t/t7812-grep-icase-non-ascii.sh

   1 #!/bin/sh
   2
   3 test_description='grep icase on non-English locales'
   4
   5 . ./lib-gettext.sh
   6
   7 test_expect_success GETTEXT_LOCALE 'setup' '
   8         test_write_lines "TILRAUN: Halló Heimur!" >file &&
   9         git add file &&
  10         LC_ALL="$is_IS_locale" &&
  11         export LC_ALL
  12 '
  13
  14 test_have_prereq GETTEXT_LOCALE &&
  15 test-tool regex "HALLÓ" "Halló" ICASE &&
  16 test_set_prereq REGEX_LOCALE
  17
  18 test_expect_success REGEX_LOCALE 'grep literal string, no -F' '
  19         git grep -i "TILRAUN: Halló Heimur!" &&
  20         git grep -i "TILRAUN: HALLÓ HEIMUR!"
  21 '
  22
  23 test_expect_success GETTEXT_LOCALE,PCRE 'grep pcre utf-8 icase' '
  24         git grep --perl-regexp    "TILRAUN: H.lló Heimur!" &&
  25         git grep --perl-regexp -i "TILRAUN: H.lló Heimur!" &&
  26         git grep --perl-regexp -i "TILRAUN: H.LLÓ HEIMUR!"
  27 '
  28
  29 test_expect_success GETTEXT_LOCALE,PCRE 'grep pcre utf-8 string with "+"' '
  30         test_write_lines "TILRAUN: Hallóó Heimur!" >file2 &&
  31         git add file2 &&
  32         git grep -l --perl-regexp "TILRAUN: H.lló+ Heimur!" >actual &&
  33         echo file >expected &&
  34         echo file2 >>expected &&
  35         test_cmp expected actual
  36 '
  37
  38 test_expect_success REGEX_LOCALE 'grep literal string, with -F' '
  39         git grep -i -F "TILRAUN: Halló Heimur!" &&
  40         git grep -i -F "TILRAUN: HALLÓ HEIMUR!"
  41 '
  42
  43 test_expect_success REGEX_LOCALE 'grep string with regex, with -F' '
  44         test_write_lines "TILRAUN: Halló Heimur [abc]!" >file3 &&
  45         git add file3 &&
  46         git grep -i -F "TILRAUN: Halló Heimur [abc]!" file3
  47 '
  48
  49 test_expect_success REGEX_LOCALE 'pickaxe -i on non-ascii' '
  50         git commit -m first &&
  51         git log --format=%f -i -S"TILRAUN: HALLÓ HEIMUR!" >actual &&
  52         echo first >expected &&
  53         test_cmp expected actual
  54 '
  55
  56 test_expect_success GETTEXT_LOCALE,PCRE 'log --author with an ascii pattern on UTF-8 data' '
  57         cat >expected <<-\EOF &&
  58         Author: <BOLD;RED>À Ú Thor<RESET> <author@example.com>
  59         EOF
  60         test_write_lines "forth" >file4 &&
  61         git add file4 &&
  62         git commit --author="À Ú Thor <author@example.com>" -m sécond &&
  63         git log -1 --color=always --perl-regexp --author=".*Thor" >log &&
  64         grep Author log >actual.raw &&
  65         test_decode_color <actual.raw >actual &&
  66         test_cmp expected actual
  67 '
  68
  69 test_expect_success GETTEXT_LOCALE,PCRE 'log --committer with an ascii pattern on ISO-8859-1 data' '
  70         cat >expected <<-\EOF &&
  71         Commit:     Ç<BOLD;RED> O Mîtter <committer@example.com><RESET>
  72         EOF
  73         test_write_lines "fifth" >file5 &&
  74         git add file5 &&
  75         GIT_COMMITTER_NAME="Ç O Mîtter" &&
  76         GIT_COMMITTER_EMAIL="committer@example.com" &&
  77         git -c i18n.commitEncoding=latin1 commit -m thïrd &&
  78         git -c i18n.logOutputEncoding=latin1 log -1 --pretty=fuller --color=always --perl-regexp --committer=" O.*" >log &&
  79         grep Commit: log >actual.raw &&
  80         test_decode_color <actual.raw >actual &&
  81         test_cmp expected actual
  82 '
  83
  84 test_expect_success GETTEXT_LOCALE,PCRE 'log --grep with an ascii pattern on UTF-8 data' '
  85         cat >expected <<-\EOF &&
  86             sé<BOLD;RED>con<RESET>d
  87         EOF
  88         git log -1 --color=always --perl-regexp --grep="con" >log &&
  89         grep con log >actual.raw &&
  90         test_decode_color <actual.raw >actual &&
  91         test_cmp expected actual
  92 '
  93
  94 test_expect_success GETTEXT_LOCALE,PCRE 'log --grep with an ascii pattern on ISO-8859-1 data' '
  95         cat >expected <<-\EOF &&
  96             <BOLD;RED>thïrd<RESET>
  97         EOF
  98         git -c i18n.logOutputEncoding=latin1 log -1 --color=always --perl-regexp --grep="th.*rd" >log &&
  99         grep "th.*rd" log >actual.raw &&
 100         test_decode_color <actual.raw >actual &&
 101         test_cmp expected actual
 102 '
 103
 104 test_expect_success GETTEXT_LOCALE,LIBPCRE2 'PCRE v2: setup invalid UTF-8 data' '
 105         printf "\\200\\n" >invalid-0x80 &&
 106         echo "ævar" >expected &&
 107         cat expected >>invalid-0x80 &&
 108         git add invalid-0x80 &&
 109
 110         # Test for PCRE2_MATCH_INVALID_UTF bug
 111         # https://bugs.exim.org/show_bug.cgi?id=2642
 112         printf "\\345Aæ\\n" >invalid-0xe5 &&
 113         git add invalid-0xe5
 114 '
 115
 116 test_expect_success GETTEXT_LOCALE,LIBPCRE2 'PCRE v2: grep ASCII from invalid UTF-8 data' '
 117         git grep -h "var" invalid-0x80 >actual &&
 118         test_cmp expected actual &&
 119         git grep -h "(*NO_JIT)var" invalid-0x80 >actual &&
 120         test_cmp expected actual
 121 '
 122
 123 test_expect_success GETTEXT_LOCALE,LIBPCRE2 'PCRE v2: grep ASCII from invalid UTF-8 data (PCRE2 bug #2642)' '
 124         git grep -h "Aæ" invalid-0xe5 >actual &&
 125         test_cmp invalid-0xe5 actual &&
 126         git grep -h "(*NO_JIT)Aæ" invalid-0xe5 >actual &&
 127         test_cmp invalid-0xe5 actual
 128 '
 129
 130 test_expect_success GETTEXT_LOCALE,LIBPCRE2 'PCRE v2: grep non-ASCII from invalid UTF-8 data' '
 131         git grep -h "æ" invalid-0x80 >actual &&
 132         test_cmp expected actual &&
 133         git grep -h "(*NO_JIT)æ" invalid-0x80 >actual &&
 134         test_cmp expected actual
 135 '
 136
 137 test_expect_success GETTEXT_LOCALE,LIBPCRE2 'PCRE v2: grep non-ASCII from invalid UTF-8 data (PCRE2 bug #2642)' '
 138         git grep -h "Aæ" invalid-0xe5 >actual &&
 139         test_cmp invalid-0xe5 actual &&
 140         git grep -h "(*NO_JIT)Aæ" invalid-0xe5 >actual &&
 141         test_cmp invalid-0xe5 actual
 142 '
 143
 144 test_lazy_prereq PCRE2_MATCH_INVALID_UTF '
 145         test-tool pcre2-config has-PCRE2_MATCH_INVALID_UTF
 146 '
 147
 148 test_expect_success GETTEXT_LOCALE,LIBPCRE2 'PCRE v2: grep non-ASCII from invalid UTF-8 data with -i' '
 149         test_might_fail git grep -hi "Æ" invalid-0x80 >actual &&
 150         test_might_fail git grep -hi "(*NO_JIT)Æ" invalid-0x80 >actual
 151 '
 152
 153 test_expect_success GETTEXT_LOCALE,LIBPCRE2,PCRE2_MATCH_INVALID_UTF 'PCRE v2: grep non-ASCII from invalid UTF-8 data with -i' '
 154         git grep -hi "Æ" invalid-0x80 >actual &&
 155         test_cmp expected actual &&
 156         git grep -hi "(*NO_JIT)Æ" invalid-0x80 >actual &&
 157         test_cmp expected actual
 158 '
 159
 160 test_expect_success GETTEXT_LOCALE,LIBPCRE2,PCRE2_MATCH_INVALID_UTF 'PCRE v2: grep non-ASCII from invalid UTF-8 data with -i (PCRE2 bug #2642)' '
 161         git grep -hi "Æ" invalid-0xe5 >actual &&
 162         test_cmp invalid-0xe5 actual &&
 163         git grep -hi "(*NO_JIT)Æ" invalid-0xe5 >actual &&
 164         test_cmp invalid-0xe5 actual &&
 165
 166         # Only the case of grepping the ASCII part in a way that
 167         # relies on -i fails
 168         git grep -hi "aÆ" invalid-0xe5 >actual &&
 169         test_cmp invalid-0xe5 actual &&
 170         git grep -hi "(*NO_JIT)aÆ" invalid-0xe5 >actual &&
 171         test_cmp invalid-0xe5 actual
 172 '
 173
 174 test_done