To run CorpusSearch:

for automatic output file (command.out)
...CorpusSearch <command.q> <input-files>
for output file with your choice of name (my_name.out)
...CorpusSearch <command.q> <input-files> -out my_name.out

Query file names must end in ".q". Output file names must end in ".out". Preference file name must end in ".prf" and be located in the same directory as the query being run.

some commonly used search functions
X exists(a node labeled X exists anywhere in token)
X precedes Y(X precedes Y without overlapping)
X iPrecedes Y(X immediately precedes Y)
X dominates Y(X dominates Y with any length path between them)
X domsWords #(X dominates # of words)
X iDominates Y(X immediately dominates Y)
X iDomsLast Y(X immediately dominates Y as last child)
X iDomsMod Y Z(X dominates Z with at most Y's intervening)
X iDomsNumber # Y(X immediately dominates Y as first, second, etc. child)
X iDomsOnly Y(X immediately dominates Y as only child)
X iDomsTotal #(X immediately dominates # of daughters)
X inID(X occurs in ID node)
X isRoot(X is the root node of the token's parse tree)
X sameIndex Y(X and Y have the same numerical index)
CODING column # X(X occurs in coding node in column #)
logical operators
A AND B(predicates A and B - possibly complex - must both be true of hits)
NOT A(expression A - possibly complex - must be false of hits)
A OR B (at least one of expressions A and B must be true of hits)
!(negate one argument to a predicate)
|(link alternative matches in an argument to a predicate)

wild cards:
*matches any character
#matches any digit(s) (0, 1, ... 9)

command-file components:

print_complement:falsequery:required: no default
print_indices:falsenode:required: no default
remove_nodes:falseadd_to_ignore:adds labels to default
ur_text_only:falsedefine: <my.def>includes definition file


includes remark in search output file