Java Reference
In-Depth Information
procedure
C
ompress
()
for
i
=
1
to
N
×
M
do
V
.
entry
[
i
]
14
←
de f ault
foreach
row
∈{
1
,
2
,...,
N
}
do
15
R
[
row
]
←
F
ind
S
hift
(
row
)
for
j
=
1
to
M
do
if
T
[
row
,
j
]
de f ault
then
place
←
R
[
row
]
+
j
V
.
entry
[
place
]
←
T
[
row
,
j
]
V
.
fromrow
[
place
]
←
row
call
T
runc
(
V
)
16
end
function
F
ind
S
hift
(
row
)
returns
Integer
return
N
×
M
−
M
(
row
,
shi f t
)
min
shi f t
=−
M
+
F
its
1
end
function
F
its
(
row
,
shi f t
)
returns
Boolean
for
j
=
1
to
M
do
if
T
[
row
,
j
]
de f ault
and not
R
oom
I
n
V(
shi f t
+
j
)
17
then return
(
false
)
return
(
true
)
end
function
R
oom
I
n
V(
where
)
returns
Boolean
if
where
≥
1
then
if
V
.
entry
[
where
]
=
de f ault
then return
(
true
)
return
(
false
)
end
procedure
T
(
V
)
for
i
=
N
×
M
downto
1
do
if
V
.
entry
[
i
]
runc
de f ault
then
/
Retain
V
[1
...
i
]
/
return
()
end
Figure 5.23: Compression algorithm.
