-perm doesn't accept some valid symbolic modes

[tavianator]

$ ./target/debug/find -perm +rwx
Error: invalid mode '+rwx'
$ ./target/debug/find -perm u+rwX
Error: invalid mode 'u+rwX'
$ ./target/debug/find -perm u=g
Error: invalid mode 'u=g'

[mcharsley]

Thanks for the bug. I don't suppose you have ready access to a description of how those operate? GNU find's documentation is rather lacking in this area :-( the best I can find is chmod's description of symbolic modes, not all of which are relevant to find... I'm assuming +rwx is the equivalent of ugo+rwx and u=g means "where user permissions = group permissions"?

…

On Wed, Jun 14, 2017 at 11:26 PM, Tavian Barnes ***@***.***> wrote: $ ./target/debug/find -perm +rwx Error: invalid mode '+rwx' $ ./target/debug/find -perm u+rwX Error: invalid mode 'u+rwX' $ ./target/debug/find -perm u=g Error: invalid mode 'u=g' — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#26>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ALbIF0C4Ett_PXtaVJsAL0l4JW4BFi7jks5sEF4JgaJpZM4N6hEW> .

[tavianator]

Luckily in this case we have the POSIX documentation:

The mode argument is used to represent file mode bits. It shall be identical in format to the symbolic_mode operand described in chmod, and shall be interpreted as follows. To start, a template shall be assumed with all file mode bits cleared. An op symbol of '+' shall set the appropriate mode bits in the template; '-' shall clear the appropriate bits; '=' shall set the appropriate mode bits, without regard to the contents of the file mode creation mask of the process. The op symbol of '-' cannot be the first character of mode; this avoids ambiguity with the optional leading <hyphen-minus>. Since the initial mode is all bits off, there are not any symbolic modes that need to use '-' as the first character.

In other words, the syntax is exactly the same as chmod. Since chmod symbolic modes actually specify a change to some existing permissions, you should interpret them as starting from 0000.

The linked chmod docs go on to describe the exact grammar.

I'm assuming +rwx is the equivalent of ugo+rwx

Yes, +rwx == a+rwx == ugo+rwx. See the lines starting with "If who is not specified, ..." in the chmod docs.

and u=g means "where user permissions = group permissions"

Not exactly, u=g means "set the user permissions to the same as the group permissions." These are normally 0 because that's what we're starting from, but something like u=rw,g=u-w is the same as u=rw,g=r.

[e12e]

Just a hint, while you can't just look at the GNU sources (without risking having to change the license), you could probably look at bsd find for help?

https://github.com/freebsd/freebsd/blob/f5d95e1f8d32db4ccccfd5ad9cecb21ed07a695d/usr.bin/find/function.c#L1277

On a similar note, I saw that the maintainer of GNU coreutils mentioned on a reddit[r] discussion of a similar project in go[g] (no find port, though) - that the GNU coreutils have an extensive test suite. The same seems to hold true for Gnu findutils as well:

http://git.savannah.gnu.org/gitweb/?p=findutils.git;a=tree;f=find/testsuite;h=b3f19586ab00fcc897be2ee17a8966cf445983f2;hb=HEAD

A set of tests written with Dejagnu[d].

While there's probably no license problems running this project through GNU tests (but care would have to be taken re-distributing Gnu tests) -- I see that the FreeBSD project has a test suite[w] using kyua[k] and ATF[a] - apparently inspired by the NetBSD test-suite. But FreeBSD doesn't have any tests for find (that I can see), but there are a handful for NetBSD (but not for symbolic permissions):

https://github.com/IIJ-NetBSD/netbsd-src/blob/master/tests/usr.bin/find/t_find.sh

NetBSD find handling of permissions:

https://github.com/IIJ-NetBSD/netbsd-src/blob/0bcc2a01826fc41bf3f1b6d9fb95f157b3db607d/usr.bin/find/function.c#L1513

All in all, I'd say there'd be a use for a project dedicated to making a POSIX test-suite as a form of executable specification/documentation for this stuff - but that's at least a separate issue, probably a separate project.

The Open Group apparently has such a test suite available, and as far as I can tell it's free for open source projects - but presumably not Free/redistributable (every developer/contributor would need to acquire a personal license?):

http://www.opengroup.org/testing/testsuites/vscpcts2016.htm

Apologies if this "drive-by essay" is a bit out of scope, but it appeared that at least looking at Net/FreeBSD might be of some help on some of these thornier syntax/behaviors. It would probably also be great to reach some consensus on test suites - both for BSD in particular and Free nix-likes in general.

From a cursory look, I'd probably start with the NetBSD/FreeBSD tests - the GNU tests are quite extensive, but would appear to benefit from some modernization.

[a] https://github.com/jmmv/atf/
[d] https://www.gnu.org/software/dejagnu/
[g] https://github.com/aisola/go-coreutils
[k] https://github.com/jmmv/kyua/
[r] https://www.reddit.com/r/golang/comments/27ggqy/a_go_language_implementation_of_the_gnu_coreutils/
[w] https://wiki.freebsd.org/TestSuite

[tavianator]

@e12e thanks for all those links!

I've also written what I believe is a correct implementation as part of a find-compatible tool: https://github.com/tavianator/bfs/blob/1.1/parse.c#L1416

I also have a test suite for it that you can run against other tools with ./tests.sh --bfs=/path/to/find (pass --gnu or just --posix to run only those parts of the test suite). In fact that's how I found most of the bugs I've reported :) https://github.com/tavianator/bfs/blob/1.1/tests.sh

[mcharsley]

Thanks all. That's really helpful. Now all I need is enough time to actually work on the project :-)

…

On Wed, Jul 19, 2017 at 1:04 AM, Tavian Barnes ***@***.***> wrote: @e12e <https://github.com/e12e> thanks for all those links! I've also written what I believe is a correct implementation as part of a find-compatible tool: https://github.com/tavianator/bfs/blob/ 0fc17759476ca38f55b479a12d88f8a7614b7a3b/parse.c#L1382 I also have a test suite for it that you can run against other tools with ./tests.sh --bfs=/path/to/find (pass --gnu or just --posix to run only those parts of the test suite). In fact that's how I found most of the bugs I've reported :) https://github.com/tavianator/bfs/blob/ 0fc17759476ca38f55b479a12d88f8a7614b7a3b/tests.sh — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#26 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ALbIFxfeUyGcJAvsFaG22PEZp8058Ybhks5sPUgXgaJpZM4N6hEW> .

[tavianator]

Also uutils/coreutils has an implementation: https://github.com/uutils/coreutils/blob/99af79a7fab9f43a5ab393637c96de513daab5e3/src/chmod/chmod.rs#L222