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         echo "*.utf16lebom text working-tree-encoding=UTF-16LE-BOM" >>.gitattributes &&
  15         printf "$text" >test.utf8.raw &&
  16         printf "$text" | iconv -f UTF-8 -t UTF-16 >test.utf16.raw &&
  17         printf "$text" | iconv -f UTF-8 -t UTF-32 >test.utf32.raw &&
  18         printf "\377\376"                         >test.utf16lebom.raw &&
  19         printf "$text" | iconv -f UTF-8 -t UTF-32LE >>test.utf16lebom.raw &&
  20
  21         # Line ending tests
  22         printf "one\ntwo\nthree\n" >lf.utf8.raw &&
  23         printf "one\r\ntwo\r\nthree\r\n" >crlf.utf8.raw &&
  24
  25         # BOM tests
  26         printf "\0a\0b\0c"                         >nobom.utf16be.raw &&
  27         printf "a\0b\0c\0"                         >nobom.utf16le.raw &&
  28         printf "\376\777\0a\0b\0c"                 >bebom.utf16be.raw &&
  29         printf "\777\376a\0b\0c\0"                 >lebom.utf16le.raw &&
  30         printf "\0\0\0a\0\0\0b\0\0\0c"             >nobom.utf32be.raw &&
  31         printf "a\0\0\0b\0\0\0c\0\0\0"             >nobom.utf32le.raw &&
  32         printf "\0\0\376\777\0\0\0a\0\0\0b\0\0\0c" >bebom.utf32be.raw &&
  33         printf "\777\376\0\0a\0\0\0b\0\0\0c\0\0\0" >lebom.utf32le.raw &&
  34
  35         # Add only UTF-16 file, we will add the UTF-32 file later
  36         cp test.utf16.raw test.utf16 &&
  37         cp test.utf32.raw test.utf32 &&
  38         cp test.utf16lebom.raw test.utf16lebom &&
  39         git add .gitattributes test.utf16 test.utf16lebom &&
  40         git commit -m initial
  41 '
  42
  43 test_expect_success 'ensure UTF-8 is stored in Git' '
  44         test_when_finished "rm -f test.utf16.git" &&
  45
  46         git cat-file -p :test.utf16 >test.utf16.git &&
  47         test_cmp_bin test.utf8.raw test.utf16.git
  48 '
  49
  50 test_expect_success 're-encode to UTF-16 on checkout' '
  51         test_when_finished "rm -f test.utf16.raw" &&
  52
  53         rm test.utf16 &&
  54         git checkout test.utf16 &&
  55         test_cmp_bin test.utf16.raw test.utf16
  56 '
  57
  58 test_expect_success 're-encode to UTF-16-LE-BOM on checkout' '
  59         rm test.utf16lebom &&
  60         git checkout test.utf16lebom &&
  61         test_cmp_bin test.utf16lebom.raw test.utf16lebom
  62 '
  63
  64 test_expect_success 'check $GIT_DIR/info/attributes support' '
  65         test_when_finished "rm -f test.utf32.git" &&
  66         test_when_finished "git reset --hard HEAD" &&
  67
  68         echo "*.utf32 text working-tree-encoding=utf-32" >.git/info/attributes &&
  69         git add test.utf32 &&
  70
  71         git cat-file -p :test.utf32 >test.utf32.git &&
  72         test_cmp_bin test.utf8.raw test.utf32.git
  73 '
  74
  75 for i in 16 32
  76 do
  77         test_expect_success "check prohibited UTF-${i} BOM" '
  78                 test_when_finished "git reset --hard HEAD" &&
  79
  80                 echo "*.utf${i}be text working-tree-encoding=utf-${i}be" >>.gitattributes &&
  81                 echo "*.utf${i}le text working-tree-encoding=utf-${i}LE" >>.gitattributes &&
  82
  83                 # Here we add a UTF-16 (resp. UTF-32) files with BOM (big/little-endian)
  84                 # but we tell Git to treat it as UTF-16BE/UTF-16LE (resp. UTF-32).
  85                 # In these cases the BOM is prohibited.
  86                 cp bebom.utf${i}be.raw bebom.utf${i}be &&
  87                 test_must_fail git add bebom.utf${i}be 2>err.out &&
  88                 test_i18ngrep "fatal: BOM is prohibited .* utf-${i}be" err.out &&
  89                 test_i18ngrep "use UTF-${i} as working-tree-encoding" err.out &&
  90
  91                 cp lebom.utf${i}le.raw lebom.utf${i}be &&
  92                 test_must_fail git add lebom.utf${i}be 2>err.out &&
  93                 test_i18ngrep "fatal: BOM is prohibited .* utf-${i}be" err.out &&
  94                 test_i18ngrep "use UTF-${i} as working-tree-encoding" err.out &&
  95
  96                 cp bebom.utf${i}be.raw bebom.utf${i}le &&
  97                 test_must_fail git add bebom.utf${i}le 2>err.out &&
  98                 test_i18ngrep "fatal: BOM is prohibited .* utf-${i}LE" err.out &&
  99                 test_i18ngrep "use UTF-${i} as working-tree-encoding" err.out &&
 100
 101                 cp lebom.utf${i}le.raw lebom.utf${i}le &&
 102                 test_must_fail git add lebom.utf${i}le 2>err.out &&
 103                 test_i18ngrep "fatal: BOM is prohibited .* utf-${i}LE" err.out &&
 104                 test_i18ngrep "use UTF-${i} as working-tree-encoding" err.out
 105         '
 106
 107         test_expect_success "check required UTF-${i} BOM" '
 108                 test_when_finished "git reset --hard HEAD" &&
 109
 110                 echo "*.utf${i} text working-tree-encoding=utf-${i}" >>.gitattributes &&
 111
 112                 cp nobom.utf${i}be.raw nobom.utf${i} &&
 113                 test_must_fail git add nobom.utf${i} 2>err.out &&
 114                 test_i18ngrep "fatal: BOM is required .* utf-${i}" err.out &&
 115                 test_i18ngrep "use UTF-${i}BE or UTF-${i}LE" err.out &&
 116
 117                 cp nobom.utf${i}le.raw nobom.utf${i} &&
 118                 test_must_fail git add nobom.utf${i} 2>err.out &&
 119                 test_i18ngrep "fatal: BOM is required .* utf-${i}" err.out &&
 120                 test_i18ngrep "use UTF-${i}BE or UTF-${i}LE" err.out
 121         '
 122
 123         test_expect_success "eol conversion for UTF-${i} encoded files on checkout" '
 124                 test_when_finished "rm -f crlf.utf${i}.raw lf.utf${i}.raw" &&
 125                 test_when_finished "git reset --hard HEAD^" &&
 126
 127                 cat lf.utf8.raw | iconv -f UTF-8 -t UTF-${i} >lf.utf${i}.raw &&
 128                 cat crlf.utf8.raw | iconv -f UTF-8 -t UTF-${i} >crlf.utf${i}.raw &&
 129                 cp crlf.utf${i}.raw eol.utf${i} &&
 130
 131                 cat >expectIndexLF <<-EOF &&
 132                         i/lf    w/-text attr/text               eol.utf${i}
 133                 EOF
 134
 135                 git add eol.utf${i} &&
 136                 git commit -m eol &&
 137
 138                 # UTF-${i} with CRLF (Windows line endings)
 139                 rm eol.utf${i} &&
 140                 git -c core.eol=crlf checkout eol.utf${i} &&
 141                 test_cmp_bin crlf.utf${i}.raw eol.utf${i} &&
 142
 143                 # Although the file has CRLF in the working tree,
 144                 # ensure LF in the index
 145                 git ls-files --eol eol.utf${i} >actual &&
 146                 test_cmp expectIndexLF actual &&
 147
 148                 # UTF-${i} with LF (Unix line endings)
 149                 rm eol.utf${i} &&
 150                 git -c core.eol=lf checkout eol.utf${i} &&
 151                 test_cmp_bin lf.utf${i}.raw eol.utf${i} &&
 152
 153                 # The file LF in the working tree, ensure LF in the index
 154                 git ls-files --eol eol.utf${i} >actual &&
 155                 test_cmp expectIndexLF actual
 156         '
 157 done
 158
 159 test_expect_success 'check unsupported encodings' '
 160         test_when_finished "git reset --hard HEAD" &&
 161
 162         echo "*.set text working-tree-encoding" >.gitattributes &&
 163         printf "set" >t.set &&
 164         test_must_fail git add t.set 2>err.out &&
 165         test_i18ngrep "true/false are no valid working-tree-encodings" err.out &&
 166
 167         echo "*.unset text -working-tree-encoding" >.gitattributes &&
 168         printf "unset" >t.unset &&
 169         git add t.unset &&
 170
 171         echo "*.empty text working-tree-encoding=" >.gitattributes &&
 172         printf "empty" >t.empty &&
 173         git add t.empty &&
 174
 175         echo "*.garbage text working-tree-encoding=garbage" >.gitattributes &&
 176         printf "garbage" >t.garbage &&
 177         test_must_fail git add t.garbage 2>err.out &&
 178         test_i18ngrep "failed to encode" err.out
 179 '
 180
 181 test_expect_success 'error if encoding round trip is not the same during refresh' '
 182         BEFORE_STATE=$(git rev-parse HEAD) &&
 183         test_when_finished "git reset --hard $BEFORE_STATE" &&
 184
 185         # Add and commit a UTF-16 file but skip the "working-tree-encoding"
 186         # filter. Consequently, the in-repo representation is UTF-16 and not
 187         # UTF-8. This simulates a Git version that has no working tree encoding
 188         # support.
 189         echo "*.utf16le text working-tree-encoding=utf-16le" >.gitattributes &&
 190         echo "hallo" >nonsense.utf16le &&
 191         TEST_HASH=$(git hash-object --no-filters -w nonsense.utf16le) &&
 192         git update-index --add --cacheinfo 100644 $TEST_HASH nonsense.utf16le &&
 193         COMMIT=$(git commit-tree -p $(git rev-parse HEAD) -m "plain commit" $(git write-tree)) &&
 194         git update-ref refs/heads/master $COMMIT &&
 195
 196         test_must_fail git checkout HEAD^ 2>err.out &&
 197         test_i18ngrep "error: .* overwritten by checkout:" err.out
 198 '
 199
 200 test_expect_success 'error if encoding garbage is already in Git' '
 201         BEFORE_STATE=$(git rev-parse HEAD) &&
 202         test_when_finished "git reset --hard $BEFORE_STATE" &&
 203
 204         # Skip the UTF-16 filter for the added file
 205         # This simulates a Git version that has no checkoutEncoding support
 206         cp nobom.utf16be.raw nonsense.utf16 &&
 207         TEST_HASH=$(git hash-object --no-filters -w nonsense.utf16) &&
 208         git update-index --add --cacheinfo 100644 $TEST_HASH nonsense.utf16 &&
 209         COMMIT=$(git commit-tree -p $(git rev-parse HEAD) -m "plain commit" $(git write-tree)) &&
 210         git update-ref refs/heads/master $COMMIT &&
 211
 212         git diff 2>err.out &&
 213         test_i18ngrep "error: BOM is required" err.out
 214 '
 215
 216 test_lazy_prereq ICONV_SHIFT_JIS '
 217         iconv -f UTF-8 -t SHIFT-JIS </dev/null
 218 '
 219
 220 test_expect_success ICONV_SHIFT_JIS 'check roundtrip encoding' '
 221         test_when_finished "rm -f roundtrip.shift roundtrip.utf16" &&
 222         test_when_finished "git reset --hard HEAD" &&
 223
 224         text="hallo there!\nroundtrip test here!" &&
 225         printf "$text" | iconv -f UTF-8 -t SHIFT-JIS >roundtrip.shift &&
 226         printf "$text" | iconv -f UTF-8 -t UTF-16 >roundtrip.utf16 &&
 227         echo "*.shift text working-tree-encoding=SHIFT-JIS" >>.gitattributes &&
 228
 229         # SHIFT-JIS encoded files are round-trip checked by default...
 230         GIT_TRACE=1 git add .gitattributes roundtrip.shift 2>&1 |
 231                 grep "Checking roundtrip encoding for SHIFT-JIS" &&
 232         git reset &&
 233
 234         # ... unless we overwrite the Git config!
 235         ! GIT_TRACE=1 git -c core.checkRoundtripEncoding=garbage \
 236                 add .gitattributes roundtrip.shift 2>&1 |
 237                 grep "Checking roundtrip encoding for SHIFT-JIS" &&
 238         git reset &&
 239
 240         # UTF-16 encoded files should not be round-trip checked by default...
 241         ! GIT_TRACE=1 git add roundtrip.utf16 2>&1 |
 242                 grep "Checking roundtrip encoding for UTF-16" &&
 243         git reset &&
 244
 245         # ... unless we tell Git to check it!
 246         GIT_TRACE=1 git -c core.checkRoundtripEncoding="UTF-16, UTF-32" \
 247                 add roundtrip.utf16 2>&1 |
 248                 grep "Checking roundtrip encoding for utf-16" &&
 249         git reset &&
 250
 251         # ... unless we tell Git to check it!
 252         # (here we also check that the casing of the encoding is irrelevant)
 253         GIT_TRACE=1 git -c core.checkRoundtripEncoding="UTF-32, utf-16" \
 254                 add roundtrip.utf16 2>&1 |
 255                 grep "Checking roundtrip encoding for utf-16" &&
 256         git reset
 257 '
 258
 259 test_done