wview(self, __vid, rset=None, __fallback_vid=None, **kwargs):
"""shortcut to self.view method automatically passing self.w as argument
"""
self._cw.view(__vid, rset, __fallback_vid, w=self.w, **kwargs)
def whead(self, data):
self._cw.html_headers.write(data)
def wdata(self, data):
"""simple helper that escapes `data` and writes into `self.w`"""
self.w(xml_escape(data))
def html_headers(self):
"""return a list of html headers (eg something to be inserted between
<head> and </head> of the returned page
by default return a meta tag to disable robot indexation of the page
"""
return [NOINDEX]
def page_title(self):
"""returns a title according to the result set - used for the
title in the HTML header
"""
vtitle = self._cw.form.get('vtitle')
if vtitle:
return self._cw._(vtitle)
# class defined title will only be used if the resulting title doesn't
# seem clear enough
vtitle = getattr(self, 'title', None) or u''
if vtitle:
vtitle = self._cw._(vtitle)
rset = self.cw_rset
if rset and rset.rowcount:
if rset.rowcount == 1:
try:
entity = rset.complete_entity(0, 0)
# use long_title to get context information if any
clabel = entity.dc_long_title()
except NotAnEntity:
clabel = display_name(self._cw, rset.description[0][0])
clabel = u'%s (%s)' % (clabel, vtitle)
else :
etypes = rset.column_types(0)
if len(etypes) == 1:
etype = iter(etypes).next()
clabel = display_name(self._cw, etype, 'plural')
else :
clabel = u'#[*] (%s)' % vtitle
else:
clabel = vtitle
return u'%s (%s)' % (clabel, self._cw.property_value('ui.site-title'))
@deprecated('[3.10] use vreg["etypes"].etype_class(etype).cw_create_url(req)')
def create_url(self, etype, **kwargs):
""" return the url of the entity creation form for a given entity type"""
return self._cw.vreg["etypes"].etype_class(etype).cw_create_url(
self._cw, **kwargs)
def field(self, label, value, row=True, show_label=True, w=None, tr=True,
table=False):
"""read-only field"""
if w is None:
w = self.w
if table:
w(u'<tr class="entityfield">')
else:
w(u'<div class="entityfield">')
if show_label and label:
if tr:
label = display_name(self._cw, label)
if table:
w(u'<th>%s</th>' % label)
else:
w(u'<span class="label">%s</span> ' % label)
if table:
if not (show_label and label):
w(u'<td colspan="2">%s</td></tr>' % value)
else:
w(u'<td>%s</td></tr>' % value)
else:
w(u'<span>%s</span></div>' % value)
# concrete views base classes #################################################
class EntityView(View):
"""base class for views applying on an entity (i.e. uniform result set)"""
__select__ = non_final_entity()
category = _('entityview')
def call(self, **kwargs):
if self.cw_rset is None:
# * cw_extra_kwargs is the place where extra selection arguments are
# stored
# * when calling req.view('somevid', entity=entity), 'entity' ends
# up in cw_extra_kwargs and kwargs
#
# handle that to avoid a TypeError with a sanity check
#
# Notice that could probably be avoided by handling entity_call in
# .render
entity = self.cw_extra_kwargs.pop('entity')
if 'entity' in kwargs:
assert kwargs.pop('entity') is entity
self.entity_call(entity, **kwargs)
else:
super(EntityView, self).call(**kwargs)
def cell_call(self, row, col, **kwargs):
self.entity_call(self.cw_rset.get_entity(row, col), **kwargs)
def entity_call(self, entity, **kwargs):
raise NotImplementedError('%r %r' % (self.__regid__, self.__class__))
class StartupView(View):
"""base class for views which doesn't need a particular result set to be
displayed (so they can always be displayed !)
"""
__select__ = none_rset()
category = _('startupview')
def html_headers(self):
"""return a list of html headers (eg something to be inserted between
<head> and </head> of the returned page
by default startup views are indexed
"""
return []
class EntityStartupView(EntityView):
"""base class for entity views which may also be applied to None
result set (usually a default rql is provided by the view class)
"""
__select__ = none_rset() | non_final_entity()
default_rql = None
def __init__(self, req, rset=None, **kwargs):
super(EntityStartupView, self).__init__(req, rset=rset, **kwargs)
if rset is None:
# this instance is not in the "entityview" category
self.category = 'startupview'
def startup_rql(self):
"""return some rql to be executed if the result set is None"""
return self.default_rql
def no_entities(self, **kwargs):
"""override to display something when no entities were found"""
pass
def call(self, **kwargs):
"""override call to execute rql returned by the .startup_rql method if
necessary
"""
rset = self.cw_rset
if rset is None:
rset = self.cw_rset = self._cw.execute(self.startup_rql())
if rset:
for i in xrange(len(rset)):
self.wview(self.__regid__, rset, row=i, **kwargs)
else:
self.no_entities(**kwargs)
class AnyRsetView(View):
"""base class for views applying on any non empty result sets"""
__select__ = nonempty_rset()
category = _('anyrsetview')
def columns_labels(self, mainindex=0, tr=True):
"""compute the label of the rset colums
The logic is based on :meth:`~rql.stmts.Union.get_description`.
:param mainindex: The index of the main variable. This is an hint to get
more accurate label for various situation
:type mainindex: int
:param tr: Should the label be translated ?
:type tr: boolean
"""
if tr:
translate = partial(display_name, self._cw)
else:
translate = lambda val, *args,**kwargs: val
# XXX [0] because of missing Union support
rql_syntax_tree = self.cw_rset.syntax_tree()
rqlstdescr = rql_syntax_tree.get_description(mainindex, translate)[0]
labels = []
for colidx, label in enumerate(rqlstdescr):
labels.append(self.column_label(colidx, label, translate))
return labels
def column_label(self, colidx, default, translate_func=None):
"""return the label of a specified columns index
Overwrite me if you need to compute specific label.
:param colidx: The index of the column the call computes a label for.
:type colidx: int
:param default: Default value. If ``"Any"`` the default value will be
recomputed as coma separated list for all possible
etypes name.
:type colidx: string
:param translate_func: A function used to translate name.
:type colidx: function
"""
label = default
if label == 'Any':
etypes = self.cw_rset.column_types(colidx)
if translate_func is not None:
etypes = map(translate_func, etypes)
label = u','.join(etypes)
return label
# concrete template base classes ##############################################
class MainTemplate(View):
"""main template are primary access point to render a full HTML page.
There is usually at least a regular main template and a simple fallback
one to display error if the first one failed
"""
@property
def doctype(self):
if self._cw.xhtml_browser():
return STRICT_DOCTYPE
return STRICT_DOCTYPE_NOEXT
def set_stream(self, w=None):
if self.w is not None:
return
if w is None:
if self.binary:
self._stream = stream = StringIO()
else:
self._stream = stream = HTMLStream(self._cw)
w = stream.write
else:
stream = None
self.w = w
return stream
def write_doctype(self, xmldecl=True):
assert isinstance(self._stream, HTMLStream)
self._stream.doctype = self.doctype
if not xmldecl:
self._stream.xmldecl = u''
def linkable(self):
return False
# concrete component base classes #############################################
class ReloadableMixIn(object):
"""simple mixin for reloadable parts of UI"""
def user_callback(self, cb, args, msg=None, nonify=False):
"""register the given user callback and return an url to call it ready to be
inserted in html
"""
self._cw.add_js('cubicweb.ajax.js')
if nonify:
_cb = cb
def cb(*args):
_cb(*args)
cbname = self._cw.register_onetime_callback(cb, *args)
return self.build_js(cbname, xml_escape(msg or ''))
def build_update_js_call(self, cbname, msg):
rql = self.cw_rset.printable_rql()
return "javascript: %s" % js.userCallbackThenUpdateUI(
cbname, self.__regid__, rql, msg, self.__registry__, self.domid)
def build_reload_js_call(self, cbname, msg):
return "javascript: %s" % js.userCallbackThenReloadPage(cbname, msg)
build_js = build_update_js_call # expect updatable component by default
@property
def domid(self):
return domid(self.__regid__)
@deprecated('[3.10] use .domid property')
def div_id(self):
return self.domid
class Component(ReloadableMixIn, View):
"""base class for components"""
__registry__ = 'components'
__select__ = yes()
# XXX huummm, much probably useless (should be...)
htmlclass = 'mainRelated'
@property
def cssclass(self):
return '%s %s' % (self.htmlclass, domid(self.__regid__))
# XXX should rely on ReloadableMixIn.domid
@property
def domid(self):
return '%sComponent' % domid(self.__regid__)
@deprecated('[3.10] use .cssclass property')
def div_class(self):
return self.cssclass
class Adapter(AppObject):
"""base class for adapters"""
__registry__ = 'adapters'
def implements_adapter_compat(iface):
def _pre39_compat(func):
def decorated(self, *args, **kwargs):
entity = self.entity
if hasattr(entity, func.__name__):
warn('[3.9] %s method is deprecated, define it on a custom '
'%s for %s instead' % (func.__name__, iface,
entity.__class__),
DeprecationWarning)
member = getattr(entity, func.__name__)
if callable(member):
return member(*args, **kwargs)
return member
return func(self, *args, **kwargs)
decorated.decorated = func
return decorated
return _pre39_compat
def unwrap_adapter_compat(cls):
parent = cls.__bases__[0]
for member_name in dir(parent):
member = getattr(parent, member_name)
if isinstance(member, types.MethodType) and hasattr(member.im_func, 'decorated') and not member_name in cls.__dict__:
method = new.instancemethod(member.im_func.decorated, None, cls)
setattr(cls, member_name, method)
class auto_unwrap_bw_compat(type):
def __new__(mcs, name, bases, classdict):
cls = type.__new__(mcs, name, bases, classdict)
if not classdict.get('__needs_bw_compat__'):
unwrap_adapter_compat(cls)
return cls
class EntityAdapter(Adapter):
"""base class for entity adapters (eg adapt an entity to an interface)"""
__metaclass__ = auto_unwrap_bw_compat
def __init__(self, _cw, **kwargs):
try:
self.entity = kwargs.pop('entity')
except KeyError:
self.entity = kwargs['rset'].get_entity(kwargs.get('row') or 0,
kwargs.get('col') or 0)
Adapter.__init__(self, _cw, **kwargs)