diff options
author | Kiyoshi Aman <kiyoshi.aman+adelie@gmail.com> | 2019-02-01 22:55:37 +0000 |
---|---|---|
committer | Kiyoshi Aman <kiyoshi.aman+adelie@gmail.com> | 2019-02-03 18:22:05 -0600 |
commit | 5b57d28ffb6e1ef86b50f7d05d977826eae89bfe (patch) | |
tree | 154a22fe556b49e6927197336f8bf91b12eacd5e /usr.bin/m4 | |
download | userland-5b57d28ffb6e1ef86b50f7d05d977826eae89bfe.tar.gz userland-5b57d28ffb6e1ef86b50f7d05d977826eae89bfe.tar.bz2 userland-5b57d28ffb6e1ef86b50f7d05d977826eae89bfe.tar.xz userland-5b57d28ffb6e1ef86b50f7d05d977826eae89bfe.zip |
initial population
Diffstat (limited to 'usr.bin/m4')
39 files changed, 6630 insertions, 0 deletions
diff --git a/usr.bin/m4/NOTES b/usr.bin/m4/NOTES new file mode 100644 index 0000000..d60f80e --- /dev/null +++ b/usr.bin/m4/NOTES @@ -0,0 +1,64 @@ +m4 - macro processor + +PD m4 is based on the macro tool distributed with the software +tools (VOS) package, and described in the "SOFTWARE TOOLS" and +"SOFTWARE TOOLS IN PASCAL" books. It has been expanded to include +most of the command set of SysV m4, the standard UN*X macro processor. + +Since both PD m4 and UN*X m4 are based on SOFTWARE TOOLS macro, +there may be certain implementation similarities between +the two. The PD m4 was produced without ANY references to m4 +sources. + +written by: Ozan S. Yigit + +References: + + Software Tools distribution: macro + + Kernighan, Brian W. and P. J. Plauger, SOFTWARE + TOOLS IN PASCAL, Addison-Wesley, Mass. 1981 + + Kernighan, Brian W. and P. J. Plauger, SOFTWARE + TOOLS, Addison-Wesley, Mass. 1976 + + Kernighan, Brian W. and Dennis M. Ritchie, + THE M4 MACRO PROCESSOR, Unix Programmer's Manual, + Seventh Edition, Vol. 2, Bell Telephone Labs, 1979 + + System V man page for M4 + + +Implementation Notes: + +[1] PD m4 uses a different (and simpler) stack mechanism than the one + described in Software Tools and Software Tools in Pascal books. + The triple stack thing is replaced with a single stack containing + the call frames and the arguments. Each frame is back-linked to a + previous stack frame, which enables us to rewind the stack after + each nested call is completed. Each argument is a character pointer + to the beginning of the argument string within the string space. + The only exceptions to this are (*) arg 0 and arg 1, which are + the macro definition and macro name strings, stored dynamically + for the hash table. + + . . + | . | <-- sp | . | + +-------+ +-----+ + | arg 3 ------------------------------->| str | + +-------+ | . | + | arg 2 --------------+ . + +-------+ | + * | | | + +-------+ | +-----+ + | plev | <-- fp +---------------->| str | + +-------+ | . | + | type | . + +-------+ + | prcf -----------+ plev: paren level + +-------+ | type: call type + | . | | prcf: prev. call frame + . | + +-------+ | + | <----------+ + +-------+ diff --git a/usr.bin/m4/PSD.doc/Makefile b/usr.bin/m4/PSD.doc/Makefile new file mode 100644 index 0000000..9de1f83 --- /dev/null +++ b/usr.bin/m4/PSD.doc/Makefile @@ -0,0 +1,10 @@ +# $NetBSD: Makefile,v 1.4 2014/07/05 19:22:04 dholland Exp $ +# +# @(#)Makefile 8.1 (Berkeley) 6/8/93 + +SECTION=psd +ARTICLE=m4 +SRCS= m4.ms +MACROS= -msU + +.include <bsd.doc.mk> diff --git a/usr.bin/m4/TEST/ack.m4 b/usr.bin/m4/TEST/ack.m4 new file mode 100644 index 0000000..2784c45 --- /dev/null +++ b/usr.bin/m4/TEST/ack.m4 @@ -0,0 +1,41 @@ +# $NetBSD: ack.m4,v 1.4 1995/09/28 05:37:54 tls Exp $ +# +# Copyright (c) 1989, 1993 +# The Regents of the University of California. All rights reserved. +# +# This code is derived from software contributed to Berkeley by +# Ozan Yigit. +# +# Redistribution and use in source and binary forms, with or without +# modification, are permitted provided that the following conditions +# are met: +# 1. Redistributions of source code must retain the above copyright +# notice, this list of conditions and the following disclaimer. +# 2. Redistributions in binary form must reproduce the above copyright +# notice, this list of conditions and the following disclaimer in the +# documentation and/or other materials provided with the distribution. +# 3. All advertising materials mentioning features or use of this software +# must display the following acknowledgement: +# This product includes software developed by the University of +# California, Berkeley and its contributors. +# 4. Neither the name of the University nor the names of its contributors +# may be used to endorse or promote products derived from this software +# without specific prior written permission. +# +# THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND +# ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +# ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE +# FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL +# DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS +# OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) +# HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT +# LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY +# OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF +# SUCH DAMAGE. +# +# @(#)ack.m4 8.1 (Berkeley) 6/6/93 +# + +define(ack, `ifelse($1,0,incr($2),$2,0,`ack(DECR($1),1)', +`ack(DECR($1), ack($1,DECR($2)))')') diff --git a/usr.bin/m4/TEST/hanoi.m4 b/usr.bin/m4/TEST/hanoi.m4 new file mode 100644 index 0000000..6f73431 --- /dev/null +++ b/usr.bin/m4/TEST/hanoi.m4 @@ -0,0 +1,46 @@ +# $NetBSD: hanoi.m4,v 1.4 1995/09/28 05:37:56 tls Exp $ +# +# Copyright (c) 1989, 1993 +# The Regents of the University of California. All rights reserved. +# +# This code is derived from software contributed to Berkeley by +# Ozan Yigit. +# +# Redistribution and use in source and binary forms, with or without +# modification, are permitted provided that the following conditions +# are met: +# 1. Redistributions of source code must retain the above copyright +# notice, this list of conditions and the following disclaimer. +# 2. Redistributions in binary form must reproduce the above copyright +# notice, this list of conditions and the following disclaimer in the +# documentation and/or other materials provided with the distribution. +# 3. All advertising materials mentioning features or use of this software +# must display the following acknowledgement: +# This product includes software developed by the University of +# California, Berkeley and its contributors. +# 4. Neither the name of the University nor the names of its contributors +# may be used to endorse or promote products derived from this software +# without specific prior written permission. +# +# THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND +# ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +# ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE +# FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL +# DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS +# OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) +# HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT +# LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY +# OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF +# SUCH DAMAGE. +# +# @(#)hanoi.m4 8.1 (Berkeley) 6/6/93 +# + +define(hanoi, `trans(A, B, C, $1)') + +define(moved,`move disk from $1 to $2 +') + +define(trans, `ifelse($4,1,`moved($1,$2)', + `trans($1,$3,$2,DECR($4))moved($1,$2)trans($3,$2,$1,DECR($4))')') diff --git a/usr.bin/m4/TEST/hash.m4 b/usr.bin/m4/TEST/hash.m4 new file mode 100644 index 0000000..5de2f79 --- /dev/null +++ b/usr.bin/m4/TEST/hash.m4 @@ -0,0 +1,56 @@ +# $NetBSD: hash.m4,v 1.4 1995/09/28 05:37:58 tls Exp $ +# +# Copyright (c) 1989, 1993 +# The Regents of the University of California. All rights reserved. +# +# This code is derived from software contributed to Berkeley by +# Ozan Yigit. +# +# Redistribution and use in source and binary forms, with or without +# modification, are permitted provided that the following conditions +# are met: +# 1. Redistributions of source code must retain the above copyright +# notice, this list of conditions and the following disclaimer. +# 2. Redistributions in binary form must reproduce the above copyright +# notice, this list of conditions and the following disclaimer in the +# documentation and/or other materials provided with the distribution. +# 3. All advertising materials mentioning features or use of this software +# must display the following acknowledgement: +# This product includes software developed by the University of +# California, Berkeley and its contributors. +# 4. Neither the name of the University nor the names of its contributors +# may be used to endorse or promote products derived from this software +# without specific prior written permission. +# +# THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND +# ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +# ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE +# FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL +# DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS +# OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) +# HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT +# LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY +# OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF +# SUCH DAMAGE. +# +# @(#)hash.m4 8.1 (Berkeley) 6/6/93 +# + +dnl This probably will not run on any m4 that cannot +dnl handle char constants in eval. +dnl +changequote(<,>) define(HASHVAL,99) dnl +define(hash,<eval(str(substr($1,1),0)%HASHVAL)>) dnl +define(str, + <ifelse($1,",$2, + <str(substr(<$1>,1),<eval($2+'substr($1,0,1)')>)>) + >) dnl +define(KEYWORD,<$1,hash($1),>) dnl +define(TSTART, +<struct prehash { + char *keyword; + int hashval; +} keytab[] = {>) dnl +define(TEND,< "",0 +};>) dnl diff --git a/usr.bin/m4/TEST/math.m4 b/usr.bin/m4/TEST/math.m4 new file mode 100644 index 0000000..dff3937 --- /dev/null +++ b/usr.bin/m4/TEST/math.m4 @@ -0,0 +1,182 @@ +dnl $NetBSD: math.m4,v 1.2 2005/10/06 17:38:09 drochner Exp $ +dnl FreeBSD: /repoman/r/ncvs/src/usr.bin/m4/TEST/math.m4,v 1.1 2004/05/01 03:27:05 smkelly Exp +dnl A regression test for m4 C operators (ksb,petef) +dnl If you think you have a short-circuiting m4, run us m4 -DSHORCIRCUIT=yes +dnl +dnl first level of precedence +ifelse(eval(-7),-7,,`failed - +')dnl +ifelse(eval(- -2),2,,`failed - +')dnl +ifelse(eval(!0),1,,`failed ! +')dnl +ifelse(eval(!7),0,,`failed ! +')dnl +ifelse(eval(~-1),0,,`failed ~ +')dnl +dnl next level of precedence +ifelse(eval(3*5),15,,`failed * +')dnl +ifelse(eval(3*0),0,,`failed * +')dnl +ifelse(eval(11/2),5,,`failed / +')dnl +ifelse(eval(1/700),0,,`failed / +')dnl +ifelse(eval(10%5),0,,`failed % +')dnl +ifelse(eval(2%5),2,,`failed % +')dnl +ifelse(eval(2%-1),0,,`failed % +')dnl +dnl next level of precedence +ifelse(eval(2+2),4,,`failed + +')dnl +ifelse(eval(2+-2),0,,`failed + +')dnl +ifelse(eval(2- -2),4,,`failed - +')dnl +ifelse(eval(2-2),0,,`failed - +')dnl +dnl next level of precedence +ifelse(eval(1<<4),16,,`failed << +')dnl +ifelse(eval(16>>4),1,,`failed >> +')dnl +dnl next level of precedence +ifelse(eval(4<4),0,,`failed < +')dnl +ifelse(eval(4<5),1,,`failed < +')dnl +ifelse(eval(4<3),0,,`failed < +')dnl +ifelse(eval(4>4),0,,`failed > +')dnl +ifelse(eval(4>5),0,,`failed > +')dnl +ifelse(eval(4>3),1,,`failed > +')dnl +ifelse(eval(4<=4),1,,`failed <= +')dnl +ifelse(eval(4<=5),1,,`failed <= +')dnl +ifelse(eval(4<=3),0,,`failed <= +')dnl +ifelse(eval(4>=4),1,,`failed >= +')dnl +ifelse(eval(4>=5),0,,`failed >= +')dnl +ifelse(eval(4>=3),1,,`failed >= +')dnl +dnl next level of precedence +ifelse(eval(1==1),1,,`failed == +')dnl +ifelse(eval(1==-1),0,,`failed == +')dnl +ifelse(eval(1!=1),0,,`failed != +')dnl +ifelse(eval(1!=2),1,,`failed != +')dnl +dnl next level of precedence +ifelse(eval(3&5),1,,`failed & +')dnl +ifelse(eval(8&7),0,,`failed & +')dnl +dnl next level of precedence +ifelse(eval(1^1),0,,`failed ^ +')dnl +ifelse(eval(21^5),16,,`failed ^ +')dnl +dnl next level of precedence +ifelse(eval(1|1),1,,`failed | +')dnl +ifelse(eval(21|5),21,,`failed | +')dnl +ifelse(eval(100|1),101,,`failed | +')dnl +dnl next level of precedence +ifelse(eval(1&&1),1,,`failed && +')dnl +ifelse(eval(0&&1),0,,`failed && +')dnl +ifelse(eval(1&&0),0,,`failed && +')dnl +ifelse(SHORTCIRCUIT,`yes',`ifelse(eval(0&&10/0),0,,`failed && shortcircuit +')')dnl +dnl next level of precedence +ifelse(eval(1||1),1,,`failed || +')dnl +ifelse(eval(1||0),1,,`failed || +')dnl +ifelse(eval(0||0),0,,`failed || +')dnl +ifelse(SHORTCIRCUIT,`yes',`ifelse(eval(1||10/0),1,,`failed || shortcircuit +')')dnl +dnl next level of precedence +ifelse(eval(0 ? 2 : 5),5,,`failed ?: +')dnl +ifelse(eval(1 ? 2 : 5),2,,`failed ?: +')dnl +ifelse(SHORTCIRCUIT,`yes',`ifelse(eval(0 ? 10/0 : 7),7,,`failed ?: shortcircuit +')')dnl +ifelse(SHORTCIRCUIT,`yes',`ifelse(eval(1 ? 7 : 10/0),7,,`failed ?: shortcircuit +')')dnl +dnl operator precedence +ifelse(eval(!0*-2),-2,,`precedence wrong, ! * +')dnl +ifelse(eval(~8/~2),3,,`precedence wrong ~ / +')dnl +ifelse(eval(~-20%7),5,,`precedence wrong ~ % +')dnl +ifelse(eval(3*2+100),106,,`precedence wrong * + +')dnl +ifelse(eval(3+2*100),203,,`precedence wrong + * +')dnl +ifelse(eval(2%5-6/3),0,,`precedence wrong % - +')dnl +ifelse(eval(2/5-5%3),-2,,`precedence wrong / - +')dnl +ifelse(eval(2+5%5+1),3,,`precedence wrong % + +')dnl +ifelse(eval(7+9<<1),32,,`precedence wrong + << +')dnl +ifelse(eval(35-3>>2),8,,`precedence wrong - >> +')dnl +ifelse(eval(9<10<<5),1,,`precedence wrong << < +')dnl +ifelse(eval(9>10<<5),0,,`precedence wrong << > +')dnl +ifelse(eval(32>>2<32),1,,`precedence wrong >> < +')dnl +ifelse(eval(9<=10<<5),1,,`precedence wrong << < +')dnl +ifelse(eval(5<<1<=20>>1),1,,`precedence wrong << <= +')dnl +ifelse(eval(5<<1>=20>>1),1,,`precedence wrong << >= +')dnl +ifelse(eval(0<7==5>=5),1,,`precedence wrong < == +')dnl +ifelse(eval(0<7!=5>=5),0,,`precedence wrong < != +')dnl +ifelse(eval(0>7==5>=5),0,,`precedence wrong > == +')dnl +ifelse(eval(0>7!=5>=5),1,,`precedence wrong > != +')dnl +ifelse(eval(1&7==7),1,,`precedence wrong & == +')dnl +ifelse(eval(0&7!=6),0,,`precedence wrong & != +')dnl +ifelse(eval(9&1|5),5,,`precedence wrong & | +')dnl +ifelse(eval(9&1^5),4,,`precedence wrong & ^ +')dnl +ifelse(eval(9^1|5),13,,`precedence wrong ^ | +')dnl +ifelse(eval(5|0&&1),1,,`precedence wrong | && +')dnl +ifelse(eval(5&&0||0&&5||5),1,,`precedence wrong && || +')dnl +ifelse(eval(0 || 1 ? 0 : 1),0,,`precedence wrong || ?: +')dnl +ifelse(eval(5&&(0||0)&&(5||5)),0,,`precedence wrong || parens +')dnl diff --git a/usr.bin/m4/TEST/sqroot.m4 b/usr.bin/m4/TEST/sqroot.m4 new file mode 100644 index 0000000..6fc8d2c --- /dev/null +++ b/usr.bin/m4/TEST/sqroot.m4 @@ -0,0 +1,46 @@ +# $NetBSD: sqroot.m4,v 1.4 1995/09/28 05:38:01 tls Exp $ +# +# Copyright (c) 1989, 1993 +# The Regents of the University of California. All rights reserved. +# +# This code is derived from software contributed to Berkeley by +# Ozan Yigit. +# +# Redistribution and use in source and binary forms, with or without +# modification, are permitted provided that the following conditions +# are met: +# 1. Redistributions of source code must retain the above copyright +# notice, this list of conditions and the following disclaimer. +# 2. Redistributions in binary form must reproduce the above copyright +# notice, this list of conditions and the following disclaimer in the +# documentation and/or other materials provided with the distribution. +# 3. All advertising materials mentioning features or use of this software +# must display the following acknowledgement: +# This product includes software developed by the University of +# California, Berkeley and its contributors. +# 4. Neither the name of the University nor the names of its contributors +# may be used to endorse or promote products derived from this software +# without specific prior written permission. +# +# THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND +# ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +# ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE +# FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL +# DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS +# OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) +# HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT +# LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY +# OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF +# SUCH DAMAGE. +# +# @(#)sqroot.m4 8.1 (Berkeley) 6/6/93 +# + +define(square_root, + `ifelse(eval($1<0),1,negative-square-root, + `square_root_aux($1, 1, eval(($1+1)/2))')') +define(square_root_aux, + `ifelse($3, $2, $3, + $3, eval($1/$2), $3, + `square_root_aux($1, $3, eval(($3+($1/$3))/2))')') diff --git a/usr.bin/m4/TEST/string.m4 b/usr.bin/m4/TEST/string.m4 new file mode 100644 index 0000000..a14f14a --- /dev/null +++ b/usr.bin/m4/TEST/string.m4 @@ -0,0 +1,46 @@ +# $NetBSD: string.m4,v 1.4 1995/09/28 05:38:03 tls Exp $ +# +# Copyright (c) 1989, 1993 +# The Regents of the University of California. All rights reserved. +# +# This code is derived from software contributed to Berkeley by +# Ozan Yigit. +# +# Redistribution and use in source and binary forms, with or without +# modification, are permitted provided that the following conditions +# are met: +# 1. Redistributions of source code must retain the above copyright +# notice, this list of conditions and the following disclaimer. +# 2. Redistributions in binary form must reproduce the above copyright +# notice, this list of conditions and the following disclaimer in the +# documentation and/or other materials provided with the distribution. +# 3. All advertising materials mentioning features or use of this software +# must display the following acknowledgement: +# This product includes software developed by the University of +# California, Berkeley and its contributors. +# 4. Neither the name of the University nor the names of its contributors +# may be used to endorse or promote products derived from this software +# without specific prior written permission. +# +# THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND +# ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +# ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE +# FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL +# DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS +# OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) +# HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT +# LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY +# OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF +# SUCH DAMAGE. +# +# @(#)string.m4 8.1 (Berkeley) 6/6/93 +# + +define(string,`integer $1(len(substr($2,1))) +str($1,substr($2,1),0) +data $1(len(substr($2,1)))/EOS/ +') + +define(str,`ifelse($2,",,data $1(incr($3))/`LET'substr($2,0,1)/ +`str($1,substr($2,1),incr($3))')') diff --git a/usr.bin/m4/TEST/test.m4 b/usr.bin/m4/TEST/test.m4 new file mode 100644 index 0000000..5b19807 --- /dev/null +++ b/usr.bin/m4/TEST/test.m4 @@ -0,0 +1,244 @@ +# $NetBSD: test.m4,v 1.4 1995/09/28 05:38:05 tls Exp $ +# +# Copyright (c) 1989, 1993 +# The Regents of the University of California. All rights reserved. +# +# This code is derived from software contributed to Berkeley by +# Ozan Yigit. +# +# Redistribution and use in source and binary forms, with or without +# modification, are permitted provided that the following conditions +# are met: +# 1. Redistributions of source code must retain the above copyright +# notice, this list of conditions and the following disclaimer. +# 2. Redistributions in binary form must reproduce the above copyright +# notice, this list of conditions and the following disclaimer in the +# documentation and/or other materials provided with the distribution. +# 3. All advertising materials mentioning features or use of this software +# must display the following acknowledgement: +# This product includes software developed by the University of +# California, Berkeley and its contributors. +# 4. Neither the name of the University nor the names of its contributors +# may be used to endorse or promote products derived from this software +# without specific prior written permission. +# +# THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND +# ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +# ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE +# FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL +# DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS +# OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) +# HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT +# LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY +# OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF +# SUCH DAMAGE. +# +# @(#)test.m4 8.1 (Berkeley) 6/6/93 +# + +# test file for mp (not comprehensive) +# +# v7 m4 does not have `decr'. +# +define(DECR,`eval($1-1)') +# +# include string macros +# +include(string.m4) +# +# create some fortrash strings for an even uglier language +# +string(TEXT, "text") +string(DATA, "data") +string(BEGIN, "begin") +string(END, "end") +string(IF, "if") +string(THEN, "then") +string(ELSE, "else") +string(CASE, "case") +string(REPEAT, "repeat") +string(WHILE, "while") +string(DEFAULT, "default") +string(UNTIL, "until") +string(FUNCTION, "function") +string(PROCEDURE, "procedure") +string(EXTERNAL, "external") +string(FORWARD, "forward") +string(TYPE, "type") +string(VAR, "var") +string(CONST, "const") +string(PROGRAM, "program") +string(INPUT, "input") +string(OUTPUT, "output") +# +divert(2) +diversion #1 +divert(3) +diversion #2 +divert(4) +diversion #3 +divert(5) +diversion #4 +divert(0) +define(abc,xxx) +ifdef(`abc',defined,undefined) +# +# v7 m4 does this wrong. The right output is +# this is A vEry lon sEntEnCE +# see m4 documentation for translit. +# +translit(`this is a very long sentence', abcdefg, ABCDEF) +# +# include towers-of-hanoi +# +include(hanoi.m4) +# +# some reasonable set of disks +# +hanoi(6) +# +# include ackermann's function +# +include(ack.m4) +# +# something like (3,3) will blow away un*x m4. +# +ack(2,3) +# +# include a square_root function for fixed nums +# +include(sqroot.m4) +# +# some square roots. +# +square_root(15) +square_root(100) +square_root(-4) +square_root(21372) +# +# some textual material for enjoyment. +# +[taken from the 'Clemson University Computer Newsletter', + September 1981, pp. 6-7] + +I am a wizard in the magical Kingdom of Transformation and I +slay dragons for a living. Actually, I am a systems programmer. +One of the problems with systems programming is explaining to +non-computer enthusiasts what that is. All of the terms I use to +describe my job are totally meaningless to them. Usually my response +to questions about my work is to say as little as possible. For +instance, if someone asks what happened at work this week, I say +"Nothing much" and then I change the subject. + +With the assistance of my brother, a mechanical engineer, I have devised +an analogy that everyone can understand. The analogy describes the +"Kingdom of Transformation" where travelers wander and are magically +transformed. This kingdom is the computer and the travelers are information. +The purpose of the computer is to change information to a more meaningful +forma. The law of conservation applies here: The computer never creates +and never intentionally destroys data. With no further ado, let us travel +to the Kingdom of Transformation: + +In a land far, far away, there is a magical kingdom called the Kingdom of +Transformation. A king rules over this land and employs a Council of +Wizardry. The main purpose of this kingdom is to provide a way for +neighboring kingdoms to transform citizens into more useful citizens. This +is done by allowing the citizens to enter the kingdom at one of its ports +and to travel any of the many routes in the kingdom. They are magically +transformed along the way. The income of the Kingdom of Transformation +comes from the many toll roads within its boundaries. + +The Kingdom of Transformation was created when several kingdoms got +together and discovered a mutual need for new talents and abilities for +citizens. They employed CTK, Inc. (Creators of Transformation, Inc.) to +create this kingdom. CTK designed the country, its transportation routes, +and its laws of transformation, and created the major highway system. + +Hazards +======= + +Because magic is not truly controllable, CTK invariably, but unknowingly, +creates dragons. Dragons are huge fire-breathing beasts which sometimes +injure or kill travelers. Fortunately, they do not travel, but always +remain near their den. + +Other hazards also exist which are potentially harmful. As the roads +become older and more weatherbeaten, pot-holes will develop, trees will +fall on travelers, etc. CTK maintenance men are called to fix these +problems. + +Wizards +======= + +The wizards play a major role in creating and maintaining the kingdom but +get little credit for their work because it is performed secretly. The +wizards do not wan the workers or travelers to learn their incantations +because many laws would be broken and chaos would result. + +CTK's grand design is always general enough to be applicable in many +different situations. As a result, it is often difficult to use. The +first duty of the wizards is to tailor the transformation laws so as to be +more beneficial and easier to use in their particular environment. + +After creation of the kingdom, a major duty of the wizards is to search for +and kill dragons. If travelers do not return on time or if they return +injured, the ruler of the country contacts the wizards. If the wizards +determine that the injury or death occurred due to the traveler's +negligence, they provide the traveler's country with additional warnings. +If not, they must determine if the cause was a road hazard or a dragon. If +the suspect a road hazard, they call in a CTK maintenance man to locate the +hazard and to eliminate it, as in repairing the pothole in the road. If +they think that cause was a dragon, then they must find and slay it. + +The most difficult part of eliminating a dragon is finding it. Sometimes +the wizard magically knows where the dragon's lair it, but often the wizard +must send another traveler along the same route and watch to see where he +disappears. This sounds like a failsafe method for finding dragons (and a +suicide mission for thr traveler) but the second traveler does not always +disappear. Some dragons eat any traveler who comes too close; others are +very picky. + +The wizards may call in CTK who designed the highway system and +transformation laws to help devise a way to locate the dragon. CTK also +helps provide the right spell or incantation to slay the dragon. (There is +no general spell to slay dragons; each dragon must be eliminated with a +different spell.) + +Because neither CTK nor wizards are perfect, spells to not always work +correctly. At best, nothing happens when the wrong spell is uttered. At +worst, the dragon becomes a much larger dragon or multiplies into several +smaller ones. In either case, new spells must be found. + +If all existing dragons are quiet (i.e. have eaten sufficiently), wizards +have time to do other things. They hide in castles and practice spells and +incatations. They also devise shortcuts for travelers and new laws of +transformation. + +Changes in the Kingdom +====================== + +As new transformation kingdoms are created and old ones are maintained, +CTK, Inc. is constantly learning new things. It learns ways to avoid +creating some of the dragons that they have previously created. It also +discovers new and better laws of transformation. As a result, CTK will +periodically create a new grand design which is far better than the old. +The wizards determine when is a good time to implement this new design. +This is when the tourist season is slow or when no important travelers +(VIPs) are to arrive. The kingdom must be closed for the actual +implementation and is leter reopened as a new and better place to go. + +A final question you might ask is what happens when the number of tourists +becomes too great for the kingdom to handle in a reasonable period of time +(i.e., the tourist lines at the ports are too long). The Kingdom of +Transformation has three options: (1) shorten the paths that a tourist must +travel, or (2) convince CTK to develop a faster breed of horses so that the +travelers can finish sooner, or (3) annex more territories so that the +kingdom can handle more travelers. + +Thus ends the story of the Kingdom of Transformation. I hope this has +explained my job to you: I slay dragons for a living. + +# +#should do an automatic undivert.. +# diff --git a/usr.bin/m4/eval.c b/usr.bin/m4/eval.c new file mode 100644 index 0000000..94150b1 --- /dev/null +++ b/usr.bin/m4/eval.c @@ -0,0 +1,1053 @@ +/* $OpenBSD: eval.c,v 1.66 2008/08/21 21:01:47 espie Exp $ */ +/* $NetBSD: eval.c,v 1.27 2018/07/30 22:58:09 kre Exp $ */ + +/* + * Copyright (c) 1989, 1993 + * The Regents of the University of California. All rights reserved. + * + * This code is derived from software contributed to Berkeley by + * Ozan Yigit at York University. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + */ + +/* + * eval.c + * Facility: m4 macro processor + * by: oz + */ +#if HAVE_NBTOOL_CONFIG_H +#include "nbtool_config.h" +#endif +#include <sys/cdefs.h> +__RCSID("$NetBSD: eval.c,v 1.27 2018/07/30 22:58:09 kre Exp $"); + +#include <sys/types.h> +#include <ctype.h> +#include <err.h> +#include <errno.h> +#include <limits.h> +#include <unistd.h> +#include <stdio.h> +#include <stdlib.h> +#include <stddef.h> +#include <stdint.h> +#include <string.h> +#include <inttypes.h> +#include <fcntl.h> +#include "mdef.h" +#include "stdd.h" +#include "extern.h" +#include "pathnames.h" + +static void dodefn(const char *); +static void dopushdef(const char *, const char *); +static void dodump(const char *[], int); +static void dotrace(const char *[], int, int); +static void doifelse(const char *[], int); +static int doincl(const char *); +static int dopaste(const char *); +static void dochq(const char *[], int); +static void dochc(const char *[], int); +static void dom4wrap(const char *); +static void dodiv(int); +static void doundiv(const char *[], int); +static void dosub(const char *[], int); +static void map(char *, const char *, const char *, const char *); +static const char *handledash(char *, char *, const char *); +static void expand_builtin(const char *[], int, int); +static void expand_macro(const char *[], int); +static void dump_one_def(const char *, struct macro_definition *); + +unsigned long expansion_id; + +/* + * eval - eval all macros and builtins calls + * argc - number of elements in argv. + * argv - element vector : + * argv[0] = definition of a user + * macro or NULL if built-in. + * argv[1] = name of the macro or + * built-in. + * argv[2] = parameters to user-defined + * . macro or built-in. + * . + * + * A call in the form of macro-or-builtin() will result in: + * argv[0] = nullstr + * argv[1] = macro-or-builtin + * argv[2] = nullstr + * + * argc is 3 for macro-or-builtin() and 2 for macro-or-builtin + */ +void +eval(const char *argv[], int argc, int td, int is_traced) +{ + size_t mark = SIZE_MAX; + + expansion_id++; + if (td & RECDEF) + m4errx(1, "expanding recursive definition for %s.", argv[1]); + if (is_traced) + mark = trace(argv, argc, infile+ilevel); + if (td == MACRTYPE) + expand_macro(argv, argc); + else + expand_builtin(argv, argc, td); + if (mark != SIZE_MAX) + finish_trace(mark); +} + +/* + * expand_builtin - evaluate built-in macros. + */ +void +expand_builtin(const char *argv[], int argc, int td) +{ + int c, n; + int ac; + static int sysval = 0; + +#ifdef DEBUG + printf("argc = %d\n", argc); + for (n = 0; n < argc; n++) + printf("argv[%d] = %s\n", n, argv[n]); + fflush(stdout); +#endif + + /* + * if argc == 3 and argv[2] is null, then we + * have macro-or-builtin() type call. We adjust + * argc to avoid further checking.. + */ + /* we keep the initial value for those built-ins that differentiate + * between builtin() and builtin. + */ + ac = argc; + + if (argc == 3 && !*(argv[2]) && !mimic_gnu) + argc--; + + switch (td & TYPEMASK) { + + case DEFITYPE: + if (argc > 2) + dodefine(argv[2], (argc > 3) ? argv[3] : null); + break; + + case PUSDTYPE: + if (argc > 2) + dopushdef(argv[2], (argc > 3) ? argv[3] : null); + break; + + case DUMPTYPE: + dodump(argv, argc); + break; + + case TRACEONTYPE: + dotrace(argv, argc, 1); + break; + + case TRACEOFFTYPE: + dotrace(argv, argc, 0); + break; + + case EXPRTYPE: + /* + * doexpr - evaluate arithmetic + * expression + */ + { + int base = 10; + int maxdigits = 0; + int e; + + if (argc > 3) { + base = strtoi(argv[3], NULL, 0, 2, 36, &e); + if (e) { + m4errx(1, "expr: base %s invalid.", argv[3]); + } + } + if (argc > 4) { + maxdigits = strtoi(argv[4], NULL, 0, 0, INT_MAX, &e); + if (e) { + m4errx(1, "expr: maxdigits %s invalid.", argv[4]); + } + } + if (argc > 2) + pbnumbase(expr(argv[2]), base, maxdigits); + break; + } + + case IFELTYPE: + if (argc > 4) + doifelse(argv, argc); + break; + + case IFDFTYPE: + /* + * doifdef - select one of two + * alternatives based on the existence of + * another definition + */ + if (argc > 3) { + if (lookup_macro_definition(argv[2]) != NULL) + pbstr(argv[3]); + else if (argc > 4) + pbstr(argv[4]); + } + break; + + case LENGTYPE: + /* + * dolen - find the length of the + * argument + */ + pbnum((argc > 2) ? strlen(argv[2]) : 0); + break; + + case INCRTYPE: + /* + * doincr - increment the value of the + * argument + */ + if (argc > 2) + pbnum(atoi(argv[2]) + 1); + break; + + case DECRTYPE: + /* + * dodecr - decrement the value of the + * argument + */ + if (argc > 2) + pbnum(atoi(argv[2]) - 1); + break; + + case SYSCTYPE: + /* + * dosys - execute system command + */ + if (argc > 2) { + fflush(stdout); + sysval = system(argv[2]); + } + break; + + case SYSVTYPE: + /* + * dosysval - return value of the last + * system call. + * + */ + pbnum(sysval); + break; + + case ESYSCMDTYPE: + if (argc > 2) + doesyscmd(argv[2]); + break; + case INCLTYPE: + if (argc > 2) + if (!doincl(argv[2])) + err(1, "%s at line %lu: include(%s)", + CURRENT_NAME, CURRENT_LINE, argv[2]); + break; + + case SINCTYPE: + if (argc > 2) + (void) doincl(argv[2]); + break; +#ifdef EXTENDED + case PASTTYPE: + if (argc > 2) + if (!dopaste(argv[2])) + err(1, "%s at line %lu: paste(%s)", + CURRENT_NAME, CURRENT_LINE, argv[2]); + break; + + case SPASTYPE: + if (argc > 2) + (void) dopaste(argv[2]); + break; + case FORMATTYPE: + doformat(argv, argc); + break; +#endif + case CHNQTYPE: + dochq(argv, ac); + break; + + case CHNCTYPE: + dochc(argv, argc); + break; + + case SUBSTYPE: + /* + * dosub - select substring + * + */ + if (argc > 3) + dosub(argv, argc); + break; + + case SHIFTYPE: + /* + * doshift - push back all arguments + * except the first one (i.e. skip + * argv[2]) + */ + if (argc > 3) { + for (n = argc - 1; n > 3; n--) { + pbstr(rquote); + pbstr(argv[n]); + pbstr(lquote); + pushback(COMMA); + } + pbstr(rquote); + pbstr(argv[3]); + pbstr(lquote); + } + break; + + case DIVRTYPE: + if (argc > 2 && (n = atoi(argv[2])) != 0) + dodiv(n); + else { + active = stdout; + oindex = 0; + } + break; + + case UNDVTYPE: + doundiv(argv, argc); + break; + + case DIVNTYPE: + /* + * dodivnum - return the number of + * current output diversion + */ + pbnum(oindex); + break; + + case UNDFTYPE: + /* + * doundefine - undefine a previously + * defined macro(s) or m4 keyword(s). + */ + if (argc > 2) + for (n = 2; n < argc; n++) + macro_undefine(argv[n]); + break; + + case POPDTYPE: + /* + * dopopdef - remove the topmost + * definitions of macro(s) or m4 + * keyword(s). + */ + if (argc > 2) + for (n = 2; n < argc; n++) + macro_popdef(argv[n]); + break; + + case MKTMTYPE: + /* + * dotemp - create a temporary file + */ + if (argc > 2) { + int fd; + char *temp; + + temp = xstrdup(argv[2]); + + fd = mkstemp(temp); + if (fd == -1) + err(1, + "%s at line %lu: couldn't make temp file %s", + CURRENT_NAME, CURRENT_LINE, argv[2]); + close(fd); + pbstr(temp); + free(temp); + } + break; + + case TRNLTYPE: + /* + * dotranslit - replace all characters in + * the source string that appears in the + * "from" string with the corresponding + * characters in the "to" string. + */ + if (argc > 3) { + char *temp; + + temp = xalloc(strlen(argv[2])+1, NULL); + if (argc > 4) + map(temp, argv[2], argv[3], argv[4]); + else + map(temp, argv[2], argv[3], null); + pbstr(temp); + free(temp); + } else if (argc > 2) + pbstr(argv[2]); + break; + + case INDXTYPE: + /* + * doindex - find the index of the second + * argument string in the first argument + * string. -1 if not present. + */ + pbnum((argc > 3) ? indx(argv[2], argv[3]) : -1); + break; + + case ERRPTYPE: + /* + * doerrp - print the arguments to stderr + * file + */ + if (argc > 2) { + for (n = 2; n < argc; n++) + fprintf(stderr, "%s%s", + mimic_gnu && n == 2 ? "" : " ", + argv[n]); + if (!mimic_gnu) + fprintf(stderr, "\n"); + } + break; + + case DNLNTYPE: + /* + * dodnl - eat-up-to and including + * newline + */ + while ((c = gpbc()) != '\n' && c != EOF) + ; + break; + + case M4WRTYPE: + /* + * dom4wrap - set up for + * wrap-up/wind-down activity + */ + if (argc > 2) + dom4wrap(argv[2]); + break; + + case EXITTYPE: + /* + * doexit - immediate exit from m4. + */ + killdiv(); + exit((argc > 2) ? atoi(argv[2]) : 0); + break; + + case DEFNTYPE: + if (argc > 2) + for (n = 2; n < argc; n++) + dodefn(argv[n]); + break; + + case INDIRTYPE: /* Indirect call */ + if (argc > 2) + doindir(argv, argc); + break; + + case BUILTINTYPE: /* Builtins only */ + if (argc > 2) + dobuiltin(argv, argc); + break; + + case PATSTYPE: + if (argc > 2) + dopatsubst(argv, argc); + break; + case REGEXPTYPE: + if (argc > 2) + doregexp(argv, argc); + break; + case LINETYPE: + doprintlineno(infile+ilevel); + break; + case FILENAMETYPE: + doprintfilename(infile+ilevel); + break; + case SELFTYPE: + pbstr(rquote); + pbstr(argv[1]); + pbstr(lquote); + break; + default: + m4errx(1, "eval: major botch."); + break; + } +} + +/* + * expand_macro - user-defined macro expansion + */ +void +expand_macro(const char *argv[], int argc) +{ + const char *t; + const char *p; + int n; + int argno; + + t = argv[0]; /* defn string as a whole */ + p = t; + while (*p) + p++; + p--; /* last character of defn */ + while (p > t) { + if (*(p - 1) != ARGFLAG) + PUSHBACK(*p); + else { + switch (*p) { + + case '#': + pbnum(argc - 2); + break; + case '0': + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + case '8': + case '9': + argno = *p - '0'; + if (mimic_gnu) { + const unsigned char *q = + (const unsigned char *)p; + while (isdigit(*++q)) { + bp--; + argno = argno * 10 + *q - '0'; + } + } + if (argno < argc - 1) + pbstr(argv[argno + 1]); + break; + case '*': + if (argc > 2) { + for (n = argc - 1; n > 2; n--) { + pbstr(argv[n]); + pushback(COMMA); + } + pbstr(argv[2]); + } + break; + case '@': + if (argc > 2) { + for (n = argc - 1; n > 2; n--) { + pbstr(rquote); + pbstr(argv[n]); + pbstr(lquote); + pushback(COMMA); + } + pbstr(rquote); + pbstr(argv[2]); + pbstr(lquote); + } + break; + default: + PUSHBACK(*p); + PUSHBACK('$'); + break; + } + p--; + } + p--; + } + if (p == t) /* do last character */ + PUSHBACK(*p); +} + + +/* + * dodefine - install definition in the table + */ +void +dodefine(const char *name, const char *defn) +{ + if (!*name && !mimic_gnu) + m4errx(1, "null definition."); + else + macro_define(name, defn); +} + +/* + * dodefn - push back a quoted definition of + * the given name. + */ +static void +dodefn(const char *name) +{ + struct macro_definition *p; + + if ((p = lookup_macro_definition(name)) != NULL) { + if ((p->type & TYPEMASK) == MACRTYPE) { + pbstr(rquote); + pbstr(p->defn); + pbstr(lquote); + } else { + pbstr(p->defn); + pbstr(BUILTIN_MARKER); + } + } +} + +/* + * dopushdef - install a definition in the hash table + * without removing a previous definition. Since + * each new entry is entered in *front* of the + * hash bucket, it hides a previous definition from + * lookup. + */ +static void +dopushdef(const char *name, const char *defn) +{ + if (!*name && !mimic_gnu) + m4errx(1, "null definition."); + else + macro_pushdef(name, defn); +} + +/* + * dump_one_def - dump the specified definition. + */ +static void +dump_one_def(const char *name, struct macro_definition *p) +{ + if (!traceout) + traceout = stderr; + if (mimic_gnu) { + if ((p->type & TYPEMASK) == MACRTYPE) + fprintf(traceout, "%s:\t%s\n", name, p->defn); + else { + fprintf(traceout, "%s:\t<%s>\n", name, p->defn); + } + } else + fprintf(traceout, "`%s'\t`%s'\n", name, p->defn); +} + +/* + * dodumpdef - dump the specified definitions in the hash + * table to stderr. If nothing is specified, the entire + * hash table is dumped. + */ +static void +dodump(const char *argv[], int argc) +{ + int n; + struct macro_definition *p; + + if (argc > 2) { + for (n = 2; n < argc; n++) + if ((p = lookup_macro_definition(argv[n])) != NULL) + dump_one_def(argv[n], p); + } else + macro_for_all(dump_one_def); +} + +/* + * dotrace - mark some macros as traced/untraced depending upon on. + */ +static void +dotrace(const char *argv[], int argc, int on) +{ + int n; + + if (argc > 2) { + for (n = 2; n < argc; n++) + mark_traced(argv[n], on); + } else + mark_traced(NULL, on); +} + +/* + * doifelse - select one of two alternatives - loop. + */ +static void +doifelse(const char *argv[], int argc) +{ + cycle { + if (argc < 5) + m4errx(1, "wrong number of args for ifelse"); + if (STREQ(argv[2], argv[3])) + pbstr(argv[4]); + else if (argc == 6) + pbstr(argv[5]); + else if (argc > 6) { + argv += 3; + argc -= 3; + continue; + } + break; + } +} + +/* + * doinclude - include a given file. + */ +static int +doincl(const char *ifile) +{ +#ifndef REAL_FREEZE + if (thawing) + return 1; +#endif + if (ilevel + 1 == MAXINP) + m4errx(1, "too many include files."); + if (fopen_trypath(infile+ilevel+1, ifile) != NULL) { + ilevel++; + bbase[ilevel] = bufbase = bp; + return (1); + } else + return (0); +} + +#ifdef EXTENDED +/* + * dopaste - include a given file without any + * macro processing. + */ +static int +dopaste(const char *pfile) +{ + FILE *pf; + int c; + + if ((pf = fopen(pfile, "r")) != NULL) { + if (synch_lines) + fprintf(active, "#line 1 \"%s\"\n", pfile); + while ((c = getc(pf)) != EOF) + putc(c, active); + (void) fclose(pf); + emit_synchline(); + return (1); + } else + return (0); +} +#endif + +/* + * dochq - change quote characters + */ +static void +dochq(const char *argv[], int ac) +{ + if (ac == 2) { + lquote[0] = LQUOTE; lquote[1] = EOS; + rquote[0] = RQUOTE; rquote[1] = EOS; + } else { + strlcpy(lquote, argv[2], sizeof(lquote)); + if (ac > 3) { + strlcpy(rquote, argv[3], sizeof(rquote)); + } else { + rquote[0] = ECOMMT; rquote[1] = EOS; + } + } +} + +/* + * dochc - change comment characters + */ +static void +dochc(const char *argv[], int argc) +{ +/* XXX Note that there is no difference between no argument and a single + * empty argument. + */ + if (argc == 2) { + scommt[0] = EOS; + ecommt[0] = EOS; + } else { + strlcpy(scommt, argv[2], sizeof(scommt)); + if (argc == 3) { + ecommt[0] = ECOMMT; ecommt[1] = EOS; + } else { + strlcpy(ecommt, argv[3], sizeof(ecommt)); + } + } +} + +/* + * dom4wrap - expand text at EOF + */ +static void +dom4wrap(const char *text) +{ + if (wrapindex >= maxwraps) { + if (maxwraps == 0) + maxwraps = 16; + else + maxwraps *= 2; + m4wraps = xrealloc(m4wraps, maxwraps * sizeof(*m4wraps), + "too many m4wraps"); + } + m4wraps[wrapindex++] = xstrdup(text); +} + +/* + * dodivert - divert the output to a temporary file + */ +static void +dodiv(int n) +{ + int fd; + + oindex = n; + if (n >= maxout) { + if (mimic_gnu) + resizedivs(n + 10); + else + n = 0; /* bitbucket */ + } + + if (n < 0) + n = 0; /* bitbucket */ + if (outfile[n] == NULL) { + char fname[] = _PATH_DIVNAME; + + if ((fd = mkstemp(fname)) < 0 || + (outfile[n] = fdopen(fd, "w+")) == NULL) + err(1, "%s: cannot divert", fname); + if (unlink(fname) == -1) + err(1, "%s: cannot unlink", fname); + } + active = outfile[n]; +} + +/* + * doundivert - undivert a specified output, or all + * other outputs, in numerical order. + */ +static void +doundiv(const char *argv[], int argc) +{ + int ind; + int n; + + if (argc > 2) { + for (ind = 2; ind < argc; ind++) { + int e; + n = strtoi(argv[ind], NULL, 0, 1, INT_MAX, &e); + if (e) { + if (errno == EINVAL && mimic_gnu) + getdivfile(argv[ind]); + } else { + if (n < maxout && outfile[n] != NULL) + getdiv(n); + } + } + } + else + for (n = 1; n < maxout; n++) + if (outfile[n] != NULL) + getdiv(n); +} + +/* + * dosub - select substring + */ +static void +dosub(const char *argv[], int argc) +{ + const char *ap, *fc, *k; + int nc; + + ap = argv[2]; /* target string */ +#ifdef EXPR + fc = ap + expr(argv[3]); /* first char */ +#else + fc = ap + atoi(argv[3]); /* first char */ +#endif + nc = strlen(fc); + if (argc >= 5) +#ifdef EXPR + nc = min(nc, expr(argv[4])); +#else + nc = min(nc, atoi(argv[4])); +#endif + if (fc >= ap && fc < ap + strlen(ap)) + for (k = fc + nc - 1; k >= fc; k--) + pushback(*k); +} + +/* + * map: + * map every character of s1 that is specified in from + * into s3 and replace in s. (source s1 remains untouched) + * + * This is a standard implementation of map(s,from,to) function of ICON + * language. Within mapvec, we replace every character of "from" with + * the corresponding character in "to". If "to" is shorter than "from", + * than the corresponding entries are null, which means that those + * characters dissapear altogether. Furthermore, imagine + * map(dest, "sourcestring", "srtin", "rn..*") type call. In this case, + * `s' maps to `r', `r' maps to `n' and `n' maps to `*'. Thus, `s' + * ultimately maps to `*'. In order to achieve this effect in an efficient + * manner (i.e. without multiple passes over the destination string), we + * loop over mapvec, starting with the initial source character. if the + * character value (dch) in this location is different than the source + * character (sch), sch becomes dch, once again to index into mapvec, until + * the character value stabilizes (i.e. sch = dch, in other words + * mapvec[n] == n). Even if the entry in the mapvec is null for an ordinary + * character, it will stabilize, since mapvec[0] == 0 at all times. At the + * end, we restore mapvec* back to normal where mapvec[n] == n for + * 0 <= n <= 127. This strategy, along with the restoration of mapvec, is + * about 5 times faster than any algorithm that makes multiple passes over + * destination string. + */ +static void +map(char *dest, const char *src, const char *from, const char *to) +{ + const char *tmp; + unsigned char sch, dch; + unsigned char found[256]; + static char frombis[257]; + static char tobis[257]; + static unsigned char mapvec[256] = { + 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, + 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, + 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, + 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, + 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, + 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, + 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, + 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, + 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, + 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, + 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, + 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, + 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, + 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, + 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, + 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, + 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, + 246, 247, 248, 249, 250, 251, 252, 253, 254, 255 + }; + + if (*src) { + if (mimic_gnu) { + /* + * expand character ranges on the fly + */ + from = handledash(frombis, frombis + 256, from); + to = handledash(tobis, tobis + 256, to); + } + tmp = from; + /* + * create a mapping between "from" and + * "to" + */ + memset(found, 0, sizeof(found)); + for (; (sch = (unsigned char)*from) != '\0'; from++) { + if (!mimic_gnu || !found[sch]) { + found[sch] = 1; + mapvec[sch] = *to; + } + if (*to) + to++; + } + + if (mimic_gnu) { + for (; (sch = (unsigned char)*src) != '\0'; src++) { + if (!found[sch]) + *dest++ = sch; + else if ((dch = mapvec[sch]) != '\0') + *dest++ = dch; + } + } else { + while (*src) { + sch = (unsigned char)(*src++); + dch = mapvec[sch]; + while (dch != sch) { + sch = dch; + dch = mapvec[sch]; + } + if ((*dest = (char)dch)) + dest++; + } + } + /* + * restore all the changed characters + */ + while (*tmp) { + mapvec[(unsigned char)(*tmp)] = (unsigned char)(*tmp); + tmp++; + } + } + *dest = '\0'; +} + + +/* + * handledash: + * use buffer to copy the src string, expanding character ranges + * on the way. + */ +static const char * +handledash(char *buffer, char *end, const char *src) +{ + char *p; + + p = buffer; + while(*src) { + if (src[1] == '-' && src[2]) { + unsigned char i; + if ((unsigned char)src[0] <= (unsigned char)src[2]) { + for (i = (unsigned char)src[0]; + i <= (unsigned char)src[2]; i++) { + *p++ = i; + if (p == end) { + *p = '\0'; + return buffer; + } + } + } else { + for (i = (unsigned char)src[0]; + i >= (unsigned char)src[2]; i--) { + *p++ = i; + if (p == end) { + *p = '\0'; + return buffer; + } + } + } + src += 3; + } else + *p++ = *src++; + if (p == end) + break; + } + *p = '\0'; + return buffer; +} diff --git a/usr.bin/m4/expr.c b/usr.bin/m4/expr.c new file mode 100644 index 0000000..a6c901f --- /dev/null +++ b/usr.bin/m4/expr.c @@ -0,0 +1,50 @@ +/* $NetBSD: expr.c,v 1.19 2009/10/26 21:11:28 christos Exp $ */ +/* $OpenBSD: expr.c,v 1.17 2006/01/20 23:10:19 espie Exp $ */ +/* + * Copyright (c) 2004 Marc Espie <espie@cvs.openbsd.org> + * + * Permission to use, copy, modify, and distribute this software for any + * purpose with or without fee is hereby granted, provided that the above + * copyright notice and this permission notice appear in all copies. + * + * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES + * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR + * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES + * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN + * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF + * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. + */ +#if HAVE_NBTOOL_CONFIG_H +#include "nbtool_config.h" +#endif +#include <sys/cdefs.h> +__RCSID("$NetBSD: expr.c,v 1.19 2009/10/26 21:11:28 christos Exp $"); +#include <stdint.h> +#include <stdio.h> +#include <stddef.h> +#include "mdef.h" +#include "extern.h" + +int32_t end_result; +const char *copy_toeval; + +extern void yy_scan_string(const char *); +extern int yyparse(void); +extern int yyerror(const char *); + +int +yyerror(const char *msg) +{ + fprintf(stderr, "m4: %s in expr %s\n", msg, copy_toeval); + return(0); +} + +int +expr(const char *toeval) +{ + copy_toeval = toeval; + yy_scan_string(toeval); + yyparse(); + return end_result; +} diff --git a/usr.bin/m4/extern.h b/usr.bin/m4/extern.h new file mode 100644 index 0000000..1cb9de1 --- /dev/null +++ b/usr.bin/m4/extern.h @@ -0,0 +1,190 @@ +/* $OpenBSD: extern.h,v 1.49 2009/10/14 17:19:47 sthen Exp $ */ +/* $NetBSD: extern.h,v 1.19 2016/01/16 18:30:57 christos Exp $ */ + +/*- + * Copyright (c) 1991, 1993 + * The Regents of the University of California. All rights reserved. + * + * This code is derived from software contributed to Berkeley by + * Ozan Yigit at York University. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + * + * @(#)extern.h 8.1 (Berkeley) 6/6/93 + */ + +/* eval.c */ +extern void eval(const char *[], int, int, int); +extern void dodefine(const char *, const char *); +extern unsigned long expansion_id; + +/* expr.c */ +extern int expr(const char *); + +/* gnum4.c */ +extern void addtoincludepath(const char *); +extern struct input_file *fopen_trypath(struct input_file *, const char *); +extern void doindir(const char *[], int); +extern void dobuiltin(const char *[], int); +extern void dopatsubst(const char *[], int); +extern void doregexp(const char *[], int); + +extern void doprintlineno(struct input_file *); +extern void doprintfilename(struct input_file *); + +extern void doesyscmd(const char *); +extern void getdivfile(const char *); +extern void doformat(const char *[], int); +#ifdef REAL_FREEZE +extern void freeze_state(const char *); +extern void thaw_state(const char *); +#endif + + +/* look.c */ + +#define FLAG_UNTRACED 0 +#define FLAG_TRACED 1 +#define FLAG_NO_TRACE 2 + +extern void init_macros(void); +extern ndptr lookup(const char *); +extern void mark_traced(const char *, int); +extern struct ohash macros; + +extern struct macro_definition *lookup_macro_definition(const char *); +extern void macro_define(const char *, const char *); +extern void macro_pushdef(const char *, const char *); +extern void macro_popdef(const char *); +extern void macro_undefine(const char *); +extern void setup_builtin(const char *, unsigned int); +extern void macro_for_all(void (*)(const char *, struct macro_definition *)); +#define macro_getdef(p) ((p)->d) +#define macro_name(p) ((p)->name) +#define macro_builtin_type(p) ((p)->builtin_type) +#define is_traced(p) ((p)->trace_flags == FLAG_NO_TRACE ? (trace_flags & TRACE_ALL) : (p)->trace_flags) + +extern ndptr macro_getbuiltin(const char *); +#ifdef REAL_FREEZE +extern void dump_state(FILE *); +extern void restore_state(FILE *); +#endif + +/* main.c */ +extern void outputstr(const char *); +extern void do_emit_synchline(void); +#define emit_synchline() do { if (synch_lines) do_emit_synchline(); } while(0) + +/* misc.c */ +extern void chrsave(int); +extern char *compute_prevep(void); +extern void getdiv(int); +extern ptrdiff_t indx(const char *, const char *); +extern void initspaces(void); +extern void killdiv(void); +extern void pbnum(int); +extern void pbnumbase(int, int, int); +extern void pbunsigned(unsigned long); +extern void pbstr(const char *); +extern void pushback(int); +extern void *xalloc(size_t, const char *fmt, ...) __printflike(2, 3); +extern void *xrealloc(void *, size_t, const char *fmt, ...) + __printflike(3, 4); +extern char *xstrdup(const char *); +extern void resizedivs(int); +extern size_t buffer_mark(void); +extern void dump_buffer(FILE *, size_t); +extern void __dead m4errx(int, const char *, ...) __printflike(2, 3); + +extern int obtain_char(struct input_file *); +extern void set_input(struct input_file *, FILE *, const char *); +extern void release_input(struct input_file *); + +/* speeded-up versions of chrsave/pushback */ +#define PUSHBACK(c) \ + do { \ + if (bp >= endpbb) \ + enlarge_bufspace(); \ + *bp++ = (c); \ + } while(0) + +#define CHRSAVE(c) \ + do { \ + if (ep >= endest) \ + enlarge_strspace(); \ + *ep++ = (c); \ + } while(0) + +/* and corresponding exposure for local symbols */ +extern void enlarge_bufspace(void); +extern void enlarge_strspace(void); +extern unsigned char *endpbb; +extern char *endest; + +/* trace.c */ +extern unsigned int trace_flags; +#define TRACE_ALL 512 +extern void trace_file(const char *); +extern size_t trace(const char **, int, struct input_file *); +extern void finish_trace(size_t); +extern void set_trace_flags(const char *); +extern FILE *traceout; + +extern ndptr hashtab[]; /* hash table for macros etc. */ +extern stae *mstack; /* stack of m4 machine */ +extern char *sstack; /* shadow stack, for string space extension */ +extern FILE *active; /* active output file pointer */ +extern struct input_file infile[];/* input file stack (0=stdin) */ +extern FILE **outfile; /* diversion array(0=bitbucket) */ +extern int maxout; /* maximum number of diversions */ +extern int fp; /* m4 call frame pointer */ +extern int ilevel; /* input file stack pointer */ +extern int oindex; /* diversion index. */ +extern int sp; /* current m4 stack pointer */ +extern unsigned char *bp; /* first available character */ +extern unsigned char *buf; /* push-back buffer */ +extern unsigned char *bufbase; /* buffer base for this ilevel */ +extern unsigned char *bbase[]; /* buffer base per ilevel */ +extern char ecommt[MAXCCHARS+1];/* end character for comment */ +extern char *ep; /* first free char in strspace */ +extern char lquote[MAXCCHARS+1];/* left quote character (`) */ +extern char **m4wraps; /* m4wrap string default. */ +extern int maxwraps; /* size of m4wraps array */ +extern int wrapindex; /* current index in m4wraps */ +extern int fatal_warnings; /* exit on warning */ +extern int quiet; /* no warnings */ +extern int nesting_limit; /* macro expansion nesting limit */ +#ifndef REAL_FREEZE +extern FILE *freezef; /* copy of input */ +extern int thawing; /* don't process includes during thaw */ +#endif + +extern const char *null; /* as it says.. just a null. */ +extern char rquote[MAXCCHARS+1];/* right quote character (') */ +extern char scommt[MAXCCHARS+1];/* start character for comment */ +extern int synch_lines; /* line synchronisation directives */ + +extern int mimic_gnu; /* behaves like gnu-m4 */ +extern int prefix_builtins; /* prefix builtin macros with m4_ */ diff --git a/usr.bin/m4/gnum4.c b/usr.bin/m4/gnum4.c new file mode 100644 index 0000000..a450a99 --- /dev/null +++ b/usr.bin/m4/gnum4.c @@ -0,0 +1,817 @@ +/* $NetBSD: gnum4.c,v 1.10 2016/01/16 16:59:18 christos Exp $ */ +/* $OpenBSD: gnum4.c,v 1.39 2008/08/21 21:01:04 espie Exp $ */ + +/* + * Copyright (c) 1999 Marc Espie + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + */ + +/* + * functions needed to support gnu-m4 extensions, including a fake freezing + */ +#if HAVE_NBTOOL_CONFIG_H +#include "nbtool_config.h" +#endif +#include <sys/cdefs.h> +__RCSID("$NetBSD: gnum4.c,v 1.10 2016/01/16 16:59:18 christos Exp $"); + +#include <sys/param.h> +#include <sys/types.h> +#include <sys/wait.h> +#include <ctype.h> +#include <err.h> +#include <paths.h> +#include <regex.h> +#include <stddef.h> +#include <stdlib.h> +#include <stdio.h> +#include <string.h> +#include <errno.h> +#include <unistd.h> +#include "mdef.h" +#include "stdd.h" +#include "extern.h" + + +int mimic_gnu = 0; +#ifndef SIZE_T_MAX +#define SIZE_T_MAX (size_t)~0ull +#endif + +/* + * Support for include path search + * First search in the current directory. + * If not found, and the path is not absolute, include path kicks in. + * First, -I options, in the order found on the command line. + * Then M4PATH env variable + */ + +struct path_entry { + char *name; + struct path_entry *next; +} *first, *last; + +static struct path_entry *new_path_entry(const char *); +static void ensure_m4path(void); +static struct input_file *dopath(struct input_file *, const char *); + +static struct path_entry * +new_path_entry(const char *dirname) +{ + struct path_entry *n; + + n = malloc(sizeof(struct path_entry)); + if (!n) + errx(1, "out of memory"); + n->name = strdup(dirname); + if (!n->name) + errx(1, "out of memory"); + n->next = 0; + return n; +} + +void +addtoincludepath(const char *dirname) +{ + struct path_entry *n; + + n = new_path_entry(dirname); + + if (last) { + last->next = n; + last = n; + } + else + last = first = n; +} + +static void +ensure_m4path(void) +{ + static int envpathdone = 0; + char *envpath; + char *sweep; + char *path; + + if (envpathdone) + return; + envpathdone = TRUE; + envpath = getenv("M4PATH"); + if (!envpath) + return; + /* for portability: getenv result is read-only */ + envpath = strdup(envpath); + if (!envpath) + errx(1, "out of memory"); + for (sweep = envpath; + (path = strsep(&sweep, ":")) != NULL;) + addtoincludepath(path); + free(envpath); +} + +static +struct input_file * +dopath(struct input_file *i, const char *filename) +{ + char path[MAXPATHLEN]; + struct path_entry *pe; + FILE *f; + + for (pe = first; pe; pe = pe->next) { + snprintf(path, sizeof(path), "%s/%s", pe->name, filename); + if ((f = fopen(path, "r")) != 0) { + set_input(i, f, path); + return i; + } + } + return NULL; +} + +struct input_file * +fopen_trypath(struct input_file *i, const char *filename) +{ + FILE *f; + + f = fopen(filename, "r"); + if (f != NULL) { + set_input(i, f, filename); + return i; + } + if (filename[0] == '/') + return NULL; + + ensure_m4path(); + + return dopath(i, filename); +} + +void +doindir(const char *argv[], int argc) +{ + ndptr n; + struct macro_definition *p; + + n = lookup(argv[2]); + if (n == NULL || (p = macro_getdef(n)) == NULL) + m4errx(1, "indir: undefined macro %s.", argv[2]); + argv[1] = p->defn; + + eval(argv+1, argc-1, p->type, is_traced(n)); +} + +void +dobuiltin(const char *argv[], int argc) +{ + ndptr p; + + argv[1] = NULL; + p = macro_getbuiltin(argv[2]); + if (p != NULL) + eval(argv+1, argc-1, macro_builtin_type(p), is_traced(p)); + else + m4errx(1, "unknown builtin %s.", argv[2]); +} + + +/* We need some temporary buffer space, as pb pushes BACK and substitution + * proceeds forward... */ +static char *buffer; +static size_t bufsize = 0; +static size_t current = 0; + +static void addchars(const char *, size_t); +static void addchar(int); +static char *twiddle(const char *); +static char *getstring(void); +static void exit_regerror(int, const char *, regex_t *) __dead; +static void do_subst(const char *, const char *, regex_t *, const char *, + regmatch_t *); +static void do_regexpindex(const char *, const char *, regex_t *, regmatch_t *); +static void do_regexp(const char *, const char *, regex_t *, const char *, regmatch_t *); +static void add_sub(size_t, const char *, regex_t *, regmatch_t *); +static void add_replace(const char *, regex_t *, const char *, regmatch_t *); +#define addconstantstring(s) addchars((s), sizeof(s)-1) + +static void +addchars(const char *c, size_t n) +{ + if (n == 0) + return; + while (current + n > bufsize) { + if (bufsize == 0) + bufsize = 1024; + else + bufsize *= 2; + buffer = xrealloc(buffer, bufsize, NULL); + } + memcpy(buffer+current, c, n); + current += n; +} + +static void +addchar(int c) +{ + if (current +1 > bufsize) { + if (bufsize == 0) + bufsize = 1024; + else + bufsize *= 2; + buffer = xrealloc(buffer, bufsize, NULL); + } + buffer[current++] = c; +} + +static char * +getstring(void) +{ + addchar('\0'); + current = 0; + return buffer; +} + + +static void +exit_regerror(int er, const char *pat, regex_t *re) +{ + size_t errlen; + char *errbuf; + + errlen = regerror(er, re, NULL, 0); + errbuf = xalloc(errlen, + "malloc in regerror: %lu", (unsigned long)errlen); + regerror(er, re, errbuf, errlen); + m4errx(1, "regular expression error: %s for: `%s'", errbuf, pat); +} + +static void +add_sub(size_t n, const char *string, regex_t *re, regmatch_t *pm) +{ + if (n > re->re_nsub) { + if (!quiet) + warnx("No subexpression %zu", n); + if (fatal_warnings) + exit(EXIT_FAILURE); + } + /* Subexpressions that did not match are + * not an error. */ + else if (pm[n].rm_so != -1 && + pm[n].rm_eo != -1) { + addchars(string + pm[n].rm_so, + pm[n].rm_eo - pm[n].rm_so); + } +} + +/* Add replacement string to the output buffer, recognizing special + * constructs and replacing them with substrings of the original string. + */ +static void +add_replace(const char *string, regex_t *re, const char *replace, regmatch_t *pm) +{ + const char *p; + + for (p = replace; *p != '\0'; p++) { + if (*p == '&' && !mimic_gnu) { + add_sub(0, string, re, pm); + continue; + } + if (*p == '\\') { + if (p[1] == '\\') { + addchar(p[1]); + p++; + continue; + } + if (p[1] == '&') { + if (mimic_gnu) + add_sub(0, string, re, pm); + else + addchar(p[1]); + p++; + continue; + } + if (isdigit((unsigned char)p[1])) { + add_sub(*(++p) - '0', string, re, pm); + continue; + } + } + addchar(*p); + } +} + +static void +do_subst(const char *pat, const char *string, regex_t *re, const char *replace, + regmatch_t *pm) +{ + int error; + int flags = 0; + const char *last_match = NULL; + + while ((error = regexec(re, string, re->re_nsub+1, pm, flags)) == 0) { + if (pm[0].rm_eo != 0) { + if (string[pm[0].rm_eo-1] == '\n') + flags = 0; + else + flags = REG_NOTBOL; + } + + /* NULL length matches are special... We use the `vi-mode' + * rule: don't allow a NULL-match at the last match + * position. + */ + if (pm[0].rm_so == pm[0].rm_eo && + string + pm[0].rm_so == last_match) { + if (*string == '\0') + return; + addchar(*string); + if (*string++ == '\n') + flags = 0; + else + flags = REG_NOTBOL; + continue; + } + last_match = string + pm[0].rm_so; + addchars(string, pm[0].rm_so); + add_replace(string, re, replace, pm); + string += pm[0].rm_eo; + buffer[current] = '\0'; + } + while (*string) + addchar(*string++); + if (error != REG_NOMATCH) + exit_regerror(error, pat, re); + pbstr(string); +} + +static void +do_regexp(const char *pat, const char *string, regex_t *re, const char *replace, + regmatch_t *pm) +{ + int error; + + switch(error = regexec(re, string, re->re_nsub+1, pm, 0)) { + case 0: + add_replace(string, re, replace, pm); + pbstr(getstring()); + break; + case REG_NOMATCH: + break; + default: + exit_regerror(error, pat, re); + } +} + +static void +do_regexpindex(const char *pat, const char *string, regex_t *re, regmatch_t *pm) +{ + int error; + + switch(error = regexec(re, string, re->re_nsub+1, pm, 0)) { + case 0: + pbunsigned(pm[0].rm_so); + break; + case REG_NOMATCH: + pbnum(-1); + break; + default: + exit_regerror(error, pat, re); + } +} + +/* In Gnu m4 mode, parentheses for backmatch don't work like POSIX 1003.2 + * says. So we twiddle with the regexp before passing it to regcomp. + */ +static char * +twiddle(const char *p) +{ + /* + at start of regexp is a normal character for Gnu m4 */ + if (*p == '^') { + addchar(*p); + p++; + } + if (*p == '+') { + addchar('\\'); + } + /* This could use strcspn for speed... */ + while (*p != '\0') { + if (*p == '\\') { + switch(p[1]) { + case '(': + case ')': + case '|': + addchar(p[1]); + break; + case 'w': + addconstantstring("[_a-zA-Z0-9]"); + break; + case 'W': + addconstantstring("[^_a-zA-Z0-9]"); + break; + case '<': + addconstantstring("[[:<:]]"); + break; + case '>': + addconstantstring("[[:>:]]"); + break; + default: + addchars(p, 2); + break; + } + p+=2; + continue; + } + if (*p == '(' || *p == ')' || *p == '|') + addchar('\\'); + + addchar(*p); + p++; + } + return getstring(); +} + +static int +checkempty(const char *argv[], int argc) +{ + const char *s; + size_t len; + + if (argc != 3 && argv[3][0] != '\0') + return 0; + + if (argc == 3) { + if (!quiet) + warnx("Too few arguments to patsubst"); + if (fatal_warnings) + exit(EXIT_FAILURE); + } + + if (argv[4] && argc > 4) + len = strlen(argv[4]); + else + len = 0; + for (s = argv[2]; *s != '\0'; s++) { + addchars(argv[4], len); + addchar(*s); + } + return 1; +} + +/* patsubst(string, regexp, opt replacement) */ +/* argv[2]: string + * argv[3]: regexp + * argv[4]: opt rep + */ +void +dopatsubst(const char *argv[], int argc) +{ + if (argc < 3) { + if (!quiet) + warnx("Too few arguments to patsubst"); + if (fatal_warnings) + exit(EXIT_FAILURE); + return; + } + /* special case: empty regexp */ + if (!checkempty(argv, argc)) { + + const char *pat; + int error; + regex_t re; + regmatch_t *pmatch; + int mode = REG_EXTENDED; + size_t l = strlen(argv[3]); + + if (!mimic_gnu || + (argv[3][0] == '^') || + (l > 0 && argv[3][l-1] == '$')) + mode |= REG_NEWLINE; + + pat = mimic_gnu ? twiddle(argv[3]) : argv[3]; + error = regcomp(&re, pat, mode); + if (error != 0) + exit_regerror(error, pat, &re); + + pmatch = xalloc(sizeof(regmatch_t) * (re.re_nsub+1), NULL); + do_subst(pat, argv[2], &re, + argc > 4 && argv[4] != NULL ? argv[4] : "", pmatch); + free(pmatch); + regfree(&re); + } + pbstr(getstring()); +} + +void +doregexp(const char *argv[], int argc) +{ + int error; + regex_t re; + regmatch_t *pmatch; + const char *pat; + + if (argc < 3) { + if (!quiet) + warnx("Too few arguments to regexp"); + if (fatal_warnings) + exit(EXIT_FAILURE); + return; + } + if (checkempty(argv, argc)) { + return; + } + + pat = mimic_gnu ? twiddle(argv[3]) : argv[3]; + error = regcomp(&re, pat, REG_EXTENDED); + if (error != 0) + exit_regerror(error, pat, &re); + + pmatch = xalloc(sizeof(regmatch_t) * (re.re_nsub+1), NULL); + if (argv[4] == NULL || argc == 4) + do_regexpindex(pat, argv[2], &re, pmatch); + else + do_regexp(pat, argv[2], &re, argv[4], pmatch); + free(pmatch); + regfree(&re); +} + +void +doformat(const char *argv[], int argc) +{ + const char *format = argv[2]; + int pos = 3; + int left_padded; + long width; + size_t l; + const char *thisarg; + char temp[2]; + size_t extra; + + while (*format != 0) { + if (*format != '%') { + addchar(*format++); + continue; + } + + format++; + if (*format == '%') { + addchar(*format++); + continue; + } + if (*format == 0) { + addchar('%'); + break; + } + + if (*format == '*') { + format++; + if (pos >= argc) + m4errx(1, + "Format with too many format specifiers."); + width = strtol(argv[pos++], NULL, 10); + } else { + char *eformat; + width = strtol(format, &eformat, 10); + format = eformat; + } + if (width < 0) { + left_padded = 1; + width = -width; + } else { + left_padded = 0; + } + if (*format == '.') { + format++; + if (*format == '*') { + format++; + if (pos >= argc) + m4errx(1, + "Format with too many format specifiers."); + extra = strtol(argv[pos++], NULL, 10); + } else { + char *eformat; + extra = strtol(format, &eformat, 10); + format = eformat; + } + } else { + extra = SIZE_T_MAX; + } + if (pos >= argc) + m4errx(1, "Format with too many format specifiers."); + switch(*format) { + case 's': + thisarg = argv[pos++]; + break; + case 'c': + temp[0] = strtoul(argv[pos++], NULL, 10); + temp[1] = 0; + thisarg = temp; + break; + default: + m4errx(1, "Unsupported format specification: %s.", + argv[2]); + } + format++; + l = strlen(thisarg); + if (l > extra) + l = extra; + if (!left_padded) { + while (l < (size_t)width--) + addchar(' '); + } + addchars(thisarg, l); + if (left_padded) { + while (l < (size_t)width--) + addchar(' '); + } + } + pbstr(getstring()); +} + +void +doesyscmd(const char *cmd) +{ + int p[2]; + pid_t pid, cpid; + const char *argv[4]; + int cc; + int status; + + /* Follow gnu m4 documentation: first flush buffers. */ + fflush(NULL); + + argv[0] = "sh"; + argv[1] = "-c"; + argv[2] = cmd; + argv[3] = NULL; + + /* Just set up standard output, share stderr and stdin with m4 */ + if (pipe(p) == -1) + err(1, "bad pipe"); + switch(cpid = fork()) { + case -1: + err(1, "bad fork"); + /* NOTREACHED */ + case 0: + (void) close(p[0]); + (void) dup2(p[1], 1); + (void) close(p[1]); + execv(_PATH_BSHELL, __UNCONST(argv)); + exit(1); + default: + /* Read result in two stages, since m4's buffer is + * pushback-only. */ + (void) close(p[1]); + do { + char result[BUFSIZE]; + cc = read(p[0], result, sizeof result); + if (cc > 0) + addchars(result, cc); + } while (cc > 0 || (cc == -1 && errno == EINTR)); + + (void) close(p[0]); + while ((pid = wait(&status)) != cpid && pid >= 0) + continue; + pbstr(getstring()); + } +} + +void +getdivfile(const char *name) +{ + FILE *f; + int c; + + f = fopen(name, "r"); + if (!f) + return; + + while ((c = getc(f))!= EOF) + putc(c, active); + (void) fclose(f); +} + +#ifdef REAL_FREEZE +void +freeze_state(const char *fname) +{ + FILE *f; + + if ((f = fopen(fname, "wb")) == NULL) + m4errx(EXIT_FAILURE, "Can't open output freeze file `%s' (%s)", + fname, strerror(errno)); + fprintf(f, "# This is a frozen state file generated by %s\nV1\n", + getprogname()); + fprintf(f, "Q%zu,%zu\n%s%s\n", strlen(lquote), strlen(rquote), + lquote, rquote); + fprintf(f, "C%zu,%zu\n%s%s\n", strlen(scommt), strlen(ecommt), + scommt, ecommt); + dump_state(f); + /* XXX: diversions? */ + fprintf(f, "D-1,0\n"); + fprintf(f, "# End of frozen state file\n"); + fclose(f); +} + +void +thaw_state(const char *fname) +{ + char *name = NULL; + size_t nl, namelen = 0; + char *defn = NULL; + size_t dl, defnlen = 0; + size_t lineno = 0; + char line[1024], *ptr, type; + FILE *f; + + if ((f = fopen(fname, "rb")) == NULL) + m4errx(EXIT_FAILURE, "Can't open frozen file `%s' (%s)", + fname, strerror(errno)); + +#define GET() if (fgets(line, (int)sizeof(line), f) == NULL) goto out +#define GETSTR(s, l) if (fread(s, 1, l, f) != l) goto out; else s[l] = '\0' + + GET(); /* comment */ + GET(); /* version */ + if ((ptr = strrchr(line, '\n')) != NULL) + *ptr = '\0'; + if (strcmp(line, "V1") != 0) + m4errx(EXIT_FAILURE, "Bad frozen version `%s'", line); + + for (;;) { + GET(); + lineno++; + switch (*line) { + case '\n': + continue; + case '#': + free(name); + free(defn); + fclose(f); + return; + default: + if (sscanf(line, "%c%zu,%zu\n", &type, &nl, &dl) != 3) + m4errx(EXIT_FAILURE, "%s, %zu: Bad line `%s'", + fname, lineno, line); + break; + } + + switch (type) { + case 'Q': + if (nl >= sizeof(lquote) || dl >= sizeof(rquote)) + m4errx(EXIT_FAILURE, "%s, %zu: Quote too long", + fname, lineno); + GETSTR(lquote, nl); + GETSTR(rquote, dl); + break; + + case 'C': + if (nl >= sizeof(scommt) || dl >= sizeof(ecommt)) + m4errx(EXIT_FAILURE, "%s, %zu: Comment too long", + fname, lineno); + GETSTR(scommt, nl); + GETSTR(ecommt, dl); + break; + + case 'T': + case 'F': + if (nl >= namelen) + name = xrealloc(name, namelen = nl + 1, + "name grow"); + if (dl >= defnlen) + defn = xrealloc(defn, defnlen = dl + 1, + "defn grow"); + GETSTR(name, nl); + GETSTR(defn, dl); + macro_pushdef(name, defn); + break; + + case 'D': + /* XXX: Not implemented */ + break; + + default: + m4errx(EXIT_FAILURE, "%s, %zu: Unknown type %c", + fname, lineno,type); + } + } +out: + m4errx(EXIT_FAILURE, "Unexprected end of file in `%s'", fname); +} +#endif diff --git a/usr.bin/m4/lib/ohash.h b/usr.bin/m4/lib/ohash.h new file mode 100644 index 0000000..8ccdaf9 --- /dev/null +++ b/usr.bin/m4/lib/ohash.h @@ -0,0 +1,73 @@ +#ifndef OHASH_H +#define OHASH_H +/* $OpenBSD: ohash.h,v 1.8 2005/12/29 18:54:47 jaredy Exp $ */ +/* ex:ts=8 sw=4: + */ + +/* Copyright (c) 1999, 2004 Marc Espie <espie@openbsd.org> + * + * Permission to use, copy, modify, and distribute this software for any + * purpose with or without fee is hereby granted, provided that the above + * copyright notice and this permission notice appear in all copies. + * + * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES + * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR + * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES + * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN + * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF + * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. + */ + +/* Open hashing support. + * Open hashing was chosen because it is much lighter than other hash + * techniques, and more efficient in most cases. + */ + +struct ohash_info { + ptrdiff_t key_offset; + void *data; /* user data */ + void *(*halloc)(size_t, void *); + void (*hfree)(void *, size_t, void *); + void *(*alloc)(size_t, void *); +}; + +struct _ohash_record; + +struct ohash { + struct _ohash_record *t; + struct ohash_info info; + unsigned int size; + unsigned int total; + unsigned int deleted; +}; + +/* For this to be tweakable, we use small primitives, and leave part of the + * logic to the client application. e.g., hashing is left to the client + * application. We also provide a simple table entry lookup that yields + * a hashing table index (opaque) to be used in find/insert/remove. + * The keys are stored at a known position in the client data. + */ +__BEGIN_DECLS +void ohash_init(struct ohash *, unsigned, struct ohash_info *); +void ohash_delete(struct ohash *); + +unsigned int ohash_lookup_interval(struct ohash *, const char *, + const char *, u_int32_t); +unsigned int ohash_lookup_memory(struct ohash *, const char *, + size_t, u_int32_t); +void *ohash_find(struct ohash *, unsigned int); +void *ohash_remove(struct ohash *, unsigned int); +void *ohash_insert(struct ohash *, unsigned int, void *); +void *ohash_first(struct ohash *, unsigned int *); +void *ohash_next(struct ohash *, unsigned int *); +unsigned int ohash_entries(struct ohash *); + +void *ohash_create_entry(struct ohash_info *, const char *, const char **); +u_int32_t ohash_interval(const char *, const char **); + +unsigned int ohash_qlookupi(struct ohash *, const char *, const char **); +unsigned int ohash_qlookup(struct ohash *, const char *); +__END_DECLS +#endif + diff --git a/usr.bin/m4/lib/ohash_create_entry.c b/usr.bin/m4/lib/ohash_create_entry.c new file mode 100644 index 0000000..3bdc9ab --- /dev/null +++ b/usr.bin/m4/lib/ohash_create_entry.c @@ -0,0 +1,38 @@ +/* $OpenBSD: ohash_create_entry.c,v 1.2 2004/06/22 20:00:16 espie Exp $ */ +/* ex:ts=8 sw=4: + */ + +/* Copyright (c) 1999, 2004 Marc Espie <espie@openbsd.org> + * + * Permission to use, copy, modify, and distribute this software for any + * purpose with or without fee is hereby granted, provided that the above + * copyright notice and this permission notice appear in all copies. + * + * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES + * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR + * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES + * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN + * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF + * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. + */ + +#include "ohash_int.h" + +/* This handles the common case of variable length keys, where the + * key is stored at the end of the record. + */ +void * +ohash_create_entry(struct ohash_info *i, const char *start, const char **end) +{ + char *p; + + if (!*end) + *end = start + strlen(start); + p = (i->alloc)(i->key_offset + (*end - start) + 1, i->data); + if (p) { + memcpy(p+i->key_offset, start, *end-start); + p[i->key_offset + (*end - start)] = '\0'; + } + return (void *)p; +} diff --git a/usr.bin/m4/lib/ohash_delete.c b/usr.bin/m4/lib/ohash_delete.c new file mode 100644 index 0000000..12a9c0c --- /dev/null +++ b/usr.bin/m4/lib/ohash_delete.c @@ -0,0 +1,31 @@ +/* $OpenBSD: ohash_delete.c,v 1.2 2004/06/22 20:00:16 espie Exp $ */ +/* ex:ts=8 sw=4: + */ + +/* Copyright (c) 1999, 2004 Marc Espie <espie@openbsd.org> + * + * Permission to use, copy, modify, and distribute this software for any + * purpose with or without fee is hereby granted, provided that the above + * copyright notice and this permission notice appear in all copies. + * + * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES + * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR + * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES + * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN + * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF + * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. + */ + +#include "ohash_int.h" +/* hash_delete only frees the hash structure. Use hash_first/hash_next + * to free entries as well. */ +void +ohash_delete(struct ohash *h) +{ + (h->info.hfree)(h->t, sizeof(struct _ohash_record) * h->size, + h->info.data); +#ifndef NDEBUG + h->t = NULL; +#endif +} diff --git a/usr.bin/m4/lib/ohash_do.c b/usr.bin/m4/lib/ohash_do.c new file mode 100644 index 0000000..2b8467d --- /dev/null +++ b/usr.bin/m4/lib/ohash_do.c @@ -0,0 +1,111 @@ +/* $OpenBSD: ohash_do.c,v 1.4 2004/06/22 20:00:16 espie Exp $ */ +/* ex:ts=8 sw=4: + */ + +/* Copyright (c) 1999, 2004 Marc Espie <espie@openbsd.org> + * + * Permission to use, copy, modify, and distribute this software for any + * purpose with or without fee is hereby granted, provided that the above + * copyright notice and this permission notice appear in all copies. + * + * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES + * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR + * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES + * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN + * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF + * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. + */ + +#include "ohash_int.h" + +static void ohash_resize(struct ohash *); + +static void +ohash_resize(struct ohash *h) +{ + struct _ohash_record *n; + unsigned int ns, j; + unsigned int i, incr; + + if (4 * h->deleted < h->total) + ns = h->size << 1; + else if (3 * h->deleted > 2 * h->total) + ns = h->size >> 1; + else + ns = h->size; + if (ns < MINSIZE) + ns = MINSIZE; +#ifdef STATS_HASH + STAT_HASH_EXPAND++; + STAT_HASH_SIZE += ns - h->size; +#endif + n = (h->info.halloc)(sizeof(struct _ohash_record) * ns, h->info.data); + if (!n) + return; + + for (j = 0; j < h->size; j++) { + if (h->t[j].p != NULL && h->t[j].p != DELETED) { + i = h->t[j].hv % ns; + incr = ((h->t[j].hv % (ns - 2)) & ~1) + 1; + while (n[i].p != NULL) { + i += incr; + if (i >= ns) + i -= ns; + } + n[i].hv = h->t[j].hv; + n[i].p = h->t[j].p; + } + } + (h->info.hfree)(h->t, sizeof(struct _ohash_record) * h->size, + h->info.data); + h->t = n; + h->size = ns; + h->total -= h->deleted; + h->deleted = 0; +} + +void * +ohash_remove(struct ohash *h, unsigned int i) +{ + void *result = __UNCONST(h->t[i].p); + + if (result == NULL || result == DELETED) + return NULL; + +#ifdef STATS_HASH + STAT_HASH_ENTRIES--; +#endif + h->t[i].p = DELETED; + h->deleted++; + if (h->deleted >= MINDELETED && 4 * h->deleted > h->total) + ohash_resize(h); + return result; +} + +void * +ohash_find(struct ohash *h, unsigned int i) +{ + if (h->t[i].p == DELETED) + return NULL; + else + return __UNCONST(h->t[i].p); +} + +void * +ohash_insert(struct ohash *h, unsigned int i, void *p) +{ +#ifdef STATS_HASH + STAT_HASH_ENTRIES++; +#endif + if (h->t[i].p == DELETED) { + h->deleted--; + h->t[i].p = p; + } else { + h->t[i].p = p; + /* Arbitrary resize boundary. Tweak if not efficient enough. */ + if (++h->total * 4 > h->size * 3) + ohash_resize(h); + } + return p; +} diff --git a/usr.bin/m4/lib/ohash_entries.c b/usr.bin/m4/lib/ohash_entries.c new file mode 100644 index 0000000..683e204 --- /dev/null +++ b/usr.bin/m4/lib/ohash_entries.c @@ -0,0 +1,26 @@ +/* $OpenBSD: ohash_entries.c,v 1.2 2004/06/22 20:00:16 espie Exp $ */ +/* ex:ts=8 sw=4: + */ + +/* Copyright (c) 1999, 2004 Marc Espie <espie@openbsd.org> + * + * Permission to use, copy, modify, and distribute this software for any + * purpose with or without fee is hereby granted, provided that the above + * copyright notice and this permission notice appear in all copies. + * + * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES + * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR + * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES + * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN + * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF + * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. + */ + +#include "ohash_int.h" + +unsigned int +ohash_entries(struct ohash *h) +{ + return h->total - h->deleted; +} diff --git a/usr.bin/m4/lib/ohash_enum.c b/usr.bin/m4/lib/ohash_enum.c new file mode 100644 index 0000000..966daa1 --- /dev/null +++ b/usr.bin/m4/lib/ohash_enum.c @@ -0,0 +1,36 @@ +/* $OpenBSD: ohash_enum.c,v 1.3 2004/06/22 20:00:16 espie Exp $ */ +/* ex:ts=8 sw=4: + */ + +/* Copyright (c) 1999, 2004 Marc Espie <espie@openbsd.org> + * + * Permission to use, copy, modify, and distribute this software for any + * purpose with or without fee is hereby granted, provided that the above + * copyright notice and this permission notice appear in all copies. + * + * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES + * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR + * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES + * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN + * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF + * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. + */ + +#include "ohash_int.h" + +void * +ohash_first(struct ohash *h, unsigned int *pos) +{ + *pos = 0; + return ohash_next(h, pos); +} + +void * +ohash_next(struct ohash *h, unsigned int *pos) +{ + for (; *pos < h->size; (*pos)++) + if (h->t[*pos].p != DELETED && h->t[*pos].p != NULL) + return __UNCONST(h->t[(*pos)++].p); + return NULL; +} diff --git a/usr.bin/m4/lib/ohash_init.3 b/usr.bin/m4/lib/ohash_init.3 new file mode 100644 index 0000000..a0939a8 --- /dev/null +++ b/usr.bin/m4/lib/ohash_init.3 @@ -0,0 +1,229 @@ +.\" $OpenBSD: ohash_init.3,v 1.14 2007/05/31 19:19:30 jmc Exp $ +.\" Copyright (c) 1999 Marc Espie <espie@openbsd.org> +.\" +.\" Permission to use, copy, modify, and distribute this software for any +.\" purpose with or without fee is hereby granted, provided that the above +.\" copyright notice and this permission notice appear in all copies. +.\" +.\" THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES +.\" WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF +.\" MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR +.\" ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES +.\" WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN +.\" ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF +.\" OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. +.\" +.Dd $Mdocdate: May 31 2007 $ +.Dt OPEN_HASH 3 +.Os +.Sh NAME +.Nm ohash_init , +.Nm ohash_delete , +.Nm ohash_lookup_interval , +.Nm ohash_lookup_memory , +.Nm ohash_find , +.Nm ohash_remove , +.Nm ohash_insert , +.Nm ohash_first , +.Nm ohash_next , +.Nm ohash_entries +.Nd light-weight open hashing +.Sh SYNOPSIS +.Fd #include <stdint.h> +.Fd #include <stddef.h> +.Fd #include <ohash.h> +.Ft void +.Fn ohash_init "struct ohash *h" "unsigned int size" "struct ohash_info *info" +.Ft void +.Fn ohash_delete "struct ohash *h" +.Ft "unsigned int" +.Fn ohash_lookup_interval "struct ohash *h" "const char *start" "const char *end" "uint32_t hv" +.Ft "unsigned int" +.Fn ohash_lookup_memory "struct ohash *h" "const char *k" "size_t s" "uint32_t hv" +.Ft void * +.Fn ohash_find "struct ohash *h" "unsigned int i" +.Ft void * +.Fn ohash_remove "struct ohash *h" "unsigned int i" +.Ft void * +.Fn ohash_insert "struct ohash *h" "unsigned int i" "void *p" +.Ft void * +.Fn ohash_first "struct ohash *h" "unsigned int *i" +.Ft void * +.Fn ohash_next "struct ohash *h" "unsigned int *i" +.Ft "unsigned int" +.Fn ohash_entries "struct ohash *h" +.Sh DESCRIPTION +These functions have been designed as a fast, extensible alternative to +the usual hash table functions. +They provide storage and retrieval of records indexed by keys, +where a key is a contiguous sequence of bytes at a fixed position in +each record. +Keys can either be NUL-terminated strings or fixed-size memory areas. +All functions take a pointer to an ohash structure as the +.Fa h +function argument. +Storage for this structure should be provided by user code. +.Pp +.Fn ohash_init +initializes the table to store roughly 2 to the power +.Fa size +elements. +.Fa info +holds the position of the key in each record, and two pointers to +.Xr calloc 3 +and +.Xr free 3 Ns -like +functions, to use for managing the table internal storage. +.Pp +.Fn ohash_delete +frees storage internal to +.Fa h . +Elements themselves should be freed by the user first, using for instance +.Fn ohash_first +and +.Fn ohash_next . +.Pp +.Fn ohash_lookup_interval +and +.Fn ohash_lookup_memory +are the basic look-up element functions. +The hashing function result is provided by the user as +.Fa hv . +These return a +.Qq slot +in the ohash table +.Fa h , +to be used with +.Fn ohash_find , +.Fn ohash_insert , +or +.Fn ohash_remove . +This slot is only valid up to the next call to +.Fn ohash_insert +or +.Fn ohash_remove . +.Pp +.Fn ohash_lookup_interval +handles string-like keys. +.Fn ohash_lookup_interval +assumes the key is the interval between +.Fa start +and +.Fa end , +exclusive, +though the actual elements stored in the table should only contain +NUL-terminated keys. +.Pp +.Fn ohash_lookup_memory +assumes the key is the memory area starting at +.Fa k +of size +.Fa s . +All bytes are significant in key comparison. +.Pp +.Fn ohash_find +retrieves an element from a slot +.Fa i +returned by the +.Fn ohash_lookup* +functions. +It returns +.Dv NULL +if the slot is empty. +.Pp +.Fn ohash_insert +inserts a new element +.Fa p +at slot +.Fa i . +Slot +.Fa i +must be empty and element +.Fa p +must have a key corresponding to the +.Fn ohash_lookup* +call. +.Pp +.Fn ohash_remove +removes the element at slot +.Fa i . +It returns the removed element, for user code to dispose of, or +.Dv NULL +if the slot was empty. +.Pp +.Fn ohash_first +and +.Fn ohash_next +can be used to access all elements in an ohash table, like this: +.Bd -literal -offset indent +for (n = ohash_first(h, &i); n != NULL; n = ohash_next(h, &i)) + do_something_with(n); +.Ed +.Pp +.Fa i +points to an auxiliary unsigned integer used to record the current position +in the ohash table. +Those functions are safe to use even while entries are added to/removed +from the table, but in such a case they don't guarantee that new entries +will be returned. +As a special case, they can safely be used to free elements in the table. +.Pp +.Fn ohash_entries +returns the number of elements in the hash table. +.Sh STORAGE HANDLING +Only +.Fn ohash_init , +.Fn ohash_insert , +.Fn ohash_remove +and +.Fn ohash_delete +may call the user-supplied memory functions. +It is the responsibility of the user memory allocation code to verify +that those calls did not fail. +.Pp +If memory allocation fails, +.Fn ohash_init +returns a useless hash table. +.Fn ohash_insert +and +.Fn ohash_remove +still perform the requested operation, but the returned table should be +considered read-only. +It can still be accessed by +.Fn ohash_lookup* , +.Fn ohash_find , +.Fn ohash_first +and +.Fn ohash_next +to dump relevant information to disk before aborting. +.Sh THREAD SAFETY +The open hashing functions are not thread-safe by design. +In particular, in a threaded environment, there is no guarantee that a +.Qq slot +will not move between a +.Fn ohash_lookup* +and a +.Fn ohash_find , +.Fn ohash_insert +or +.Fn ohash_remove +call. +.Pp +Multi-threaded applications should explicitly protect ohash table access. +.Sh SEE ALSO +.Xr ohash_interval 3 +.Rs +.%A Donald E. Knuth +.%B The Art of Computer Programming +.%V Vol. 3 +.%P pp 506-550 +.%D 1973 +.Re +.Sh STANDARDS +Those functions are completely non-standard and should be avoided in +portable programs. +.Sh HISTORY +Those functions were designed and written for +.Ox +.Xr make 1 +by Marc Espie in 1999. diff --git a/usr.bin/m4/lib/ohash_init.c b/usr.bin/m4/lib/ohash_init.c new file mode 100644 index 0000000..4d24fa4 --- /dev/null +++ b/usr.bin/m4/lib/ohash_init.c @@ -0,0 +1,41 @@ +/* $OpenBSD: ohash_init.c,v 1.2 2004/06/22 20:00:16 espie Exp $ */ +/* ex:ts=8 sw=4: + */ + +/* Copyright (c) 1999, 2004 Marc Espie <espie@openbsd.org> + * + * Permission to use, copy, modify, and distribute this software for any + * purpose with or without fee is hereby granted, provided that the above + * copyright notice and this permission notice appear in all copies. + * + * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES + * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR + * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES + * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN + * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF + * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. + */ + +#include "ohash_int.h" + +void +ohash_init(struct ohash *h, unsigned int size, struct ohash_info *info) +{ + h->size = 1UL << size; + if (h->size < MINSIZE) + h->size = MINSIZE; +#ifdef STATS_HASH + STAT_HASH_CREATION++; + STAT_HASH_SIZE += h->size; +#endif + /* Copy info so that caller may free it. */ + h->info.key_offset = info->key_offset; + h->info.halloc = info->halloc; + h->info.hfree = info->hfree; + h->info.alloc = info->alloc; + h->info.data = info->data; + h->t = (h->info.halloc)(sizeof(struct _ohash_record) * h->size, + h->info.data); + h->total = h->deleted = 0; +} diff --git a/usr.bin/m4/lib/ohash_int.h b/usr.bin/m4/lib/ohash_int.h new file mode 100644 index 0000000..88c818b --- /dev/null +++ b/usr.bin/m4/lib/ohash_int.h @@ -0,0 +1,23 @@ +/* $OpenBSD: ohash_int.h,v 1.3 2006/01/16 15:52:25 espie Exp $ */ + +#if HAVE_NBTOOL_CONFIG_H +#include "nbtool_config.h" +#endif + +#include <stddef.h> +#include <stdint.h> +#include <stdlib.h> +#include <string.h> +#include "ohash.h" + +struct _ohash_record { + u_int32_t hv; + const char *p; +}; + +#define DELETED ((const char *)h) +#define NONE (h->size) + +/* Don't bother changing the hash table if the change is small enough. */ +#define MINSIZE (1UL << 4) +#define MINDELETED 4 diff --git a/usr.bin/m4/lib/ohash_interval.3 b/usr.bin/m4/lib/ohash_interval.3 new file mode 100644 index 0000000..2e56ca5 --- /dev/null +++ b/usr.bin/m4/lib/ohash_interval.3 @@ -0,0 +1,90 @@ +.\" $OpenBSD: ohash_interval.3,v 1.11 2007/05/31 19:19:30 jmc Exp $ +.\" Copyright (c) 2001 Marc Espie <espie@openbsd.org> +.\" +.\" Permission to use, copy, modify, and distribute this software for any +.\" purpose with or without fee is hereby granted, provided that the above +.\" copyright notice and this permission notice appear in all copies. +.\" +.\" THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES +.\" WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF +.\" MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR +.\" ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES +.\" WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN +.\" ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF +.\" OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. +.\" +.Dd $Mdocdate: May 31 2007 $ +.Dt OPEN_HASH_HELPER 3 +.Os +.Sh NAME +.Nm ohash_interval , +.Nm ohash_create_entry , +.Nm ohash_qlookup , +.Nm ohash_qlookupi +.Nd helper functions for open hashing +.Sh SYNOPSIS +.Fd #include <stdint.h> +.Fd #include <stddef.h> +.Fd #include <ohash.h> +.Ft u_int32_t +.Fn ohash_interval "const char *start" "const char **pend" +.Ft "void *" +.Fn ohash_create_entry "struct ohash_info *info" "const char *start" "const char **pend" +.Ft "unsigned int" +.Fn ohash_qlookupi "struct ohash *h" "const char *start" "const char **pend" +.Ft "unsigned int" +.Fn ohash_qlookup "struct ohash *h" "const char *start" +.Sh DESCRIPTION +These functions are commonly used to simplify open hashing usage, and use +similar conventions. +They operate indifferently on NUL-terminated strings +.Po +by setting +.Fa *pend += +.Dv NULL +.Pc +or memory ranges +.Po +delimited by +.Fa start +and +.Fa *pend +.Pc . +For NUL-terminated strings, as a side effect, those functions +set +.Fa *pend +to the terminating NUL byte. +.Pp +.Fn ohash_interval +is a simple hashing function that yields good results on common data sets. +.Pp +.Fn ohash_create_entry +can be used to create a new record with a given key. +In that case, +the alloc field of +.Fa info +should point to a +.Xr malloc 3 Ns -like +function to allocate the storage. +.Pp +.Fn ohash_qlookupi +is a wrapper function that simply calls +.Fn ohash_interval +and +.Fn ohash_lookup_interval . +.Pp +.Fn ohash_qlookup +is a variation on +.Fn ohash_qlookupi +designed for NUL-terminated strings. +.Sh SEE ALSO +.Xr ohash_init 3 +.Sh STANDARDS +Those functions are completely non-standard and should be avoided in +portable programs. +.Sh HISTORY +Those functions were designed and written for +.Ox +.Xr make 1 +by Marc Espie in 1999. diff --git a/usr.bin/m4/lib/ohash_interval.c b/usr.bin/m4/lib/ohash_interval.c new file mode 100644 index 0000000..2434d3b --- /dev/null +++ b/usr.bin/m4/lib/ohash_interval.c @@ -0,0 +1,36 @@ +/* $OpenBSD: ohash_interval.c,v 1.3 2006/01/16 15:52:25 espie Exp $ */ +/* ex:ts=8 sw=4: + */ + +/* Copyright (c) 1999, 2004 Marc Espie <espie@openbsd.org> + * + * Permission to use, copy, modify, and distribute this software for any + * purpose with or without fee is hereby granted, provided that the above + * copyright notice and this permission notice appear in all copies. + * + * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES + * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR + * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES + * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN + * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF + * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. + */ + +#include "ohash_int.h" + +uint32_t +ohash_interval(const char *s, const char **e) +{ + uint32_t k; + + if (!*e) + *e = s + strlen(s); + if (s == *e) + k = 0; + else + k = *s++; + while (s != *e) + k = ((k << 2) | (k >> 30)) ^ *s++; + return k; +} diff --git a/usr.bin/m4/lib/ohash_lookup_interval.c b/usr.bin/m4/lib/ohash_lookup_interval.c new file mode 100644 index 0000000..e497801 --- /dev/null +++ b/usr.bin/m4/lib/ohash_lookup_interval.c @@ -0,0 +1,68 @@ +/* $OpenBSD: ohash_lookup_interval.c,v 1.3 2006/01/16 15:52:25 espie Exp $ */ +/* ex:ts=8 sw=4: + */ + +/* Copyright (c) 1999, 2004 Marc Espie <espie@openbsd.org> + * + * Permission to use, copy, modify, and distribute this software for any + * purpose with or without fee is hereby granted, provided that the above + * copyright notice and this permission notice appear in all copies. + * + * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES + * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR + * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES + * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN + * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF + * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. + */ + +#include "ohash_int.h" + +unsigned int +ohash_lookup_interval(struct ohash *h, const char *start, const char *end, + uint32_t hv) +{ + unsigned int i, incr; + unsigned int empty; + +#ifdef STATS_HASH + STAT_HASH_LOOKUP++; +#endif + empty = NONE; + i = hv % h->size; + incr = ((hv % (h->size-2)) & ~1) + 1; + while (h->t[i].p != NULL) { +#ifdef STATS_HASH + STAT_HASH_LENGTH++; +#endif + if (h->t[i].p == DELETED) { + if (empty == NONE) + empty = i; + } else if (h->t[i].hv == hv && + strncmp(h->t[i].p+h->info.key_offset, start, + end - start) == 0 && + (h->t[i].p+h->info.key_offset)[end-start] == '\0') { + if (empty != NONE) { + h->t[empty].hv = hv; + h->t[empty].p = h->t[i].p; + h->t[i].p = DELETED; + return empty; + } else { +#ifdef STATS_HASH + STAT_HASH_POSITIVE++; +#endif + return i; + } + } + i += incr; + if (i >= h->size) + i -= h->size; + } + + /* Found an empty position. */ + if (empty != NONE) + i = empty; + h->t[i].hv = hv; + return i; +} diff --git a/usr.bin/m4/lib/ohash_lookup_memory.c b/usr.bin/m4/lib/ohash_lookup_memory.c new file mode 100644 index 0000000..93ded5a --- /dev/null +++ b/usr.bin/m4/lib/ohash_lookup_memory.c @@ -0,0 +1,64 @@ +/* $OpenBSD: ohash_lookup_memory.c,v 1.3 2006/01/16 15:52:25 espie Exp $ */ +/* ex:ts=8 sw=4: + */ + +/* Copyright (c) 1999, 2004 Marc Espie <espie@openbsd.org> + * + * Permission to use, copy, modify, and distribute this software for any + * purpose with or without fee is hereby granted, provided that the above + * copyright notice and this permission notice appear in all copies. + * + * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES + * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR + * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES + * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN + * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF + * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. + */ + +#include "ohash_int.h" + +unsigned int +ohash_lookup_memory(struct ohash *h, const char *k, size_t size, uint32_t hv) +{ + unsigned int i, incr; + unsigned int empty; + +#ifdef STATS_HASH + STAT_HASH_LOOKUP++; +#endif + empty = NONE; + i = hv % h->size; + incr = ((hv % (h->size-2)) & ~1) + 1; + while (h->t[i].p != NULL) { +#ifdef STATS_HASH + STAT_HASH_LENGTH++; +#endif + if (h->t[i].p == DELETED) { + if (empty == NONE) + empty = i; + } else if (h->t[i].hv == hv && + memcmp(h->t[i].p+h->info.key_offset, k, size) == 0) { + if (empty != NONE) { + h->t[empty].hv = hv; + h->t[empty].p = h->t[i].p; + h->t[i].p = DELETED; + return empty; + } else { +#ifdef STATS_HASH + STAT_HASH_POSITIVE++; +#endif + } return i; + } + i += incr; + if (i >= h->size) + i -= h->size; + } + + /* Found an empty position. */ + if (empty != NONE) + i = empty; + h->t[i].hv = hv; + return i; +} diff --git a/usr.bin/m4/lib/ohash_qlookup.c b/usr.bin/m4/lib/ohash_qlookup.c new file mode 100644 index 0000000..58b3454 --- /dev/null +++ b/usr.bin/m4/lib/ohash_qlookup.c @@ -0,0 +1,27 @@ +/* $OpenBSD: ohash_qlookup.c,v 1.2 2004/06/22 20:00:17 espie Exp $ */ +/* ex:ts=8 sw=4: + */ + +/* Copyright (c) 1999, 2004 Marc Espie <espie@openbsd.org> + * + * Permission to use, copy, modify, and distribute this software for any + * purpose with or without fee is hereby granted, provided that the above + * copyright notice and this permission notice appear in all copies. + * + * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES + * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR + * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES + * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN + * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF + * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. + */ + +#include "ohash_int.h" + +unsigned int +ohash_qlookup(struct ohash *h, const char *s) +{ + const char *e = NULL; + return ohash_qlookupi(h, s, &e); +} diff --git a/usr.bin/m4/lib/ohash_qlookupi.c b/usr.bin/m4/lib/ohash_qlookupi.c new file mode 100644 index 0000000..2acec4e --- /dev/null +++ b/usr.bin/m4/lib/ohash_qlookupi.c @@ -0,0 +1,29 @@ +/* $OpenBSD: ohash_qlookupi.c,v 1.2 2004/06/22 20:00:17 espie Exp $ */ +/* ex:ts=8 sw=4: + */ + +/* Copyright (c) 1999, 2004 Marc Espie <espie@openbsd.org> + * + * Permission to use, copy, modify, and distribute this software for any + * purpose with or without fee is hereby granted, provided that the above + * copyright notice and this permission notice appear in all copies. + * + * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES + * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR + * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES + * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN + * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF + * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. + */ + +#include "ohash_int.h" + +unsigned int +ohash_qlookupi(struct ohash *h, const char *s, const char **e) +{ + u_int32_t hv; + + hv = ohash_interval(s, e); + return ohash_lookup_interval(h, s, *e, hv); +} diff --git a/usr.bin/m4/lib/strtonum.c b/usr.bin/m4/lib/strtonum.c new file mode 100644 index 0000000..8d11712 --- /dev/null +++ b/usr.bin/m4/lib/strtonum.c @@ -0,0 +1,73 @@ +/* $OpenBSD: strtonum.c,v 1.6 2004/08/03 19:38:01 millert Exp $ */ + +/* + * Copyright (c) 2004 Ted Unangst and Todd Miller + * All rights reserved. + * + * Permission to use, copy, modify, and distribute this software for any + * purpose with or without fee is hereby granted, provided that the above + * copyright notice and this permission notice appear in all copies. + * + * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES + * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR + * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES + * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN + * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF + * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. + */ +#if HAVE_NBTOOL_CONFIG_H +#include "nbtool_config.h" +#endif +#include <sys/cdefs.h> +__RCSID("$NetBSD: strtonum.c,v 1.2 2009/10/26 21:14:18 christos Exp $"); +#include <errno.h> +#include <limits.h> +#include <stdlib.h> + +#define INVALID 1 +#define TOOSMALL 2 +#define TOOLARGE 3 + +long long +strtonum(const char *numstr, long long minval, long long maxval, + const char **errstrp); +long long +strtonum(const char *numstr, long long minval, long long maxval, + const char **errstrp) +{ + long long ll = 0; + char *ep; + int error = 0; + struct errval { + const char *errstr; + int err; + } ev[4] = { + { NULL, 0 }, + { "invalid", EINVAL }, + { "too small", ERANGE }, + { "too large", ERANGE }, + }; + + ev[0].err = errno; + errno = 0; + if (minval > maxval) + error = INVALID; + else { + ll = strtoll(numstr, &ep, 10); + if (numstr == ep || *ep != '\0') + error = INVALID; + else if ((ll == LLONG_MIN && errno == ERANGE) || ll < minval) + error = TOOSMALL; + else if ((ll == LLONG_MAX && errno == ERANGE) || ll > maxval) + error = TOOLARGE; + } + if (errstrp != NULL) + *errstrp = ev[error].errstr; + errno = ev[error].err; + if (error) + ll = 0; + + return (ll); +} + diff --git a/usr.bin/m4/look.c b/usr.bin/m4/look.c new file mode 100644 index 0000000..de2c282 --- /dev/null +++ b/usr.bin/m4/look.c @@ -0,0 +1,315 @@ +/* $NetBSD: look.c,v 1.13 2016/01/16 17:00:07 christos Exp $ */ +/* $OpenBSD: look.c,v 1.21 2009/10/14 17:23:17 sthen Exp $ */ + +/* + * Copyright (c) 1989, 1993 + * The Regents of the University of California. All rights reserved. + * + * This code is derived from software contributed to Berkeley by + * Ozan Yigit at York University. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + */ + +/* + * look.c + * Facility: m4 macro processor + * by: oz + */ +#if HAVE_NBTOOL_CONFIG_H +#include "nbtool_config.h" +#endif +#include <sys/cdefs.h> +__RCSID("$NetBSD: look.c,v 1.13 2016/01/16 17:00:07 christos Exp $"); +#include <stdio.h> +#include <stdlib.h> +#include <stdint.h> +#include <stddef.h> +#include <string.h> +#include <ohash.h> +#include "mdef.h" +#include "stdd.h" +#include "extern.h" + +static void *hash_alloc(size_t, void *); +static void hash_free(void *, size_t, void *); +static void *element_alloc(size_t, void *); +static void setup_definition(struct macro_definition *, const char *, + const char *); + +static struct ohash_info macro_info = { + offsetof(struct ndblock, name), + NULL, hash_alloc, hash_free, element_alloc }; + +struct ohash macros; + +/* Support routines for hash tables. */ +void * +hash_alloc(size_t s, void *u UNUSED) +{ + void *storage = xalloc(s, "hash alloc"); + if (storage) + memset(storage, 0, s); + return storage; +} + +void +hash_free(void *p, size_t s UNUSED, void *u UNUSED) +{ + free(p); +} + +void * +element_alloc(size_t s, void *u UNUSED) +{ + return xalloc(s, "element alloc"); +} + +void +init_macros(void) +{ + ohash_init(¯os, 10, ¯o_info); +} + +/* + * find name in the hash table + */ +ndptr +lookup(const char *name) +{ + return ohash_find(¯os, ohash_qlookup(¯os, name)); +} + +struct macro_definition * +lookup_macro_definition(const char *name) +{ + ndptr p; + + p = ohash_find(¯os, ohash_qlookup(¯os, name)); + if (p) + return p->d; + else + return NULL; +} + +static void +setup_definition(struct macro_definition *d, const char *defn, const char *name) +{ + ndptr p; + + if (strncmp(defn, BUILTIN_MARKER, sizeof(BUILTIN_MARKER)-1) == 0 && + (p = macro_getbuiltin(defn+sizeof(BUILTIN_MARKER)-1)) != NULL) { + d->type = macro_builtin_type(p); + d->defn = xstrdup(defn+sizeof(BUILTIN_MARKER)-1); + } else { + if (!*defn) + d->defn = xstrdup(null); + else + d->defn = xstrdup(defn); + d->type = MACRTYPE; + } + if (STREQ(name, defn)) + d->type |= RECDEF; +} + +static ndptr +create_entry(const char *name) +{ + const char *end = NULL; + unsigned int i; + ndptr n; + + i = ohash_qlookupi(¯os, name, &end); + n = ohash_find(¯os, i); + if (n == NULL) { + n = ohash_create_entry(¯o_info, name, &end); + ohash_insert(¯os, i, n); + n->trace_flags = FLAG_NO_TRACE; + n->builtin_type = MACRTYPE; + n->d = NULL; + } + return n; +} + +void +macro_define(const char *name, const char *defn) +{ + ndptr n = create_entry(name); + if (n->d != NULL) { + if (n->d->defn != null) + free(n->d->defn); + } else { + n->d = xalloc(sizeof(struct macro_definition), NULL); + n->d->next = NULL; + } + setup_definition(n->d, defn, name); +} + +void +macro_pushdef(const char *name, const char *defn) +{ + ndptr n; + struct macro_definition *d; + + n = create_entry(name); + d = xalloc(sizeof(struct macro_definition), NULL); + d->next = n->d; + n->d = d; + setup_definition(n->d, defn, name); +} + +void +macro_undefine(const char *name) +{ + ndptr n = lookup(name); + if (n != NULL) { + struct macro_definition *r, *r2; + + for (r = n->d; r != NULL; r = r2) { + r2 = r->next; + if (r->defn != null) + free(r->defn); + free(r); + } + n->d = NULL; + } +} + +void +macro_popdef(const char *name) +{ + ndptr n = lookup(name); + + if (n != NULL) { + struct macro_definition *r = n->d; + if (r != NULL) { + n->d = r->next; + if (r->defn != null) + free(r->defn); + free(r); + } + } +} + +void +macro_for_all(void (*f)(const char *, struct macro_definition *)) +{ + ndptr n; + unsigned int i; + + for (n = ohash_first(¯os, &i); n != NULL; + n = ohash_next(¯os, &i)) + if (n->d != NULL) + f(n->name, n->d); +} + +void +setup_builtin(const char *name, unsigned int type) +{ + ndptr n; + char *name2; + + if (prefix_builtins) { + name2 = xalloc(strlen(name)+3+1, NULL); + memcpy(name2, "m4_", 3); + memcpy(name2 + 3, name, strlen(name)+1); + } else + name2 = xstrdup(name); + + n = create_entry(name2); + n->builtin_type = type; + n->d = xalloc(sizeof(struct macro_definition), NULL); + n->d->defn = name2; + n->d->type = type; + n->d->next = NULL; +} + +void +mark_traced(const char *name, int on) +{ + ndptr p; + unsigned int i; + + if (name == NULL) { + if (on) + trace_flags |= TRACE_ALL; + else + trace_flags &= ~TRACE_ALL; + for (p = ohash_first(¯os, &i); p != NULL; + p = ohash_next(¯os, &i)) + p->trace_flags = FLAG_NO_TRACE; + } else { + p = create_entry(name); + p->trace_flags = on; + } +} + +ndptr +macro_getbuiltin(const char *name) +{ + ndptr p; + + p = lookup(name); + if (p == NULL || p->builtin_type == MACRTYPE) + return NULL; + else + return p; +} + +#ifdef REAL_FREEZE +static void +recurse(FILE *f, ndptr n, struct macro_definition *d) +{ + if (d->next != NULL) + recurse(f, n, d->next); + + // skip built-ins, because it is cheaper to do so + // and initialize them manually + if (d->type & (NOARGS|NEEDARGS)) + return; + fprintf(f, "%c%zu,%zu\n%s%s\n", + (d->type & (NOARGS|NEEDARGS)) ? 'F' : 'T', + strlen(n->name), strlen(d->defn), + n->name, d->defn); +} + +static void +dump_entry(FILE *f, ndptr n) +{ + if (n->d == NULL) + return; + recurse(f, n, n->d); +} + +void +dump_state(FILE *f) +{ + ndptr n; + unsigned int i; + for (n = ohash_first(¯os, &i); n != NULL; + n = ohash_next(¯os, &i)) + dump_entry(f, n); +} +#endif diff --git a/usr.bin/m4/m4.1 b/usr.bin/m4/m4.1 new file mode 100644 index 0000000..f033cee --- /dev/null +++ b/usr.bin/m4/m4.1 @@ -0,0 +1,522 @@ +.\" $NetBSD: m4.1,v 1.27 2016/01/17 11:24:28 wiz Exp $ +.\" @(#) $OpenBSD: m4.1,v 1.56 2009/10/14 17:19:47 sthen Exp $ +.\" +.\" Copyright (c) 1989, 1993 +.\" The Regents of the University of California. All rights reserved. +.\" +.\" This code is derived from software contributed to Berkeley by +.\" Ozan Yigit at York University. +.\" +.\" Redistribution and use in source and binary forms, with or without +.\" modification, are permitted provided that the following conditions +.\" are met: +.\" 1. Redistributions of source code must retain the above copyright +.\" notice, this list of conditions and the following disclaimer. +.\" 2. Redistributions in binary form must reproduce the above copyright +.\" notice, this list of conditions and the following disclaimer in the +.\" documentation and/or other materials provided with the distribution. +.\" 3. Neither the name of the University nor the names of its contributors +.\" may be used to endorse or promote products derived from this software +.\" without specific prior written permission. +.\" +.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND +.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +.\" ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE +.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL +.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS +.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) +.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT +.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY +.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF +.\" SUCH DAMAGE. +.\" +.Dd January 16, 2015 +.Dt M4 1 +.Os +.Sh NAME +.Nm m4 +.Nd macro language processor +.Sh SYNOPSIS +.Nm m4 +.Op Fl EGgiPQsv +.Oo +.Sm off +.Fl D Ar name Op No = Ar value +.Sm on +.Oc +.Op Fl d Ar flags +.Op Fl e Ar filename +.Op Fl F Ar filename +.Op Fl I Ar dirname +.Op Fl L Ar number +.Op Fl o Ar filename +.Op Fl R Ar filename +.Op Fl t Ar macro +.Op Fl U Ns Ar name +.Op Ar +.Sh DESCRIPTION +The +.Nm m4 +utility is a macro processor that can be used as a front end to any +language (e.g., C, ratfor, fortran, lex, and yacc). +If no input files are given, +.Nm m4 +reads from the standard input, +otherwise files specified on the command line are +processed in the given order. +Input files can be regular files, files in the m4 include paths, or a +single dash +.Pq Sq - , +denoting standard input. +.Nm m4 +writes +the processed text to the standard output, unless told otherwise. +.Pp +Macro calls have the form name(argument1[, argument2, ..., argumentN]). +.Pp +There cannot be any space following the macro name and the open +parenthesis +.Sq \&( . +If the macro name is not followed by an open +parenthesis it is processed with no arguments. +.Pp +Macro names consist of a leading alphabetic or underscore +possibly followed by alphanumeric or underscore characters, e.g., +valid macro names match the pattern +.Dq [a-zA-Z_][a-zA-Z0-9_]* . +.Pp +In arguments to macros, leading unquoted space, tab, and newline +.Pq Sq \en +characters are ignored. +To quote strings, use left and right single quotes +.Po e.g.,\ \& +.Sq "\ this is a string with a leading space" +.Pc . +You can change the quote characters with the +.Ic changequote +built-in macro. +.Pp +Most built-ins don't make any sense without arguments, and hence are not +recognized as special when not followed by an open parenthesis. +.Pp +The options are as follows: +.Bl -tag -width Ds +.It Fl D , Fl Fl define Ar name Ns Op Pf = Ns Ar value +Define the symbol +.Ar name +to have some value (or +.Dv NULL ) . +.It Fl d , Fl Fl debug Ar "flags" +Set trace flags. +.Ar flags +may hold the following: +.Bl -tag -width Ds +.It Ar a +print macro arguments. +.It Ar c +print macro expansion over several lines. +.It Ar e +print result of macro expansion. +.It Ar f +print filename location. +.It Ar l +print line number. +.It Ar q +quote arguments and expansion with the current quotes. +.It Ar t +start with all macros traced. +.It Ar x +number macro expansions. +.It Ar V +turn on all options. +.El +.Pp +By default, trace is set to +.Qq eq . +.It Fl E , Fl Fl fatal-warnings +Warnings make +.Nm +exit. +.It Fl e , Fl Fl error-output Ar filename +Redirect error output to filename. +.It Fl F , Fl Fl freeze-state Ar filename +Save the input state to +.Ar filename . +.It Fl G , Fl Fl traditional +Disable GNU-m4 extensions. +.It Fl g , Fl Fl gnu +Activate GNU-m4 compatibility mode. +In this mode, translit handles simple character +ranges (e.g., a-z), regular expressions mimic emacs behavior, +multiple m4wrap calls are handled as a stack, +the number of diversions is unlimited, +empty names for macro definitions are allowed, +and eval understands +.Sq 0rbase:value +numbers. +.It Fl Fl help +Print help message and exit. +.It Fl I , Fl Fl include Ar "dirname" +Add directory +.Ar dirname +to the include path. +.It Fl i , Fl Fl interactive +Set unbuffered output, disable tty signals. +.It Fl L , Fl Fl nesting-limit +Set the nesting limit in macro expansions. +This is unimplemented and unlimited. +.It Fl o Ar filename +Send trace output to +.Ar filename . +.It Fl P , Fl Fl prefix-builtins +Prefix all built-in macros with +.Sq m4_ . +For example, instead of writing +.Ic define , +use +.Ic m4_define . +.It Fl Q , Fl Fl quiet , Fl Fl silent +Don't print warnings. +.It Fl R , Fl Fl reload-state Ar filename +Reload a previously saved state from +.Ar filename . +.It Fl s , Fl Fl synclines +Output line synchronization directives, suitable for +.Xr cpp 1 . +.It Fl t , Fl Fl trace Ar macro +Turn tracing on for +.Ar macro . +.It Fl U , Fl Fl undefine Ar "name" +Undefine the symbol +.Ar name . +.It Fl v , Fl Fl version +Print the version and exit. +.El +.Sh SYNTAX +.Nm m4 +provides the following built-in macros. +They may be redefined, losing their original meaning. +Return values are null unless otherwise stated. +.Bl -tag -width changequote +.It Fn builtin name +Calls a built-in by its +.Fa name , +overriding possible redefinitions. +.It Fn changecom startcomment endcomment +Changes the start comment and end comment sequences. +Comment sequences may be up to five characters long. +The default values are the hash sign +and the newline character. +.Bd -literal -offset indent +# This is a comment +.Ed +.Pp +With no arguments, comments are turned off. +With one single argument, the end comment sequence is set +to the newline character. +.It Fn changequote beginquote endquote +Defines the open quote and close quote sequences. +Quote sequences may be up to five characters long. +The default values are the backquote character and the quote +character. +.Bd -literal -offset indent +`Here is a quoted string' +.Ed +.Pp +With no arguments, the default quotes are restored. +With one single argument, the close quote sequence is set +to the newline character. +.It Fn decr arg +Decrements the argument +.Fa arg +by 1. +The argument +.Fa arg +must be a valid numeric string. +.It Fn define name value +Define a new macro named by the first argument +.Fa name +to have the +value of the second argument +.Fa value . +Each occurrence of +.Sq $n +(where +.Ar n +is 0 through 9) is replaced by the +.Ar n Ns 'th +argument. +.Sq $0 +is the name of the calling macro. +Undefined arguments are replaced by a null string. +.Sq $# +is replaced by the number of arguments; +.Sq $* +is replaced by all arguments comma separated; +.Sq $@ +is the same as +.Sq $* +but all arguments are quoted against further expansion. +.It Fn defn name ... +Returns the quoted definition for each argument. +This can be used to rename +macro definitions (even for built-in macros). +.It Fn divert num +There are 10 output queues (numbered 0-9). +At the end of processing +.Nm m4 +concatenates all the queues in numerical order to produce the +final output. +Initially the output queue is 0. +The divert +macro allows you to select a new output queue (an invalid argument +passed to divert causes output to be discarded). +.It Ic divnum +Returns the current output queue number. +.It Ic dnl +Discard input characters up to and including the next newline. +.It Fn dumpdef name ... +Prints the names and definitions for the named items, or for everything +if no arguments are passed. +.It Fn errprint msg +Prints the first argument on the standard error output stream. +.It Fn esyscmd cmd +Passes its first argument to a shell and returns the shell's standard output. +Note that the shell shares its standard input and standard error with +.Nm m4 . +.It Fn eval expr[,radix[,minimum]] +Computes the first argument as an arithmetic expression using 32-bit +arithmetic. +Operators are the standard C ternary, arithmetic, logical, +shift, relational, bitwise, and parentheses operators. +You can specify +octal, decimal, and hexadecimal numbers as in C. +The optional second argument +.Fa radix +specifies the radix for the result and the optional third argument +.Fa minimum +specifies the minimum number of digits in the result. +.It Fn expr expr +This is an alias for +.Ic eval . +.It Fn format formatstring arg1 ... +Returns +.Fa formatstring +with escape sequences substituted with +.Fa arg1 +and following arguments, in a way similar to +.Xr printf 3 . +This built-in is only available in GNU-m4 compatibility mode, and the only +parameters implemented are there for autoconf compatibility: +left-padding flag, an optional field width, a maximum field width, +*-specified field widths, and the %s and %c data type. +.It Fn ifdef name yes no +If the macro named by the first argument is defined then return the second +argument, otherwise the third. +If there is no third argument, the value is +.Dv NULL . +The word +.Qq unix +is predefined. +.It Fn ifelse a b yes ... +If the first argument +.Fa a +matches the second argument +.Fa b +then +.Fn ifelse +returns +the third argument +.Fa yes . +If the match fails the three arguments are +discarded and the next three arguments are used until there is +zero or one arguments left, either this last argument or +.Dv NULL +is returned if no other matches were found. +.It Fn include name +Returns the contents of the file specified in the first argument. +If the file is not found as is, look through the include path: +first the directories specified with +.Fl I +on the command line, then the environment variable +.Ev M4PATH , +as a colon-separated list of directories. +Include aborts with an error message if the file cannot be included. +.It Fn incr arg +Increments the argument by 1. +The argument must be a valid numeric string. +.It Fn index string substring +Returns the index of the second argument in the first argument (e.g., +.Ic index(the quick brown fox jumped, fox) +returns 16). +If the second +argument is not found index returns \-1. +.It Fn indir macro arg1 ... +Indirectly calls the macro whose name is passed as the first argument, +with the remaining arguments passed as first, ... arguments. +.It Fn len arg +Returns the number of characters in the first argument. +Extra arguments +are ignored. +.It Fn m4exit code +Immediately exits with the return value specified by the first argument, +0 if none. +.It Fn m4wrap todo +Allows you to define what happens at the final +.Dv EOF , +usually for cleanup purposes (e.g., +.Ic m4wrap("cleanup(tempfile)") +causes the macro cleanup to be +invoked after all other processing is done). +.Pp +Multiple calls to +.Fn m4wrap +get inserted in sequence at the final +.Dv EOF . +.It Fn maketemp template +Invokes +.Xr mkstemp 3 +on the first argument, and returns the modified string. +This can be used to create unique +temporary file names. +.It Fn paste file +Includes the contents of the file specified by the first argument without +any macro processing. +Aborts with an error message if the file cannot be +included. +.It Fn patsubst string regexp replacement +Substitutes a regular expression in a string with a replacement string. +Usual substitution patterns apply: an ampersand +.Pq Sq \&& +is replaced by the string matching the regular expression. +The string +.Sq \e# , +where +.Sq # +is a digit, is replaced by the corresponding back-reference. +.It Fn popdef arg ... +Restores the +.Ic pushdef Ns ed +definition for each argument. +.It Fn pushdef macro def +Takes the same arguments as +.Ic define , +but it saves the definition on a +stack for later retrieval by +.Fn popdef . +.It Fn regexp string regexp replacement +Finds a regular expression in a string. +If no further arguments are given, +it returns the first match position or \-1 if no match. +If a third argument +is provided, it returns the replacement string, with sub-patterns replaced. +.It Fn shift arg1 ... +Returns all but the first argument, the remaining arguments are +quoted and pushed back with commas in between. +The quoting +nullifies the effect of the extra scan that will subsequently be +performed. +.It Fn sinclude file +Similar to +.Ic include , +except it ignores any errors. +.It Fn spaste file +Similar to +.Fn paste , +except it ignores any errors. +.It Fn substr string offset length +Returns a substring of the first argument starting at the offset specified +by the second argument and the length specified by the third argument. +If no third argument is present it returns the rest of the string. +.It Fn syscmd cmd +Passes the first argument to the shell. +Nothing is returned. +.It Ic sysval +Returns the return value from the last +.Ic syscmd . +.It Fn traceon arg ... +Enables tracing of macro expansions for the given arguments, or for all +macros if no argument is given. +.It Fn traceoff arg ... +Disables tracing of macro expansions for the given arguments, or for all +macros if no argument is given. +.It Fn translit string mapfrom mapto +Transliterate the characters in the first argument from the set +given by the second argument to the set given by the third. +You cannot use +.Xr tr 1 +style abbreviations. +.It Fn undefine name1 ... +Removes the definition for the macros specified by its arguments. +.It Fn undivert arg ... +Flushes the named output queues (or all queues if no arguments). +.It Ic unix +A pre-defined macro for testing the OS platform. +.It Ic __line__ +Returns the current file's line number. +.It Ic __file__ +Returns the current file's name. +.El +.Sh STANDARDS +The +.Nm +utility is compliant with the +.St -p1003.1-2008 +specification. +.Pp +The flags +.Op Fl dgIot +and the macros +.Ic builtin , +.Ic esyscmd , +.Ic expr , +.Ic format , +.Ic indir , +.Ic paste , +.Ic patsubst , +.Ic regexp , +.Ic spaste , +.Ic unix , +.Ic __line__ , +and +.Ic __file__ +are extensions to that specification. +.Pp +The output format of tracing and of +.Ic dumpdef +are not specified in any standard, +are likely to change and should not be relied upon. +The current format of tracing is closely modelled on +.Nm gnu-m4 , +to allow +.Nm autoconf +to work. +.Pp +The built-ins +.Ic pushdef +and +.Ic popdef +handle macro definitions as a stack. +However, +.Ic define +interacts with the stack in an undefined way. +In this implementation, +.Ic define +replaces the top-most definition only. +Other implementations may erase all definitions on the stack instead. +.Pp +All built-ins do expand without arguments in many other +.Nm m4 . +.Pp +Many other +.Nm +have dire size limitations with respect to buffer sizes. +.Sh AUTHORS +.An -nosplit +.An Ozan Yigit Aq Mt oz@sis.yorku.ca +and +.An Richard A. O'Keefe Aq Mt ok@goanna.cs.rmit.OZ.AU . +.Pp +GNU-m4 compatibility extensions by +.An Marc Espie Aq Mt espie@cvs.openbsd.org . diff --git a/usr.bin/m4/main.c b/usr.bin/m4/main.c new file mode 100644 index 0000000..cba65ad --- /dev/null +++ b/usr.bin/m4/main.c @@ -0,0 +1,811 @@ +/* $OpenBSD: main.c,v 1.77 2009/10/14 17:19:47 sthen Exp $ */ +/* $NetBSD: main.c,v 1.46 2016/01/23 14:24:43 christos Exp $ */ + +/*- + * Copyright (c) 1989, 1993 + * The Regents of the University of California. All rights reserved. + * + * This code is derived from software contributed to Berkeley by + * Ozan Yigit at York University. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + */ + +/* + * main.c + * Facility: m4 macro processor + * by: oz + */ +#if HAVE_NBTOOL_CONFIG_H +#include "nbtool_config.h" +#endif +#include <sys/cdefs.h> +__RCSID("$NetBSD: main.c,v 1.46 2016/01/23 14:24:43 christos Exp $"); +#include <assert.h> +#include <signal.h> +#include <getopt.h> +#include <err.h> +#include <errno.h> +#include <unistd.h> +#include <stdio.h> +#include <ctype.h> +#include <string.h> +#include <stddef.h> +#include <stdint.h> +#include <stdlib.h> +#include <ohash.h> +#include "mdef.h" +#include "stdd.h" +#include "extern.h" +#include "pathnames.h" + +ndptr hashtab[HASHSIZE]; /* hash table for macros etc. */ +stae *mstack; /* stack of m4 machine */ +char *sstack; /* shadow stack, for string space extension */ +static size_t STACKMAX; /* current maximum size of stack */ +int sp; /* current m4 stack pointer */ +int fp; /* m4 call frame pointer */ +struct input_file infile[MAXINP];/* input file stack (0=stdin) */ +FILE **outfile; /* diversion array(0=bitbucket)*/ +int maxout; +FILE *active; /* active output file pointer */ +int ilevel = 0; /* input file stack pointer */ +int oindex = 0; /* diversion index.. */ +const char *null = ""; /* as it says.. just a null.. */ +char **m4wraps = NULL; /* m4wraps array. */ +int maxwraps = 0; /* size of m4wraps array */ +int wrapindex = 0; /* current offset in m4wraps */ +char lquote[MAXCCHARS+1] = {LQUOTE}; /* left quote character (`) */ +char rquote[MAXCCHARS+1] = {RQUOTE}; /* right quote character (') */ +char scommt[MAXCCHARS+1] = {SCOMMT}; /* start character for comment */ +char ecommt[MAXCCHARS+1] = {ECOMMT}; /* end character for comment */ +int synch_lines = 0; /* line synchronisation for C preprocessor */ +int prefix_builtins = 0; /* -P option to prefix builtin keywords */ +int fatal_warnings = 0; /* -E option to exit on warnings */ +int quiet = 0; /* -Q option to silence warnings */ +int nesting_limit = -1; /* -L for nesting limit */ +const char *freeze = NULL; /* -F to freeze state */ +const char *reload = NULL; /* -R to reload state */ +#ifndef REAL_FREEZE +FILE *freezef = NULL; +int thawing = 0; +#endif + +struct keyblk { + const char *knam; /* keyword name */ + int ktyp; /* keyword type */ +}; + +struct keyblk keywrds[] = { /* m4 keywords to be installed */ + { "include", INCLTYPE }, + { "sinclude", SINCTYPE }, + { "define", DEFITYPE }, + { "defn", DEFNTYPE }, + { "divert", DIVRTYPE | NOARGS }, + { "expr", EXPRTYPE }, + { "eval", EXPRTYPE }, + { "substr", SUBSTYPE }, + { "ifelse", IFELTYPE }, + { "ifdef", IFDFTYPE }, + { "len", LENGTYPE }, + { "incr", INCRTYPE }, + { "decr", DECRTYPE }, + { "dnl", DNLNTYPE | NOARGS }, + { "changequote", CHNQTYPE | NOARGS }, + { "changecom", CHNCTYPE | NOARGS }, + { "index", INDXTYPE }, +#ifdef EXTENDED + { "paste", PASTTYPE }, + { "spaste", SPASTYPE }, + /* Newer extensions, needed to handle gnu-m4 scripts */ + { "indir", INDIRTYPE}, + { "builtin", BUILTINTYPE}, + { "patsubst", PATSTYPE}, + { "regexp", REGEXPTYPE}, + { "esyscmd", ESYSCMDTYPE}, + { "__file__", FILENAMETYPE | NOARGS}, + { "__line__", LINETYPE | NOARGS}, +#endif + { "popdef", POPDTYPE }, + { "pushdef", PUSDTYPE }, + { "dumpdef", DUMPTYPE | NOARGS }, + { "shift", SHIFTYPE | NOARGS }, + { "translit", TRNLTYPE }, + { "undefine", UNDFTYPE }, + { "undivert", UNDVTYPE | NOARGS }, + { "divnum", DIVNTYPE | NOARGS }, + { "maketemp", MKTMTYPE }, + { "errprint", ERRPTYPE | NOARGS }, + { "m4wrap", M4WRTYPE | NOARGS }, + { "m4exit", EXITTYPE | NOARGS }, + { "syscmd", SYSCTYPE }, + { "sysval", SYSVTYPE | NOARGS }, + { "traceon", TRACEONTYPE | NOARGS }, + { "traceoff", TRACEOFFTYPE | NOARGS }, + +#if defined(unix) || defined(__unix__) + { "unix", SELFTYPE | NOARGS }, +#else +#ifdef vms + { "vms", SELFTYPE | NOARGS }, +#endif +#endif +}; + +#define MAXKEYS (sizeof(keywrds)/sizeof(struct keyblk)) + +#define MAXRECORD 50 +static struct position { + char *name; + unsigned long line; +} quotes[MAXRECORD], paren[MAXRECORD]; + +static void record(struct position *, int); +static void dump_stack(struct position *, int); + +static void macro(void); +static void initkwds(void); +static ndptr inspect(int, char *); +static int do_look_ahead(int, const char *); +static void reallyoutputstr(const char *); +static void reallyputchar(int); + +static void enlarge_stack(void); +static void help(void); + +static void +usage(FILE *f) +{ + fprintf(f, "Usage: %s [-EGgiPQsv] [-Dname[=value]] [-d flags] " + "[-I dirname] [-o filename] [-L limit]\n" + "\t[-t macro] [-Uname] [file ...]\n", getprogname()); +} + +__dead static void +onintr(int signo) +{ + char intrmessage[] = "m4: interrupted.\n"; + write(STDERR_FILENO, intrmessage, sizeof(intrmessage)-1); + _exit(1); +} + +#define OPT_HELP 1 + +struct option longopts[] = { + { "debug", optional_argument, 0, 'd' }, + { "define", required_argument, 0, 'D' }, + { "error-output", required_argument, 0, 'e' }, + { "fatal-warnings", no_argument, 0, 'E' }, + { "freeze-state", required_argument, 0, 'F' }, + { "gnu", no_argument, 0, 'g' }, + { "help", no_argument, 0, OPT_HELP }, + { "include", required_argument, 0, 'I' }, + { "interactive", no_argument, 0, 'i' }, + { "nesting-limit", required_argument, 0, 'L' }, + { "prefix-builtins", no_argument, 0, 'P' }, + { "quiet", no_argument, 0, 'Q' }, + { "reload-state", required_argument, 0, 'R' }, + { "silent", no_argument, 0, 'Q' }, + { "synclines", no_argument, 0, 's' }, + { "trace", required_argument, 0, 't' }, + { "traditional", no_argument, 0, 'G' }, + { "undefine", required_argument, 0, 'U' }, + { "version", no_argument, 0, 'v' }, +#ifdef notyet + { "arglength", required_argument, 0, 'l' }, + { "debugfile", optional_argument, 0, OPT_DEBUGFILE }, + { "hashsize", required_argument, 0, 'H' }, + { "warn-macro-sequence",optional_argument, 0, OPT_WARN_SEQUENCE }, +#endif + { 0, 0, 0, 0 }, +}; + +int +main(int argc, char *argv[]) +{ + int c; + int n; + char *p; + + setprogname(argv[0]); + + if (signal(SIGINT, SIG_IGN) != SIG_IGN) + signal(SIGINT, onintr); + + init_macros(); + initspaces(); + STACKMAX = INITSTACKMAX; + + mstack = (stae *)xalloc(sizeof(stae) * STACKMAX, NULL); + sstack = (char *)xalloc(STACKMAX, NULL); + + maxout = 0; + outfile = NULL; + resizedivs(MAXOUT); + + while ((c = getopt_long(argc, argv, "D:d:e:EF:GgI:iL:o:PR:Qst:U:v", + longopts, NULL)) != -1) + switch(c) { + case 'D': /* define something..*/ + for (p = optarg; *p; p++) + if (*p == '=') + break; + if (*p) + *p++ = EOS; + dodefine(optarg, p); + break; + case 'd': + set_trace_flags(optarg); + break; + case 'E': + fatal_warnings++; + break; + case 'e': + if (freopen(optarg, "w+", stderr) == NULL) + err(EXIT_FAILURE, "Can't redirect errors to `%s'", + optarg); + break; + case 'F': + freeze = optarg; +#ifndef REAL_FREEZE + if ((freezef = fopen(freeze, "w")) == NULL) + err(EXIT_FAILURE, "Can't open `%s'", freeze); +#endif + break; + case 'I': + addtoincludepath(optarg); + break; + case 'i': + setvbuf(stdout, NULL, _IONBF, 0); + signal(SIGINT, SIG_IGN); + break; + case 'G': + mimic_gnu = 0; + break; + case 'g': + mimic_gnu = 1; + break; + case 'L': + nesting_limit = atoi(optarg); + break; + case 'o': + trace_file(optarg); + break; + case 'P': + prefix_builtins = 1; + break; + case 'Q': + quiet++; + break; + case 'R': + reload = optarg; + break; + case 's': + synch_lines = 1; + break; + case 't': + mark_traced(optarg, 1); + break; + case 'U': /* undefine... */ + macro_popdef(optarg); + break; + case 'v': + fprintf(stderr, "%s version %d\n", getprogname(), + VERSION); + return EXIT_SUCCESS; + case OPT_HELP: + help(); + return EXIT_SUCCESS; + case '?': + default: + usage(stderr); + return EXIT_FAILURE; + } + +#ifdef REDIRECT + /* + * This is meant only for debugging; it makes all output + * go to a known file, even if the command line options + * send it elsewhere. It should not be turned of in production code. + */ + if (freopen("/tmp/m4", "w+", stderr) == NULL) + err(EXIT_FAILURE, "Can't redirect errors to `%s'", + "/tmp/m4"); +#endif + argc -= optind; + argv += optind; + + + initkwds(); + if (mimic_gnu) + setup_builtin("format", FORMATTYPE); + + active = stdout; /* default active output */ + bbase[0] = bufbase; + + if (reload) { +#ifdef REAL_FREEZE + thaw_state(reload); +#else + if (fopen_trypath(infile, reload) == NULL) + err(1, "Can't open `%s'", reload); + sp = -1; + fp = 0; + thawing = 1; + macro(); + thawing = 0; + release_input(infile); +#endif + } + + if (!argc) { + sp = -1; /* stack pointer initialized */ + fp = 0; /* frame pointer initialized */ + set_input(infile+0, stdin, "stdin"); + /* default input (naturally) */ + macro(); + } else + for (; argc--; ++argv) { + p = *argv; + if (p[0] == '-' && p[1] == EOS) + set_input(infile, stdin, "stdin"); + else if (fopen_trypath(infile, p) == NULL) + err(1, "%s", p); + sp = -1; + fp = 0; + macro(); + release_input(infile); + } + + if (wrapindex) { + int i; + + ilevel = 0; /* in case m4wrap includes.. */ + bufbase = bp = buf; /* use the entire buffer */ + if (mimic_gnu) { + while (wrapindex != 0) { + for (i = 0; i < wrapindex; i++) + pbstr(m4wraps[i]); + wrapindex =0; + macro(); + } + } else { + for (i = 0; i < wrapindex; i++) { + pbstr(m4wraps[i]); + macro(); + } + } + } + + if (active != stdout) + active = stdout; /* reset output just in case */ + for (n = 1; n < maxout; n++) /* default wrap-up: undivert */ + if (outfile[n] != NULL) + getdiv(n); + /* remove bitbucket if used */ + if (outfile[0] != NULL) { + (void) fclose(outfile[0]); + } + +#ifdef REAL_FREEZE + if (freeze) + freeze_state(freeze); +#else + if (freezef) + fclose(freezef); +#endif + + return 0; +} + +/* + * Look ahead for `token'. + * (on input `t == token[0]') + * Used for comment and quoting delimiters. + * Returns 1 if `token' present; copied to output. + * 0 if `token' not found; all characters pushed back + */ +static int +do_look_ahead(int t, const char *token) +{ + int i; + + assert((unsigned char)t == (unsigned char)token[0]); + + for (i = 1; *++token; i++) { + t = gpbc(); + if (t == EOF || (unsigned char)t != (unsigned char)*token) { + pushback(t); + while (--i) + pushback(*--token); + return 0; + } + } + return 1; +} + +#define LOOK_AHEAD(t, token) (t != EOF && \ + (unsigned char)(t)==(unsigned char)(token)[0] && \ + do_look_ahead(t,token)) + +/* + * macro - the work horse.. + */ +static void +macro(void) +{ + char token[MAXTOK+1]; + int t, l; + ndptr p; + int nlpar; + + cycle { + t = gpbc(); + + if (LOOK_AHEAD(t,lquote)) { /* strip quotes */ + nlpar = 0; + record(quotes, nlpar++); + /* + * Opening quote: scan forward until matching + * closing quote has been found. + */ + do { + + l = gpbc(); + if (LOOK_AHEAD(l,rquote)) { + if (--nlpar > 0) + outputstr(rquote); + } else if (LOOK_AHEAD(l,lquote)) { + record(quotes, nlpar++); + outputstr(lquote); + } else if (l == EOF) { + if (!quiet) { + if (nlpar == 1) + warnx("unclosed quote:"); + else + warnx( + "%d unclosed quotes:", + nlpar); + dump_stack(quotes, nlpar); + } + exit(EXIT_FAILURE); + } else { + if (nlpar > 0) { + if (sp < 0) + reallyputchar(l); + else + CHRSAVE(l); + } + } + } + while (nlpar != 0); + } else if (sp < 0 && LOOK_AHEAD(t, scommt)) { + reallyoutputstr(scommt); + + for(;;) { + t = gpbc(); + if (LOOK_AHEAD(t, ecommt)) { + reallyoutputstr(ecommt); + break; + } + if (t == EOF) + break; + reallyputchar(t); + } + } else if (t == '_' || isalpha(t)) { + p = inspect(t, token); + if (p != NULL) + pushback(l = gpbc()); + if (p == NULL || (l != LPAREN && + (macro_getdef(p)->type & NEEDARGS) != 0)) + outputstr(token); + else { + /* + * real thing.. First build a call frame: + */ + pushf(fp); /* previous call frm */ + pushf(macro_getdef(p)->type); /* type of the call */ + pushf(is_traced(p)); + pushf(0); /* parenthesis level */ + fp = sp; /* new frame pointer */ + /* + * now push the string arguments: + */ + pushs1(macro_getdef(p)->defn); /* defn string */ + pushs1((char *)macro_name(p)); /* macro name */ + pushs(ep); /* start next..*/ + + if (l != LPAREN && PARLEV == 0) { + /* no bracks */ + chrsave(EOS); + + if ((size_t)sp == STACKMAX) + errx(1, "internal stack overflow"); + eval((const char **) mstack+fp+1, 2, + CALTYP, TRACESTATUS); + + ep = PREVEP; /* flush strspace */ + sp = PREVSP; /* previous sp.. */ + fp = PREVFP; /* rewind stack...*/ + } + } + } else if (t == EOF) { + if (sp > -1 && ilevel <= 0) { + if (!quiet) { + warnx("unexpected end of input, " + "unclosed parenthesis:"); + dump_stack(paren, PARLEV); + } + exit(EXIT_FAILURE); + } + if (ilevel <= 0) + break; /* all done thanks.. */ + release_input(infile+ilevel--); + emit_synchline(); + bufbase = bbase[ilevel]; + continue; + } else if (sp < 0) { /* not in a macro at all */ + reallyputchar(t); /* output directly.. */ + } + + else switch(t) { + + case LPAREN: + if (PARLEV > 0) + chrsave(t); + while (isspace(l = gpbc())) /* skip blank, tab, nl.. */ + if (PARLEV > 0) + chrsave(l); + pushback(l); + record(paren, PARLEV++); + break; + + case RPAREN: + if (--PARLEV > 0) + chrsave(t); + else { /* end of argument list */ + chrsave(EOS); + + if ((size_t)sp == STACKMAX) + errx(1, "internal stack overflow"); + + eval((const char **) mstack+fp+1, sp-fp, + CALTYP, TRACESTATUS); + + ep = PREVEP; /* flush strspace */ + sp = PREVSP; /* previous sp.. */ + fp = PREVFP; /* rewind stack...*/ + } + break; + + case COMMA: + if (PARLEV == 1) { + chrsave(EOS); /* new argument */ + while (isspace(l = gpbc())) + ; + pushback(l); + pushs(ep); + } else + chrsave(t); + break; + + default: + if (LOOK_AHEAD(t, scommt)) { + char *q; + for (q = scommt; *q; q++) + chrsave(*q); + for(;;) { + t = gpbc(); + if (LOOK_AHEAD(t, ecommt)) { + for (q = ecommt; *q; q++) + chrsave(*q); + break; + } + if (t == EOF) + break; + CHRSAVE(t); + } + } else + CHRSAVE(t); /* stack the char */ + break; + } + } +} + +/* + * output string directly, without pushing it for reparses. + */ +void +outputstr(const char *s) +{ + if (sp < 0) + reallyoutputstr(s); + else + while (*s) + CHRSAVE(*s++); +} + +void +reallyoutputstr(const char *s) +{ + if (synch_lines) { + while (*s) { + fputc(*s, active); + if (*s++ == '\n') { + infile[ilevel].synch_lineno++; + if (infile[ilevel].synch_lineno != + infile[ilevel].lineno) + do_emit_synchline(); + } + } + } else + fputs(s, active); +} + +void +reallyputchar(int c) +{ + putc(c, active); + if (synch_lines && c == '\n') { + infile[ilevel].synch_lineno++; + if (infile[ilevel].synch_lineno != infile[ilevel].lineno) + do_emit_synchline(); + } +} + +/* + * build an input token.. + * consider only those starting with _ or A-Za-z. + */ +static ndptr +inspect(int c, char *tp) +{ + char *name = tp; + char *etp = tp+MAXTOK; + ndptr p; + + *tp++ = c; + + while ((isalnum(c = gpbc()) || c == '_') && tp < etp) + *tp++ = c; + if (c != EOF) + PUSHBACK(c); + *tp = EOS; + /* token is too long, it won't match anything, but it can still + * be output. */ + if (tp == ep) { + outputstr(name); + while (isalnum(c = gpbc()) || c == '_') { + if (sp < 0) + reallyputchar(c); + else + CHRSAVE(c); + } + *name = EOS; + return NULL; + } + + p = ohash_find(¯os, ohash_qlookupi(¯os, name, (void *)&tp)); + if (p == NULL) + return NULL; + if (macro_getdef(p) == NULL) + return NULL; + return p; +} + +/* + * initkwds - initialise m4 keywords as fast as possible. + * This very similar to install, but without certain overheads, + * such as calling lookup. Malloc is not used for storing the + * keyword strings, since we simply use the static pointers + * within keywrds block. + */ +static void +initkwds(void) +{ + unsigned int type; + size_t i; + + for (i = 0; i < MAXKEYS; i++) { + type = keywrds[i].ktyp; + if ((keywrds[i].ktyp & NOARGS) == 0) + type |= NEEDARGS; + setup_builtin(keywrds[i].knam, type); + } +} + +static void +record(struct position *t, int lev) +{ + if (lev < MAXRECORD) { + t[lev].name = CURRENT_NAME; + t[lev].line = CURRENT_LINE; + } +} + +static void +dump_stack(struct position *t, int lev) +{ + int i; + + for (i = 0; i < lev; i++) { + if (i == MAXRECORD) { + fprintf(stderr, " ...\n"); + break; + } + fprintf(stderr, " %s at line %lu\n", + t[i].name, t[i].line); + } +} + + +static void +enlarge_stack(void) +{ + STACKMAX += STACKMAX/2; + mstack = xrealloc(mstack, sizeof(stae) * STACKMAX, + "Evaluation stack overflow (%lu)", + (unsigned long)STACKMAX); + sstack = xrealloc(sstack, STACKMAX, + "Evaluation stack overflow (%lu)", + (unsigned long)STACKMAX); +} + +static const struct { + const char *n; + const char *d; +} nd [] = { +{ "-d, --debug[=flags]", "set debug flags" }, +{ "-D, --define=name[=value]", "define macro" }, +{ "-e, --error-output=file", "send error output to file" }, +{ "-E, --fatal-warnings", "exit on warnings" }, +{ "-F, --freeze-state=file", "save state to file" }, +{ "-g, --gnu", "enable gnu extensions" }, +{ " --help", "print this message and exit" }, +{ "-I, --include=file", "include file" }, +{ "-i, --interactive", "unbuffer output, ignore tty signals" }, +{ "-L, --nesting-limit=num", "macro expansion nesting limit (unimpl)" }, +{ "-P, --prefix-builtins", "prefix builtins with m4_" }, +{ "-Q, --quiet", "don't print warnings" }, +{ "-R, --reload-state=file", "restore state from file" }, +{ "-Q, --silent", "don't print warnings" }, +{ "-s, --synclines", "output line directives for cpp(1)" }, +{ "-t, --trace=macro", "trace the named macro" }, +{ "-G, --traditional", "disable gnu extensions" }, +{ "-U, --undefine=name", "undefine the named macro" }, +{ "-v, --version", "print the version number and exit" }, +}; + +static void +help(void) +{ + size_t i; + fprintf(stdout, "%s version %d\n\n", getprogname(), VERSION); + usage(stdout); + + fprintf(stdout, "\nThe long options are:\n"); + for (i = 0; i < __arraycount(nd); i++) + fprintf(stdout, "\t%-25.25s\t%s\n", nd[i].n, nd[i].d); +} diff --git a/usr.bin/m4/mdef.h b/usr.bin/m4/mdef.h new file mode 100644 index 0000000..f80c208 --- /dev/null +++ b/usr.bin/m4/mdef.h @@ -0,0 +1,237 @@ +/* $OpenBSD: mdef.h,v 1.29 2006/03/20 20:27:45 espie Exp $ */ +/* $NetBSD: mdef.h,v 1.17 2016/01/16 18:31:29 christos Exp $ */ + +/* + * Copyright (c) 1989, 1993 + * The Regents of the University of California. All rights reserved. + * + * This code is derived from software contributed to Berkeley by + * Ozan Yigit at York University. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + * + * @(#)mdef.h 8.1 (Berkeley) 6/6/93 + */ + +#ifdef __GNUC__ +# define UNUSED __attribute__((__unused__)) +#else +# define UNUSED +#endif + +#define MACRTYPE 1 +#define DEFITYPE 2 +#define EXPRTYPE 3 +#define SUBSTYPE 4 +#define IFELTYPE 5 +#define LENGTYPE 6 +#define CHNQTYPE 7 +#define SYSCTYPE 8 +#define UNDFTYPE 9 +#define INCLTYPE 10 +#define SINCTYPE 11 +#define PASTTYPE 12 +#define SPASTYPE 13 +#define INCRTYPE 14 +#define IFDFTYPE 15 +#define PUSDTYPE 16 +#define POPDTYPE 17 +#define SHIFTYPE 18 +#define DECRTYPE 19 +#define DIVRTYPE 20 +#define UNDVTYPE 21 +#define DIVNTYPE 22 +#define MKTMTYPE 23 +#define ERRPTYPE 24 +#define M4WRTYPE 25 +#define TRNLTYPE 26 +#define DNLNTYPE 27 +#define DUMPTYPE 28 +#define CHNCTYPE 29 +#define INDXTYPE 30 +#define SYSVTYPE 31 +#define EXITTYPE 32 +#define DEFNTYPE 33 +#define SELFTYPE 34 +#define INDIRTYPE 35 +#define BUILTINTYPE 36 +#define PATSTYPE 37 +#define FILENAMETYPE 38 +#define LINETYPE 39 +#define REGEXPTYPE 40 +#define ESYSCMDTYPE 41 +#define TRACEONTYPE 42 +#define TRACEOFFTYPE 43 +#define FORMATTYPE 44 + +#define BUILTIN_MARKER "__builtin_" + +#define TYPEMASK 0xff /* Keep bits really corresponding to a type. */ +#define RECDEF 0x100 /* Pure recursive def, don't expand it */ +#define NOARGS 0x200 /* builtin needs no args */ +#define NEEDARGS 0x400 /* mark builtin that need args with this */ + +/* + * m4 special characters + */ + +#define ARGFLAG '$' +#define LPAREN '(' +#define RPAREN ')' +#define LQUOTE '`' +#define RQUOTE '\'' +#define COMMA ',' +#define SCOMMT '#' +#define ECOMMT '\n' + +#ifdef msdos +#define system(str) (-1) +#endif + +/* + * other important constants + */ + +#define EOS '\0' +#define MAXINP 10 /* maximum include files */ +#define MAXOUT 10 /* maximum # of diversions */ +#define BUFSIZE 4096 /* starting size of pushback buffer */ +#define INITSTACKMAX 4096 /* starting size of call stack */ +#define STRSPMAX 4096 /* starting size of string space */ +#define MAXTOK 512 /* maximum chars in a tokn */ +#define HASHSIZE 199 /* maximum size of hashtab */ +#define MAXCCHARS 5 /* max size of comment/quote delim */ + +#define ALL 1 +#define TOP 0 + +#ifndef TRUE +#define TRUE 1 +#endif +#ifndef FALSE +#define FALSE 0 +#endif +#define cycle for(;;) + +/* + * m4 data structures + */ + +typedef struct ndblock *ndptr; + +struct macro_definition { + struct macro_definition *next; + char *defn; /* definition.. */ + unsigned int type; /* type of the entry.. */ +}; + + +struct ndblock { /* hashtable structure */ + unsigned int builtin_type; + unsigned int trace_flags; + struct macro_definition *d; + char name[1]; /* entry name.. */ +}; + +typedef union { /* stack structure */ + int sfra; /* frame entry */ + char *sstr; /* string entry */ +} stae; + +struct input_file { + FILE *file; + char *name; + unsigned long lineno; + unsigned long synch_lineno; /* used for -s */ + int c; +}; + +#define CURRENT_NAME (infile[ilevel].name) +#define CURRENT_LINE (infile[ilevel].lineno) +#define TOKEN_LINE(f) (f->lineno - (f->c == '\n' ? 1 : 0)) + +/* + * macros for readibility and/or speed + * + * gpbc() - get a possibly pushed-back character + * pushf() - push a call frame entry onto stack + * pushs() - push a string pointer onto stack + */ +#define gpbc() (bp > bufbase) ? *--bp : obtain_char(infile+ilevel) +#define pushf(x) \ + do { \ + if ((size_t)++sp == STACKMAX) \ + enlarge_stack();\ + mstack[sp].sfra = (x); \ + sstack[sp] = 0; \ + } while (0) + +#define pushs(x) \ + do { \ + if ((size_t)++sp == STACKMAX) \ + enlarge_stack();\ + mstack[sp].sstr = (x); \ + sstack[sp] = 1; \ + } while (0) + +#define pushs1(x) \ + do { \ + if ((size_t)++sp == STACKMAX) \ + enlarge_stack();\ + mstack[sp].sstr = (x); \ + sstack[sp] = 0; \ + } while (0) + +/* + * . . + * | . | <-- sp | . | + * +-------+ +-----+ + * | arg 3 ----------------------->| str | + * +-------+ | . | + * | arg 2 ---PREVEP-----+ . + * +-------+ | + * . | | | + * +-------+ | +-----+ + * | plev | PARLEV +-------->| str | + * +-------+ | . | + * | type | CALTYP . + * +-------+ + * | prcf ---PREVFP--+ + * +-------+ | + * | . | PREVSP | + * . | + * +-------+ | + * | <----------+ + * +-------+ + * + */ +#define PARLEV (mstack[fp].sfra) +#define CALTYP (mstack[fp-2].sfra) +#define TRACESTATUS (mstack[fp-1].sfra) +#define PREVEP (mstack[fp+3].sstr) +#define PREVSP (fp-4) +#define PREVFP (mstack[fp-3].sfra) + +#define VERSION 20150116 diff --git a/usr.bin/m4/misc.c b/usr.bin/m4/misc.c new file mode 100644 index 0000000..d7bc8cf --- /dev/null +++ b/usr.bin/m4/misc.c @@ -0,0 +1,414 @@ +/* $OpenBSD: misc.c,v 1.41 2009/10/14 17:19:47 sthen Exp $ */ +/* $NetBSD: misc.c,v 1.24 2016/01/16 17:01:22 christos Exp $ */ + +/* + * Copyright (c) 1989, 1993 + * The Regents of the University of California. All rights reserved. + * + * This code is derived from software contributed to Berkeley by + * Ozan Yigit at York University. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + */ +#if HAVE_NBTOOL_CONFIG_H +#include "nbtool_config.h" +#endif +#include <sys/cdefs.h> +__RCSID("$NetBSD: misc.c,v 1.24 2016/01/16 17:01:22 christos Exp $"); +#include <sys/types.h> +#include <errno.h> +#include <unistd.h> +#include <stdarg.h> +#include <stdio.h> +#include <stdlib.h> +#include <stddef.h> +#include <string.h> +#include <err.h> +#include "mdef.h" +#include "stdd.h" +#include "extern.h" +#include "pathnames.h" + + +char *ep; /* first free char in strspace */ +static char *strspace; /* string space for evaluation */ +char *endest; /* end of string space */ +static size_t strsize = STRSPMAX; +static size_t bufsize = BUFSIZE; + +unsigned char *buf; /* push-back buffer */ +unsigned char *bufbase; /* the base for current ilevel */ +unsigned char *bbase[MAXINP]; /* the base for each ilevel */ +unsigned char *bp; /* first available character */ +unsigned char *endpbb; /* end of push-back buffer */ + + +/* + * find the index of second str in the first str. + */ +ptrdiff_t +indx(const char *s1, const char *s2) +{ + char *t; + + t = strstr(s1, s2); + if (t == NULL) + return (-1); + else + return (t - s1); +} +/* + * pushback - push character back onto input + */ +void +pushback(int c) +{ + if (c == EOF) + return; + if (bp >= endpbb) + enlarge_bufspace(); + *bp++ = c; +} + +/* + * pbstr - push string back onto input + * pushback is replicated to improve + * performance. + */ +void +pbstr(const char *s) +{ + size_t n; + + n = strlen(s); + while ((size_t)(endpbb - bp) <= n) + enlarge_bufspace(); + while (n > 0) + *bp++ = s[--n]; +} + +/* + * pbnum - convert number to string, push back on input. + */ +void +pbnum(int n) +{ + pbnumbase(n, 10, 0); +} + +void +pbnumbase(int n, int base, int d) +{ + static char digits[36] = "0123456789abcdefghijklmnopqrstuvwxyz"; + int num; + int printed = 0; + + if (base > 36) + m4errx(1, "base %d > 36: not supported.", base); + + if (base < 2) + m4errx(1, "bad base %d for conversion.", base); + + num = (n < 0) ? -n : n; + do { + pushback(digits[num % base]); + printed++; + } + while ((num /= base) > 0); + + if (n < 0) + printed++; + while (printed++ < d) + pushback('0'); + + if (n < 0) + pushback('-'); +} + +/* + * pbunsigned - convert unsigned long to string, push back on input. + */ +void +pbunsigned(unsigned long n) +{ + do { + pushback(n % 10 + '0'); + } + while ((n /= 10) > 0); +} + +void +initspaces(void) +{ + int i; + + strspace = xalloc(strsize+1, NULL); + ep = strspace; + endest = strspace+strsize; + buf = (unsigned char *)xalloc(bufsize, NULL); + bufbase = buf; + bp = buf; + endpbb = buf + bufsize; + for (i = 0; i < MAXINP; i++) + bbase[i] = buf; +} + +void +enlarge_strspace(void) +{ + char *newstrspace; + int i; + + strsize *= 2; + newstrspace = malloc(strsize + 1); + if (!newstrspace) + errx(1, "string space overflow"); + memcpy(newstrspace, strspace, strsize/2); + for (i = 0; i <= sp; i++) + if (sstack[i]) + mstack[i].sstr = (mstack[i].sstr - strspace) + + newstrspace; + ep = (ep-strspace) + newstrspace; + free(strspace); + strspace = newstrspace; + endest = strspace + strsize; +} + +void +enlarge_bufspace(void) +{ + unsigned char *newbuf; + int i; + + bufsize += bufsize/2; + newbuf = xrealloc(buf, bufsize, "too many characters pushed back"); + for (i = 0; i < MAXINP; i++) + bbase[i] = (bbase[i]-buf)+newbuf; + bp = (bp-buf)+newbuf; + bufbase = (bufbase-buf)+newbuf; + buf = newbuf; + endpbb = buf+bufsize; +} + +/* + * chrsave - put single char on string space + */ +void +chrsave(int c) +{ + if (ep >= endest) + enlarge_strspace(); + *ep++ = c; +} + +/* + * read in a diversion file, and dispose it. + */ +void +getdiv(int n) +{ + int c; + + if (active == outfile[n]) + m4errx(1, "undivert: diversion still active."); + rewind(outfile[n]); + while ((c = getc(outfile[n])) != EOF) + putc(c, active); + (void) fclose(outfile[n]); + outfile[n] = NULL; +} + +/* + * killdiv - get rid of the diversion files + */ +void +killdiv(void) +{ + int n; + + for (n = 0; n < maxout; n++) + if (outfile[n] != NULL) { + (void) fclose(outfile[n]); + } +} + +void +m4errx(int exval, const char *fmt, ...) +{ + fprintf(stderr, "%s: ", getprogname()); + fprintf(stderr, "%s at line %lu: ", CURRENT_NAME, CURRENT_LINE); + if (fmt != NULL) { + va_list ap; + + va_start(ap, fmt); + vfprintf(stderr, fmt, ap); + va_end(ap); + } + fprintf(stderr, "\n"); + exit(exval); +} + +/* + * resizedivs: allocate more diversion files */ +void +resizedivs(int n) +{ + int i; + + outfile = (FILE **)xrealloc(outfile, sizeof(FILE *) * n, + "too many diverts %d", n); + for (i = maxout; i < n; i++) + outfile[i] = NULL; + maxout = n; +} + +void * +xalloc(size_t n, const char *fmt, ...) +{ + void *p = malloc(n); + + if (p == NULL) { + if (fmt == NULL) + err(1, "malloc"); + else { + va_list va; + + va_start(va, fmt); + verr(1, fmt, va); + va_end(va); + } + } + return p; +} + +void * +xrealloc(void *old, size_t n, const char *fmt, ...) +{ + char *p = realloc(old, n); + + if (p == NULL) { + free(old); + if (fmt == NULL) + err(1, "realloc"); + else { + va_list va; + + va_start(va, fmt); + verr(1, fmt, va); + va_end(va); + } + } + return p; +} + +char * +xstrdup(const char *s) +{ + char *p = strdup(s); + if (p == NULL) + err(1, "strdup"); + return p; +} + +int +obtain_char(struct input_file *f) +{ + if (f->c == EOF) + return EOF; + + f->c = fgetc(f->file); +#ifndef REAL_FREEZE + if (freezef) + fputc(f->c, freezef); +#endif + if (f->c == '\n') + f->lineno++; + + return f->c; +} + +void +set_input(struct input_file *f, FILE *real, const char *name) +{ + f->file = real; + f->lineno = 1; + f->c = 0; + f->name = xstrdup(name); + emit_synchline(); +} + +void +do_emit_synchline(void) +{ + fprintf(active, "#line %lu \"%s\"\n", + infile[ilevel].lineno, infile[ilevel].name); + infile[ilevel].synch_lineno = infile[ilevel].lineno; +} + +void +release_input(struct input_file *f) +{ + if (f->file != stdin) + fclose(f->file); + f->c = EOF; + /* + * XXX can't free filename, as there might still be + * error information pointing to it. + */ +} + +void +doprintlineno(struct input_file *f) +{ + pbunsigned(TOKEN_LINE(f)); +} + +void +doprintfilename(struct input_file *f) +{ + pbstr(rquote); + pbstr(f->name); + pbstr(lquote); +} + +/* + * buffer_mark/dump_buffer: allows one to save a mark in a buffer, + * and later dump everything that was added since then to a file. + */ +size_t +buffer_mark(void) +{ + return bp - buf; +} + + +void +dump_buffer(FILE *f, size_t m) +{ + unsigned char *s; + + for (s = bp; (size_t)(s - buf) > m;) + fputc(*--s, f); +} diff --git a/usr.bin/m4/parser.y b/usr.bin/m4/parser.y new file mode 100644 index 0000000..f8256ef --- /dev/null +++ b/usr.bin/m4/parser.y @@ -0,0 +1,86 @@ +%{ +/* $NetBSD: parser.y,v 1.3 2015/01/04 18:31:09 joerg Exp $ */ +/* $OpenBSD: parser.y,v 1.6 2008/08/21 21:00:14 espie Exp $ */ +/* + * Copyright (c) 2004 Marc Espie <espie@cvs.openbsd.org> + * + * Permission to use, copy, modify, and distribute this software for any + * purpose with or without fee is hereby granted, provided that the above + * copyright notice and this permission notice appear in all copies. + * + * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES + * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR + * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES + * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN + * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF + * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. + */ +#if HAVE_NBTOOL_CONFIG_H +#include "nbtool_config.h" +#endif +#include <sys/cdefs.h> +__RCSID("$NetBSD: parser.y,v 1.3 2015/01/04 18:31:09 joerg Exp $"); +#include <stdint.h> +#define YYSTYPE int32_t +extern int32_t end_result; +extern int yylex(void); +extern int yyerror(const char *); +%} +%token NUMBER +%token ERROR +%left LOR +%left LAND +%left '|' +%left '^' +%left '&' +%left EQ NE +%left '<' LE '>' GE +%left LSHIFT RSHIFT +%left '+' '-' +%left '*' '/' '%' +%right UMINUS UPLUS '!' '~' + +%% + +top : expr { end_result = $1; } + ; +expr : expr '+' expr { $$ = $1 + $3; } + | expr '-' expr { $$ = $1 - $3; } + | expr '*' expr { $$ = $1 * $3; } + | expr '/' expr { + if ($3 == 0) { + yyerror("division by zero"); + exit(1); + } + $$ = $1 / $3; + } + | expr '%' expr { + if ($3 == 0) { + yyerror("modulo zero"); + exit(1); + } + $$ = $1 % $3; + } + | expr LSHIFT expr { $$ = $1 << $3; } + | expr RSHIFT expr { $$ = $1 >> $3; } + | expr '<' expr { $$ = $1 < $3; } + | expr '>' expr { $$ = $1 > $3; } + | expr LE expr { $$ = $1 <= $3; } + | expr GE expr { $$ = $1 >= $3; } + | expr EQ expr { $$ = $1 == $3; } + | expr NE expr { $$ = $1 != $3; } + | expr '&' expr { $$ = $1 & $3; } + | expr '^' expr { $$ = $1 ^ $3; } + | expr '|' expr { $$ = $1 | $3; } + | expr LAND expr { $$ = $1 && $3; } + | expr LOR expr { $$ = $1 || $3; } + | '(' expr ')' { $$ = $2; } + | '-' expr %prec UMINUS { $$ = -$2; } + | '+' expr %prec UPLUS { $$ = $2; } + | '!' expr { $$ = !$2; } + | '~' expr { $$ = ~$2; } + | NUMBER + ; +%% + diff --git a/usr.bin/m4/pathnames.h b/usr.bin/m4/pathnames.h new file mode 100644 index 0000000..ef788e0 --- /dev/null +++ b/usr.bin/m4/pathnames.h @@ -0,0 +1,40 @@ +/* $NetBSD: pathnames.h,v 1.15 2011/03/08 23:55:19 riz Exp $ */ + +/* + * Copyright (c) 1989, 1993 + * The Regents of the University of California. All rights reserved. + * + * This code is derived from software contributed to Berkeley by + * Ozan Yigit at York University. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + * + * @(#)pathnames.h 8.1 (Berkeley) 6/6/93 + */ + +/* + * Definitions of diversion files. + */ +#define _PATH_DIVNAME "/tmp/m4.0XXXXXXXXXX" /* unix diversion files */ diff --git a/usr.bin/m4/stdd.h b/usr.bin/m4/stdd.h new file mode 100644 index 0000000..0cbd681 --- /dev/null +++ b/usr.bin/m4/stdd.h @@ -0,0 +1,54 @@ +/* $NetBSD: stdd.h,v 1.4 2003/08/07 11:14:34 agc Exp $ */ + +/*- + * Copyright (c) 1991, 1993 + * The Regents of the University of California. All rights reserved. + * + * This code is derived from software contributed to Berkeley by + * Ozan Yigit at York University. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + * + * @(#)stdd.h 8.1 (Berkeley) 6/6/93 + */ + +/* + * standard defines + */ + +#define max(a,b) ((a) > (b)? (a): (b)) +#define min(a,b) ((a) < (b)? (a): (b)) + +#define iswhite(c) ((c) == ' ' || (c) == '\t') + +/* + * STREQ is an optimised strcmp(a,b)==0 + * STREQN is an optimised strncmp(a,b,n)==0; assumes n > 0 + */ +#define STREQ(a, b) ((a)[0] == (b)[0] && strcmp(a, b) == 0) +#define STREQN(a, b, n) ((a)[0] == (b)[0] && strncmp(a, b, n) == 0) + +#define YES 1 +#define NO 0 diff --git a/usr.bin/m4/tokenizer.l b/usr.bin/m4/tokenizer.l new file mode 100644 index 0000000..fc4ebf1 --- /dev/null +++ b/usr.bin/m4/tokenizer.l @@ -0,0 +1,108 @@ +%{ +/* $NetBSD: tokenizer.l,v 1.6 2012/03/20 20:34:58 matt Exp $ */ +/* $OpenBSD: tokenizer.l,v 1.6 2008/08/21 21:00:14 espie Exp $ */ +/* + * Copyright (c) 2004 Marc Espie <espie@cvs.openbsd.org> + * + * Permission to use, copy, modify, and distribute this software for any + * purpose with or without fee is hereby granted, provided that the above + * copyright notice and this permission notice appear in all copies. + * + * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES + * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR + * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES + * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN + * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF + * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. + */ +#if HAVE_NBTOOL_CONFIG_H +#include "nbtool_config.h" +#endif +#include "parser.h" +__RCSID("$NetBSD: tokenizer.l,v 1.6 2012/03/20 20:34:58 matt Exp $"); +#include <stdlib.h> +#include <errno.h> +#include <stdint.h> +#include <limits.h> + +extern int mimic_gnu; +extern int32_t yylval; +extern int yylex(void); +extern int yywrap(void); + +int32_t number(void); +int32_t parse_radix(void); + +%} + +%option nounput +%option noinput + +delim [ \t\n] +ws {delim}+ +hex 0[xX][0-9a-fA-F]+ +oct 0[0-7]* +dec [1-9][0-9]* +radix 0[rR][0-9]+:[0-9a-zA-Z]+ + +%% +{ws} {/* just skip it */} +{hex}|{oct}|{dec} { yylval = number(); return(NUMBER); } +{radix} { if (mimic_gnu) { + yylval = parse_radix(); return(NUMBER); + } else { + return(ERROR); + } + } +"<=" { return(LE); } +">=" { return(GE); } +"<<" { return(LSHIFT); } +">>" { return(RSHIFT); } +"==" { return(EQ); } +"!=" { return(NE); } +"&&" { return(LAND); } +"||" { return(LOR); } +. { return yytext[0]; } +%% + +int32_t +number(void) +{ + long l; + + errno = 0; + l = strtol(yytext, NULL, 0); + if (((l == LONG_MAX || l == LONG_MIN) && errno == ERANGE) || + l > INT32_MAX || l < INT32_MIN) { + fprintf(stderr, "m4: numeric overflow in expr: %s\n", yytext); + } + return l; +} + +int32_t +parse_radix(void) +{ + long base; + char *next; + long l; + + l = 0; + base = strtol(yytext+2, &next, 0); + if (base > 36 || next == NULL) { + fprintf(stderr, "m4: error in number %s\n", yytext); + } else { + next++; + while (*next != 0) { + if (*next >= '0' && *next <= '9') + l = base * l + *next - '0'; + else if (*next >= 'a' && *next <= 'z') + l = base * l + *next - 'a' + 10; + else if (*next >= 'A' && *next <= 'Z') + l = base * l + *next - 'A' + 10; + next++; + } + } + return l; +} + diff --git a/usr.bin/m4/trace.c b/usr.bin/m4/trace.c new file mode 100644 index 0000000..bb9397f --- /dev/null +++ b/usr.bin/m4/trace.c @@ -0,0 +1,203 @@ +/* $NetBSD: trace.c,v 1.8 2012/03/20 20:34:58 matt Exp $ */ +/* $OpenBSD: trace.c,v 1.15 2006/03/24 08:03:44 espie Exp $ */ +/* + * Copyright (c) 2001 Marc Espie. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE OPENBSD PROJECT AND CONTRIBUTORS + * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE OPENBSD + * PROJECT OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ +#if HAVE_NBTOOL_CONFIG_H +#include "nbtool_config.h" +#endif +#include <sys/cdefs.h> +__RCSID("$NetBSD: trace.c,v 1.8 2012/03/20 20:34:58 matt Exp $"); + +#include <sys/types.h> +#include <err.h> +#include <stddef.h> +#include <stdint.h> +#include <stdio.h> +#include <stdlib.h> +#include "mdef.h" +#include "stdd.h" +#include "extern.h" + +FILE *traceout; + +#define TRACE_ARGS 1 +#define TRACE_EXPANSION 2 +#define TRACE_QUOTE 4 +#define TRACE_FILENAME 8 +#define TRACE_LINENO 16 +#define TRACE_CONT 32 +#define TRACE_ID 64 +#define TRACE_NEWFILE 128 /* not implemented yet */ +#define TRACE_INPUT 256 /* not implemented yet */ + +static unsigned int letter_to_flag(int); +static void print_header(struct input_file *); +static int frame_level(void); + + +unsigned int trace_flags = TRACE_QUOTE | TRACE_EXPANSION; + +void +trace_file(const char *name) +{ + + if (traceout && traceout != stderr) + fclose(traceout); + traceout = fopen(name, "w"); + if (!traceout) + err(1, "can't open %s", name); +} + +static unsigned int +letter_to_flag(int c) +{ + switch(c) { + case 'a': + return TRACE_ARGS; + case 'e': + return TRACE_EXPANSION; + case 'q': + return TRACE_QUOTE; + case 'c': + return TRACE_CONT; + case 'x': + return TRACE_ID; + case 'f': + return TRACE_FILENAME; + case 'l': + return TRACE_LINENO; + case 'p': + return TRACE_NEWFILE; + case 'i': + return TRACE_INPUT; + case 't': + return TRACE_ALL; + case 'V': + return ~0; + default: + return 0; + } +} + +void +set_trace_flags(const char *s) +{ + char mode = 0; + unsigned int f = 0; + + if (*s == '+' || *s == '-') + mode = *s++; + while (*s) + f |= letter_to_flag(*s++); + switch(mode) { + case 0: + trace_flags = f; + break; + case '+': + trace_flags |= f; + break; + case '-': + trace_flags &= ~f; + break; + } +} + +static int +frame_level(void) +{ + int level; + int framep; + + for (framep = fp, level = 0; framep != 0; + level++,framep = mstack[framep-3].sfra) + ; + return level; +} + +static void +print_header(struct input_file *inp) +{ + fprintf(traceout, "m4trace:"); + if (trace_flags & TRACE_FILENAME) + fprintf(traceout, "%s:", inp->name); + if (trace_flags & TRACE_LINENO) + fprintf(traceout, "%lu:", TOKEN_LINE(inp)); + fprintf(traceout, " -%d- ", frame_level()); + if (trace_flags & TRACE_ID) + fprintf(traceout, "id %lu: ", expansion_id); +} + +size_t +trace(const char *argv[], int argc, struct input_file *inp) +{ + if (!traceout) + traceout = stderr; + print_header(inp); + if (trace_flags & TRACE_CONT) { + fprintf(traceout, "%s ...\n", argv[1]); + print_header(inp); + } + fprintf(traceout, "%s", argv[1]); + if ((trace_flags & TRACE_ARGS) && argc > 2) { + char delim[3]; + int i; + + delim[0] = LPAREN; + delim[1] = EOS; + for (i = 2; i < argc; i++) { + fprintf(traceout, "%s%s%s%s", delim, + (trace_flags & TRACE_QUOTE) ? lquote : "", + argv[i], + (trace_flags & TRACE_QUOTE) ? rquote : ""); + delim[0] = COMMA; + delim[1] = ' '; + delim[2] = EOS; + } + fprintf(traceout, "%c", RPAREN); + } + if (trace_flags & TRACE_CONT) { + fprintf(traceout, " -> ???\n"); + print_header(inp); + fprintf(traceout, argc > 2 ? "%s(...)" : "%s", argv[1]); + } + if (trace_flags & TRACE_EXPANSION) + return buffer_mark(); + else { + fprintf(traceout, "\n"); + return SIZE_MAX; + } +} + +void +finish_trace(size_t mark) +{ + fprintf(traceout, " -> "); + if (trace_flags & TRACE_QUOTE) + fprintf(traceout, "%s", lquote); + dump_buffer(traceout, mark); + if (trace_flags & TRACE_QUOTE) + fprintf(traceout, "%s", rquote); + fprintf(traceout, "\n"); +} |