Appeal 2007-2127
Reexamination Control No. 90/006,621
Compiler Design (1977), page 6 (Br. 93). Lexical analyzers for compliers
necessarily determine when the last character of a keyword, identifier,
operator, or punctuation mark has been inputted as part of the process of
separating the symbols into "tokens." Id. The "tokens" can be considered to
be "groups," so there is inherent written description support for claim 78.
Since a computer program symbol can be a "word" (i.e., a "keyword"), there
is written description support for claims 66, 76, and 79 because these claims
do not require that the input code only consists of words. This reason for the
rejection of claims 66, 76, 78, and 79 is reversed.
Although a computer program token can be a "word" or an
"identifier," not every token is "word" or an "identifier" because there must
be identifiers, operators, and punctuation marks. Therefore, we find no
written description support for the limitation of "grouping the characters into
sequences each constituting an identifier" (claim 63) and a "lexical analyzer
to form the characters into sequences each constituting a word" (claim 65).
Because Patent Owner does not define "identifier," he may be equating the
term with a "word," although in the computer art, it has a special meaning as
variable name. These claims are trying to indirectly claim that the data are
words of a language. The rejection of claims 63 and 65 is affirmed.
We do not find where Patent Owner addresses the limitation of a
"syntactic analyzer thread to determine whether said identifiers are
interrelated in accordance with predetermined rules of grammar" (claim 64).
While programming language syntactic analyzers apply predetermined rules
of grammar, it is not known whether they determine whether the identifiers
are related or if Patent Owner is trying to indirectly claim syntactic analysis
148
Page: Previous 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 Next
Last modified: September 9, 2013