def test_parsestringurl():
# Base URLs should end up in the location info of the resulting XML tree
node = parse.tree(b"gurk",
parse.SGMLOP(),
parse.NS(),
parse.Node(),
validate=True)
assert str(node[0].startloc.url) == "STRING"
node = parse.tree(parse.String(b"gurk", url="root:gurk.xmlxsc"),
parse.SGMLOP(), parse.NS(), parse.Node())
assert str(node[0].startloc.url) == "root:gurk.xmlxsc"
def test_url(recwarn): # The ``recwarn`` argument silences the ``RequiredAttrMissingWarning`` node = parse.tree(b"<?url root:images/gurk.gif?>", parse.SGMLOP(), parse.NS(html), parse.Node()) assert node.bytes(base="root:about/us.html") == b"../images/gurk.gif" node = parse.tree(b'<img src="root:images/gurk.gif"/>', parse.Expat(), parse.NS(html), parse.Node()) assert node.bytes(base="root:about/us.html") == b'<img src="../images/gurk.gif" />'
def check(input, output):
node = parse.tree(input,
parse.SGMLOP(),
parse.NS(a),
parse.Node(),
validate=True)
node = node.walknodes(a)[0]
assert str(node.attrs.title) == output
def check(input, output):
node = parse.tree(f'<a title="{input}">{input}</a>'.encode("utf-8"),
parse.SGMLOP(),
parse.NS(a.xmlns),
parse.Node(pool=xsc.Pool(a, bar, foo, chars)),
validate=True)
node = node.walknodes(a)[0]
assert str(node) == output
assert str(node.attrs.title) == output
def test_parselocationsgmlop():
# sgmlop doesn't provide any location info, so check only the URL
node = parse.tree(b"<z>gurk&hurz*hinz٦hunz</z>",
parse.SGMLOP(),
parse.NS(doc),
parse.Node(),
validate=True)
assert len(node) == 1
assert len(node[0]) == 1
assert str(node[0][0].startloc.url) == "STRING"
assert node[0][0].startloc.line is None
assert node[0][0].startloc.col is None
def test_sgmlop_no_multiple_text_events():
# Test that we don't get consecutive text events with sgmlop
i = parse.events(
parse.Iter(b"<a>gurk & hurz & hinz & kunz</a>"),
parse.SGMLOP())
assert next(i) == ("url", url.URL("ITER"))
assert next(i) == ("enterstarttag", "a")
assert next(i) == ("leavestarttag", "a")
assert next(i) == ("text", "gurk & hurz & hinz & kunz")
assert next(i) == ("endtag", "a")
with pytest.raises(StopIteration):
next(i)
def xsc2txt(instream, outstream, title, width):
e = parse.tree(parse.Stream(instream), parse.SGMLOP(), parse.NS(doc),
parse.Node(pool=xsc.docpool()))
if title is None:
title = xsc.Null
else:
title = doc.title(title)
e = html.html(html.body(doc.section(title, e)))
e = e.conv()
outstream.write(html.astext(e, width=width))
def test_multipleparsecalls():
def check(parser):
for i in range(3):
try:
parse.tree(b"<>gurk",
parser,
parse.NS(html),
parse.Node(),
validate=True)
except Exception:
pass
for j in range(3):
assert parse.tree(b"<a>gurk</a>", parser, parse.NS(html),
parse.Node()).string() == "<a>gurk</a>"
# A Parser instance should be able to parse multiple XML sources, even when some of the parse calls fail
check(parse.SGMLOP())
check(parse.Expat())
def getdoc(thing, format):
if thing.__doc__ is None:
return xsc.Null
# Remove indentation
lines = textwrap.dedent(thing.__doc__).split("\n")
# remove empty lines
while lines and not lines[0]:
del lines[0]
while lines and not lines[-1]:
del lines[-1]
text = "\n".join(lines)
modulename = _getmodulename(thing)
if inspect.ismethod(thing):
base = f"METHOD-DOCSTRING({modulename}.{thing.__class__.__name__}.{thing.__qualname__})"
elif isinstance(thing, property):
base = f"PROPERTY-DOCSTRING({modulename}.{thing})"
elif inspect.isfunction(thing):
base = f"FUNCTION-DOCSTRING({modulename}.{thing.__qualname__})"
elif inspect.isclass(thing):
base = f"CLASS-DOCSTRING({modulename}.{thing.__qualname__})"
elif inspect.ismodule(thing):
base = f"MODULE-DOCSTRING({modulename})"
else:
base = "DOCSTRING"
lformat = format.lower()
if lformat == "plaintext":
return xsc.Text(text)
elif lformat == "restructuredtext":
from ll.xist.ns import rest, doc
return rest.fromstring(text, base=base).conv(target=doc)
elif lformat == "xist":
from ll.xist.ns import doc
node = parse.tree(parse.String(text), parse.SGMLOP(), parse.NS(doc),
parse.Node(pool=xsc.docpool(), base=base))
if not node[
p]: # optimization: one paragraph docstrings don't need a <p> element.
node = p(node)
if inspect.ismethod(thing):
# Use the original method instead of the decorator
realthing = thing
while hasattr(realthing, "__wrapped__"):
realthing = realthing.__wrapped__
for ref in node.walknodes(pyref):
if "module" not in ref.attrs:
ref["module"] = _getmodulename(realthing)
if "class_" not in ref.attrs:
ref["class_"] = thing.__self__.__class__.__name__
if "method" not in ref.attrs:
ref["method"] = thing.__name__
elif inspect.isfunction(thing):
# Use the original method instead of the decorator
while hasattr(thing, "__wrapped__"):
thing = thing.__wrapped__
for ref in node.walknodes(pyref):
if "module" not in ref.attrs:
ref["module"] = _getmodulename(thing)
elif inspect.isclass(thing):
for ref in node.walknodes(pyref):
if "module" not in ref.attrs:
ref["module"] = _getmodulename(thing)
if "class_" not in ref.attrs:
ref["class_"] = thing.__name__
elif inspect.ismodule(thing):
for ref in node.walknodes(pyref):
if "module" not in ref.attrs:
ref["module"] = thing.__name__
return node
else:
raise ValueError(f"unsupported __docformat__ {format!r}")
def test_frag():
e = parse.tree(b"das ist <b>klaus</b>. das ist <b>erich</b>",
parse.SGMLOP(), parse.NS(html), parse.Node())
assert "".join(map(str, e.walknodes(e // html.b))) == "klauserich"