| 1 | #!/bin/sh
|
|---|
| 2 |
|
|---|
| 3 | # SUSv3 compliant sed tests.
|
|---|
| 4 | # Copyright 2005 by Rob Landley <rob@landley.net>
|
|---|
| 5 | # Licensed under GPLv2, see file LICENSE in this source tree.
|
|---|
| 6 |
|
|---|
| 7 | . ./testing.sh
|
|---|
| 8 |
|
|---|
| 9 | # testing "description" "commands" "result" "infile" "stdin"
|
|---|
| 10 |
|
|---|
| 11 | # Corner cases
|
|---|
| 12 | testing "sed no files (stdin)" 'sed ""' "hello\n" "" "hello\n"
|
|---|
| 13 | testing "sed explicit stdin" 'sed "" -' "hello\n" "" "hello\n"
|
|---|
| 14 | testing "sed handles empty lines" "sed -e 's/\$/@/'" "@\n" "" "\n"
|
|---|
| 15 | testing "sed stdin twice" 'sed "" - -' "hello" "" "hello"
|
|---|
| 16 |
|
|---|
| 17 | # Trailing EOF.
|
|---|
| 18 | # Match $, at end of each file or all files?
|
|---|
| 19 |
|
|---|
| 20 | # -e corner cases
|
|---|
| 21 | # without -e
|
|---|
| 22 | # multiple -e
|
|---|
| 23 | # interact with a
|
|---|
| 24 | # -eee arg1 arg2 arg3
|
|---|
| 25 | # -f corner cases
|
|---|
| 26 | # -e -f -e
|
|---|
| 27 | # -n corner cases
|
|---|
| 28 | # no newline at EOF?
|
|---|
| 29 | # -r corner cases
|
|---|
| 30 | # Just make sure it works.
|
|---|
| 31 | # -i corner cases:
|
|---|
| 32 | # sed -i -
|
|---|
| 33 | # permissions
|
|---|
| 34 | # -i on a symlink
|
|---|
| 35 | # on a directory
|
|---|
| 36 | # With $ last-line test
|
|---|
| 37 | # Continue with \
|
|---|
| 38 | # End of script with trailing \
|
|---|
| 39 |
|
|---|
| 40 | # command list
|
|---|
| 41 | testing "sed accepts blanks before command" "sed -e '1 d'" "" "" ""
|
|---|
| 42 | testing "sed accepts newlines in -e" "sed -e 'i\
|
|---|
| 43 | 1
|
|---|
| 44 | a\
|
|---|
| 45 | 3'" "1\n2\n3\n" "" "2\n"
|
|---|
| 46 | testing "sed accepts multiple -e" "sed -e 'i\' -e '1' -e 'a\' -e '3'" \
|
|---|
| 47 | "1\n2\n3\n" "" "2\n"
|
|---|
| 48 |
|
|---|
| 49 | # substitutions
|
|---|
| 50 | testing "sed -n" "sed -n -e s/foo/bar/ -e s/bar/baz/" "" "" "foo\n"
|
|---|
| 51 | testing "sed with empty match" "sed 's/z*//g'" "string\n" "" "string\n"
|
|---|
| 52 | testing "sed s//p" "sed -e s/foo/bar/p -e s/bar/baz/p" "bar\nbaz\nbaz\n" \
|
|---|
| 53 | "" "foo\n"
|
|---|
| 54 | testing "sed -n s//p" "sed -ne s/abc/def/p" "def\n" "" "abc\n"
|
|---|
| 55 | testing "sed s//g (exhaustive)" "sed -e 's/[[:space:]]*/,/g'" ",1,2,3,4,5,\n" \
|
|---|
| 56 | "" "12345\n"
|
|---|
| 57 | testing "sed s arbitrary delimiter" "sed -e 's woo boing '" "boing\n" "" "woo\n"
|
|---|
| 58 | testing "sed s chains" "sed -e s/foo/bar/ -e s/bar/baz/" "baz\n" "" "foo\n"
|
|---|
| 59 | testing "sed s chains2" "sed -e s/foo/bar/ -e s/baz/nee/" "bar\n" "" "foo\n"
|
|---|
| 60 | testing "sed s [delimiter]" "sed -e 's@[@]@@'" "onetwo" "" "one@two"
|
|---|
| 61 | testing "sed s with \\t (GNU ext)" "sed 's/\t/ /'" "one two" "" "one\ttwo"
|
|---|
| 62 |
|
|---|
| 63 | # branch
|
|---|
| 64 | testing "sed b (branch)" "sed -e 'b one;p;: one'" "foo\n" "" "foo\n"
|
|---|
| 65 | testing "sed b (branch with no label jumps to end)" "sed -e 'b;p'" \
|
|---|
| 66 | "foo\n" "" "foo\n"
|
|---|
| 67 |
|
|---|
| 68 | # test and branch
|
|---|
| 69 | testing "sed t (test/branch)" "sed -e 's/a/1/;t one;p;: one;p'" \
|
|---|
| 70 | "1\n1\nb\nb\nb\nc\nc\nc\n" "" "a\nb\nc\n"
|
|---|
| 71 | testing "sed t (test/branch clears test bit)" "sed -e 's/a/b/;:loop;t loop'" \
|
|---|
| 72 | "b\nb\nc\n" "" "a\nb\nc\n"
|
|---|
| 73 | testing "sed T (!test/branch)" "sed -e 's/a/1/;T notone;p;: notone;p'" \
|
|---|
| 74 | "1\n1\n1\nb\nb\nc\nc\n" "" "a\nb\nc\n"
|
|---|
| 75 |
|
|---|
| 76 | test x"$SKIP_KNOWN_BUGS" = x"" && {
|
|---|
| 77 | # Normal sed end-of-script doesn't print "c" because n flushed the pattern
|
|---|
| 78 | # space. If n hits EOF, pattern space is empty when script ends.
|
|---|
| 79 | # Query: how does this interact with no newline at EOF?
|
|---|
| 80 | testing "sed n (flushes pattern space, terminates early)" "sed -e 'n;p'" \
|
|---|
| 81 | "a\nb\nb\nc\n" "" "a\nb\nc\n"
|
|---|
| 82 | }
|
|---|
| 83 | # non-GNU sed: N does _not_ flush pattern space, therefore c is eaten @ script end
|
|---|
| 84 | # GNU sed: N flushes pattern space, therefore c is printed too @ script end
|
|---|
| 85 | testing "sed N (flushes pattern space (GNU behavior))" "sed -e 'N;p'" \
|
|---|
| 86 | "a\nb\na\nb\nc\n" "" "a\nb\nc\n"
|
|---|
| 87 |
|
|---|
| 88 | testing "sed N test2" "sed ':a;N;s/\n/ /;ta'" \
|
|---|
| 89 | "a b c\n" "" "a\nb\nc\n"
|
|---|
| 90 |
|
|---|
| 91 | testing "sed N test3" "sed 'N;s/\n/ /'" \
|
|---|
| 92 | "a b\nc\n" "" "a\nb\nc\n"
|
|---|
| 93 |
|
|---|
| 94 | testing "sed address match newline" 'sed "/b/N;/b\\nc/i woo"' \
|
|---|
| 95 | "a\nwoo\nb\nc\nd\n" "" "a\nb\nc\nd\n"
|
|---|
| 96 |
|
|---|
| 97 | # Multiple lines in pattern space
|
|---|
| 98 | testing "sed N (stops at end of input) and P (prints to first newline only)" \
|
|---|
| 99 | "sed -n 'N;P;p'" "a\na\nb\n" "" "a\nb\nc\n"
|
|---|
| 100 |
|
|---|
| 101 | # Hold space
|
|---|
| 102 | testing "sed G (append hold space to pattern space)" 'sed G' "a\n\nb\n\nc\n\n" \
|
|---|
| 103 | "" "a\nb\nc\n"
|
|---|
| 104 | #testing "sed g/G (swap/append hold and patter space)"
|
|---|
| 105 | #testing "sed g (swap hold/pattern space)"
|
|---|
| 106 |
|
|---|
| 107 | testing "sed d ends script iteration" \
|
|---|
| 108 | "sed -e '/ook/d;s/ook/ping/p;i woot'" "" "" "ook\n"
|
|---|
| 109 | testing "sed d ends script iteration (2)" \
|
|---|
| 110 | "sed -e '/ook/d;a\' -e 'bang'" "woot\nbang\n" "" "ook\nwoot\n"
|
|---|
| 111 |
|
|---|
| 112 | # Multiple files, with varying newlines and NUL bytes
|
|---|
| 113 | test x"$SKIP_KNOWN_BUGS" = x"" && {
|
|---|
| 114 | testing "sed embedded NUL" "sed -e 's/woo/bang/'" "\0bang\0woo\0" "" \
|
|---|
| 115 | "\0woo\0woo\0"
|
|---|
| 116 | }
|
|---|
| 117 | testing "sed embedded NUL g" "sed -e 's/woo/bang/g'" "bang\0bang\0" "" \
|
|---|
| 118 | "woo\0woo\0"
|
|---|
| 119 | test x"$SKIP_KNOWN_BUGS" = x"" && {
|
|---|
| 120 | $ECHO -e "/woo/a he\0llo" > sed.commands
|
|---|
| 121 | testing "sed NUL in command" "sed -f sed.commands" "woo\nhe\0llo\n" "" "woo"
|
|---|
| 122 | rm sed.commands
|
|---|
| 123 | }
|
|---|
| 124 |
|
|---|
| 125 | # sed has funky behavior with newlines at the end of file. Test lots of
|
|---|
| 126 | # corner cases with the optional newline appending behavior.
|
|---|
| 127 |
|
|---|
| 128 | testing "sed normal newlines" "sed -e 's/woo/bang/' input -" "bang\nbang\n" \
|
|---|
| 129 | "woo\n" "woo\n"
|
|---|
| 130 | testing "sed leave off trailing newline" "sed -e 's/woo/bang/' input -" \
|
|---|
| 131 | "bang\nbang" "woo\n" "woo"
|
|---|
| 132 | testing "sed autoinsert newline" "sed -e 's/woo/bang/' input -" "bang\nbang" \
|
|---|
| 133 | "woo" "woo"
|
|---|
| 134 | testing "sed empty file plus cat" "sed -e 's/nohit//' input -" "one\ntwo" \
|
|---|
| 135 | "" "one\ntwo"
|
|---|
| 136 | testing "sed cat plus empty file" "sed -e 's/nohit//' input -" "one\ntwo" \
|
|---|
| 137 | "one\ntwo" ""
|
|---|
| 138 | test x"$SKIP_KNOWN_BUGS" = x"" && {
|
|---|
| 139 | testing "sed append autoinserts newline" "sed -e '/woot/a woo' -" \
|
|---|
| 140 | "woot\nwoo\n" "" "woot"
|
|---|
| 141 | }
|
|---|
| 142 | testing "sed insert doesn't autoinsert newline" "sed -e '/woot/i woo' -" \
|
|---|
| 143 | "woo\nwoot" "" "woot"
|
|---|
| 144 | testing "sed print autoinsert newlines" "sed -e 'p' -" "one\none" "" "one"
|
|---|
| 145 | testing "sed print autoinsert newlines two files" "sed -e 'p' input -" \
|
|---|
| 146 | "one\none\ntwo\ntwo" "one" "two"
|
|---|
| 147 | testing "sed noprint, no match, no newline" "sed -ne 's/woo/bang/' input" \
|
|---|
| 148 | "" "no\n" ""
|
|---|
| 149 | testing "sed selective matches with one nl" "sed -ne 's/woo/bang/p' input -" \
|
|---|
| 150 | "a bang\nc bang\n" "a woo\nb no" "c woo\nd no"
|
|---|
| 151 | testing "sed selective matches insert newline" \
|
|---|
| 152 | "sed -ne 's/woo/bang/p' input -" "a bang\nb bang\nd bang" \
|
|---|
| 153 | "a woo\nb woo" "c no\nd woo"
|
|---|
| 154 | testing "sed selective matches noinsert newline" \
|
|---|
| 155 | "sed -ne 's/woo/bang/p' input -" "a bang\nb bang" "a woo\nb woo" \
|
|---|
| 156 | "c no\nd no"
|
|---|
| 157 | test x"$SKIP_KNOWN_BUGS" = x"" && {
|
|---|
| 158 | testing "sed clusternewline" \
|
|---|
| 159 | "sed -e '/one/a 111' -e '/two/i 222' -e p input -" \
|
|---|
| 160 | "one\none\n111\n222\ntwo\ntwo" "one" "two"
|
|---|
| 161 | }
|
|---|
| 162 | testing "sed subst+write" \
|
|---|
| 163 | "sed -e 's/i/z/' -e 'woutputw' input -; $ECHO -n X; cat outputw" \
|
|---|
| 164 | "thzngy\nagaznXthzngy\nagazn" "thingy" "again"
|
|---|
| 165 | rm outputw
|
|---|
| 166 | testing "sed trailing NUL" \
|
|---|
| 167 | "sed 's/i/z/' input -" \
|
|---|
| 168 | "a\0b\0\nc" "a\0b\0" "c"
|
|---|
| 169 | testing "sed escaped newline in command" \
|
|---|
| 170 | "sed 's/a/z\\
|
|---|
| 171 | z/' input" \
|
|---|
| 172 | "z\nz" "a" ""
|
|---|
| 173 |
|
|---|
| 174 | # Test end-of-file matching behavior
|
|---|
| 175 |
|
|---|
| 176 | testing "sed match EOF" "sed -e '"'$p'"'" "hello\nthere\nthere" "" \
|
|---|
| 177 | "hello\nthere"
|
|---|
| 178 | testing "sed match EOF two files" "sed -e '"'$p'"' input -" \
|
|---|
| 179 | "one\ntwo\nthree\nfour\nfour" "one\ntwo" "three\nfour"
|
|---|
| 180 | # sed match EOF inline: gnu sed 4.1.5 outputs this:
|
|---|
| 181 | #00000000 6f 6e 65 0a 6f 6f 6b 0a 6f 6f 6b 0a 74 77 6f 0a |one.ook.ook.two.|
|
|---|
| 182 | #00000010 0a 74 68 72 65 65 0a 6f 6f 6b 0a 6f 6f 6b 0a 66 |.three.ook.ook.f|
|
|---|
| 183 | #00000020 6f 75 72 |our|
|
|---|
| 184 | # which looks buggy to me.
|
|---|
| 185 | $ECHO -ne "three\nfour" > input2
|
|---|
| 186 | testing "sed match EOF inline" \
|
|---|
| 187 | "sed -e '"'$i ook'"' -i input input2 && cat input input2" \
|
|---|
| 188 | "one\nook\ntwothree\nook\nfour" "one\ntwo" ""
|
|---|
| 189 | rm input2
|
|---|
| 190 |
|
|---|
| 191 | # Test lie-to-autoconf
|
|---|
| 192 |
|
|---|
| 193 | testing "sed lie-to-autoconf" "sed --version | grep -o 'GNU sed version '" \
|
|---|
| 194 | "GNU sed version \n" "" ""
|
|---|
| 195 |
|
|---|
| 196 | # Jump to nonexistent label
|
|---|
| 197 | test x"$SKIP_KNOWN_BUGS" = x"" && {
|
|---|
| 198 | # Incompatibility: illegal jump is not detected if input is ""
|
|---|
| 199 | # (that is, no lines at all). GNU sed 4.1.5 complains even in this case
|
|---|
| 200 | testing "sed nonexistent label" "sed -e 'b walrus' 2>/dev/null || echo yes" \
|
|---|
| 201 | "yes\n" "" ""
|
|---|
| 202 | }
|
|---|
| 203 |
|
|---|
| 204 | testing "sed backref from empty s uses range regex" \
|
|---|
| 205 | "sed -e '/woot/s//eep \0 eep/'" "eep woot eep" "" "woot"
|
|---|
| 206 |
|
|---|
| 207 | testing "sed backref from empty s uses range regex with newline" \
|
|---|
| 208 | "sed -e '/woot/s//eep \0 eep/'" "eep woot eep\n" "" "woot\n"
|
|---|
| 209 |
|
|---|
| 210 | # -i with no filename
|
|---|
| 211 |
|
|---|
| 212 | touch ./- # Detect gnu failure mode here.
|
|---|
| 213 | testing "sed -i with no arg [GNUFAIL]" "sed -e '' -i 2> /dev/null || echo yes" \
|
|---|
| 214 | "yes\n" "" ""
|
|---|
| 215 | rm ./- # Clean up
|
|---|
| 216 |
|
|---|
| 217 | testing "sed s/xxx/[/" "sed -e 's/xxx/[/'" "[\n" "" "xxx\n"
|
|---|
| 218 |
|
|---|
| 219 | # Ponder this a bit more, why "woo not found" from gnu version?
|
|---|
| 220 | #testing "sed doesn't substitute in deleted line" \
|
|---|
| 221 | # "sed -e '/ook/d;s/ook//;t woo;a bang;'" "bang" "" "ook\n"
|
|---|
| 222 |
|
|---|
| 223 | # This makes both seds very unhappy. Why?
|
|---|
| 224 | #testing "sed -g (exhaustive)" "sed -e 's/[[:space:]]*/,/g'" ",1,2,3,4,5," \
|
|---|
| 225 | # "" "12345"
|
|---|
| 226 |
|
|---|
| 227 | # testing "description" "commands" "result" "infile" "stdin"
|
|---|
| 228 |
|
|---|
| 229 | testing "sed n command must reset 'substituted' bit" \
|
|---|
| 230 | "sed 's/1/x/;T;n;: next;s/3/y/;t quit;n;b next;: quit;q'" \
|
|---|
| 231 | "0\nx\n2\ny\n" "" "0\n1\n2\n3\n"
|
|---|
| 232 |
|
|---|
| 233 | testing "sed d does not break n,m matching" \
|
|---|
| 234 | "sed -n '1d;1,3p'" \
|
|---|
| 235 | "second\nthird\n" "" "first\nsecond\nthird\nfourth\n"
|
|---|
| 236 |
|
|---|
| 237 | testing "sed d does not break n,regex matching" \
|
|---|
| 238 | "sed -n '1d;1,/hir/p'" \
|
|---|
| 239 | "second\nthird\n" "" "first\nsecond\nthird\nfourth\n"
|
|---|
| 240 |
|
|---|
| 241 | testing "sed d does not break n,regex matching #2" \
|
|---|
| 242 | "sed -n '1,5d;1,/hir/p'" \
|
|---|
| 243 | "second2\nthird2\n" "" \
|
|---|
| 244 | "first\nsecond\nthird\nfourth\n""first2\nsecond2\nthird2\nfourth2\n"
|
|---|
| 245 |
|
|---|
| 246 | testing "sed 2d;2,1p (gnu compat)" \
|
|---|
| 247 | "sed -n '2d;2,1p'" \
|
|---|
| 248 | "third\n" "" \
|
|---|
| 249 | "first\nsecond\nthird\nfourth\n"
|
|---|
| 250 |
|
|---|
| 251 | # Regex means: "match / at BOL or nothing, then one or more not-slashes".
|
|---|
| 252 | # The bug was that second slash in /usr/lib was treated as "at BOL" too.
|
|---|
| 253 | testing "sed beginning (^) matches only once" \
|
|---|
| 254 | "sed 's,\(^/\|\)[^/][^/]*,>\0<,g'" \
|
|---|
| 255 | ">/usr</>lib<\n" "" \
|
|---|
| 256 | "/usr/lib\n"
|
|---|
| 257 |
|
|---|
| 258 | testing "sed c" \
|
|---|
| 259 | "sed 'crepl'" \
|
|---|
| 260 | "repl\nrepl\n" "" \
|
|---|
| 261 | "first\nsecond\n"
|
|---|
| 262 |
|
|---|
| 263 | testing "sed nested {}s" \
|
|---|
| 264 | "sed '/asd/ { p; /s/ { s/s/c/ }; p; q }'" \
|
|---|
| 265 | "qwe\nasd\nacd\nacd\n" "" \
|
|---|
| 266 | "qwe\nasd\nzxc\n"
|
|---|
| 267 |
|
|---|
| 268 | testing "sed a cmd ended by double backslash" \
|
|---|
| 269 | "sed -e '/| one /a \\
|
|---|
| 270 | | three \\\\' -e '/| one-/a \\
|
|---|
| 271 | | three-* \\\\'" \
|
|---|
| 272 | ' | one \\
|
|---|
| 273 | | three \\
|
|---|
| 274 | | two \\
|
|---|
| 275 | ' '' \
|
|---|
| 276 | ' | one \\
|
|---|
| 277 | | two \\
|
|---|
| 278 | '
|
|---|
| 279 |
|
|---|
| 280 | # first three lines are deleted; 4th line is matched and printed by "2,3" and by "4" ranges
|
|---|
| 281 | testing "sed with N skipping lines past ranges on next cmds" \
|
|---|
| 282 | "sed -n '1{N;N;d};1p;2,3p;3p;4p'" \
|
|---|
| 283 | "4\n4\n" "" "1\n2\n3\n4\n"
|
|---|
| 284 |
|
|---|
| 285 | testing "sed -i with address modifies all files, not only first" \
|
|---|
| 286 | "cp input input2; sed -i -e '1s/foo/bar/' input input2 && cat input input2; rm input2" \
|
|---|
| 287 | "bar\nbar\n" "foo\n" ""
|
|---|
| 288 |
|
|---|
| 289 | testing "sed understands \r" \
|
|---|
| 290 | "sed 's/r/\r/'" \
|
|---|
| 291 | "\rrr\n" "" "rrr\n"
|
|---|
| 292 |
|
|---|
| 293 | testing "sed -i finishes ranges correctly" \
|
|---|
| 294 | "sed '1,2d' -i input; echo \$?; cat input" \
|
|---|
| 295 | "0\n3\n4\n" "1\n2\n3\n4\n" ""
|
|---|
| 296 |
|
|---|
| 297 | testing "sed zero chars match/replace advances correctly 1" \
|
|---|
| 298 | "sed 's/l*/@/g'" \
|
|---|
| 299 | "@h@e@o@\n" "" "helllo\n"
|
|---|
| 300 |
|
|---|
| 301 | testing "sed zero chars match/replace advances correctly 2" \
|
|---|
| 302 | "sed 's [^ .]* x g'" \
|
|---|
| 303 | "x x.x\n" "" " a.b\n"
|
|---|
| 304 |
|
|---|
| 305 | testing "sed zero chars match/replace logic must not falsely trigger here 1" \
|
|---|
| 306 | "sed 's/a/A/g'" \
|
|---|
| 307 | "_AAA1AA\n" "" "_aaa1aa\n"
|
|---|
| 308 |
|
|---|
| 309 | testing "sed zero chars match/replace logic must not falsely trigger here 2" \
|
|---|
| 310 | "sed 's/ *$/_/g'" \
|
|---|
| 311 | "qwerty_\n" "" "qwerty\n"
|
|---|
| 312 |
|
|---|
| 313 | # testing "description" "commands" "result" "infile" "stdin"
|
|---|
| 314 |
|
|---|
| 315 | exit $FAILCOUNT
|
|---|
