path: root/testsuite/sed.tests (plain)
blob: 05c00a99b8a441a7a6a5e24637743bdcba9df757

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	testing "sed n (flushes pattern space, terminates early)" "sed -e 'n;p'" \
77	"a\nb\nb\nc\n" "" "a\nb\nc\n"
78
79	# non-GNU sed: N does _not_ flush pattern space, therefore c is eaten @ script end
80	# GNU sed: N flushes pattern space, therefore c is printed too @ script end
81	testing "sed N (flushes pattern space (GNU behavior))" "sed -e 'N;p'" \
82	"a\nb\na\nb\nc\n" "" "a\nb\nc\n"
83
84	testing "sed N test2" "sed ':a;N;s/\n/ /;ta'" \
85	"a b c\n" "" "a\nb\nc\n"
86
87	testing "sed N test3" "sed 'N;s/\n/ /'" \
88	"a b\nc\n" "" "a\nb\nc\n"
89
90	testing "sed address match newline" 'sed "/b/N;/b\\nc/i woo"' \
91	"a\nwoo\nb\nc\nd\n" "" "a\nb\nc\nd\n"
92
93	# Multiple lines in pattern space
94	testing "sed N (stops at end of input) and P (prints to first newline only)" \
95	"sed -n 'N;P;p'" "a\na\nb\n" "" "a\nb\nc\n"
96
97	# Hold space
98	testing "sed G (append hold space to pattern space)" 'sed G' "a\n\nb\n\nc\n\n" \
99	"" "a\nb\nc\n"
100	#testing "sed g/G (swap/append hold and patter space)"
101	#testing "sed g (swap hold/pattern space)"
102
103	testing "sed d ends script iteration" \
104	"sed -e '/ook/d;s/ook/ping/p;i woot'" "" "" "ook\n"
105	testing "sed d ends script iteration (2)" \
106	"sed -e '/ook/d;a\' -e 'bang'" "woot\nbang\n" "" "ook\nwoot\n"
107
108	# Multiple files, with varying newlines and NUL bytes
109	test x"$SKIP_KNOWN_BUGS" = x"" && {
110	testing "sed embedded NUL" "sed -e 's/woo/bang/'" "\0bang\0woo\0" "" \
111	"\0woo\0woo\0"
112	}
113	testing "sed embedded NUL g" "sed -e 's/woo/bang/g'" "bang\0bang\0" "" \
114	"woo\0woo\0"
115	test x"$SKIP_KNOWN_BUGS" = x"" && {
116	$ECHO -e "/woo/a he\0llo" > sed.commands
117	testing "sed NUL in command" "sed -f sed.commands" "woo\nhe\0llo\n" "" "woo"
118	rm sed.commands
119	}
120
121	# sed has funky behavior with newlines at the end of file. Test lots of
122	# corner cases with the optional newline appending behavior.
123
124	testing "sed normal newlines" "sed -e 's/woo/bang/' input -" "bang\nbang\n" \
125	"woo\n" "woo\n"
126	testing "sed leave off trailing newline" "sed -e 's/woo/bang/' input -" \
127	"bang\nbang" "woo\n" "woo"
128	testing "sed autoinsert newline" "sed -e 's/woo/bang/' input -" "bang\nbang" \
129	"woo" "woo"
130	testing "sed empty file plus cat" "sed -e 's/nohit//' input -" "one\ntwo" \
131	"" "one\ntwo"
132	testing "sed cat plus empty file" "sed -e 's/nohit//' input -" "one\ntwo" \
133	"one\ntwo" ""
134	testing "sed append autoinserts newline" "sed -e '/woot/a woo' -" \
135	"woot\nwoo\n" "" "woot"
136	testing "sed append autoinserts newline 2" "sed -e '/oot/a woo' - input" \
137	"woot\nwoo\nboot\nwoo\n" "boot" "woot"
138	testing "sed append autoinserts newline 3" "sed -e '/oot/a woo' -i input && cat input" \
139	"boot\nwoo\n" "boot" ""
140	testing "sed insert doesn't autoinsert newline" "sed -e '/woot/i woo' -" \
141	"woo\nwoot" "" "woot"
142	testing "sed print autoinsert newlines" "sed -e 'p' -" "one\none" "" "one"
143	testing "sed print autoinsert newlines two files" "sed -e 'p' input -" \
144	"one\none\ntwo\ntwo" "one" "two"
145	testing "sed noprint, no match, no newline" "sed -ne 's/woo/bang/' input" \
146	"" "no\n" ""
147	testing "sed selective matches with one nl" "sed -ne 's/woo/bang/p' input -" \
148	"a bang\nc bang\n" "a woo\nb no" "c woo\nd no"
149	testing "sed selective matches insert newline" \
150	"sed -ne 's/woo/bang/p' input -" "a bang\nb bang\nd bang" \
151	"a woo\nb woo" "c no\nd woo"
152	testing "sed selective matches noinsert newline" \
153	"sed -ne 's/woo/bang/p' input -" "a bang\nb bang" "a woo\nb woo" \
154	"c no\nd no"
155	testing "sed clusternewline" \
156	"sed -e '/one/a 111' -e '/two/i 222' -e p input -" \
157	"one\none\n111\n222\ntwo\ntwo" "one" "two"
158	testing "sed subst+write" \
159	"sed -e 's/i/z/' -e 'woutputw' input -; $ECHO -n X; cat outputw" \
160	"thzngy\nagaznXthzngy\nagazn" "thingy" "again"
161	rm outputw
162	testing "sed trailing NUL" \
163	"sed 's/i/z/' input -" \
164	"a\0b\0\nc" "a\0b\0" "c"
165	testing "sed escaped newline in command" \
166	"sed 's/a/z\\
167	z/' input" \
168	"z\nz" "a" ""
169
170	# Test end-of-file matching behavior
171
172	testing "sed match EOF" "sed -e '"'$p'"'" "hello\nthere\nthere" "" \
173	"hello\nthere"
174	testing "sed match EOF two files" "sed -e '"'$p'"' input -" \
175	"one\ntwo\nthree\nfour\nfour" "one\ntwo" "three\nfour"
176	# sed match EOF inline: gnu sed 4.1.5 outputs this:
177	#00000000 6f 6e 65 0a 6f 6f 6b 0a 6f 6f 6b 0a 74 77 6f 0a |one.ook.ook.two.|
178	#00000010 0a 74 68 72 65 65 0a 6f 6f 6b 0a 6f 6f 6b 0a 66 |.three.ook.ook.f|
179	#00000020 6f 75 72 |our|
180	# which looks buggy to me.
181	$ECHO -ne "three\nfour" > input2
182	testing "sed match EOF inline" \
183	"sed -e '"'$i ook'"' -i input input2 && cat input input2" \
184	"one\nook\ntwothree\nook\nfour" "one\ntwo" ""
185	rm input2
186
187	# Test lie-to-autoconf
188
189	testing "sed lie-to-autoconf" "sed --version | grep -o 'GNU sed version '" \
190	"GNU sed version \n" "" ""
191
192	# Jump to nonexistent label
193	test x"$SKIP_KNOWN_BUGS" = x"" && {
194	# Incompatibility: illegal jump is not detected if input is ""
195	# (that is, no lines at all). GNU sed 4.1.5 complains even in this case
196	testing "sed nonexistent label" "sed -e 'b walrus' 2>/dev/null || echo yes" \
197	"yes\n" "" ""
198	}
199
200	testing "sed backref from empty s uses range regex" \
201	"sed -e '/woot/s//eep \0 eep/'" "eep woot eep" "" "woot"
202
203	testing "sed backref from empty s uses range regex with newline" \
204	"sed -e '/woot/s//eep \0 eep/'" "eep woot eep\n" "" "woot\n"
205
206	# -i with no filename
207
208	touch ./- # Detect gnu failure mode here.
209	testing "sed -i with no arg [GNUFAIL]" "sed -e '' -i 2> /dev/null || echo yes" \
210	"yes\n" "" ""
211	rm ./- # Clean up
212
213	testing "sed s/xxx/[/" "sed -e 's/xxx/[/'" "[\n" "" "xxx\n"
214
215	# Ponder this a bit more, why "woo not found" from gnu version?
216	#testing "sed doesn't substitute in deleted line" \
217	# "sed -e '/ook/d;s/ook//;t woo;a bang;'" "bang" "" "ook\n"
218
219	# This makes both seds very unhappy. Why?
220	#testing "sed -g (exhaustive)" "sed -e 's/[[:space:]]*/,/g'" ",1,2,3,4,5," \
221	# "" "12345"
222
223	# testing "description" "commands" "result" "infile" "stdin"
224
225	testing "sed n command must reset 'substituted' bit" \
226	"sed 's/1/x/;T;n;: next;s/3/y/;t quit;n;b next;: quit;q'" \
227	"0\nx\n2\ny\n" "" "0\n1\n2\n3\n"
228
229	testing "sed d does not break n,m matching" \
230	"sed -n '1d;1,3p'" \
231	"second\nthird\n" "" "first\nsecond\nthird\nfourth\n"
232
233	testing "sed d does not break n,regex matching" \
234	"sed -n '1d;1,/hir/p'" \
235	"second\nthird\n" "" "first\nsecond\nthird\nfourth\n"
236
237	testing "sed d does not break n,regex matching #2" \
238	"sed -n '1,5d;1,/hir/p'" \
239	"second2\nthird2\n" "" \
240	"first\nsecond\nthird\nfourth\n""first2\nsecond2\nthird2\nfourth2\n"
241
242	testing "sed 2d;2,1p (gnu compat)" \
243	"sed -n '2d;2,1p'" \
244	"third\n" "" \
245	"first\nsecond\nthird\nfourth\n"
246
247	# Regex means: "match / at BOL or nothing, then one or more not-slashes".
248	# The bug was that second slash in /usr/lib was treated as "at BOL" too.
249	testing "sed beginning (^) matches only once" \
250	"sed 's,\(^/\|\)[^/][^/]*,>\0<,g'" \
251	">/usr</>lib<\n" "" \
252	"/usr/lib\n"
253
254	testing "sed c" \
255	"sed 'crepl'" \
256	"repl\nrepl\n" "" \
257	"first\nsecond\n"
258
259	testing "sed nested {}s" \
260	"sed '/asd/ { p; /s/ { s/s/c/ }; p; q }'" \
261	"qwe\nasd\nacd\nacd\n" "" \
262	"qwe\nasd\nzxc\n"
263
264	testing "sed a cmd ended by double backslash" \
265	"sed -e '/| one /a \\
266	| three \\\\' -e '/| one-/a \\
267	| three-* \\\\'" \
268	' | one \\
269	| three \\
270	| two \\
271	' '' \
272	' | one \\
273	| two \\
274	'
275
276	testing "sed a cmd understands \\n,\\t,\\r" \
277	"sed '/1/a\\\\t\\rzero\\none\\\\ntwo\\\\\\nthree'" \
278	"\
279	line1
280	\t\rzero
281	one\\\\ntwo\\
282	three
283	" "" "line1\n"
284
285	testing "sed i cmd understands \\n,\\t,\\r" \
286	"sed '/1/i\\\\t\\rzero\\none\\\\ntwo\\\\\\nthree'" \
287	"\
288	\t\rzero
289	one\\\\ntwo\\
290	three
291	line1
292	" "" "line1\n"
293
294	# first three lines are deleted; 4th line is matched and printed by "2,3" and by "4" ranges
295	testing "sed with N skipping lines past ranges on next cmds" \
296	"sed -n '1{N;N;d};1p;2,3p;3p;4p'" \
297	"4\n4\n" "" "1\n2\n3\n4\n"
298
299	testing "sed -i with address modifies all files, not only first" \
300	"cp input input2; sed -i -e '1s/foo/bar/' input input2 && cat input input2; rm input2" \
301	"bar\nbar\n" "foo\n" ""
302
303	testing "sed understands \r" \
304	"sed 's/r/\r/'" \
305	"\rrr\n" "" "rrr\n"
306
307	testing "sed -i finishes ranges correctly" \
308	"sed '1,2d' -i input; echo \$?; cat input" \
309	"0\n3\n4\n" "1\n2\n3\n4\n" ""
310
311	testing "sed zero chars match/replace advances correctly 1" \
312	"sed 's/l*/@/g'" \
313	"@h@e@o@\n" "" "helllo\n"
314
315	testing "sed zero chars match/replace advances correctly 2" \
316	"sed 's [^ .]* x g'" \
317	"x x.x\n" "" " a.b\n"
318
319	testing "sed zero chars match/replace logic must not falsely trigger here 1" \
320	"sed 's/a/A/g'" \
321	"_AAA1AA\n" "" "_aaa1aa\n"
322
323	testing "sed zero chars match/replace logic must not falsely trigger here 2" \
324	"sed 's/ *$/_/g'" \
325	"qwerty_\n" "" "qwerty\n"
326
327	testing "sed /\$_in_regex/ should not match newlines, only end-of-line" \
328	"sed ': testcont; /\\\\$/{ =; N; b testcont }'" \
329	"\
330	this is a regular line
331	2
332	line with \\
333	continuation
334	more regular lines
335	5
336	line with \\
337	continuation
338	" \
339	"" "\
340	this is a regular line
341	line with \\
342	continuation
343	more regular lines
344	line with \\
345	continuation
346	"
347
348	testing "sed s///NUM test" \
349	"sed -e 's/a/b/2; s/a/c/g'" \
350	"cb\n" "" "aa\n"
351
352	testing "sed /regex/,N{...} addresses work" \
353	"sed /^2/,2{d}" \
354	"1\n3\n4\n5\n" \
355	"" \
356	"1\n2\n3\n4\n5\n"
357
358	testing "sed /regex/,+N{...} addresses work" \
359	"sed /^2/,+2{d}" \
360	"1\n5\n" \
361	"" \
362	"1\n2\n3\n4\n5\n"
363
364	testing "sed /regex/,+N{...} -i works" \
365	"cat - >input2; sed /^4/,+2{d} -i input input2; echo \$?; cat input input2; rm input2" \
366	"0\n""1\n2\n3\n7\n8\n""1\n2\n7\n8\n" \
367	"1\n2\n3\n4\n5\n6\n7\n8\n" \
368	"1\n2\n4\n5\n6\n7\n8\n" \
369
370	# GNU sed 4.2.1 would also accept "/^4/,+{d}" with the same meaning, we don't
371	testing "sed /regex/,+0{...} -i works" \
372	"cat - >input2; sed /^4/,+0{d} -i input input2; echo \$?; cat input input2; rm input2" \
373	"0\n""1\n2\n3\n5\n6\n7\n8\n""1\n2\n5\n6\n7\n8\n" \
374	"1\n2\n3\n4\n5\n6\n7\n8\n" \
375	"1\n2\n4\n5\n6\n7\n8\n" \
376
377	# GNU sed 4.2.1 would also accept "/^4/,+d" with the same meaning, we don't
378	testing "sed /regex/,+0<cmd> -i works" \
379	"cat - >input2; sed /^4/,+0d -i input input2; echo \$?; cat input input2; rm input2" \
380	"0\n""1\n2\n3\n5\n6\n7\n8\n""1\n2\n5\n6\n7\n8\n" \
381	"1\n2\n3\n4\n5\n6\n7\n8\n" \
382	"1\n2\n4\n5\n6\n7\n8\n" \
383
384	testing "sed 's///w FILE'" \
385	"sed 's/qwe/ZZZ/wz'; cat z; rm z" \
386	"123\nZZZ\nasd\n""ZZZ\n" \
387	"" \
388	"123\nqwe\nasd\n"
389
390	# testing "description" "commands" "result" "infile" "stdin"
391
392	exit $FAILCOUNT
393