#
67039 |
| 20-Apr-1994 |
bostic |
P1003.2b/D9 says: If the end of the input is reached, the n command shall branch to the end of the script and quit without starting a new cycle |
#
67038 |
| 20-Apr-1994 |
bostic |
change the replacement algoricthm to only replace when: (1) the previous replacement was for an empty match, or (2) the current replacement is for a non-empty match. this makes empty RE's work as the
change the replacement algoricthm to only replace when: (1) the previous replacement was for an empty match, or (2) the current replacement is for a non-empty match. this makes empty RE's work as they did historically From: "Chris G. Demetriou" <cgd@sun-lamp.cs.berkeley.edu>
show more ...
|
#
67036 |
| 20-Apr-1994 |
bostic |
The `D' command in sed was leaving junk at the end of the pattern space. From: Charles Hannum <mycroft@sun-lamp.cs.berkeley.edu> |
#
66817 |
| 16-Apr-1994 |
bostic |
update copyright notice |
#
66120 |
| 16-Feb-1994 |
bostic |
sed -e 's/$/ \\/g' /etc/passwd loops forever break out of the "g" loop when the regular expression being substituted is of zero length From: Diomidis Spinellis <dspin@leon.nrcps.ariadne-t.gr> |
#
62229 |
| 06-Jun-1993 |
bostic |
4.4BSD snapshot (revision 8.1); add 1993 to copyright |
#
60120 |
| 18-May-1993 |
elan |
Strnchr() -> memchr(). |
#
59077 |
| 14-Apr-1993 |
torek |
fix REs over completely empty spaces (test case: echo '' | sed 'x;s/^.*/foo/') |
#
59073 |
| 14-Apr-1993 |
elan |
make sed work on 8-bit objects TK checking in for Elan Amir |
#
57733 |
| 27-Jan-1993 |
elan |
Initial version that will work with binary files. |
#
57572 |
| 15-Jan-1993 |
elan |
Fixed bug in cspace -- precdence of == is less that '+'. |
#
57421 |
| 06-Jan-1993 |
bostic |
"echo '1+1' | sed -f /usr/src/usr.bin/sed/math.sed" drops core from Donn at BSDI |
#
56949 |
| 02-Dec-1992 |
bostic |
branch should succeed if any substitution has succeeded, not just the last one; from Diomidis Spinellis <dds@doc.ic.ac.uk> |
#
56657 |
| 01-Nov-1992 |
bostic |
unneeded set |
#
56093 |
| 30-Aug-1992 |
bostic |
don't access entries multiple times; from Diomidis |
#
56091 |
| 29-Aug-1992 |
bostic |
re_off is a size_t, cspace is an external function |
#
56079 |
| 28-Aug-1992 |
bostic |
MUST copy data from stdio buffer, change interface to mf_fgets; this simplifies the interface to cspace, so clean that up, too; make match (pmatch) a single global structure, cleaning up regexec_e();
MUST copy data from stdio buffer, change interface to mf_fgets; this simplifies the interface to cspace, so clean that up, too; make match (pmatch) a single global structure, cleaning up regexec_e(); fix default RE calculation for re_nsub > maxbref; fix default RE bugs
show more ...
|
#
56070 |
| 27-Aug-1992 |
bostic |
let's not be tricky... |
#
56064 |
| 27-Aug-1992 |
bostic |
don't close fd's twice. |
#
56019 |
| 24-Aug-1992 |
bostic |
make empty RE's work, some minor cleanups, too |
#
56005 |
| 24-Aug-1992 |
bostic |
fix the 'd' and 'D' flags for the new "pd" variable |
#
55997 |
| 23-Aug-1992 |
bostic |
date and time created 92/08/23 18:26:16 by bostic |