aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2012-12-23Elie De Brauwer pointed out that xputs() isn't reliably reporting errors ↵Rob Landley
because there's no flush. Rather than change the output granularity, flush before exit and check errors there. (We still need xputc() doing it so "yes" doesn't continue forever.)
2012-12-23Whitespace.Rob Landley
2012-12-16Extend seq to support -f (format string) and -s (string separator) as ↵Elie De Brauwer
requested by LSB
2012-12-22Fix tac to handle the "abc\ndef" case properlyElie De Brauwer
2012-12-23Update mkstatus.py, roadmap.html, and status.html.Rob Landley
2012-12-22Refactoring broke revElie De Brauwer
2012-12-19Web page update for release.Rob Landley
2012-12-19Minor cleanups.Rob Landley
2012-12-16Adding implementation of rev including testsElie De Brauwer
2012-12-15Adding tests for tac based on cat testsElie De Brauwer
2012-12-15Simplify tac, the newline checking is already done in the get_line() functionElie De Brauwer
2012-12-16Missing letter in help text.Rob Landley
2012-12-15Added tag 0.4.2 for changeset 8947c0d35e58Rob Landley
2012-12-12The Linux kernel "make clean" calls rm -f with no arguments, which ↵0.4.2Rob Landley
apparently is not an error.
2012-12-10Add stat submission to new "pending" directory, along with infrastructure to ↵Rob Landley
support pending.
2012-12-10essat's flag not to follow symlinks isn't in the system call, and if libc is ↵Rob Landley
supposed to implement this as a wrapper uClibc gets it wrong. So use the stat info about symlinks instead. (Doesn't check the parent directory, but if that's read only we can't delete the file anyway so prompting is moot.)
2012-12-09Meddle.Rob Landley
The <1 has to come first in the option string, normalize whitespace, sprintf of %d maxes out at -2 billion ala 12 bytes with null terminator so we don't need a length check in a 4k buffer, use the "%*s" feature of printf to prepend whitespace for us, take advantage of c99 defining ! to return 0 or 1.
2012-12-08Adding -s (single shot) and -o (omit pids) options to pidofElie De Brauwer
2012-12-09Set executable bits on two test scripts.Rob Landley
2012-12-08Adding sha1sum tests based on RFC 3174Elie De Brauwer
2012-12-08Adding tests for catElie De Brauwer
2012-12-08Add README file to each toys/ directory, teach build infrastructure to get ↵Rob Landley
fancy name from that. This means adding new directories shouldn't require touching build scripts.
2012-12-08Complicate the rm -i behavior to do what posix specifies.Rob Landley
2012-12-08Make yesno() always read from stdin and write to stderr. (If we need to find ↵Rob Landley
our tty, open /dev/tty, but existing users don't.)
2012-12-07New rm command.Rob Landley
2012-12-06Update link to posix in docs (open group broke their website).Rob Landley
2012-12-06Have dirtree_add_node() set parent so error message can provide full path.Rob Landley
2012-12-03Use sched_{set,get}affinity directly so musl doesn't have to provide wrappers.Rob Landley
2012-12-01Minor whitespace tweak.Rob Landley
2012-12-01First guess at what internationalization support for expand would look like.Rob Landley
2012-12-01Generate status page with new pubs.opengroup.org urls.Rob Landley
2012-12-01Remove unused structure definition.Rob Landley
2012-12-01The open group changed their website stupidly, so opengroup.org/onlinepubs ↵Rob Landley
forwards to a dead server and you have to use pubs.opengroup.org/onlinepubs now. Change the hello template to note the new location. (Waiting to see if they fix it upstream before touching every command in toys/posix.)
2012-12-01Add header that musl libc needs.Rob Landley
2012-12-01Fix ancient glibc workaround to force fstatat64Rob Landley
2012-12-01Minor cleanup: unify two codepaths that do the same thing.Rob Landley
2012-12-01Would the compiler like to warn me about declaring two variables with the ↵Rob Landley
same name in the same function? No? Carry on then...
2012-12-01Felix Janda pointed out that the r in mbrtowc() stands for "restartable" so ↵Rob Landley
it's already buffering the partial data we feed it, so rolling back most of the last commit to wc.
2012-12-01Re-add backspace support to expand (oops) and fix test to expect data plus ↵Rob Landley
backspace characters rather than the chracters cancelling each other out before output.
2012-11-30Largeish rewrite of expand, mostly described on the mailing list.Rob Landley
2012-11-28md5sum tests from RFC 1321Rob Landley
2012-11-28Use the specified compiler for preprocessing.Rob Landley
2012-11-28The previous wc -m didn't handle multibyte characters that crossed a buffer ↵Rob Landley
boundary, so take a guess at making that work. (I haven't got a test case for this. I also don't know how to handle invalid sequences so just don't count them.)
2012-11-27login: Avoid gcc to drop exit condition because of "always false condition".Jonathan Clairembault
2012-11-27First round of expand cleanups.Rob Landley
2012-11-23expand: handle backspace.Jonathan Clairembault
2012-11-23Add expand command as described in POSIX-2008.Jonathan Clairembault
Erratum: Do not handle backspace.
2012-11-26Squash "type punned pointer" nonsense with a bigger hammer. (Sheesh.)Rob Landley
2012-11-26The headers of uClibc and older glibc got unhappy with O_DIRECTORY and some ↵Rob Landley
other stuff. Fix it up in portability.h.
2012-11-26Cleanup i18n support (#ifdefectomy, move global init to process launch). ↵Rob Landley
Teach make.sh to emit "#define FLAG_x 0" for options inside disabled USE macros so we can unconditionally refer to them.