def getTreeBuilder(treeType, implementation=None, **kwargs):
"""Get a TreeBuilder class for various types of tree with built-in support
treeType - the name of the tree type required (case-insensitive). Supported
values are "simpletree", "dom", "etree" and "beautifulsoup"
"simpletree" - a built-in DOM-ish tree type with support for some
more pythonic idioms.
"dom" - A generic builder for DOM implementations, defaulting to
a xml.dom.minidom based implementation for the sake of
backwards compatibility (as releases up until 0.10 had a
builder called "dom" that was a minidom implemenation).
"etree" - A generic builder for tree implementations exposing an
elementtree-like interface (known to work with
ElementTree, cElementTree and lxml.etree).
"beautifulsoup" - Beautiful soup (if installed)
implementation - (Currently applies to the "etree" and "dom" tree types). A
module implementing the tree type e.g.
xml.etree.ElementTree or lxml.etree."""
treeType = treeType.lower()
if treeType not in treeBuilderCache:
if treeType == "dom":
from . import dom
# XXX: Keep backwards compatibility by using minidom if no implementation is given
if implementation == None:
from xml.dom import minidom
implementation = minidom
# XXX: NEVER cache here, caching is done in the dom submodule
return dom.getDomModule(implementation, **kwargs).TreeBuilder
elif treeType == "simpletree":
from . import simpletree
treeBuilderCache[treeType] = simpletree.TreeBuilder
elif treeType == "beautifulsoup":
from . import soup
treeBuilderCache[treeType] = soup.TreeBuilder
elif treeType == "lxml":
from . import etree_lxml
treeBuilderCache[treeType] = etree_lxml.TreeBuilder
elif treeType == "etree":
if implementation is None:
try:
import xml.etree.cElementTree as etree
except:
import xml.etree.ElementTree as etree
implementation = etree
from . import etree
# XXX: NEVER cache here, caching is done in the etree submodule
return etree.getETreeModule(implementation, **kwargs).TreeBuilder
return treeBuilderCache.get(treeType)
def getTreeBuilder(treeType, implementation=None, **kwargs):
"""Get a TreeBuilder class for various types of tree with built-in support
treeType - the name of the tree type required (case-insensitive). Supported
values are "simpletree", "dom", "etree" and "beautifulsoup"
"simpletree" - a built-in DOM-ish tree type with support for some
more pythonic idioms.
"dom" - A generic builder for DOM implementations, defaulting to
a xml.dom.minidom based implementation for the sake of
backwards compatibility (as releases up until 0.10 had a
builder called "dom" that was a minidom implemenation).
"etree" - A generic builder for tree implementations exposing an
elementtree-like interface (known to work with
ElementTree, cElementTree and lxml.etree).
"beautifulsoup" - Beautiful soup (if installed)
implementation - (Currently applies to the "etree" and "dom" tree types). A
module implementing the tree type e.g.
xml.etree.ElementTree or lxml.etree."""
treeType = treeType.lower()
if treeType not in treeBuilderCache:
if treeType == "dom":
from . import dom
# XXX: Keep backwards compatibility by using minidom if no implementation is given
if implementation == None:
from xml.dom import minidom
implementation = minidom
# XXX: NEVER cache here, caching is done in the dom submodule
return dom.getDomModule(implementation, **kwargs).TreeBuilder
elif treeType == "simpletree":
from . import simpletree
treeBuilderCache[treeType] = simpletree.TreeBuilder
elif treeType == "beautifulsoup":
from . import soup
treeBuilderCache[treeType] = soup.TreeBuilder
elif treeType == "lxml":
from . import etree_lxml
treeBuilderCache[treeType] = etree_lxml.TreeBuilder
elif treeType == "etree":
if implementation is None:
try:
import xml.etree.cElementTree as etree
except:
import xml.etree.ElementTree as etree
implementation = etree
from . import etree
# XXX: NEVER cache here, caching is done in the etree submodule
return etree.getETreeModule(implementation, **kwargs).TreeBuilder
return treeBuilderCache.get(treeType)
def getTreeWalker(treeType, implementation=None, **kwargs):
"""Get a TreeWalker class for various types of tree with built-in support
treeType - the name of the tree type required (case-insensitive). Supported
values are "simpletree", "dom", "etree" and "beautifulsoup"
"simpletree" - a built-in DOM-ish tree type with support for some
more pythonic idioms.
"dom" - The xml.dom.minidom DOM implementation
"pulldom" - The xml.dom.pulldom event stream
"etree" - A generic walker for tree implementations exposing an
elementtree-like interface (known to work with
ElementTree, cElementTree and lxml.etree).
"lxml" - Optimized walker for lxml.etree
"beautifulsoup" - Beautiful soup (if installed)
"genshi" - a Genshi stream
implementation - (Currently applies to the "etree" tree type only). A module
implementing the tree type e.g. xml.etree.ElementTree or
cElementTree."""
treeType = treeType.lower()
if treeType not in treeWalkerCache:
if treeType in ("dom", "pulldom", "simpletree"):
mod = __import__(treeType, globals())
treeWalkerCache[treeType] = mod.TreeWalker
elif treeType == "genshi":
from . import genshistream
treeWalkerCache[treeType] = genshistream.TreeWalker
elif treeType == "beautifulsoup":
from . import soup
treeWalkerCache[treeType] = soup.TreeWalker
elif treeType == "lxml":
from . import lxmletree
treeWalkerCache[treeType] = lxmletree.TreeWalker
elif treeType == "etree":
from . import etree
if implementation is None:
try:
from xml.etree import cElementTree as etree
except:
from xml.etree import ElementTree as etree
implementation = etree
# XXX: NEVER cache here, caching is done in the etree submodule
return etree.getETreeModule(implementation, **kwargs).TreeWalker
return treeWalkerCache.get(treeType)
def getTreeWalker(treeType, implementation=None, **kwargs):
"""Get a TreeWalker class for various types of tree with built-in support
treeType - the name of the tree type required (case-insensitive). Supported
values are "simpletree", "dom", "etree" and "beautifulsoup"
"simpletree" - a built-in DOM-ish tree type with support for some
more pythonic idioms.
"dom" - The xml.dom.minidom DOM implementation
"pulldom" - The xml.dom.pulldom event stream
"etree" - A generic walker for tree implementations exposing an
elementtree-like interface (known to work with
ElementTree, cElementTree and lxml.etree).
"lxml" - Optimized walker for lxml.etree
"beautifulsoup" - Beautiful soup (if installed)
"genshi" - a Genshi stream
implementation - (Currently applies to the "etree" tree type only). A module
implementing the tree type e.g. xml.etree.ElementTree or
cElementTree."""
treeType = treeType.lower()
if treeType not in treeWalkerCache:
if treeType in ("dom", "pulldom", "simpletree"):
mod = __import__(treeType, globals())
treeWalkerCache[treeType] = mod.TreeWalker
elif treeType == "genshi":
from . import genshistream
treeWalkerCache[treeType] = genshistream.TreeWalker
elif treeType == "beautifulsoup":
from . import soup
treeWalkerCache[treeType] = soup.TreeWalker
elif treeType == "lxml":
from . import lxmletree
treeWalkerCache[treeType] = lxmletree.TreeWalker
elif treeType == "etree":
from . import etree
if implementation is None:
try:
from xml.etree import cElementTree as etree
except:
from xml.etree import ElementTree as etree
implementation = etree
# XXX: NEVER cache here, caching is done in the etree submodule
return etree.getETreeModule(implementation, **kwargs).TreeWalker
return treeWalkerCache.get(treeType)