A string in Factor is a sequence of characters. Most words from the
sequences vocabulary work on strings, in addition to the
dedicated string-handling words in splitting,
ascii, and friends.
Strings are written in double quotes. The usual escapes are supported
— \n, \t, \r, \\, \":
"hello, world" . ! => "hello, world"
"first\nsecond" print ! prints two lines
Because a string is a sequence, the sequences words
length and nth work on it directly.
length returns the number of characters:
"hello" length . ! => 5
nth takes a 0-based index and the sequence, and returns the
element at that position. For a string the element is a
character — a code point (an integer), not a one-character
substring:
0 "hello" nth . ! => 104 (CHAR: h)
4 "hello" nth . ! => 111 (CHAR: o)
To turn a single code point into a one-character string, use
1string ( char -- str ) from the strings vocabulary.
CHAR: is the parser-time form of a literal code point:
CHAR: A 1string . ! => "A"
0 "hello" nth 1string . ! => "h"
The splitting vocabulary cuts a string into pieces.
split1 cuts on the first occurrence of a separator and returns
both halves:
split1 ( seq subseq -- before after )
"foo: bar" ": " split1 .s
! => "foo"
! => "bar"
split cuts on any of the characters in a separator set, possibly
producing empty pieces:
"[ERROR]: Stack overflow" "[]" split .
! => { "" "ERROR" ": Stack overflow" }
harvest (in sequences) drops the empty pieces:
"[ERROR]: Stack overflow" "[]" split harvest .
! => { "ERROR" ": Stack overflow" }
Once a string is split into an array of pieces, first,
second, third, and fourth (in sequences)
return the piece at that position:
"apple,bee,carrot,duck" "," split first . ! => "apple"
"apple,bee,carrot,duck" "," split second . ! => "bee"
"apple,bee,carrot,duck" "," split third . ! => "carrot"
"apple,bee,carrot,duck" "," split fourth . ! => "duck"
When you want all the leading pieces on the stack at once,
first2 and first3 unpack two or three of them in one step:
"apple,bee,carrot,duck" "," split first2 .s
! => "apple"
! => "bee"
"apple,bee,carrot,duck" "," split first3 .s
! => "apple"
! => "bee"
! => "carrot"
On a string itself, first returns the first character — a
code point — rather than a substring:
"hello" first . ! => 104 (CHAR: h)
[ blank? ] trim (from sequences, with blank? from
the ascii vocabulary) removes leading and trailing whitespace:
" Disk full \r\n" [ blank? ] trim .
! => "Disk full"
>lower and >upper (from ascii) return a new string
with the case changed:
"WARNING" >lower . ! => "warning"
"hello" >upper . ! => "HELLO"
Two sequences words assemble strings:
surround wraps a string with a prefix and a suffix:
surround ( seq pre post -- new-seq )
"warning" "(" ")" surround . ! => "(warning)"
glue joins two strings with a separator between them:
glue ( s1 s2 sep -- s )
"Disk full" "(error)" " " glue . ! => "Disk full (error)"
In this exercise you'll be processing log lines.
Each log line is a string formatted as "[<LEVEL>]: <MESSAGE>".
There are three log levels: INFO, WARNING, and ERROR.
You have three tasks. Each takes a single log line off the stack.
Define message to return the log line's message. Any leading or
trailing whitespace should be removed.
"[ERROR]: Invalid operation" message .
! => "Invalid operation"
"[WARNING]: Disk almost full\r\n" message .
! => "Disk almost full"
Define log-level to return the log line's level, lowercased.
"[ERROR]: Invalid operation" log-level .
! => "error"
Define reformat to put the message first and the lowercase log level
in parentheses after it.
"[INFO]: Operation completed" reformat .
! => "Operation completed (info)"
Note: All strings in this exercise are ASCII. Later exercises will tackle Unicode-aware string handling.
Sign up to Exercism to learn and master Factor with 47 concepts163 exercises, and real human mentoring, all for free.