def _convert_command(element, elements, index, iterable, parent):
_get_prefix_element(element, parent)
params, tag, attributes = COMMANDS[element]
new_parent = eTree.SubElement(parent, tag, **attributes)
alignment = None
if element in MATRICES and (element.endswith('*') or element == r'\array'):
index += 1
alignment = elements[index]
next(iterable)
for j in range(params):
index += 1
param = elements[index]
if element == r'\left' or element == r'\right':
symbol = convert_symbol(param)
new_parent.text = param if symbol is None else '&#x{};'.format(
symbol)
elif element == r'\array':
_convert_array_content(param, new_parent, alignment)
elif element in MATRICES:
_convert_matrix_content(param, new_parent, alignment)
else:
if isinstance(param, list):
_parent = eTree.SubElement(new_parent, 'mrow')
_classify_subgroup(param, _parent)
else:
_classify(param, new_parent)
_get_postfix_element(element, parent)
if element == r'\overline':
mo = eTree.SubElement(new_parent, 'mo', stretchy='true')
mo.text = '¯'
elif element == r'\underline':
mo = eTree.SubElement(new_parent, 'mo', stretchy='true')
mo.text = '̲'
[next(iterable) for _ in range(params)]
def _classify(_element, parent):
symbol = convert_symbol(_element)
if re.match(r'\d+(.\d+)?', _element):
mn = eTree.SubElement(parent, 'mn')
mn.text = _element
elif _element in '+-*/()=':
mo = eTree.SubElement(parent, 'mo')
mo.text = _element if symbol is None else '&#x{};'.format(symbol)
elif symbol and (int(symbol, 16) in range(int('2200', 16),
int('22FF', 16) + 1)
or int(symbol, 16) in range(int('2190', 16),
int('21FF', 16) + 1)):
mo = eTree.SubElement(parent, 'mo')
mo.text = '&#x{};'.format(symbol)
elif _element.startswith('\\'):
if symbol:
mi = eTree.SubElement(parent, 'mi')
mi.text = '&#x{};'.format(symbol)
else:
e = _element.lstrip('\\')
mi = eTree.SubElement(parent, 'mi')
mi.text = e
else:
mi = eTree.SubElement(parent, 'mi')
mi.text = _element
def _convert_command(
element: str,
elements: List[Any],
index: int,
iterable: Iterator[int],
parent: Element,
):
_get_prefix_element(element, parent)
if element == r"\substack":
parent = SubElement(parent, "mstyle", scriptlevel="1")
params, tag, attributes = COMMANDS[element]
if len(elements) - 1 < params:
mo = SubElement(parent, "mo")
mo.text = element[1:]
return
new_parent = SubElement(parent, tag, attributes)
alignment = ""
if element in MATRICES and (element.endswith("*") or element == r"\array"):
index += 1
alignment = elements[index]
next(iterable)
if element in (r"\lim", r"\inf", r"\sup", r"\max", r"\min"):
limit = SubElement(new_parent, "mo")
limit.text = element[1:]
for j in range(params):
index += 1
param = elements[index]
if element == "_" and index == 1 and param == r"\sum":
new_parent.tag = "munder"
_classify(param, new_parent)
elif element == r"\left" or element == r"\right":
if param == ".":
pass
else:
symbol = convert_symbol(param)
new_parent.text = param if symbol is None else "&#x{};".format(
symbol)
elif element == r"\array":
_convert_array_content(param, new_parent, alignment)
elif element in MATRICES:
_convert_matrix_content(param, new_parent, alignment,
element == r"\substack")
else:
if isinstance(param, list):
_parent = SubElement(new_parent, "mrow")
_classify_subgroup(param, _parent)
else:
_classify(param, new_parent)
_get_postfix_element(element, parent)
if element in (r"\overline", r"\bar"):
mo = SubElement(new_parent, "mo", stretchy="true")
mo.text = "¯"
elif element == r"\underline":
mo = SubElement(new_parent, "mo", stretchy="true")
mo.text = "̲"
elif element in (r"\overrightarrow", r"\vec"):
mo = SubElement(new_parent, "mo", stretchy="true")
mo.text = "→"
[next(iterable) for _ in range(params)]
def _classify(_element: str,
parent: Element,
is_math_mode: bool = False) -> None:
symbol = convert_symbol(_element)
if re.match(r"\d+(.\d+)?", _element):
mn = SubElement(parent, "mn")
mn.text = _element
elif len(_element) and _element in "<>&":
mo = SubElement(parent, "mo")
mo.text = {"<": "<", ">": ">", "&": "&"}[_element]
elif len(_element) and _element in "+-*/()=":
mo = SubElement(parent, "mo")
mo.text = _element if symbol is None else "&#x{};".format(symbol)
if _element in "()":
mo.attrib["stretchy"] = "false"
elif (symbol and (int(symbol, 16) in range(int("2200", 16),
int("22FF", 16) + 1)
or int(symbol, 16) in range(int("2190", 16),
int("21FF", 16) + 1))
or symbol == "."):
mo = SubElement(parent, "mo")
mo.text = "&#x{};".format(symbol)
elif _element.startswith("\\"):
tag = SubElement(parent, "mo" if is_math_mode else "mi")
if symbol:
tag.text = "&#x{};".format(symbol)
elif _element in (
r"\log",
r"\ln",
r"\tan",
r"\sec",
r"\cos",
r"\sin",
r"\cot",
r"\csc",
):
tag.text = _element[1:]
else:
tag.text = _element
else:
tag = SubElement(parent, "mo" if is_math_mode else "mi")
tag.text = _element
def tokenize(latex_string: str, skip_comments: bool = True) -> Iterator[str]:
"""
Converts Latex string into tokens.
:param latex_string: Latex string.
:param skip_comments: Flag to skip comments (default=True).
"""
for match in PATTERN.finditer(latex_string):
tokens = tuple(filter(lambda x: x is not None, match.groups()))
if tokens[0].startswith(commands.MATH):
full_math = "".join(tokens)
symbol = convert_symbol(full_math)
if symbol:
yield f"&#x{symbol};"
continue
for captured in tokens:
if skip_comments and captured.startswith("%"):
break
if captured.endswith(UNITS):
yield captured.replace(" ", "")
continue
yield captured
def test_convert_symbol(latex: str, expected: str) -> None:
assert convert_symbol(latex) == expected
def tokenize(data) -> Iterator[str]:
iterable = iter(data)
buffer = ""
while True:
try:
char = next(iterable)
if char == "\\":
if buffer == "\\":
yield buffer + char
buffer = ""
continue
elif len(buffer):
yield buffer
buffer = char
try:
buffer += next(iterable)
if buffer in (r"\\", r"\[", r"\]"):
yield buffer
buffer = ""
except StopIteration:
break
elif char.isalpha():
if len(buffer):
if buffer.endswith("}"):
yield buffer
yield char
buffer = ""
elif buffer.startswith("\\"):
buffer += char
else:
yield char
elif char.isdigit():
if len(buffer):
yield buffer
buffer = char
while True:
try:
char = next(iterable)
except StopIteration:
break
if char.isspace():
yield buffer
buffer = ""
break
elif char.isdigit() or char == ".":
buffer += char
else:
if buffer.endswith("."):
yield buffer[:-1]
yield buffer[-1]
else:
yield buffer
buffer = ""
if char == "\\":
buffer = char
else:
yield char
break
elif char.isspace():
if len(buffer):
yield buffer
buffer = ""
elif char in "{}*":
# FIXME: Anything that starts with '\math' passes. There is a huge list of math symbols in
# unimathsymbols.txt and hard-coding all of them is inefficient.
if (buffer.startswith(r"\begin") or buffer.startswith(r"\end")
or buffer.startswith(r"\math")):
if buffer.endswith("}"):
yield buffer
yield char
buffer = ""
continue
elif buffer.startswith(r"\math") and char == "}":
symbol = convert_symbol(buffer + char)
if symbol:
yield "&#x{};".format(symbol)
buffer = ""
continue
buffer += char
else:
if len(buffer):
yield buffer
buffer = ""
yield char
else:
if len(buffer):
if buffer.startswith(r"\math"):
yield buffer[:-1]
yield buffer[-1]
else:
yield buffer
buffer = ""
if len(char):
yield char
except StopIteration:
break
if len(buffer):
yield buffer
def test_convert_symbol(name: str, latex: str, expected: str):
assert convert_symbol(latex) == expected
def _convert_and_append_operator(symbol, parent):
symbol = convert_symbol(symbol)
mo = eTree.SubElement(parent, 'mo')
mo.text = '&#x{};'.format(symbol)
def _convert_command(node: Node, parent: Element, font: Optional[Dict[str, Optional[str]]] = None) -> None:
command = node.token
modifier = node.modifier
if command in (commands.SUBSTACK, commands.SMALLMATRIX):
parent = SubElement(parent, "mstyle", scriptlevel="1")
elif command == commands.CASES:
lbrace = SubElement(parent, "mo", OrderedDict([("stretchy", "true"), ("fence", "true"), ("form", "prefix")]))
lbrace.text = "&#x{};".format(convert_symbol(commands.LBRACE))
elif command in (commands.DBINOM, commands.DFRAC):
parent = SubElement(parent, "mstyle", displaystyle="true", scriptlevel="0")
elif command == commands.HPHANTOM:
parent = SubElement(parent, "mpadded", height="0", depth="0")
elif command == commands.VPHANTOM:
parent = SubElement(parent, "mpadded", width="0")
elif command in (commands.TBINOM, commands.HBOX, commands.MBOX, commands.TFRAC):
parent = SubElement(parent, "mstyle", displaystyle="false", scriptlevel="0")
elif command in (commands.MOD, commands.PMOD):
SubElement(parent, "mspace", width="1em")
tag, attributes = copy.deepcopy(commands.CONVERSION_MAP[command])
if node.attributes is not None and node.token != commands.SKEW:
attributes.update(node.attributes)
if command == commands.LEFT:
parent = SubElement(parent, "mrow")
_append_prefix_element(node, parent)
alignment, column_lines = _get_alignment_and_column_lines(node.alignment)
if column_lines:
attributes["columnlines"] = column_lines
if command == commands.SUBSUP and node.children is not None and node.children[0].token == commands.GCD:
tag = "munderover"
elif command == commands.SUPERSCRIPT and modifier in (commands.LIMITS, commands.OVERBRACE):
tag = "mover"
elif command == commands.SUBSCRIPT and modifier in (commands.LIMITS, commands.UNDERBRACE):
tag = "munder"
elif command == commands.SUBSUP and modifier in (commands.LIMITS, commands.OVERBRACE, commands.UNDERBRACE):
tag = "munderover"
elif (
command in (commands.XLEFTARROW, commands.XRIGHTARROW) and node.children is not None and len(node.children) == 2
):
tag = "munderover"
element = SubElement(parent, tag, attributes)
if command in commands.LIMIT:
element.text = command[1:]
elif command in (commands.MOD, commands.PMOD):
element.text = "mod"
SubElement(parent, "mspace", width="0.333em")
elif command == commands.BMOD:
element.text = "mod"
elif command in (commands.XLEFTARROW, commands.XRIGHTARROW):
style = SubElement(element, "mstyle", scriptlevel="0")
arrow = SubElement(style, "mo")
if command == commands.XLEFTARROW:
arrow.text = "←"
elif command == commands.XRIGHTARROW:
arrow.text = "→"
elif node.text is not None:
if command == commands.MIDDLE:
element.text = "&#x{};".format(convert_symbol(node.text))
elif command == commands.HBOX:
mtext: Optional[Element] = element
for text, mode in separate_by_mode(node.text):
if mode == Mode.TEXT:
if mtext is None:
mtext = SubElement(parent, tag, attributes)
mtext.text = text.replace(" ", " ")
_set_font(mtext, "mtext", font)
mtext = None
else:
_row = SubElement(parent, "mrow")
_convert_group(iter(walk(text)), _row)
else:
if command == commands.FBOX:
element = SubElement(element, "mtext")
element.text = node.text.replace(" ", " ")
_set_font(element, "mtext", font)
elif node.delimiter is not None and command not in (commands.FRAC, commands.GENFRAC):
if node.delimiter != ".":
symbol = convert_symbol(node.delimiter)
element.text = node.delimiter if symbol is None else "&#x{};".format(symbol)
if node.children is not None:
_parent = element
if command in (commands.LEFT, commands.MOD, commands.PMOD):
_parent = parent
if command in commands.MATRICES:
if command == commands.CASES:
alignment = "l"
elif command in (commands.SPLIT, commands.ALIGN):
alignment = "rl"
_convert_matrix(iter(node.children), _parent, command, alignment=alignment)
elif command == commands.CFRAC:
for child in node.children:
p = SubElement(_parent, "mstyle", displaystyle="false", scriptlevel="0")
_convert_group(iter([child]), p, font)
elif command == commands.SIDESET:
Node(
r"\style",
children=(Node(r"\mspace", attributes={"width": "-0.167em"}),),
attributes={"scriptlevel": "0"},
),
left, right = node.children
_convert_group(iter([left]), _parent, font)
fill = SubElement(_parent, "mstyle", scriptlevel="0")
SubElement(fill, "mspace", width="-0.167em")
_convert_group(iter([right]), _parent, font)
elif command == commands.SKEW:
child = node.children[0]
new_node = Node(
token=child.token,
children=(
Node(
token=commands.BRACES,
children=(*child.children, Node(token=commands.MKERN, attributes=node.attributes)),
),
),
)
_convert_group(iter([new_node]), _parent, font)
elif command in (commands.XLEFTARROW, commands.XRIGHTARROW):
for child in node.children:
padded = SubElement(
_parent,
"mpadded",
OrderedDict(
[("width", "+0.833em"), ("lspace", "0.556em"), ("voffset", "-.2em"), ("height", "-.2em")]
),
)
_convert_group(iter([child]), padded, font)
SubElement(padded, "mspace", depth=".25em")
else:
_convert_group(iter(node.children), _parent, font)
_add_diacritic(command, element)
_append_postfix_element(node, parent)
def _convert_and_append_operator(symbol: str, parent: Element) -> None:
converted = convert_symbol(symbol)
mo = SubElement(parent, "mo")
mo.text = "&#x{};".format(converted)
def test_operator_plus(self):
self.assertEqual('0002B', symbols_parser.convert_symbol('+'))
def test_alias_command(self):
self.assertEqual('02192', symbols_parser.convert_symbol(r'\to'))
def _convert_symbol(node: Node, parent: Element, font: Optional[Dict[str, Optional[str]]] = None) -> None:
token = node.token
attributes = node.attributes or {}
symbol = convert_symbol(token)
if re.match(r"\d+(.\d+)?", token):
element = SubElement(parent, "mn", attrib=attributes)
element.text = token
_set_font(element, element.tag, font)
elif token in OPERATORS:
element = SubElement(parent, "mo", attrib=attributes)
element.text = token if symbol is None else "&#x{};".format(symbol)
if token == r"\|":
element.attrib["fence"] = "false"
if token == r"\smallint":
element.attrib["largeop"] = "false"
if token in ("(", ")", "[", "]", "|", r"\|", r"\{", r"\}", r"\surd"):
element.attrib["stretchy"] = "false"
_set_font(element, "fence", font)
else:
_set_font(element, element.tag, font)
elif (
symbol
and (
int(symbol, 16) in range(int("2200", 16), int("22FF", 16) + 1)
or int(symbol, 16) in range(int("2190", 16), int("21FF", 16) + 1)
)
or symbol == "."
):
element = SubElement(parent, "mo", attrib=attributes)
element.text = "&#x{};".format(symbol)
_set_font(element, element.tag, font)
elif token in (r"\ ", "~", commands.NOBREAKSPACE, commands.SPACE):
element = SubElement(parent, "mtext", attrib=attributes)
element.text = " "
_set_font(element, "mtext", font)
elif token == commands.NOT:
mpadded = SubElement(parent, "mpadded", width="0")
element = SubElement(mpadded, "mtext")
element.text = "⧸"
elif token in (
commands.DETERMINANT,
commands.GCD,
commands.INTOP,
commands.INJLIM,
commands.LIMINF,
commands.LIMSUP,
commands.PR,
commands.PROJLIM,
):
element = SubElement(parent, "mo", attrib={"movablelimits": "true", **attributes})
texts = {
commands.INJLIM: "inj lim",
commands.INTOP: "∫",
commands.LIMINF: "lim inf",
commands.LIMSUP: "lim sup",
commands.PROJLIM: "proj lim",
}
element.text = texts.get(token, token[1:])
_set_font(element, element.tag, font)
elif token == commands.IDOTSINT:
_parent = SubElement(parent, "mrow", attrib=attributes)
for s in ("∫", "⋯", "∫"):
element = SubElement(_parent, "mo")
element.text = s
elif token in (commands.LATEX, commands.TEX):
_parent = SubElement(parent, "mrow", attrib=attributes)
if token == commands.LATEX:
mi_l = SubElement(_parent, "mi")
mi_l.text = "L"
SubElement(_parent, "mspace", width="-.325em")
mpadded = SubElement(_parent, "mpadded", height="+.21ex", depth="-.21ex", voffset="+.21ex")
mstyle = SubElement(mpadded, "mstyle", displaystyle="false", scriptlevel="1")
mrow = SubElement(mstyle, "mrow")
mi_a = SubElement(mrow, "mi")
mi_a.text = "A"
SubElement(_parent, "mspace", width="-.17em")
_set_font(mi_l, mi_l.tag, font)
_set_font(mi_a, mi_a.tag, font)
mi_t = SubElement(_parent, "mi")
mi_t.text = "T"
SubElement(_parent, "mspace", width="-.14em")
mpadded = SubElement(_parent, "mpadded", height="-.5ex", depth="+.5ex", voffset="-.5ex")
mrow = SubElement(mpadded, "mrow")
mi_e = SubElement(mrow, "mi")
mi_e.text = "E"
SubElement(_parent, "mspace", width="-.115em")
mi_x = SubElement(_parent, "mi")
mi_x.text = "X"
_set_font(mi_t, mi_t.tag, font)
_set_font(mi_e, mi_e.tag, font)
_set_font(mi_x, mi_x.tag, font)
elif token.startswith(commands.OPERATORNAME):
element = SubElement(parent, "mo", attrib=attributes)
element.text = token[14:-1]
elif token.startswith(commands.BACKSLASH):
element = SubElement(parent, "mi", attrib=attributes)
if symbol:
element.text = "&#x{};".format(symbol)
elif token in commands.FUNCTIONS:
element.text = token[1:]
else:
element.text = token
_set_font(element, element.tag, font)
else:
element = SubElement(parent, "mi", attrib=attributes)
element.text = token
_set_font(element, element.tag, font)
def _convert_and_append_command(command: str, parent: Element, attributes: Optional[Dict[str, str]] = None) -> None:
code_point = convert_symbol(command)
mo = SubElement(parent, "mo", attributes if attributes is not None else {})
mo.text = "&#x{};".format(code_point) if code_point else command
def test_operator_plus():
assert '0002B' == convert_symbol('+')
def test_alias_command():
assert '02192' == convert_symbol(r'\to')
def _walk(tokens: Iterator[str], terminator: str = None, limit: int = 0) -> List[Node]:
group: List[Node] = []
token: str
has_available_tokens = False
for token in tokens:
has_available_tokens = True
if token == terminator:
delimiter = None
if terminator == commands.RIGHT:
delimiter = next(tokens)
group.append(Node(token=token, delimiter=delimiter))
break
elif (token == commands.RIGHT != terminator) or (token == commands.MIDDLE and terminator != commands.RIGHT):
raise ExtraLeftOrMissingRightError
elif token == commands.LEFT:
delimiter = next(tokens)
children = tuple(_walk(tokens, terminator=commands.RIGHT)) # make \right as a child of \left
if len(children) == 0 or children[-1].token != commands.RIGHT:
raise ExtraLeftOrMissingRightError
node = Node(token=token, children=children if len(children) else None, delimiter=delimiter)
elif token == commands.OPENING_BRACE:
children = tuple(_walk(tokens, terminator=commands.CLOSING_BRACE))
if len(children) and children[-1].token == commands.CLOSING_BRACE:
children = children[:-1]
node = Node(token=commands.BRACES, children=children)
elif token in (commands.SUBSCRIPT, commands.SUPERSCRIPT):
try:
previous = group.pop()
except IndexError:
previous = Node(token="") # left operand can be empty if not present
if token == previous.token == commands.SUBSCRIPT:
raise DoubleSubscriptsError
if token == previous.token == commands.SUPERSCRIPT:
raise DoubleSuperscriptsError
modifier = None
if previous.token == commands.LIMITS:
modifier = commands.LIMITS
try:
previous = group.pop()
if not previous.token.startswith("\\"): # TODO: Complete list of operators
raise LimitsMustFollowMathOperatorError
except IndexError:
raise LimitsMustFollowMathOperatorError
if token == commands.SUBSCRIPT and previous.token == commands.SUPERSCRIPT and previous.children is not None:
children = tuple(_walk(tokens, terminator=terminator, limit=1))
node = Node(
token=commands.SUBSUP,
children=(previous.children[0], *children, previous.children[1]),
modifier=previous.modifier,
)
elif (
token == commands.SUPERSCRIPT and previous.token == commands.SUBSCRIPT and previous.children is not None
):
children = tuple(_walk(tokens, terminator=terminator, limit=1))
node = Node(token=commands.SUBSUP, children=(*previous.children, *children), modifier=previous.modifier)
else:
try:
children = tuple(_walk(tokens, terminator=terminator, limit=1))
except NoAvailableTokensError:
raise MissingSuperScriptOrSubscriptError
if previous.token in (commands.OVERBRACE, commands.UNDERBRACE):
modifier = previous.token
node = Node(token=token, children=(previous, *children), modifier=modifier)
elif token == commands.APOSTROPHE:
try:
previous = group.pop()
except IndexError:
previous = Node(token="") # left operand can be empty if not present
if (
previous.token == commands.SUPERSCRIPT
and previous.children is not None
and len(previous.children) >= 2
and previous.children[1].token == commands.PRIME
):
node = Node(token=commands.SUPERSCRIPT, children=(previous.children[0], Node(token=commands.DPRIME)))
else:
node = Node(token=commands.SUPERSCRIPT, children=(previous, Node(token=commands.PRIME)))
elif token in commands.COMMANDS_WITH_TWO_PARAMETERS:
attributes = None
children = tuple(_walk(tokens, terminator=terminator, limit=2))
if token in (commands.OVERSET, commands.UNDERSET):
children = children[::-1]
node = Node(token=token, children=children, attributes=attributes)
elif token in commands.COMMANDS_WITH_ONE_PARAMETER or token.startswith(commands.MATH):
children = tuple(_walk(tokens, terminator=terminator, limit=1))
node = Node(token=token, children=children)
elif token == commands.NOT:
try:
next_node = tuple(_walk(tokens, terminator=terminator, limit=1))[0]
if next_node.token.startswith("\\"):
negated_symbol = r"\n" + next_node.token[1:]
symbol = convert_symbol(negated_symbol)
if symbol:
node = Node(token=negated_symbol)
group.append(node)
continue
node = Node(token=token)
group.extend((node, next_node))
continue
except NoAvailableTokensError:
node = Node(token=token)
elif token in (commands.XLEFTARROW, commands.XRIGHTARROW):
children = tuple(_walk(tokens, terminator=terminator, limit=1))
if children[0].token == commands.OPENING_BRACKET:
children = (
Node(
token=commands.BRACES, children=tuple(_walk(tokens, terminator=commands.CLOSING_BRACKET))[:-1]
),
*tuple(_walk(tokens, terminator=terminator, limit=1)),
)
node = Node(token=token, children=children)
elif token in (commands.HSKIP, commands.HSPACE, commands.KERN, commands.MKERN, commands.MSKIP, commands.MSPACE):
children = tuple(_walk(tokens, terminator=terminator, limit=1))
if children[0].token == commands.BRACES and children[0].children is not None:
children = children[0].children
node = Node(token=token, attributes={"width": children[0].token})
elif token == commands.COLOR:
attributes = {"mathcolor": next(tokens)}
children = tuple(_walk(tokens, terminator=terminator))
sibling = None
if len(children) and children[-1].token == terminator:
children, sibling = children[:-1], children[-1]
group.append(Node(token=token, children=children, attributes=attributes))
if sibling:
group.append(sibling)
break
elif token == commands.STYLE:
attributes = {"style": next(tokens)}
next_node = tuple(_walk(tokens, terminator=terminator, limit=1))[0]
node = next_node._replace(attributes=attributes)
elif token in (
*commands.BIG.keys(),
*commands.BIG_OPEN_CLOSE.keys(),
commands.FBOX,
commands.HBOX,
commands.MBOX,
commands.MIDDLE,
commands.TEXT,
commands.TEXTBF,
commands.TEXTIT,
commands.TEXTRM,
commands.TEXTSF,
commands.TEXTTT,
):
node = Node(token=token, text=next(tokens))
elif token == commands.HREF:
attributes = {"href": next(tokens)}
children = tuple(_walk(tokens, terminator=terminator, limit=1))
node = Node(token=token, children=children, attributes=attributes)
elif token in (
commands.ABOVE,
commands.ATOP,
commands.ABOVEWITHDELIMS,
commands.ATOPWITHDELIMS,
commands.BRACE,
commands.BRACK,
commands.CHOOSE,
commands.OVER,
):
attributes = None
delimiter = None
if token == commands.ABOVEWITHDELIMS:
delimiter = next(tokens).lstrip("\\") + next(tokens).lstrip("\\")
elif token == commands.ATOPWITHDELIMS:
attributes = {"linethickness": "0"}
delimiter = next(tokens).lstrip("\\") + next(tokens).lstrip("\\")
elif token == commands.BRACE:
delimiter = "{}"
elif token == commands.BRACK:
delimiter = "[]"
elif token == commands.CHOOSE:
delimiter = "()"
if token in (commands.ABOVE, commands.ABOVEWITHDELIMS):
dimension_node = tuple(_walk(tokens, terminator=terminator, limit=1))[0]
dimension = _get_dimension(dimension_node)
attributes = {"linethickness": dimension}
elif token in (commands.ATOP, commands.BRACE, commands.BRACK, commands.CHOOSE):
attributes = {"linethickness": "0"}
denominator = tuple(_walk(tokens, terminator=terminator))
sibling = None
if len(denominator) and denominator[-1].token == terminator:
denominator, sibling = denominator[:-1], denominator[-1]
if len(denominator) == 0:
if token in (commands.BRACE, commands.BRACK):
denominator = (Node(token=commands.BRACES, children=()),)
else:
raise DenominatorNotFoundError
if len(group) == 0:
if token in (commands.BRACE, commands.BRACK):
group = [Node(token=commands.BRACES, children=())]
else:
raise NumeratorNotFoundError
if len(denominator) > 1:
denominator = (Node(token=commands.BRACES, children=denominator),)
if len(group) == 1:
children = (*group, *denominator)
else:
children = (Node(token=commands.BRACES, children=tuple(group)), *denominator)
group = [Node(token=commands.FRAC, children=children, attributes=attributes, delimiter=delimiter)]
if sibling is not None:
group.append(sibling)
break
elif token == commands.SQRT:
root_nodes = None
next_node = tuple(_walk(tokens, limit=1))[0]
if next_node.token == commands.OPENING_BRACKET:
root_nodes = tuple(_walk(tokens, terminator=commands.CLOSING_BRACKET))[:-1]
next_node = tuple(_walk(tokens, limit=1))[0]
if len(root_nodes) > 1:
root_nodes = (Node(token=commands.BRACES, children=root_nodes),)
if root_nodes:
node = Node(token=commands.ROOT, children=(next_node, *root_nodes))
else:
node = Node(token=token, children=(next_node,))
elif token == commands.ROOT:
root_nodes = tuple(_walk(tokens, terminator=r"\of"))[:-1]
next_node = tuple(_walk(tokens, limit=1))[0]
if len(root_nodes) > 1:
root_nodes = (Node(token=commands.BRACES, children=root_nodes),)
if root_nodes:
node = Node(token=token, children=(next_node, *root_nodes))
else:
node = Node(token=token, children=(next_node, Node(token=commands.BRACES, children=())))
elif token in commands.MATRICES:
children = tuple(_walk(tokens, terminator=terminator))
sibling = None
if len(children) and children[-1].token == terminator:
children, sibling = children[:-1], children[-1]
if len(children) == 1 and children[0].token == commands.BRACES and children[0].children:
children = children[0].children
if sibling is not None:
group.extend([Node(token=token, children=children, alignment=""), sibling])
break
else:
node = Node(token=token, children=children, alignment="")
elif token == commands.GENFRAC:
delimiter = next(tokens).lstrip("\\") + next(tokens).lstrip("\\")
dimension_node, style_node = tuple(_walk(tokens, terminator=terminator, limit=2))
dimension = _get_dimension(dimension_node)
style = _get_style(style_node)
attributes = {"linethickness": dimension}
children = tuple(_walk(tokens, terminator=terminator, limit=2))
group.extend(
[Node(token=style), Node(token=token, children=children, delimiter=delimiter, attributes=attributes)]
)
break
elif token == commands.SIDESET:
left, right, operator = tuple(_walk(tokens, terminator=terminator, limit=3))
left_token, left_children = _make_subsup(left)
right_token, right_children = _make_subsup(right)
attributes = {"movablelimits": "false"}
node = Node(
token=token,
children=(
Node(
token=left_token,
children=(
Node(
token=commands.VPHANTOM,
children=(
Node(token=operator.token, children=operator.children, attributes=attributes),
),
),
*left_children,
),
),
Node(
token=right_token,
children=(
Node(token=operator.token, children=operator.children, attributes=attributes),
*right_children,
),
),
),
)
elif token == commands.SKEW:
width_node, child = tuple(_walk(tokens, terminator=terminator, limit=2))
width = width_node.token
if width == commands.BRACES:
if width_node.children is None or len(width_node.children) == 0:
raise InvalidWidthError
width = width_node.children[0].token
if not width.isdigit():
raise InvalidWidthError
node = Node(token=token, children=(child,), attributes={"width": f"{0.0555 * int(width):.3f}em"})
elif token.startswith(commands.BEGIN):
node = _get_environment_node(token, tokens)
else:
node = Node(token=token)
group.append(node)
if limit and len(group) >= limit:
break
if not has_available_tokens:
raise NoAvailableTokensError
return group