1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
|
=encoding utf8
=head1 NAME
B<jl> - pretty-print JSON logs
=head1 SYNOPSIS
B<jl> [I<options...>] [I<file>]
=head1 DESCRIPTION
B<jl> is a parser and formatter for JSON logs, making machine-readable JSON logs human readable again.
B<jl> consumes JSON logs from stdin and writes human-readable logs to stdout. To use B<jl>, just pipe your JSON logs through it:
./my-app-executable | jl
cat app-log.json | jl
B<jl> can read from a file:
jl my-app-log.json
B<jl> itself doesn't support following log files, but since it can consume from a pipe, you can use B<tail>:
tail -F app-log.json | jl
You can page B<jl>'s colorised output using B<less> with the B<-R> flag:
jl my-app-log.json | less -R
=head1 OPTIONS
A summary of options is included below.
=over
=item B<-color> B<auto>|B<yes>|B<no>
Set the color mode. The options are "auto", "yes", and "no". "auto" disables color if stdout is not a TTY (default: "auto")
=item B<-format> B<compact>|B<logfmt>
Formatter for logs. The options are "compact" and "logfmt" (default: "compact")
=item B<-truncate>[B<=false>]
Whether to truncate strings in the compact formatter (default: true)
=back
=head1 FORMATTERS
B<jl> currently supports two formatters, with plans to make the formatters customizable.
The default is B<-format compact>, which extracts only important fields from the JSON log, like B<message>, B<timestamp>, B<level>, colorises and presents them in a easy to skim way. It drops unrecognised fields from the logs.
The other option is B<-format logfmt>, which formats the JSON logs in a way that closely resembles logfmt
This option emits all fields from each log line.
Both formatters echo non-JSON log lines as-is.
=head1 LOG FORMATS
JSON application logs tend to have some core shared fields, like B<level>, B<timestamp>, and B<message> which B<jl> tries to discover and prioritise for formatting. For now, the following formats work best with B<jl>. For string fields other than B<level>, only the keys matter.
Other formats can be used after preprocessing with B<jq>.
=over
=item * Java-like:
{
"level": "error",
"timestamp": "2019-02-02 15:39:45",
"logger": "HelloWorldService",
"thread": "main",
"message": "hello world",
"exception": "java.lang.IllegalArgumentException: The world isn't here\n...stacktraces..."
}
=item * Go/Logrus-like:
{
"level": "error",
"timestamp": "2019-02-02 15:39:45",
"msg": "hello world",
"error": "hello error",
"stack": "\nhello\n\thello.go\nworld\n\tworld.go"
}
=back
=head1 COPYRIGHT
Copyright (C) 2019-2023 Yunchi Luo
License: MIT/Expat
This manual page is based on the README document by the author.
It was reformatted and adapted by Andrej Shadura L<< <andrewsh@debian.org> >>
for the Debian project, and is distributed under the same license as the
original project.
=head1 SEE ALSO
L<jq(1)>,
L<< https://blog.codeship.com/logfmt-a-log-format-thats-easy-to-read-and-write/ >>
|
