Commit graph

715 commits

Author SHA1 Message Date
Kurtis Rader
0541a34201 improve bash command history importing
Fixes #2091
2017-02-08 20:50:57 -08:00
Andrei Dziahel
85ba4e0142 Fix colors table (#3826)
* color: make brgrey really grey

The 0055 value is actually 0x2d which isn't 0x55 mentioned further and is probably a typo

* color.cpp: reformat color table

Tidy color table up and also fix hex number case for grey color. This should ease spotting errors like one from previous commit.
2017-02-08 13:49:19 -08:00
Kurtis Rader
0970cc8736 time for another make style-all 2017-02-07 21:52:35 -08:00
Kurtis Rader
616d301083 set stdout to unbuffered if attached to a tty
Fixes #3748
2017-02-07 21:01:22 -08:00
ridiculousfish
162053ed8d Make history file rewriting be more lock savvy
We now are stingier with taking history file locks - if the lock
is held too long we may just break it. But the current file save
architecture holds the lock for the duration of the save. It also
has some not-quite-right checks that can cause spurious failures in
the history stress test.

Reimplement the history save to retry. Rather than holding the lock,
rewrite the file to a temporary location and then take the lock. If
the history file has changed, start all over.

This is going to be slower under contention, but the advantage is that
the lock is only held for a brief period (stat + rename) rather than
across calls to write().

Some updated logic also fixes spurious failures that were easy to observe
when tsan was enabled. These failures were due to failing to check if the
file at the path was the same file we opened.

The next step is to move the history file saving to a background thread
to reduce the chances of it impacting user's typing.
2017-02-06 11:04:07 -08:00
ridiculousfish
f9371899a1 Including missing atomic header in history.cpp 2017-02-06 10:51:27 -08:00
ridiculousfish
56b60fab4b Improve history's save_internal_via_appending
Allow retrying, fix an issue where we trip over our own changes
by thinking the file has changed when we are responsible for changing
it, and improve some commenting
2017-02-06 10:39:07 -08:00
ridiculousfish
b91a2dd0bb Factor map_fd out of map_file
Will enable us to map an already-opened file
2017-02-06 10:09:31 -08:00
ridiculousfish
83f386347f Make file_id_for_fd return kInvalidFiledID when passed a negative fd
Rather than making fstat do this, make it explicit.
2017-02-06 10:01:33 -08:00
ridiculousfish
6eb1e31070 Add some safety to history_file_lock
Use an atomic to avoid TSAN complaints, and ensure that we don't
leave the file locked if we've decided to give up on locking.
2017-02-06 10:00:34 -08:00
ridiculousfish
e3c538a991 Simplify history_output_buffer_t
This class is used to accumulate data to be written to the history
file. It has some dubious optimizations around trying to track an
offset separately from the size of the buffer. After some investigation
these aren't helping, vector behaves fine on its own. So just make
this a simple wrapper around vector.
2017-02-06 10:00:01 -08:00
Kurtis Rader
c4f2210cc5 cache prompts and escape sequences we've seen
Cache the escape sequences we've seen when checking for those which
don't take any visual space when writing the prompt or similar strings.
This reduces the cost of determining the true cost of such strings by a
full order of magnitude if they include lots of such escape sequences.

Periodically sort the cached escape sequence lengths based on feedback
from cache hits so that we're always checking for the most likely
sequence lengths first.

Also cache the prompt layouts to avoid doing the calculations if the
prompt doesn't change.

Fixes #3793
2017-02-05 18:31:14 -08:00
Kurtis Rader
6d72d538a6 refactor escape_code_length()
Step one in addressing issue #3793
2017-02-05 18:31:14 -08:00
ridiculousfish
56161250de Make a missing --sets-mode property for a key binding do nothing
Currently, if bind is run with --mode but not --sets-mode, the
binding gets an implicit --sets-mode equivalent to the mode. This
is usually unobservable but it may matter if the mode is changed
by some internal part of the binding (e.g. set fish_bind_mode...)
then that setting will be lost after the binding is complete.
2017-02-05 02:49:45 -08:00
ridiculousfish
f48291b70c Add #include<atomic> to env_universal_common.cpp
Should fix the Linux builds
2017-01-30 13:41:17 -08:00
ridiculousfish
2a472af7b8 Add sanity checking to LRU cache 2017-01-30 12:59:01 -08:00
ridiculousfish
927a678056 Add support for sorting in LRU caches
Performs an in-place merge sort
2017-01-30 10:25:17 -08:00
ridiculousfish
2565ffab25 Make a bool atomic in env_universal_common.cpp
Fixes a race identified by thread sanitizer
2017-01-29 21:42:33 -08:00
ridiculousfish
cb70ac6932 Use a std::atomic in test_iothread
Reduces noise from thread-sanitizer
2017-01-29 21:35:22 -08:00
ridiculousfish
e8d90dbf4b Remove support for the builtin_script_t feature
This was an old experiment to compile scripts directly into the
shell itself, reducing the amount of I/O performed at startup.
It has not been used for a long time. Time to remove it.
2017-01-29 21:34:30 -08:00
ridiculousfish
6db372133d Clean up and adopt owning_lock in intern.cpp 2017-01-29 21:19:39 -08:00
ridiculousfish
f220250ada Adopt owning_lock in wgettext 2017-01-29 21:12:23 -08:00
ridiculousfish
4ac2cfba61 Adopt owning_lock in iothread.cpp 2017-01-29 21:06:46 -08:00
ridiculousfish
0fefdb458f Revert "Adopt owning_lock and some cleanup of termsize storage in common.cpp"
Tests are failing on Travis but not locally

This reverts commit c5d9e7e391.
2017-01-29 19:33:30 -08:00
ridiculousfish
8aab725782 Make autoload_t no longer virtual
Equip it instead with a function pointer that it invokes
when a command is removed
2017-01-29 18:56:55 -08:00
ridiculousfish
c5d9e7e391 Adopt owning_lock and some cleanup of termsize storage in common.cpp 2017-01-29 18:15:38 -08:00
ridiculousfish
3569987c81 Adopt owning_lock in builtin_commandline 2017-01-29 18:15:38 -08:00
ridiculousfish
10a2275c34 Adopt owning_lock in history_collection_t 2017-01-29 18:15:38 -08:00
ridiculousfish
017836cffa Adopt owning_lock for job_ids 2017-01-29 18:15:38 -08:00
ridiculousfish
eaf143dd8a Introduce owning_lock template
Will enable rust-style lock ownership semantics
2017-01-29 18:15:38 -08:00
Kurtis Rader
973097f025 style fixups 2017-01-29 18:00:48 -08:00
ridiculousfish
e52a04e341 Cleanup of LRU cache implementation
Switch to CRTP from virtual functions and improve ownership semantics.
It's no longer necessary for clients to use new and delete.
2017-01-29 12:16:42 -08:00
Kurtis Rader
5b6cc5af6d time for another make style-all
Gotta keep the entropy (i.e., disorder) from increasing.
2017-01-26 20:05:46 -08:00
ridiculousfish
05d569ee44 Simplify error case of token_type_description and keyword_description
Simply return a constant string, since in practice the error case
is not hit
2017-01-26 17:47:24 -08:00
ridiculousfish
a40f491c93 Defeat some miscellaneous warnings in the g++ build 2017-01-26 17:33:48 -08:00
ridiculousfish
e78cefd759 Defeat "enumeral mismatch" g++ warnings through typecasting 2017-01-26 17:18:38 -08:00
ridiculousfish
1efb81456b Use std::move instead of swap in a few places where it improves clarity 2017-01-26 16:14:50 -08:00
ridiculousfish
fec83fa975 Eliminate moved_ref
Use real rvalue references instead
2017-01-26 15:43:46 -08:00
ridiculousfish
1634c9df78 Make job_get_flag and job_set_flag instance methods of jobs
Makes them easier to call when you have a smart pointer
2017-01-26 15:06:58 -08:00
ridiculousfish
14fb38f952 Switch job handling to use shared pointers instead of raw pointers
Clarifies memory management around allocation of job_ts
2017-01-26 14:47:32 -08:00
ridiculousfish
1d9cc12984 Repair control-C cancellation of loops
Commit ab189a75 introduced a regression where we stop breaking out
of loops in response to a child death via a signal. Fix that regression.

Also introduces a test to help ensure we don't regress in the future.

Fixes #3780
2017-01-26 13:41:56 -08:00
ridiculousfish
2fb1d5900d Use true and false for bool, not 0 and 1 2017-01-26 12:51:22 -08:00
ridiculousfish
59a3968fd2 Switch to using unique_ptr in env_node_t
Makes our memory management of the variable stack more explicit
2017-01-26 12:51:22 -08:00
ridiculousfish
8e577b01bc Migrate "unspecified scope" logic for set into var_stack_t 2017-01-26 12:51:22 -08:00
ridiculousfish
8d2dfdf2c9 Migrate environment variable cache into var_stack_t 2017-01-26 12:03:14 -08:00
ridiculousfish
dabc34e0f9 Migrate next_scope_to_search into var_stack_t 2017-01-26 11:32:45 -08:00
ridiculousfish
3525a9d7ee Make env_node_t's destructor private
Migrate responsibility for popping the environment into var_stack_t
2017-01-26 11:20:09 -08:00
ridiculousfish
61887c061b Migrate responsibility for node creation into var_stack_t 2017-01-26 11:06:03 -08:00
ridiculousfish
e4fd1d5f5a Wrap up the variable stack in a little struct
First step towards cleaning up env.cpp's global variables
2017-01-26 10:38:55 -08:00
ridiculousfish
afe54df843 Eliminate the 'global' var_table_t variable in env.cpp
This can always be trivially found from 'global_env'
2017-01-26 10:32:31 -08:00