__all__ = ['SimpleDoc']
class DocError(Exception):
pass
class SimpleDoc(object):
"""
class generating xml/html documents using context managers
doc, tag, text = SimpleDoc().tagtext()
with tag('html'):
with tag('body', id = 'hello'):
with tag('h1'):
text('Hello world!')
print(doc.getvalue())
"""
class Tag(object):
def __init__(self, doc, name, attrs): # name is the tag name (ex: 'div')
self.doc = doc
self.name = name
self.attrs = attrs
def __enter__(self):
self.parent_tag = self.doc.current_tag
self.doc.current_tag = self
self.position = len(self.doc.result)
self.doc._append('')
def __exit__(self, tpe, value, traceback):
if value is None:
if self.attrs:
self.doc.result[self.position] = "<%s %s>" % (
self.name,
dict_to_attrs(self.attrs),
)
else:
self.doc.result[self.position] = "<%s>" % self.name
self.doc._append("%s>" % self.name)
self.doc.current_tag = self.parent_tag
class DocumentRoot(object):
class DocumentRootError(DocError, AttributeError):
# Raising an AttributeError on __getattr__ instead of just a DocError makes it compatible
# with the pickle module (some users asked for pickling of SimpleDoc instances).
# I also keep the DocError from earlier versions to avoid possible compatibility issues
# with existing code.
pass
def __getattr__(self, item):
raise SimpleDoc.DocumentRoot.DocumentRootError("DocumentRoot here. You can't access anything here.")
def __init__(self, stag_end = ' />'):
self.result = []
self.current_tag = self.__class__.DocumentRoot()
self._append = self.result.append
assert stag_end in (' />', '/>', '>')
self._stag_end = stag_end
def tag(self, tag_name, *args, **kwargs):
"""
opens a HTML/XML tag for use inside a `with` statement
the tag is closed when leaving the `with` block
HTML/XML attributes can be supplied as keyword arguments,
or alternatively as (key, value) pairs.
The values of the keyword arguments should be strings.
They are escaped for use as HTML attributes
(the " character is replaced with ")
In order to supply a "class" html attributes, you must supply a `klass` keyword
argument. This is because `class` is a reserved python keyword so you can't use it
outside of a class definition.
Example::
with tag('h1', id = 'main-title'):
text("Hello world!")
#
Hello world!
was appended to the document
with tag('td',
('data-search', 'lemon'),
('data-order', '1384'),
id = '16'
):
text('Citrus Limon')
# you get: Citrus Limon |
"""
return self.__class__.Tag(self, tag_name, _attributes(args, kwargs))
def text(self, *strgs):
"""
appends 0 or more strings to the document
the strings are escaped for use as text in html documents, that is,
& becomes &
< becomes <
> becomes >
Example::
username = 'Max'
text('Hello ', username, '!') # appends "Hello Max!" to the current node
text('16 > 4') # appends "16 > 4" to the current node
"""
for strg in strgs:
self._append(html_escape(strg))
def line(self, tag_name, text_content, *args, **kwargs):
"""
Shortcut to write tag nodes that contain only text.
For example, in order to obtain::
The 7 secrets of catchy titles
you would write::
line('h1', 'The 7 secrets of catchy titles')
which is just a shortcut for::
with tag('h1'):
text('The 7 secrets of catchy titles')
The first argument is the tag name, the second argument
is the text content of the node.
The optional arguments after that are interpreted as xml/html
attributes. in the same way as with the `tag` method.
Example::
line('a', 'Who are we?', href = '/about-us.html')
produces::
Who are we?
"""
with self.tag(tag_name, *args, **kwargs):
self.text(text_content)
def asis(self, *strgs):
"""
appends 0 or more strings to the documents
contrary to the `text` method, the strings are appended "as is"
&, < and > are NOT escaped
Example::
doc.asis('') # appends to the document
"""
for strg in strgs:
if strg is None:
raise TypeError("Expected a string, got None instead.")
# passing None by mistake was frequent enough to justify a check
# see https://github.com/leforestier/yattag/issues/20
self._append(strg)
def nl(self):
self._append('\n')
def attr(self, *args, **kwargs):
"""
sets HTML/XML attribute(s) on the current tag
HTML/XML attributes are supplied as (key, value) pairs of strings,
or as keyword arguments.
The values of the keyword arguments should be strings.
They are escaped for use as HTML attributes
(the " character is replaced with ")
Note that, instead, you can set html/xml attributes by passing them as
keyword arguments to the `tag` method.
In order to supply a "class" html attributes, you can either pass
a ('class', 'my_value') pair, or supply a `klass` keyword argument
(this is because `class` is a reserved python keyword so you can't use it
outside of a class definition).
Examples::
with tag('h1'):
text('Welcome!')
doc.attr(id = 'welcome-message', klass = 'main-title')
# you get: Welcome!
with tag('td'):
text('Citrus Limon')
doc.attr(
('data-search', 'lemon'),
('data-order', '1384')
)
# you get: Citrus Limon |
"""
self.current_tag.attrs.update(_attributes(args, kwargs))
def data(self, *args, **kwargs):
"""
sets HTML/XML data attribute(s) on the current tag
HTML/XML data attributes are supplied as (key, value) pairs of strings,
or as keyword arguments.
The values of the keyword arguments should be strings.
They are escaped for use as HTML attributes
(the " character is replaced with ")
Note that, instead, you can set html/xml data attributes by passing them as
keyword arguments to the `tag` method.
Examples::
with tag('h1'):
text('Welcome!')
doc.data(msg='welcome-message')
# you get: Welcome!
with tag('td'):
text('Citrus Limon')
doc.data(
('search', 'lemon'),
('order', '1384')
)
# you get: Citrus Limon |
"""
self.attr(
*(('data-%s' % key, value) for (key, value) in args),
**dict(('data-%s' % key, value) for (key, value) in kwargs.items())
)
def stag(self, tag_name, *args, **kwargs):
"""
appends a self closing tag to the document
html/xml attributes can be supplied as keyword arguments,
or alternatively as (key, value) pairs.
The values of the keyword arguments should be strings.
They are escaped for use as HTML attributes
(the " character is replaced with ")
Example::
doc.stag('img', src = '/salmon-plays-piano.jpg')
# appends
to the document
If you want to produce self closing tags without the ending slash (HTML5 style),
use the stag_end parameter of the SimpleDoc constructor at the creation of the
SimpleDoc instance.
Example::
>>> doc = SimpleDoc(stag_end = '>')
>>> doc.stag('br')
>>> doc.getvalue()
'
'
"""
if args or kwargs:
self._append("<%s %s%s" % (
tag_name,
dict_to_attrs(_attributes(args, kwargs)),
self._stag_end
))
else:
self._append("<%s%s" % (tag_name, self._stag_end))
def cdata(self, strg, safe = False):
"""
appends a CDATA section containing the supplied string
You don't have to worry about potential ']]>' sequences that would terminate
the CDATA section. They are replaced with ']]]]>'.
If you're sure your string does not contain ']]>', you can pass `safe = True`.
If you do that, your string won't be searched for ']]>' sequences.
"""
self._append('', ']]]]>'))
self._append(']]>')
def getvalue(self):
"""
returns the whole document as a single string
"""
return ''.join(self.result)
def tagtext(self):
"""
return a triplet composed of::
. the document itself
. its tag method
. its text method
Example::
doc, tag, text = SimpleDoc().tagtext()
with tag('h1'):
text('Hello world!')
print(doc.getvalue()) # prints Hello world!
"""
return self, self.tag, self.text
def ttl(self):
"""
returns a quadruplet composed of::
. the document itself
. its tag method
. its text method
. its line method
Example::
doc, tag, text, line = SimpleDoc().ttl()
with tag('ul', id='grocery-list'):
line('li', 'Tomato sauce', klass="priority")
line('li', 'Salt')
line('li', 'Pepper')
print(doc.getvalue())
"""
return self, self.tag, self.text, self.line
def add_class(self, *classes):
"""
adds one or many elements to the html "class" attribute of the current tag
Example::
user_logged_in = False
with tag('a', href="/nuclear-device", klass = 'small'):
if not user_logged_in:
doc.add_class('restricted-area')
text("Our new product")
print(doc.getvalue())
# prints
"""
self._set_classes(
self._get_classes().union(classes)
)
def discard_class(self, *classes):
"""
remove one or many elements from the html "class" attribute of the current
tag if they are present (do nothing if they are absent)
"""
self._set_classes(
self._get_classes().difference(classes)
)
def toggle_class(self, elem, active):
"""
if active is a truthy value, ensure elem is present inside the html
"class" attribute of the current tag, otherwise (if active is falsy)
ensure elem is absent
"""
classes = self._get_classes()
if active:
classes.add(elem)
else:
classes.discard(elem)
self._set_classes(classes)
def _get_classes(self):
try:
current_classes = self.current_tag.attrs['class']
except KeyError:
return set()
else:
return set(current_classes.split())
def _set_classes(self, classes_set):
if classes_set:
self.current_tag.attrs['class'] = ' '.join(classes_set)
else:
try:
del self.current_tag.attrs['class']
except KeyError:
pass
def html_escape(s):
if isinstance(s,(int,float)):
return str(s)
try:
return s.replace("&", "&").replace("<", "<").replace(">", ">")
except AttributeError:
raise TypeError(
"You can only insert a string, an int or a float inside a xml/html text node. "
"Got %s (type %s) instead." % (repr(s), repr(type(s)))
)
def attr_escape(s):
if isinstance(s,(int,float)):
return str(s)
try:
return s.replace("&", "&").replace("<", "<").replace('"', """)
except AttributeError:
raise TypeError(
"xml/html attributes should be passed as strings, ints or floats. "
"Got %s (type %s) instead." % (repr(s), repr(type(s)))
)
ATTR_NO_VALUE = object()
def dict_to_attrs(dct):
return ' '.join(
(key if value is ATTR_NO_VALUE
else '%s="%s"' % (key, attr_escape(value)))
for key,value in dct.items()
)
def _attributes(args, kwargs):
lst = []
for arg in args:
if isinstance(arg, tuple):
lst.append(arg)
elif isinstance(arg, str):
lst.append((arg, ATTR_NO_VALUE))
else:
raise ValueError(
"Couldn't make a XML or HTML attribute/value pair out of %s."
% repr(arg)
)
result = dict(lst)
result.update(
(('class', value) if key == 'klass' else (key, value))
for key,value in kwargs.items()
)
return result