Skip to content

C
cpython
Commit 48d358ba authored by Eli Bendersky's avatar Eli Bendersky
Browse files
Options

Issue #14007: implemented the 'element_factory' feature of TreeBuilder in 

_elementtree, with a test.
parent a0cf90e3
Hide whitespace changes
Inline Side-by-side
Showing with 90 additions and 37 deletions
Doc/library/xml.etree.elementtree.rst
View file @ 48d358ba
...@@ -716,9 +716,9 @@ TreeBuilder Objects ...@@ -716,9 +716,9 @@ TreeBuilder Objects
Generic element structure builder. This builder converts a sequence of Generic element structure builder. This builder converts a sequence of
start, data, and end method calls to a well-formed element structure. You start, data, and end method calls to a well-formed element structure. You
can use this class to build an element structure using a custom XML parser, can use this class to build an element structure using a custom XML parser,
or a parser for some other XML-like format. The *element_factory* is called or a parser for some other XML-like format. *element_factory*, when given,
to create new :class:`Element` instances when given. must be a callable accepting two positional arguments: a tag and
a dict of attributes. It is expected to return a new element instance.
.. method:: close() .. method:: close()
......
Lib/test/test_xml_etree.py
View file @ 48d358ba
...@@ -1959,6 +1959,8 @@ class TreeBuilderTest(unittest.TestCase): ...@@ -1959,6 +1959,8 @@ class TreeBuilderTest(unittest.TestCase):
' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">' ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">'
'<html>text</html>') '<html>text</html>')
sample2 = '''<toplevel>sometext</toplevel>'''
def test_dummy_builder(self): def test_dummy_builder(self):
class BaseDummyBuilder: class BaseDummyBuilder:
def close(self): def close(self):
...@@ -1993,11 +1995,19 @@ class TreeBuilderTest(unittest.TestCase): ...@@ -1993,11 +1995,19 @@ class TreeBuilderTest(unittest.TestCase):
e = parser.close() e = parser.close()
self.assertEqual(e.tag, 'html') self.assertEqual(e.tag, 'html')
# XXX in _elementtree, the constructor of TreeBuilder expects no
# arguments
@unittest.expectedFailure
def test_element_factory(self): def test_element_factory(self):
tb = ET.TreeBuilder(element_factory=lambda: ET.Element()) lst = []
def myfactory(tag, attrib):
nonlocal lst
lst.append(tag)
return ET.Element(tag, attrib)
tb = ET.TreeBuilder(element_factory=myfactory)
parser = ET.XMLParser(target=tb)
parser.feed(self.sample2)
parser.close()
self.assertEqual(lst, ['toplevel'])
@unittest.expectedFailure # XXX issue 14007 with C ElementTree @unittest.expectedFailure # XXX issue 14007 with C ElementTree
def test_doctype(self): def test_doctype(self):
......
Modules/_elementtree.c
View file @ 48d358ba
...@@ -191,6 +191,15 @@ list_join(PyObject* list) ...@@ -191,6 +191,15 @@ list_join(PyObject* list)
return result; return result;
} }
/* Is the given object an empty dictionary?
*/
static int
is_empty_dict(PyObject *obj)
{
return PyDict_CheckExact(obj) && PyDict_Size(obj) == 0;
}
/* -------------------------------------------------------------------- */ /* -------------------------------------------------------------------- */
/* the Element type */ /* the Element type */
...@@ -297,14 +306,9 @@ create_new_element(PyObject* tag, PyObject* attrib) ...@@ -297,14 +306,9 @@ create_new_element(PyObject* tag, PyObject* attrib)
self = PyObject_GC_New(ElementObject, &Element_Type); self = PyObject_GC_New(ElementObject, &Element_Type);
if (self == NULL) if (self == NULL)
return NULL; return NULL;
/* use None for empty dictionaries */
if (PyDict_CheckExact(attrib) && !PyDict_Size(attrib))
attrib = Py_None;
self->extra = NULL; self->extra = NULL;
if (attrib != Py_None) { if (attrib != Py_None && !is_empty_dict(attrib)) {
if (create_extra(self, attrib) < 0) { if (create_extra(self, attrib) < 0) {
PyObject_Del(self); PyObject_Del(self);
return NULL; return NULL;
...@@ -416,22 +420,14 @@ element_init(PyObject *self, PyObject *args, PyObject *kwds) ...@@ -416,22 +420,14 @@ element_init(PyObject *self, PyObject *args, PyObject *kwds)
self_elem = (ElementObject *)self; self_elem = (ElementObject *)self;
/* Use None for empty dictionaries */ if (attrib != Py_None && !is_empty_dict(attrib)) {
if (PyDict_CheckExact(attrib) && PyDict_Size(attrib) == 0) {
Py_INCREF(Py_None);
attrib = Py_None;
}
if (attrib != Py_None) {
if (create_extra(self_elem, attrib) < 0) { if (create_extra(self_elem, attrib) < 0) {
PyObject_Del(self_elem); PyObject_Del(self_elem);
return -1; return -1;
} }
} }
/* If create_extra needed attrib, it took a reference to it, so we can /* We own a reference to attrib here and it's no longer needed. */
* release ours anyway.
*/
Py_DECREF(attrib); Py_DECREF(attrib);
/* Replace the objects already pointed to by tag, text and tail. */ /* Replace the objects already pointed to by tag, text and tail. */
...@@ -1813,6 +1809,8 @@ typedef struct { ...@@ -1813,6 +1809,8 @@ typedef struct {
PyObject *stack; /* element stack */ PyObject *stack; /* element stack */
Py_ssize_t index; /* current stack size (0 means empty) */ Py_ssize_t index; /* current stack size (0 means empty) */
PyObject *element_factory;
/* element tracing */ /* element tracing */
PyObject *events; /* list of events, or NULL if not collecting */ PyObject *events; /* list of events, or NULL if not collecting */
PyObject *start_event_obj; /* event objects (NULL to ignore) */ PyObject *start_event_obj; /* event objects (NULL to ignore) */
...@@ -1841,6 +1839,7 @@ treebuilder_new(PyTypeObject *type, PyObject *args, PyObject *kwds) ...@@ -1841,6 +1839,7 @@ treebuilder_new(PyTypeObject *type, PyObject *args, PyObject *kwds)
t->last = (ElementObject *)Py_None; t->last = (ElementObject *)Py_None;
t->data = NULL; t->data = NULL;
t->element_factory = NULL;
t->stack = PyList_New(20); t->stack = PyList_New(20);
if (!t->stack) { if (!t->stack) {
Py_DECREF(t->this); Py_DECREF(t->this);
...@@ -1859,11 +1858,38 @@ treebuilder_new(PyTypeObject *type, PyObject *args, PyObject *kwds) ...@@ -1859,11 +1858,38 @@ treebuilder_new(PyTypeObject *type, PyObject *args, PyObject *kwds)
static int static int
treebuilder_init(PyObject *self, PyObject *args, PyObject *kwds) treebuilder_init(PyObject *self, PyObject *args, PyObject *kwds)
{ {
static char *kwlist[] = {"element_factory", NULL};
PyObject *element_factory = NULL;
TreeBuilderObject *self_tb = (TreeBuilderObject *)self;
if (!PyArg_ParseTupleAndKeywords(args, kwds, "|O:TreeBuilder", kwlist,
&element_factory)) {
return -1;
}
if (element_factory) {
Py_INCREF(element_factory);
Py_XDECREF(self_tb->element_factory);
self_tb->element_factory = element_factory;
}
return 0; return 0;
} }
static void static int
treebuilder_dealloc(TreeBuilderObject *self) treebuilder_gc_traverse(TreeBuilderObject *self, visitproc visit, void *arg)
{
Py_VISIT(self->root);
Py_VISIT(self->this);
Py_VISIT(self->last);
Py_VISIT(self->data);
Py_VISIT(self->stack);
Py_VISIT(self->element_factory);
return 0;
}
static int
treebuilder_gc_clear(TreeBuilderObject *self)
{ {
Py_XDECREF(self->end_ns_event_obj); Py_XDECREF(self->end_ns_event_obj);
Py_XDECREF(self->start_ns_event_obj); Py_XDECREF(self->start_ns_event_obj);
...@@ -1874,8 +1900,16 @@ treebuilder_dealloc(TreeBuilderObject *self) ...@@ -1874,8 +1900,16 @@ treebuilder_dealloc(TreeBuilderObject *self)
Py_XDECREF(self->data); Py_XDECREF(self->data);
Py_DECREF(self->last); Py_DECREF(self->last);
Py_DECREF(self->this); Py_DECREF(self->this);
Py_CLEAR(self->element_factory);
Py_XDECREF(self->root); Py_XDECREF(self->root);
return 0;
}
static void
treebuilder_dealloc(TreeBuilderObject *self)
{
PyObject_GC_UnTrack(self);
treebuilder_gc_clear(self);
Py_TYPE(self)->tp_free((PyObject *)self); Py_TYPE(self)->tp_free((PyObject *)self);
} }
...@@ -1904,9 +1938,14 @@ treebuilder_handle_start(TreeBuilderObject* self, PyObject* tag, ...@@ -1904,9 +1938,14 @@ treebuilder_handle_start(TreeBuilderObject* self, PyObject* tag,
self->data = NULL; self->data = NULL;
} }
node = create_new_element(tag, attrib); if (self->element_factory) {
if (!node) node = PyObject_CallFunction(self->element_factory, "OO", tag, attrib);
} else {
node = create_new_element(tag, attrib);
}
if (!node) {
return NULL; return NULL;
}
this = (PyObject*) self->this; this = (PyObject*) self->this;
...@@ -2180,10 +2219,11 @@ static PyTypeObject TreeBuilder_Type = { ...@@ -2180,10 +2219,11 @@ static PyTypeObject TreeBuilder_Type = {
0, /* tp_getattro */ 0, /* tp_getattro */
0, /* tp_setattro */ 0, /* tp_setattro */
0, /* tp_as_buffer */ 0, /* tp_as_buffer */
Py_TPFLAGS_DEFAULT | Py_TPFLAGS_BASETYPE, /* tp_flags */ Py_TPFLAGS_DEFAULT | Py_TPFLAGS_BASETYPE | Py_TPFLAGS_HAVE_GC,
/* tp_flags */
0, /* tp_doc */ 0, /* tp_doc */
0, /* tp_traverse */ (traverseproc)treebuilder_gc_traverse, /* tp_traverse */
0, /* tp_clear */ (inquiry)treebuilder_gc_clear, /* tp_clear */
0, /* tp_richcompare */ 0, /* tp_richcompare */
0, /* tp_weaklistoffset */ 0, /* tp_weaklistoffset */
0, /* tp_iter */ 0, /* tp_iter */
...@@ -2443,17 +2483,20 @@ expat_start_handler(XMLParserObject* self, const XML_Char* tag_in, ...@@ -2443,17 +2483,20 @@ expat_start_handler(XMLParserObject* self, const XML_Char* tag_in,
attrib = Py_None; attrib = Py_None;
} }
if (TreeBuilder_CheckExact(self->target)) /* If we get None, pass an empty dictionary on */
if (attrib == Py_None) {
Py_DECREF(attrib);
attrib = PyDict_New();
if (!attrib)
return;
}
if (TreeBuilder_CheckExact(self->target)) {
/* shortcut */ /* shortcut */
res = treebuilder_handle_start((TreeBuilderObject*) self->target, res = treebuilder_handle_start((TreeBuilderObject*) self->target,
tag, attrib); tag, attrib);
}
else if (self->handle_start) { else if (self->handle_start) {
if (attrib == Py_None) {
Py_DECREF(attrib);
attrib = PyDict_New();
if (!attrib)
return;
}
res = PyObject_CallFunction(self->handle_start, "OO", tag, attrib); res = PyObject_CallFunction(self->handle_start, "OO", tag, attrib);
} else } else
res = NULL; res = NULL;
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment
GitLab Nexedi Edition | About GitLab | About Nexedi | 沪ICP备2021021310号-2 | 沪ICP备2021021310号-7