def tokeniter()

in courses/machine_learning/deepdive2/structured/labs/serving/application/lib/jinja2/lexer.py [0:0]


    def tokeniter(self, source, name, filename=None, state=None):
        """This method tokenizes the text and returns the tokens in a
        generator.  Use this method if you just want to tokenize a template.
        """
        source = text_type(source)
        lines = source.splitlines()
        if self.keep_trailing_newline and source:
            for newline in ('\r\n', '\r', '\n'):
                if source.endswith(newline):
                    lines.append('')
                    break
        source = '\n'.join(lines)
        pos = 0
        lineno = 1
        stack = ['root']
        if state is not None and state != 'root':
            assert state in ('variable', 'block'), 'invalid state'
            stack.append(state + '_begin')
        else:
            state = 'root'
        statetokens = self.rules[stack[-1]]
        source_length = len(source)

        balancing_stack = []

        while 1:
            # tokenizer loop
            for regex, tokens, new_state in statetokens:
                m = regex.match(source, pos)
                # if no match we try again with the next rule
                if m is None:
                    continue

                # we only match blocks and variables if braces / parentheses
                # are balanced. continue parsing with the lower rule which
                # is the operator rule. do this only if the end tags look
                # like operators
                if balancing_stack and \
                   tokens in ('variable_end', 'block_end',
                              'linestatement_end'):
                    continue

                # tuples support more options
                if isinstance(tokens, tuple):
                    for idx, token in enumerate(tokens):
                        # failure group
                        if token.__class__ is Failure:
                            raise token(lineno, filename)
                        # bygroup is a bit more complex, in that case we
                        # yield for the current token the first named
                        # group that matched
                        elif token == '#bygroup':
                            for key, value in iteritems(m.groupdict()):
                                if value is not None:
                                    yield lineno, key, value
                                    lineno += value.count('\n')
                                    break
                            else:
                                raise RuntimeError('%r wanted to resolve '
                                                   'the token dynamically'
                                                   ' but no group matched'
                                                   % regex)
                        # normal group
                        else:
                            data = m.group(idx + 1)
                            if data or token not in ignore_if_empty:
                                yield lineno, token, data
                            lineno += data.count('\n')

                # strings as token just are yielded as it.
                else:
                    data = m.group()
                    # update brace/parentheses balance
                    if tokens == 'operator':
                        if data == '{':
                            balancing_stack.append('}')
                        elif data == '(':
                            balancing_stack.append(')')
                        elif data == '[':
                            balancing_stack.append(']')
                        elif data in ('}', ')', ']'):
                            if not balancing_stack:
                                raise TemplateSyntaxError('unexpected \'%s\'' %
                                                          data, lineno, name,
                                                          filename)
                            expected_op = balancing_stack.pop()
                            if expected_op != data:
                                raise TemplateSyntaxError('unexpected \'%s\', '
                                                          'expected \'%s\'' %
                                                          (data, expected_op),
                                                          lineno, name,
                                                          filename)
                    # yield items
                    if data or tokens not in ignore_if_empty:
                        yield lineno, tokens, data
                    lineno += data.count('\n')

                # fetch new position into new variable so that we can check
                # if there is a internal parsing error which would result
                # in an infinite loop
                pos2 = m.end()

                # handle state changes
                if new_state is not None:
                    # remove the uppermost state
                    if new_state == '#pop':
                        stack.pop()
                    # resolve the new state by group checking
                    elif new_state == '#bygroup':
                        for key, value in iteritems(m.groupdict()):
                            if value is not None:
                                stack.append(key)
                                break
                        else:
                            raise RuntimeError('%r wanted to resolve the '
                                               'new state dynamically but'
                                               ' no group matched' %
                                               regex)
                    # direct state name given
                    else:
                        stack.append(new_state)
                    statetokens = self.rules[stack[-1]]
                # we are still at the same position and no stack change.
                # this means a loop without break condition, avoid that and
                # raise error
                elif pos2 == pos:
                    raise RuntimeError('%r yielded empty string without '
                                       'stack change' % regex)
                # publish new function and start again
                pos = pos2
                break
            # if loop terminated without break we haven't found a single match
            # either we are at the end of the file or we have a problem
            else:
                # end of text
                if pos >= source_length:
                    return
                # something went wrong
                raise TemplateSyntaxError('unexpected char %r at %d' %
                                          (source[pos], pos), lineno,
                                          name, filename)