Mercurial > hg > nginx
annotate contrib/unicode2nginx/unicode-to-nginx.pl @ 8682:916a2e1d6617 quic
HTTP/3: client header validation.
A header with the name containing null, CR, LF, colon or uppercase characters,
is now considered an error. A header with the value containing null, CR or LF,
is also considered an error.
Also, header is considered invalid unless its name only contains lowercase
characters, digits, minus and optionally underscore. Such header can be
optionally ignored.
author | Roman Arutyunyan <arut@nginx.com> |
---|---|
date | Mon, 18 Jan 2021 13:43:36 +0300 |
parents | 8752257e883f |
children |
rev | line source |
---|---|
667 | 1 #!/usr/bin/perl -w |
2 | |
3 # Convert unicode mappings to nginx configuration file format. | |
4 | |
5 # You may find useful mappings in various places, including | |
6 # unicode.org official site: | |
7 # | |
8 # http://www.unicode.org/Public/MAPPINGS/VENDORS/MICSFT/WINDOWS/CP1251.TXT | |
9 # http://www.unicode.org/Public/MAPPINGS/VENDORS/MISC/KOI8-R.TXT | |
10 | |
11 # Needs perl 5.6 or later. | |
12 | |
6665
8752257e883f
Contrib: unicode2nginx compatibility with recent Perl versions.
Maxim Dounin <mdounin@mdounin.ru>
parents:
667
diff
changeset
|
13 # Written by Maxim Dounin, mdounin@mdounin.ru |
667 | 14 |
15 ############################################################################### | |
16 | |
17 require 5.006; | |
18 | |
19 while (<>) { | |
20 # Skip comments and empty lines | |
21 | |
22 next if /^#/; | |
23 next if /^\s*$/; | |
24 chomp; | |
25 | |
26 # Convert mappings | |
27 | |
28 if (/^\s*0x(..)\s*0x(....)\s*(#.*)/) { | |
29 # Mapping <from-code> <unicode-code> "#" <unicode-name> | |
30 my $cs_code = $1; | |
31 my $un_code = $2; | |
32 my $un_name = $3; | |
33 | |
34 # Produce UTF-8 sequence from character code; | |
35 | |
6665
8752257e883f
Contrib: unicode2nginx compatibility with recent Perl versions.
Maxim Dounin <mdounin@mdounin.ru>
parents:
667
diff
changeset
|
36 my $un_utf8 = join('', |
8752257e883f
Contrib: unicode2nginx compatibility with recent Perl versions.
Maxim Dounin <mdounin@mdounin.ru>
parents:
667
diff
changeset
|
37 map { sprintf("%02X", $_) } |
8752257e883f
Contrib: unicode2nginx compatibility with recent Perl versions.
Maxim Dounin <mdounin@mdounin.ru>
parents:
667
diff
changeset
|
38 unpack("U0C*", pack("U", hex($un_code))) |
8752257e883f
Contrib: unicode2nginx compatibility with recent Perl versions.
Maxim Dounin <mdounin@mdounin.ru>
parents:
667
diff
changeset
|
39 ); |
667 | 40 |
41 print " $cs_code $un_utf8 ; $un_name\n"; | |
42 | |
43 } else { | |
44 warn "Unrecognized line: '$_'"; | |
45 } | |
46 } | |
47 | |
48 ############################################################################### |