libflux/buffer: Add trimmed peek/read line variants

[chu11]

For convenience, support flux_buffer_peek_trimmed_line() and
flux_buffer_read_trimmed_line() variants, which will read lines
but strip off any trailing newline characters.

Fixes #1624

[coveralls]

Coverage decreased (-0.002%) to 79.146% when pulling ff4a7a7 on chu11:issue1624 into 29c0c6c on flux-framework:master.

[chu11]

Hit a valgrind fail, which I'm fairly certain this PR has nothing to do with. I also hit a similar one in #1564. I'm going to open an issue on it. Restarted the builder though.

�[31mFound ./t/t5000-valgrind.output�[39m
expecting success: 
	libtool e flux ${VALGRIND} \
		--tool=memcheck \
		--leak-check=full \
		--gen-suppressions=all \
		--trace-children=no \
		--child-silent-after-fork=yes \
		--num-callers=30 \
		--leak-resolution=med \
		--error-exitcode=1 \
		--suppressions=$VALGRIND_SUPPRESSIONS \
		${BROKER} --shutdown-grace=4 ${VALGRIND_WORKLOAD} 10

==12893== Memcheck, a memory error detector
==12893== Copyright (C) 2002-2013, and GNU GPL'd, by Julian Seward et al.
==12893== Using Valgrind-3.10.1 and LibVEX; rerun with -h for copyright info
==12893== Command: /home/travis/build/flux-framework/flux-core/src/broker/.libs/lt-flux-broker --shutdown-grace=4 /home/travis/build/flux-framework/flux-core/t/valgrind/valgrind-workload.sh 10
==12893== 
FLUX_URI=local:///tmp/flux-MivVCS
2018-08-31T19:35:47.995037Z connector-local.err[0]: send kvs.watch response to client 442F0: Broken pipe
==12893== 
==12893== HEAP SUMMARY:
==12893==     in use at exit: 130,944 bytes in 155 blocks
==12893==   total heap usage: 940,255 allocs, 940,100 frees, 79,491,791 bytes allocated
==12893== 
==12893== 8,620 (64 direct, 8,556 indirect) bytes in 1 blocks are definitely lost in loss record 71 of 74
==12893==    at 0x4C2CC70: calloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==12893==    by 0x4E5D92C: then_context_create (future.c:165)
==12893==    by 0x4E5D92C: flux_future_then (future.c:537)
==12893==    by 0x8D026E7: spawn_exec_handler (job.c:712)
==12893==    by 0x8D026E7: runevent_continuation (job.c:762)
==12893==    by 0x4E7B2F4: ev_invoke_pending (ev.c:3314)
==12893==    by 0x4E7E4B6: ev_run (ev.c:3717)
==12893==    by 0x4E523A2: flux_reactor_run (reactor.c:132)
==12893==    by 0x8D02B66: mod_main (job.c:943)
==12893==    by 0x40A978: module_thread (module.c:158)
==12893==    by 0x59D6183: start_thread (pthread_create.c:312)
==12893==    by 0x5EEE03C: clone (clone.S:111)
==12893== 
{
   <insert_a_suppression_name_here>
   Memcheck:Leak
   match-leak-kinds: definite
   fun:calloc
   fun:then_context_create
   fun:flux_future_then
   fun:spawn_exec_handler
   fun:runevent_continuation
   fun:ev_invoke_pending
   fun:ev_run
   fun:flux_reactor_run
   fun:mod_main
   fun:module_thread
   fun:start_thread
   fun:clone
}
==12893== 10,964 bytes in 1 blocks are definitely lost in loss record 72 of 74
==12893==    at 0x4C2AB80: malloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==12893==    by 0x5583A10: zmq::msg_t::init_size(unsigned long) (msg.cpp:77)
==12893==    by 0x52DA75C: zframe_new (zframe.c:52)
==12893==    by 0x52E4A85: zmsg_addmem (zmsg.c:293)
==12893==    by 0x4E57D21: flux_msg_copy (message.c:1324)
==12893==    by 0x8D01846: job_create_cb (job.c:423)
==12893==    by 0x4E53193: call_handler (msg_handler.c:302)
==12893==    by 0x4E5330E: dispatch_message (msg_handler.c:326)
==12893==    by 0x4E5330E: handle_cb (msg_handler.c:392)
==12893==    by 0x4E7B2F4: ev_invoke_pending (ev.c:3314)
==12893==    by 0x4E7E4B6: ev_run (ev.c:3717)
==12893==    by 0x4E523A2: flux_reactor_run (reactor.c:132)
==12893==    by 0x8D02B66: mod_main (job.c:943)
==12893==    by 0x40A978: module_thread (module.c:158)
==12893==    by 0x59D6183: start_thread (pthread_create.c:312)
==12893==    by 0x5EEE03C: clone (clone.S:111)
==12893== 
{
   <insert_a_suppression_name_here>
   Memcheck:Leak
   match-leak-kinds: definite
   fun:malloc
   fun:_ZN3zmq5msg_t9init_sizeEm
   fun:zframe_new
   fun:zmsg_addmem
   fun:flux_msg_copy
   fun:job_create_cb
   fun:call_handler
   fun:dispatch_message
   fun:handle_cb
   fun:ev_invoke_pending
   fun:ev_run
   fun:flux_reactor_run
   fun:mod_main
   fun:module_thread
   fun:start_thread
   fun:clone
}
==12893== LEAK SUMMARY:
==12893==    definitely lost: 11,028 bytes in 2 blocks
==12893==    indirectly lost: 8,548 bytes in 43 blocks
==12893==      possibly lost: 0 bytes in 0 blocks
==12893==    still reachable: 110,848 bytes in 108 blocks
==12893==         suppressed: 520 bytes in 2 blocks
==12893== Reachable blocks (those to which a pointer was found) are not shown.
==12893== To see them, rerun with: --leak-check=full --show-leak-kinds=all
==12893== 
==12893== For counts of detected and suppressed errors, rerun with: -v
==12893== ERROR SUMMARY: 2 errors from 2 contexts (suppressed: 1 from 1)
not ok 1 - valgrind reports no new errors on single broker run

# failed 1 among 1 test(s)
1..1

[grondo]

Seems good to me. At first I was thinking we should add a string of characters to trim, instead of just trimming one possible newline (e.g. any whitespace character might be useful), but that could be added later if even necessary.

LGTM!