Partial Matching of Character Items in a Vector
DESCRIPTION:
Returns, for each element of x, the position in table or unique completion whose initial substring matches that element.
USAGE: 
pmatch(x, table, nomatch=NA, duplicates.ok=F)
REQUIRED ARGUMENTS:
x:
character vector of items to be matched sequentially in table, using the partial match algorithm.
table:
character vector giving the possible values in x.
OPTIONAL ARGUMENTS:
nomatch:
the value to be returned when an item in x does not uniquely match any item in table.
duplicates.ok:
logical flag: allow duplicate values? If FALSE, duplicate values in x will correspond to nomatch; this is the way to simulate function call name matching. If the goal is to simulate character subset calculations, however, this argument should be TRUE.
VALUE:
numeric vector like x giving, for each element of x, the position in table of the first table[i] whose initial substring uniquely matches that element.
DETAILS:
Matching is done by the same algorithm used to match named arguments in S-PLUS function calls; see Section 11.3.5 of Becker, Chambers and Wilks. However, pmatch does not match empty strings, in contrast to the argument-matching algorithm, which matches unnamed arguments positionally after dealing with named arguments.

The charmatch function is a close synonym of pmatch. See the documentation for charmatch for a comparison of their differences.

SEE ALSO:
charmatch , amatch , match .
EXAMPLES: 
pmatch("Ala",state.name) # is NA (could match Alaska or Alabama)
pmatch("Alab",state.name) # is 1, matching Alabama


# given names, possibly truncated versions of state names
# produce the full state names
state.name[pmatch(names,state.name)]