t/t0028-working-tree-encoding.sh

   1 #!/bin/sh
   2
   3 test_description='working-tree-encoding conversion via gitattributes'
   4
   5 . ./test-lib.sh
   6
   7 GIT_TRACE_WORKING_TREE_ENCODING=1 && export GIT_TRACE_WORKING_TREE_ENCODING
   8
   9 test_expect_success 'setup test files' '
  10         git config core.eol lf &&
  11
  12         text="hallo there!\ncan you read me?" &&
  13         echo "*.utf16 text working-tree-encoding=utf-16" >.gitattributes &&
  14         printf "$text" >test.utf8.raw &&
  15         printf "$text" | iconv -f UTF-8 -t UTF-16 >test.utf16.raw &&
  16         printf "$text" | iconv -f UTF-8 -t UTF-32 >test.utf32.raw &&
  17
  18         # Line ending tests
  19         printf "one\ntwo\nthree\n" >lf.utf8.raw &&
  20         printf "one\r\ntwo\r\nthree\r\n" >crlf.utf8.raw &&
  21
  22         # BOM tests
  23         printf "\0a\0b\0c"                         >nobom.utf16be.raw &&
  24         printf "a\0b\0c\0"                         >nobom.utf16le.raw &&
  25         printf "\376\777\0a\0b\0c"                 >bebom.utf16be.raw &&
  26         printf "\777\376a\0b\0c\0"                 >lebom.utf16le.raw &&
  27         printf "\0\0\0a\0\0\0b\0\0\0c"             >nobom.utf32be.raw &&
  28         printf "a\0\0\0b\0\0\0c\0\0\0"             >nobom.utf32le.raw &&
  29         printf "\0\0\376\777\0\0\0a\0\0\0b\0\0\0c" >bebom.utf32be.raw &&
  30         printf "\777\376\0\0a\0\0\0b\0\0\0c\0\0\0" >lebom.utf32le.raw &&
  31
  32         # Add only UTF-16 file, we will add the UTF-32 file later
  33         cp test.utf16.raw test.utf16 &&
  34         cp test.utf32.raw test.utf32 &&
  35         git add .gitattributes test.utf16 &&
  36         git commit -m initial
  37 '
  38
  39 test_expect_success 'ensure UTF-8 is stored in Git' '
  40         test_when_finished "rm -f test.utf16.git" &&
  41
  42         git cat-file -p :test.utf16 >test.utf16.git &&
  43         test_cmp_bin test.utf8.raw test.utf16.git
  44 '
  45
  46 test_expect_success 're-encode to UTF-16 on checkout' '
  47         test_when_finished "rm -f test.utf16.raw" &&
  48
  49         rm test.utf16 &&
  50         git checkout test.utf16 &&
  51         test_cmp_bin test.utf16.raw test.utf16
  52 '
  53
  54 test_expect_success 'check $GIT_DIR/info/attributes support' '
  55         test_when_finished "rm -f test.utf32.git" &&
  56         test_when_finished "git reset --hard HEAD" &&
  57
  58         echo "*.utf32 text working-tree-encoding=utf-32" >.git/info/attributes &&
  59         git add test.utf32 &&
  60
  61         git cat-file -p :test.utf32 >test.utf32.git &&
  62         test_cmp_bin test.utf8.raw test.utf32.git
  63 '
  64
  65 for i in 16 32
  66 do
  67         test_expect_success "check prohibited UTF-${i} BOM" '
  68                 test_when_finished "git reset --hard HEAD" &&
  69
  70                 echo "*.utf${i}be text working-tree-encoding=utf-${i}be" >>.gitattributes &&
  71                 echo "*.utf${i}le text working-tree-encoding=utf-${i}LE" >>.gitattributes &&
  72
  73                 # Here we add a UTF-16 (resp. UTF-32) files with BOM (big/little-endian)
  74                 # but we tell Git to treat it as UTF-16BE/UTF-16LE (resp. UTF-32).
  75                 # In these cases the BOM is prohibited.
  76                 cp bebom.utf${i}be.raw bebom.utf${i}be &&
  77                 test_must_fail git add bebom.utf${i}be 2>err.out &&
  78                 test_i18ngrep "fatal: BOM is prohibited .* utf-${i}be" err.out &&
  79                 test_i18ngrep "use UTF-${i} as working-tree-encoding" err.out &&
  80
  81                 cp lebom.utf${i}le.raw lebom.utf${i}be &&
  82                 test_must_fail git add lebom.utf${i}be 2>err.out &&
  83                 test_i18ngrep "fatal: BOM is prohibited .* utf-${i}be" err.out &&
  84                 test_i18ngrep "use UTF-${i} as working-tree-encoding" err.out &&
  85
  86                 cp bebom.utf${i}be.raw bebom.utf${i}le &&
  87                 test_must_fail git add bebom.utf${i}le 2>err.out &&
  88                 test_i18ngrep "fatal: BOM is prohibited .* utf-${i}LE" err.out &&
  89                 test_i18ngrep "use UTF-${i} as working-tree-encoding" err.out &&
  90
  91                 cp lebom.utf${i}le.raw lebom.utf${i}le &&
  92                 test_must_fail git add lebom.utf${i}le 2>err.out &&
  93                 test_i18ngrep "fatal: BOM is prohibited .* utf-${i}LE" err.out &&
  94                 test_i18ngrep "use UTF-${i} as working-tree-encoding" err.out
  95         '
  96
  97         test_expect_success "check required UTF-${i} BOM" '
  98                 test_when_finished "git reset --hard HEAD" &&
  99
 100                 echo "*.utf${i} text working-tree-encoding=utf-${i}" >>.gitattributes &&
 101
 102                 cp nobom.utf${i}be.raw nobom.utf${i} &&
 103                 test_must_fail git add nobom.utf${i} 2>err.out &&
 104                 test_i18ngrep "fatal: BOM is required .* utf-${i}" err.out &&
 105                 test_i18ngrep "use UTF-${i}BE or UTF-${i}LE" err.out &&
 106
 107                 cp nobom.utf${i}le.raw nobom.utf${i} &&
 108                 test_must_fail git add nobom.utf${i} 2>err.out &&
 109                 test_i18ngrep "fatal: BOM is required .* utf-${i}" err.out &&
 110                 test_i18ngrep "use UTF-${i}BE or UTF-${i}LE" err.out
 111         '
 112
 113         test_expect_success "eol conversion for UTF-${i} encoded files on checkout" '
 114                 test_when_finished "rm -f crlf.utf${i}.raw lf.utf${i}.raw" &&
 115                 test_when_finished "git reset --hard HEAD^" &&
 116
 117                 cat lf.utf8.raw | iconv -f UTF-8 -t UTF-${i} >lf.utf${i}.raw &&
 118                 cat crlf.utf8.raw | iconv -f UTF-8 -t UTF-${i} >crlf.utf${i}.raw &&
 119                 cp crlf.utf${i}.raw eol.utf${i} &&
 120
 121                 cat >expectIndexLF <<-EOF &&
 122                         i/lf    w/-text attr/text               eol.utf${i}
 123                 EOF
 124
 125                 git add eol.utf${i} &&
 126                 git commit -m eol &&
 127
 128                 # UTF-${i} with CRLF (Windows line endings)
 129                 rm eol.utf${i} &&
 130                 git -c core.eol=crlf checkout eol.utf${i} &&
 131                 test_cmp_bin crlf.utf${i}.raw eol.utf${i} &&
 132
 133                 # Although the file has CRLF in the working tree,
 134                 # ensure LF in the index
 135                 git ls-files --eol eol.utf${i} >actual &&
 136                 test_cmp expectIndexLF actual &&
 137
 138                 # UTF-${i} with LF (Unix line endings)
 139                 rm eol.utf${i} &&
 140                 git -c core.eol=lf checkout eol.utf${i} &&
 141                 test_cmp_bin lf.utf${i}.raw eol.utf${i} &&
 142
 143                 # The file LF in the working tree, ensure LF in the index
 144                 git ls-files --eol eol.utf${i} >actual &&
 145                 test_cmp expectIndexLF actual
 146         '
 147 done
 148
 149 test_expect_success 'check unsupported encodings' '
 150         test_when_finished "git reset --hard HEAD" &&
 151
 152         echo "*.set text working-tree-encoding" >.gitattributes &&
 153         printf "set" >t.set &&
 154         test_must_fail git add t.set 2>err.out &&
 155         test_i18ngrep "true/false are no valid working-tree-encodings" err.out &&
 156
 157         echo "*.unset text -working-tree-encoding" >.gitattributes &&
 158         printf "unset" >t.unset &&
 159         git add t.unset &&
 160
 161         echo "*.empty text working-tree-encoding=" >.gitattributes &&
 162         printf "empty" >t.empty &&
 163         git add t.empty &&
 164
 165         echo "*.garbage text working-tree-encoding=garbage" >.gitattributes &&
 166         printf "garbage" >t.garbage &&
 167         test_must_fail git add t.garbage 2>err.out &&
 168         test_i18ngrep "failed to encode" err.out
 169 '
 170
 171 test_expect_success 'error if encoding round trip is not the same during refresh' '
 172         BEFORE_STATE=$(git rev-parse HEAD) &&
 173         test_when_finished "git reset --hard $BEFORE_STATE" &&
 174
 175         # Add and commit a UTF-16 file but skip the "working-tree-encoding"
 176         # filter. Consequently, the in-repo representation is UTF-16 and not
 177         # UTF-8. This simulates a Git version that has no working tree encoding
 178         # support.
 179         echo "*.utf16le text working-tree-encoding=utf-16le" >.gitattributes &&
 180         echo "hallo" >nonsense.utf16le &&
 181         TEST_HASH=$(git hash-object --no-filters -w nonsense.utf16le) &&
 182         git update-index --add --cacheinfo 100644 $TEST_HASH nonsense.utf16le &&
 183         COMMIT=$(git commit-tree -p $(git rev-parse HEAD) -m "plain commit" $(git write-tree)) &&
 184         git update-ref refs/heads/master $COMMIT &&
 185
 186         test_must_fail git checkout HEAD^ 2>err.out &&
 187         test_i18ngrep "error: .* overwritten by checkout:" err.out
 188 '
 189
 190 test_expect_success 'error if encoding garbage is already in Git' '
 191         BEFORE_STATE=$(git rev-parse HEAD) &&
 192         test_when_finished "git reset --hard $BEFORE_STATE" &&
 193
 194         # Skip the UTF-16 filter for the added file
 195         # This simulates a Git version that has no checkoutEncoding support
 196         cp nobom.utf16be.raw nonsense.utf16 &&
 197         TEST_HASH=$(git hash-object --no-filters -w nonsense.utf16) &&
 198         git update-index --add --cacheinfo 100644 $TEST_HASH nonsense.utf16 &&
 199         COMMIT=$(git commit-tree -p $(git rev-parse HEAD) -m "plain commit" $(git write-tree)) &&
 200         git update-ref refs/heads/master $COMMIT &&
 201
 202         git diff 2>err.out &&
 203         test_i18ngrep "error: BOM is required" err.out
 204 '
 205
 206 test_expect_success 'check roundtrip encoding' '
 207         test_when_finished "rm -f roundtrip.shift roundtrip.utf16" &&
 208         test_when_finished "git reset --hard HEAD" &&
 209
 210         text="hallo there!\nroundtrip test here!" &&
 211         printf "$text" | iconv -f UTF-8 -t SHIFT-JIS >roundtrip.shift &&
 212         printf "$text" | iconv -f UTF-8 -t UTF-16 >roundtrip.utf16 &&
 213         echo "*.shift text working-tree-encoding=SHIFT-JIS" >>.gitattributes &&
 214
 215         # SHIFT-JIS encoded files are round-trip checked by default...
 216         GIT_TRACE=1 git add .gitattributes roundtrip.shift 2>&1 |
 217                 grep "Checking roundtrip encoding for SHIFT-JIS" &&
 218         git reset &&
 219
 220         # ... unless we overwrite the Git config!
 221         ! GIT_TRACE=1 git -c core.checkRoundtripEncoding=garbage \
 222                 add .gitattributes roundtrip.shift 2>&1 |
 223                 grep "Checking roundtrip encoding for SHIFT-JIS" &&
 224         git reset &&
 225
 226         # UTF-16 encoded files should not be round-trip checked by default...
 227         ! GIT_TRACE=1 git add roundtrip.utf16 2>&1 |
 228                 grep "Checking roundtrip encoding for UTF-16" &&
 229         git reset &&
 230
 231         # ... unless we tell Git to check it!
 232         GIT_TRACE=1 git -c core.checkRoundtripEncoding="UTF-16, UTF-32" \
 233                 add roundtrip.utf16 2>&1 |
 234                 grep "Checking roundtrip encoding for utf-16" &&
 235         git reset &&
 236
 237         # ... unless we tell Git to check it!
 238         # (here we also check that the casing of the encoding is irrelevant)
 239         GIT_TRACE=1 git -c core.checkRoundtripEncoding="UTF-32, utf-16" \
 240                 add roundtrip.utf16 2>&1 |
 241                 grep "Checking roundtrip encoding for utf-16" &&
 242         git reset
 243 '
 244
 245 test_done