Commit | Line | Data |
---|---|---|
995308a3 KB |
1 | # @(#)README 5.1 (Berkeley) %G% |
2 | ||
3 | Alpha+1 release. | |
4 | Mon Jul 27 17:58:48 EDT 1992 | |
5 | henry@zoo.toronto.edu | |
6 | ||
7 | New this time: improvements to the manual pages, and an important | |
8 | extension, the REG_STARTEND option to regexec(). | |
9 | ||
10 | installation notes: | |
11 | -------- | |
12 | Limits.h, stdlib.h are fakes for ANSI headers not present on my system; | |
13 | delete them if you've got real ones. | |
14 | ||
15 | Memmove.c is a fake for an ANSI subroutine I don't have. | |
16 | ||
17 | Split.c is used for regression testing but is not part of the RE routines | |
18 | themselves. | |
19 | ||
20 | Just "make" will build regcomp.o, regexec.o, regfree.o, and regerror.o | |
21 | (the actual routines), bundle them together into a test program, and run | |
22 | a regression test on them. No output is good output. | |
23 | ||
24 | Regex.h is for /usr/include. All other .h files are internal only. | |
25 | ||
26 | For an ANSI compiler, you want to uncomment the prototype parameters in | |
27 | the function declarations in regex.h. | |
28 | ||
29 | If you have an ANSI compiler, take -Dconst= out of Makefile's CFLAGS. | |
30 | Do not take -DPOSIX_MISTAKE out. You probably want to put -DNDEBUG in, | |
31 | to get rid of all the assertion tests, for building production versions | |
32 | after you've run the regression test. (The regression test won't quite | |
33 | compile with -DNDEBUG at the moment.) | |
34 | -------- |