=> R0 = pointer to syntax string
R1 = pointer to input string
R2 = pointer to output buffer
R3 = size of output buffer
<= R3 = bytes left in output buffer
This command scans the input string, and places values in output buffer according to the syntax string.
See PRM 1, page 465.
Syntax string
string = element{, element}
element = [identifier [=alias] ] {/qualifier}
qualifier =
A : Keyword must always be given value
K : Identifier must be used
S : Switch
E : Do OS_EvaluateExpression on value
G : Do OS_GSTrans on value
Note: An element doesn't have to have a name. "," is a perfectly legal syntax string.
Input-string
example:
Syntax
"number=times/e,file/k/a,expand/s"
These input-strings will match:
"-n 10 -file myfile"
"-times 1+7 -file myfile -expand"
"-file myfile -e"
These will not:
"myfile -number 4"
"-number 20 -times 4 -file myfile"
Output buffer
For the Nth identifier (named or not), word N in output buffer contains :
Switch (/S) : non zero value if switch used
GSTransed (/G) : pointer to block (0 if unused) containing a two-byte length and then string itself (not terminated)
Evaluated (/E) : pointer to block (0 if unused) containing a one-byte type then four bytes of value (integer if type=0)
else : pointer to NULL terminated string
The results blocks pointed to are appended after the array of N words, therefore the output buffer must be at least N words long plus space to hold values/strings themselves.
[sh-index] Back to list of manuals
|
| |
