xref: /openbsd/regress/usr.bin/sed/sedtest.sh (revision 09467b48) 
1#!/bin/sh -
2#	$OpenBSD: sedtest.sh,v 1.8 2018/12/07 15:30:31 schwarze Exp $
3#
4# Copyright (c) 1992 Diomidis Spinellis.
5# Copyright (c) 1992, 1993
6#	The Regents of the University of California.  All rights reserved.
7#
8# Redistribution and use in source and binary forms, with or without
9# modification, are permitted provided that the following conditions
10# are met:
11# 1. Redistributions of source code must retain the above copyright
12#    notice, this list of conditions and the following disclaimer.
13# 2. Redistributions in binary form must reproduce the above copyright
14#    notice, this list of conditions and the following disclaimer in the
15#    documentation and/or other materials provided with the distribution.
16# 3. Neither the name of the University nor the names of its contributors
17#    may be used to endorse or promote products derived from this software
18#    without specific prior written permission.
19#
20# THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
21# ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
22# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
23# ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
24# FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
25# DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
26# OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
27# HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
28# LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
29# OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
30# SUCH DAMAGE.
31#
32#	from: @(#)sed.test	8.1 (Berkeley) 6/6/93
33#
34
35# sed Regression Tests
36#
37# The following files are created:
38# lines[1-4], script1, script2
39# Test results are stored in sed.out
40
41main()
42{
43	TEST=${1-sed}
44	TESTLOG=${2-sed.out}
45	DICT=${3-/usr/share/dict/words}
46
47	ulimit -n 256
48	rm -f lines1 lines2
49	i=1
50	while [ $i -lt 15 ]; do
51		echo "l1_$i" >> lines1
52		if [ $i -lt 10 ]; then
53			echo "l2_$i" >> lines2
54		fi
55		i=$((i + 1))
56	done
57
58	tests $TEST $TESTLOG
59}
60
61tests()
62{
63	SED=$1
64	LOG=$2
65	rm -f $LOG
66	MARK=100
67
68	exec 3>&0 4>&1 5>&2
69	exec 0</dev/null 1>/dev/null 2>/dev/null
70	test_error
71	exec 0>&3 1>&4 2>&5
72
73	exec 4>&1 5>&2
74	test_args
75	test_addr
76	echo Testing commands
77	test_group
78	test_acid
79	test_branch
80	test_pattern
81	test_print
82	test_subst
83	exec 1>&4 2>&5
84}
85
86mark()
87{
88	exec 2>&1 >>$LOG
89	test $MARK -ne 100 && echo ""
90	MARK=$((MARK + 1))
91	echo "Test $1:$MARK" | sed 's/./=/g'
92	echo "Test $1:$MARK"
93	echo "Test $1:$MARK" | sed 's/./=/g'
94}
95
96test_args()
97{
98	mark '1.1'
99	echo Testing argument parsing
100	echo First type
101	$SED 's/^/e1_/p' lines1
102	mark '1.2' ; $SED -n 's/^/e1_/p' lines1
103	mark '1.3'
104	$SED 's/^/e1_/p' <lines1
105	mark '1.4' ; $SED -n 's/^/e1_/p' <lines1
106	echo Second type
107	mark '1.4.1'
108	$SED -e '' <lines1
109	echo 's/^/s1_/p' >script1
110	echo 's/^/s2_/p' >script2
111	mark '1.5'
112	$SED -f script1 lines1
113	mark '1.6'
114	$SED -f script1 <lines1
115	mark '1.7'
116	$SED -e 's/^/e1_/p' lines1
117	mark '1.8'
118	$SED -e 's/^/e1_/p' <lines1
119	mark '1.9' ; $SED -n -f script1 lines1
120	mark '1.10' ; $SED -n -f script1 <lines1
121	mark '1.11' ; $SED -n -e 's/^/e1_/p' lines1
122	mark '1.12'
123	$SED -n -e 's/^/e1_/p' <lines1
124	mark '1.13'
125	$SED -e 's/^/e1_/p' -e 's/^/e2_/p' lines1
126	mark '1.14'
127	$SED -f script1 -f script2 lines1
128	mark '1.15'
129	$SED -e 's/^/e1_/p' -f script1 lines1
130	mark '1.16'
131	$SED -e 's/^/e1_/p' lines1 lines1
132	# POSIX D11.2:11251
133	mark '1.17' ; $SED p <lines1 lines1
134cat >script1 <<EOF
135#n
136# A comment
137
138p
139EOF
140	mark '1.18' ; $SED -f script1 <lines1 lines1
141}
142
143test_addr()
144{
145	echo Testing address ranges
146	mark '2.1' ; $SED -n -e '4p' lines1
147	mark '2.2' ; $SED -n -e '20p' lines1 lines2
148	mark '2.3' ; $SED -n -e '$p' lines1
149	mark '2.4' ; $SED -n -e '$p' lines1 lines2
150	mark '2.5' ; $SED -n -e '$a\
151hello' /dev/null
152	mark '2.6' ; $SED -n -e '$p' lines1 /dev/null lines2
153	# Should not print anything
154	mark '2.7' ; $SED -n -e '20p' lines1
155	mark '2.8' ; $SED -n -e '0p' lines1
156	mark '2.9' ; $SED -n '/l1_7/p' lines1
157	mark '2.10' ; $SED -n ' /l1_7/ p' lines1
158	mark '2.11'
159	$SED -n '\_l1\_7_p' lines1
160	mark '2.12' ; $SED -n '1,4p' lines1
161	mark '2.13' ; $SED -n '1,$p' lines1 lines2
162	mark '2.14' ; $SED -n '1,/l2_9/p' lines1 lines2
163	mark '2.15' ; $SED -n '/4/,$p' lines1 lines2
164	mark '2.16' ; $SED -n '/4/,20p' lines1 lines2
165	mark '2.17' ; $SED -n '/4/,/10/p' lines1 lines2
166	mark '2.18' ; $SED -n '/l2_3/,/l1_8/p' lines1 lines2
167	mark '2.19'
168	$SED -n '12,3p' lines1 lines2
169	mark '2.20'
170	$SED -n '/l1_7/,3p' lines1 lines2
171}
172
173test_group()
174{
175	echo Brace and other grouping
176	mark '3.1' ; $SED -e '
1774,12 {
178	s/^/^/
179	s/$/$/
180	s/_/T/
181}' lines1
182	mark '3.2' ; $SED -e '
1834,12 {
184	s/^/^/
185	/6/,/10/ {
186		s/$/$/
187		/8/ s/_/T/
188	}
189}' lines1
190	mark '3.3' ; $SED -e '
1914,12 !{
192	s/^/^/
193	/6/,/10/ !{
194		s/$/$/
195		/8/ !s/_/T/
196	}
197}' lines1
198	mark '3.4' ; $SED -e '4,12!s/^/^/' lines1
199}
200
201test_acid()
202{
203	echo Testing a c d and i commands
204	mark '4.1' ; $SED -n -e '
205s/^/before_i/p
20620i\
207inserted
208s/^/after_i/p
209' lines1 lines2
210	mark '4.2' ; $SED -n -e '
2115,12s/^/5-12/
212s/^/before_a/p
213/5-12/a\
214appended
215s/^/after_a/p
216' lines1 lines2
217	mark '4.3'
218	$SED -n -e '
219s/^/^/p
220/l1_/a\
221appended
2228,10N
223s/$/$/p
224' lines1 lines2
225	mark '4.4' ; $SED -n -e '
226c\
227hello
228' lines1
229	mark '4.5' ; $SED -n -e '
2308c\
231hello
232' lines1
233	mark '4.6' ; $SED -n -e '
2343,14c\
235hello
236' lines1
237# SunOS and GNU sed behave differently.   We follow POSIX
238#	mark '4.7' ; $SED -n -e '
239#8,3c\
240#hello
241#' lines1
242	mark '4.8' ; $SED d <lines1
243}
244
245test_branch()
246{
247	echo Testing labels and branching
248	mark '5.1' ; $SED -n -e '
249b label4
250:label3
251s/^/label3_/p
252b end
253:label4
2542,12b label1
255b label2
256:label1
257s/^/label1_/p
258b
259:label2
260s/^/label2_/p
261b label3
262:end
263' lines1
264	mark '5.2'
265	$SED -n -e '
266s/l1_/l2_/
267t ok
268b
269:ok
270s/^/tested /p
271' lines1 lines2
272# SunOS sed behaves differently here.  Clarification needed.
273#	mark '5.3' ; $SED -n -e '
274#5,8b inside
275#1,5 {
276#	s/^/^/p
277#	:inside
278#	s/$/$/p
279#}
280#' lines1
281# Check that t clears the substitution done flag
282	mark '5.4' ; $SED -n -e '
2831,8s/^/^/
284t l1
285:l1
286t l2
287s/$/$/p
288b
289:l2
290s/^/ERROR/
291' lines1
292# Check that reading a line clears the substitution done flag
293	mark '5.5'
294	$SED -n -e '
295t l2
2961,8s/^/^/p
2972,7N
298b
299:l2
300s/^/ERROR/p
301' lines1
302	mark '5.6' ; $SED 5q lines1
303	mark '5.7' ; $SED -e '
3045i\
305hello
3065q' lines1
307# Branch across block boundary
308	mark '5.8' ; $SED -e '
309{
310:b
311}
312s/l/m/
313tb' lines1
314# Check that branch commands can be followed by a semicolon and another
315# command.  POSIX does not require that this works, but GNU sed allows it.
316# Since POSIX does not require that label names can contain semicolons,
317# the extension is viable.
318	mark '5.9' ; echo "A\nB" | $SED '1b;='
319	mark '5.10' ; echo "A\nB" | $SED 's/A/C/;t;='
320	mark '5.11' ; echo "A\nB" | $SED '1bL;=;:L'
321}
322
323test_pattern()
324{
325echo Pattern space commands
326# Check that the pattern space is deleted
327	mark '6.1' ; $SED -n -e '
328c\
329changed
330p
331' lines1
332	mark '6.2' ; $SED -n -e '
3334d
334p
335' lines1
336# SunOS sed refused to print here
337#	mark '6.3' ; $SED -e '
338#N
339#N
340#N
341#D
342#P
343#4p
344#' lines1
345	mark '6.4' ; $SED -e '
3462h
3473H
3484g
3495G
3506x
3516p
3526x
3536p
354' lines1
355	mark '6.5' ; $SED -e '4n' lines1
356	mark '6.6' ; $SED -n -e '4n' lines1
357}
358
359test_print()
360{
361	echo Testing print and file routines
362	awk 'END {for (i = 1; i < 256; i++) printf("%c", i);print "\n"}' \
363		</dev/null >lines3
364	# GNU and SunOS sed behave differently here
365	mark '7.1'
366	$SED -n l lines3
367	mark '7.2' ; $SED -e '/l2_/=' lines1 lines2
368	rm -f lines4
369	mark '7.3' ; $SED -e '3,12w lines4' lines1
370	echo w results
371	cat lines4
372	mark '7.4' ; $SED -e '4r lines2' lines1
373	mark '7.5' ; $SED -e '5r /dev/dds' lines1
374	mark '7.6' ; $SED -e '6r /dev/null' lines1
375	mark '7.7'
376	sed '200q' $DICT | sed 's$.*$s/^/&/w tmpdir/&$' >script1
377	rm -rf tmpdir
378	mkdir tmpdir
379	$SED -f script1 lines1
380	cat tmpdir/*
381	rm -rf tmpdir
382	mark '7.8'
383	echo line1 > lines3
384	echo "" >> lines3
385	$SED -n -e '$p' lines3 /dev/null
386}
387
388test_subst()
389{
390	echo Testing substitution commands
391	mark '8.1' ; $SED -e 's/./X/g' lines1
392	mark '8.2' ; $SED -e 's,.,X,g' lines1
393# GNU and SunOS sed thinks we are escaping . as wildcard, not as separator
394#	mark '8.3' ; $SED -e 's.\..X.g' lines1
395# POSIX does not say that this should work
396#	mark '8.4' ; $SED -e 's/[/]/Q/' lines1
397	mark '8.4' ; $SED -e 's/[\/]/Q/' lines1
398	mark '8.5' ; $SED -e 's_\__X_' lines1
399	mark '8.6' ; $SED -e 's/./(&)/g' lines1
400	mark '8.7' ; $SED -e 's/./(\&)/g' lines1
401	mark '8.8' ; $SED -e 's/\(.\)\(.\)\(.\)/x\3x\2x\1/g' lines1
402	mark '8.9' ; $SED -e 's/_/u0\
403u1\
404u2/g' lines1
405	mark '8.10'
406	$SED -e 's/./X/4' lines1
407	rm -f lines4
408	mark '8.11' ; $SED -e 's/1/X/w lines4' lines1
409	echo s wfile results
410	cat lines4
411	mark '8.12' ; $SED -e 's/[123]/X/g' lines1
412	mark '8.13' ; $SED -e 'y/0123456789/9876543210/' lines1
413	mark '8.14' ;
414	$SED -e 'y10\123456789198765432\101' lines1
415	mark '8.15' ; $SED -e '1N;2y/\n/X/' lines1
416	mark '8.16'
417	echo 'eeefff' | $SED -e 'p' -e 's/e/X/p' -e ':x' -e 's//Y/p' -e '/f/bx'
418	echo 'various special characters as delimiters'
419# POSIX says "Within the BRE and the replacement, the BRE delimiter itself
420# can be used as a literal character if it is preceded by a <backslash>".
421# That is an ambiguous statement.  We interpret it in the sense that the
422# character is passed on literally to the RE engine, without the backslash.
423	mark '8.17' ; $SED -e 's[_[X[' lines1
424	mark '8.18' ; $SED -e 's$1\$$\$R$' lines1
425	mark '8.19' ; $SED -Ee 's(\(3|5)(\(\1)(' lines1
426	mark '8.20' ; $SED -e 's*_1\**\*R*' lines1
427	mark '8.21' ; $SED -Ee 's+_1\++\+R+' lines1
428	mark '8.22' ; $SED -e 's.1\..R\..g' lines1
429	mark '8.23' ; sed 's/_/\//' lines1 | $SED -e 's/\/1/R/'
430	mark '8.24' ; $SED -Ee 's?_1\??\?R?' lines1
431	mark '8.25' ; $SED -e 's[\[2-4][\[R][' lines1
432	mark '8.26' ; $SED -e 's^\^l^R\^^' lines1
433	mark '8.27' ; $SED -Ee 's{1\{2}{\{R}{' lines1
434	mark '8.28' ; $SED -Ee 's|2\|4|\|R|' lines1
435	echo '\ in y command'
436	mark '8.29'
437	printf 'a\\b(c\n' |
438	$SED 'y%ABCDEFGHIJKLMNOPQRSTUVWXYZ, /\\()"%abcdefghijklmnopqrstuvwxyz,------%'
439	mark '8.30'
440# Test if an unmatched line is only printed once.
441	printf 'z\n' | $SED -e 's/^a*/b/2p'
442}
443
444test_error()
445{
446	set -x
447	$SED -x && exit 1
448	$SED -f && exit 1
449	$SED -e && exit 1
450	$SED -f /dev/dds && exit 1
451	$SED p /dev/dds && exit 1
452	$SED -f /bin/sh && exit 1
453	$SED '{' && exit 1
454	$SED '{' && exit 1
455	$SED '/hello/' && exit 1
456	$SED '1,/hello/' && exit 1
457	$SED -e '-5p' && exit 1
458	$SED '/jj' && exit 1
459	$SED 'a hello' && exit 1
460	$SED 'a \ hello' && exit 1
461	$SED 'b foo' && exit 1
462	$SED 'd hello' && exit 1
463	$SED 's/aa' && exit 1
464	$SED 's/aa/' && exit 1
465	$SED 's/a/b' && exit 1
466	$SED 's/a/b/c/d' && exit 1
467	$SED 's/a/b/ 1 2' && exit 1
468	$SED 's/a/b/ 1 g' && exit 1
469	$SED 's/a/b/w' && exit 1
470	$SED 'y/aa' && exit 1
471	$SED 'y/aa/b/' && exit 1
472	$SED 'y/aa/' && exit 1
473	$SED 'y/a/b' && exit 1
474	$SED 'y/a/b/c/d' && exit 1
475	$SED '!' && exit 1
476	$SED supercalifrangolisticexprialidociussupercalifrangolisticexcius
477	set +x
478}
479
480main "$@"
481