BOOTSTRAP (kutils) functions
LIBRARY ROUTINE
kparse_file_search - search a file for a specific key
LIBRARY CALL
int
kparse_file_search(
kfile *file,
char *key,
int mode,
char **key_format)
INPUT
- file - a pointer to the transport file opened for reading
- key - The regular expression key to search for.
- mode - tells the parser which mode to work in:
KIGNORE_CASE Ignore case on a-z and A-Z
KLITERAL Use case information on
a-z and A-Z
OUTPUT
- key_format - the address of a pointer to hold the returned
key that was matched. Sufficient space for
the returned string will be allocated if you
pass in a valid pointer. Note that if this
parameter is passed in as NULL, this routine
will ignore it, and the string that key matched
will not be returned.
RETURN VALUE
return status of the search. It can be one of the following:
KPARSE_OK (parse ok, return data valid),
KPARSE_NOKEY (couldn't find key)
KPARSE_DATAERR (something was wrong with the data)
KPARSE_PARTKEY (data ended with a partial match)
KPARSE_NULLKEY (key was NULL)
KPARSE_SYNTAXKEY (key had an illegal syntax)
DESCRIPTION
This routine is a Khoros Transport Interface to the string
parser provided in the kparse_string_search routine. It
returns a malloc'ed copy of the key that was matched in the
data file. It also sets the current file position, with the
transport call kfseek(), to the character following the
matched key.
ADDITIONAL INFORMATION
The regular expression syntax supported by the parser
is: '.', '*', '[..]', '[^..]', '$', '^', and '\\'. The explainations
for each syntax is explained below.
.
Match any single character except newline
*
Match the preceding character or ranger of characters 0 or more times. This
matching includes items specified within a [...] construct.
[...] or [^...]
Matches any one character contained within the brackets. If the first
character after the is a ']', then it is included as a character to match.
If the first character after the '[' is a '^', then it will match all characters
NOT included within the []. Finally, the '-' will indicate a range of
characters. For example, [a-z] specifies all characters between and
including the ascii values 'a' and 'z'. If the '-' follows the '[', or is
right before the ']', then it is interpreted literally.
^
If it is the first character of the regular expression, it matches the
beginning of the line.
$
If it is the last character of the regular expression, it matches the end of
line. However, it does NOT put the \\n in your key when the matched key is
returned.
\e
This escapes the meaning of a special character.
EXAMPLES
none
SIDE EFFECTS
This routine mallocs data and sets the value of key_format
parameter. Thus, the user should pass in an address of
an unused character pointer. The calling routine is
responsible for freeing the space malloc'ed for
the key_format parameter.
RESTRICTIONS
It does not support the following regular expression
constructs: or'ing '|', grouping of regular expressions '()',
match one or more times '+', or match n to m times '\\{n,m\\}'.
Finally, the '\\number' and '\\(\\)' constructs have no meaning
for these routines, so they are not supported either.
Search keys and the data file should not contain the
values '\\001', '\\002', '\\003', or '\\004',
because these values are used as special search parameters
by the parser.
MODIFICATION
none
FILES
$BOOTSTRAP/objects/library/kutils/src/parse.c
SEE ALSO
kparse_string_search(kutils), kfseek(kutils)
COPYRIGHT
Copyright (C) 1993 - 1997, Khoral Research, Inc. ("KRI") All rights reserved.