article-athas-shell-redirections.mw - tgtimes - The Gopher Times
HTML git clone git://bitreich.org/tgtimes git://enlrupgkhuxnvlhsf6lc3fziv5h2hhfrinws65d7roiv6bfj7d652fid.onion/tgtimes
DIR Log
DIR Files
DIR Refs
DIR Tags
DIR README
---
article-athas-shell-redirections.mw (2204B)
---
1 .SH athas
2 Shell Redirections
3 .
4 .PP
5 Newcomers to the Unix shell quickly encounter handy tools such as
6 sed(1) and sort(1). This command prints the lines of the given file
7 to stdout, in sorted order:
8 .
9 .DS
10 $ sort numbers
11 .DE
12 .
13 .PP
14 Soon after, newcomers will also encounter shell redirection, by which
15 the output of these tools can conveniently be read from or stored in
16 files:
17 .
18 .DS
19 $ sort < numbers > numbers_sorted
20 .DE
21 .
22 .PP
23 Our new user, fascinated by the modularity of the Unix shell, may then
24 try the rather obvious possibility of having the input and output file
25 be the same:
26 .
27 .DS
28 $ sort < numbers > numbers
29 .DE
30 .
31 .PP
32 But disaster strikes: the file is empty! The user has lost their
33 precious collection of numbers - let's hope they had a backup. Losing
34 data this way is almost a rite of passage for Unix users, but let us
35 spell out the reason for those who have yet to hurt themselves this
36 way.
37 .
38 .PP
39 When the Unix shell evaluates a command, it starts by processing the
40 redirection operators - that's the '>' and '<' above. While '<' just
41 opens the file, '>' *truncates* the file in-place as it is opened for
42 reading! This means that the 'sort' process will dutifully read an
43 empty file, sort its non-existent lines, and correctly produce empty
44 output.
45 .
46 .PP
47 Some programs can be asked to write their output directly to files
48 instead of using shell redirection (sed(1) has '-i', and for sort(1)
49 we can use '-o'), but this is not a general solution, and does not
50 work for pipelines. Another solution is to use the sponge(1) tool
51 from the "moreutils" project, which stores its standard input in
52 memory before finally writing it to a file:
53 .
54 .DS
55 $ sort < numbers | sponge numbers
56 .DE
57 .
58 .PP
59 The most interesting solution is to take advantage of subshells, the
60 shell evaluation order, and Unix file systems semantics. When we
61 delete a file in Unix, it is removed from the file system, but any
62 file descriptors referencing the file remain valid. We can exploit
63 this behaviour to delete the input file *after* directing the input,
64 but *before* redirecting the output:
65 .
66 .DS
67 $ (rm numbers && sort > numbers) < numbers
68 .DE
69 .
70 .PP
71 This approach requires no dependencies and will work in any Unix
72 shell.