# |
Date |
Size |
|
Editor |
Comment |
Action |
480 |
2016-10-14 14:24:04 |
58844 |
to previous |
GreyCat |
Pitfall 49: OIFS="$IFS"; ...; IFS="$OIFS" |
view |
479 |
2016-09-24 19:04:15 |
58297 |
to previous |
geirha |
forgot the anchor |
view |
478 |
2016-09-24 18:49:28 |
58279 |
to previous |
geirha |
stray quote |
view |
477 |
2016-09-24 18:44:42 |
58280 |
to previous |
geirha |
pf 48 -- export CDPATH |
view |
476 |
2016-09-16 12:40:52 |
56968 |
to previous |
geirha |
no, the point was to show it works without the $ |
view |
475 |
2016-09-16 12:01:09 |
56969 |
to previous |
80 |
Fix a missing dereference |
view |
474 |
2016-05-14 22:17:39 |
56968 |
to previous |
tor-exit |
[ If an open bracket introduces a bracket expression as in XBD RE Bracket Expression, except that the <exclamation-mark> character ( '!' ) shall replace the <circumflex> character ( '^' ) in its |
view |
473 |
2016-03-30 12:10:15 |
56968 |
to previous |
GreyCat |
|
view |
472 |
2016-03-30 12:06:52 |
56730 |
to previous |
GreyCat |
[ x == y ] is NOT valid in posix shells. It is a bashism. I will add some clarifying language after this revert. |
view |
471 |
2016-03-30 02:06:52 |
56490 |
to previous |
aftr-5-146-249-61 |
This simply isn't true - '==' works in '['! |
view |
470 |
2016-03-08 05:41:15 |
56730 |
to previous |
ool-457d9675 |
emphasize "for x;do" more in example 24 |
view |
469 |
2016-03-02 08:11:54 |
56652 |
to previous |
lmcb-76 |
Added a comment about using find for pitfall 1. bolded the so whats the right way to do it. |
view |
468 |
2016-01-18 13:03:44 |
56502 |
to previous |
StephaneChazelas |
|
view |
467 |
2016-01-18 13:02:01 |
56497 |
to previous |
StephaneChazelas |
|
view |
466 |
2016-01-18 12:50:50 |
56163 |
to previous |
StephaneChazelas |
Mention that it also concerns symlinks to inaccessible files. |
view |
465 |
2016-01-18 12:44:39 |
56026 |
to previous |
StephaneChazelas |
errors on stderr |
view |
464 |
2016-01-11 21:25:12 |
56022 |
to previous |
GreyCat |
not perfectly... |
view |
463 |
2015-12-24 00:38:26 |
56024 |
to previous |
GreyCat |
((foo > 7)) and [[ foo -gt 7 ]] are both *correct*, in both bash and ksh. Try 'em. |
view |
462 |
2015-12-23 17:30:55 |
56026 |
to previous |
38 |
Adds missing $ signs in first part of Pitfall 7. |
view |
461 |
2015-11-26 15:34:38 |
56024 |
to previous |
GreyCat |
Completely incorrect assertion. Try this: var="-n"; echo "$var" Even if your assertion WERE true for bash >= n we would still have to mark it a pitfall for other shells/versions. |
view |
460 |
2015-11-26 08:18:12 |
55716 |
to previous |
pool-173-61-210-188 |
Removed claim that echo can't safely display a variable like var="-n Text". As of bash 4.1.7, echo $var _will_ interpret the leading -n as a flag, but echo "$var" will print exactly what's expected. |
view |
459 |
2015-11-22 21:25:54 |
56024 |
to previous |
GreyCat |
ssh (or ffmpeg) only running once in a while read loop is FAQ 89 |
view |
458 |
2015-11-22 20:48:03 |
56519 |
to previous |
199 |
|
view |
457 |
2015-09-02 15:13:49 |
56024 |
to previous |
GreyCat |
Put the UMQ right up top. Change some wording. |
view |
456 |
2015-09-02 14:47:54 |
55095 |
to previous |
GreyCat |
no, IFS=$'\n' won't work either |
view |
455 |
2015-09-02 13:05:50 |
54761 |
to previous |
cpc38-dudl11-2-0-cust78 |
|
view |
454 |
2015-08-18 08:42:11 |
54695 |
to previous |
geirha |
The description of how sed works, and even the proposed solution is wrong. |
view |
453 |
2015-08-17 23:23:13 |
55177 |
to previous |
pool-100-32-136-77 |
added pitfal on using sed to grab values from stream |
view |
452 |
2015-08-13 11:55:05 |
54695 |
to previous |
GreyCat |
No, it really was pitfall #2. Not #3. But screw it. Just put English words directly in the comment. |
view |
451 |
2015-08-13 02:32:38 |
54693 |
to previous |
llua |
|
view |
450 |
2015-04-18 22:02:23 |
54627 |
to previous |
GreyCat |
IFS=, read -ra fields <<< "$csv_line" |
view |
449 |
2015-04-18 21:51:18 |
53375 |
to previous |
GreyCat |
fix anchors |
view |
448 |
2014-11-02 11:37:32 |
53423 |
to previous |
vil69-1-82-67-51-185 |
|
view |
447 |
2014-10-10 12:10:04 |
53354 |
to previous |
pgas |
|
view |
446 |
2014-09-22 18:02:34 |
53099 |
to previous |
freedom2 |
|
view |
445 |
2014-07-05 09:08:40 |
52949 |
to previous |
MikeJonkmans |
added some words on `]` |
view |
444 |
2014-06-28 13:44:22 |
52712 |
to previous |
ormaaj |
Don't use which(1). Since [ is usually builtin, which will lie to you. |
view |
443 |
2014-06-27 14:53:19 |
52718 |
to previous |
gbibp9ph1--blueice3n7 |
|
view |
442 |
2014-06-25 20:05:28 |
52658 |
to previous |
ormaaj |
mksh now supports pipefail. |
view |
441 |
2014-06-02 14:56:14 |
52590 |
to previous |
GreyCat |
y=$(( array[$x] )) code injection |
view |
440 |
2014-04-06 16:42:55 |
51679 |
to previous |
212 |
typo fixed |
view |
439 |
2014-03-27 12:12:34 |
51680 |
to previous |
GreyCat |
indentation in PF 4 |
view |
438 |
2014-01-24 21:17:15 |
51667 |
to previous |
geirha |
just testing what the gui editor would do |
view |
437 |
2014-01-24 21:15:01 |
51739 |
to previous |
geirha |
|
view |
436 |
2014-01-24 21:08:49 |
51667 |
to previous |
GreyCat |
Please don't put blank lines between Anchor and question. Greybot parses this, and those break it. |
view |
435 |
2014-01-21 13:47:04 |
51741 |
to previous |
GreyCat |
|
view |
434 |
2014-01-21 13:44:49 |
51736 |
to previous |
GreyCat |
remove set -e stuff; I merged it into FAQ 105 |
view |
433 |
2014-01-21 11:40:30 |
52200 |
to previous |
mx1 |
|
view |
432 |
2014-01-07 18:22:44 |
51764 |
to previous |
StephaneChazelas |
removed note about quote nesting for backticks which is incorrect, add missing -P and -- for cd and dirname |
view |
431 |
2014-01-07 18:17:14 |
51877 |
to previous |
StephaneChazelas |
adding missing quotes |
view |
430 |
2013-12-16 16:46:51 |
51867 |
to previous |
94 |
Ever heard of the ext4 data loss "bug"? |
view |
429 |
2013-10-18 15:45:47 |
51740 |
to previous |
GreyCat |
PF35: same issue with [[ $foo = "glob" ]] |
view |
428 |
2013-10-13 05:41:43 |
51499 |
to previous |
76-10-140-160 |
added missing quotes around possible GLOB |
view |
427 |
2013-08-05 00:51:47 |
51456 |
to previous |
shmdmzpr03-ext |
|
view |
426 |
2013-07-03 12:33:03 |
51467 |
to previous |
fran |
Don't break out of loop in case a file has been removed before it is processed. |
view |
425 |
2013-06-06 10:52:15 |
51464 |
to previous |
host223-59-dynamic |
|
view |
424 |
2013-04-18 05:49:06 |
51457 |
to previous |
geirha |
Revert to revision 422. |
view |
423 |
2013-04-17 12:23:17 |
51458 |
to previous |
92 |
|
view |
422 |
2013-04-04 22:45:26 |
51457 |
to previous |
GreyCat |
Variables in a math context DON'T require a $ to be expanded. That was kinda the point of the "Works, but most will consider it wrong" comment. |
view |
421 |
2013-04-04 12:24:30 |
51459 |
to previous |
95 |
variable require $ prefix |
view |
420 |
2013-03-25 12:48:41 |
51457 |
to previous |
GreyCat |
pf41: (lack of) portability note |
view |
419 |
2013-03-24 11:02:48 |
51419 |
to previous |
ormaaj |
Add crappy array wordsplit abuse to pf 1 |
view |
418 |
2013-03-24 10:47:50 |
51341 |
to previous |
ormaaj |
|
view |
417 |
2013-03-24 10:44:33 |
51332 |
to previous |
ormaaj |
There's nothing wrong with this expression, people just need to be aware of its behavior. |
view |
416 |
2013-02-11 01:26:07 |
50308 |
to previous |
ormaaj |
Had to fix indenting. Also, done through pf 11 |
view |
415 |
2013-02-10 22:24:50 |
49735 |
to previous |
ormaaj |
Copy editing. Pf 1-8 done. Enough for today. |
view |
414 |
2013-02-09 17:32:16 |
46184 |
to previous |
ormaaj |
Buggy solution. Needs either -- or to not to strip the ./. |
view |
413 |
2013-01-27 06:38:12 |
46291 |
to previous |
Lhunath |
nonsense edit. |
view |
412 |
2013-01-27 04:17:04 |
46294 |
to previous |
18 |
|
view |
411 |
2013-01-15 21:19:30 |
46291 |
to previous |
GreyCat |
link to UsingFind in #1 |
view |
410 |
2013-01-03 13:11:40 |
46231 |
to previous |
GreyCat |
simplify |
view |
409 |
2013-01-03 10:11:21 |
46244 |
to previous |
ormaaj |
(($?)) && ... |
view |
408 |
2012-12-20 16:43:27 |
45589 |
to previous |
HSI-KBW-46-223-133-42 |
|
view |
407 |
2012-12-20 16:32:05 |
45565 |
to previous |
HSI-KBW-46-223-133-42 |
|
view |
406 |
2012-09-03 12:30:35 |
45306 |
to previous |
nrbg-4dbe18f3 |
Let's provoke a disaster on Debian too that can be taken seriously. Here, rm * will _always_ (by default) do a rm -i *. (Except for the superuser) |
view |
405 |
2012-07-30 09:49:49 |
45303 |
to previous |
151 |
pitfall 23: exmark='!' ; echo "Hello, world$exmark" |
view |
404 |
2012-07-21 19:36:26 |
45225 |
to previous |
178 |
Mention Substring Expansion for removing first character of a word |
view |
403 |
2012-07-21 19:19:34 |
45157 |
to previous |
178 |
|
view |
402 |
2012-06-28 15:53:52 |
45161 |
to previous |
201 |
We have to be consistent in variable naming conventions. !varcap. |
view |
401 |
2012-04-23 16:41:01 |
45161 |
to previous |
c-71-232-80-14 |
|
view |
400 |
2012-04-04 03:57:57 |
44974 |
to previous |
ormaaj |
2>&1. How'd we go without this so long? |
view |
399 |
2012-04-02 10:10:42 |
43922 |
to previous |
Lhunath |
don't take personal offense, 125, and thanks for the contribution. |
view |
398 |
2012-04-02 09:08:39 |
43955 |
to previous |
125 |
sorry for mistakes ive made. |
view |
397 |
2012-04-02 08:36:10 |
44112 |
to previous |
125 |
|
view |
396 |
2012-04-02 06:45:59 |
43955 |
to previous |
Lhunath |
|
view |
395 |
2012-04-02 06:44:15 |
43955 |
to previous |
Lhunath |
|
view |
394 |
2012-04-02 04:53:54 |
43684 |
to previous |
125 |
|
view |
393 |
2012-04-02 04:51:13 |
43679 |
to previous |
125 |
|
view |
392 |
2012-03-29 20:49:26 |
42242 |
to previous |
ormaaj |
use greycat's version. |
view |
391 |
2012-03-29 20:46:54 |
41881 |
to previous |
ormaaj |
Too long, greycat angry, so move to BraceExpansion |
view |
390 |
2012-03-29 20:45:22 |
42242 |
to previous |
GreyCat |
Strip esoteric crap out of #33 (for i in {1..$n}). If the material is important, please put it on BraceExpansion or in a separate FAQ entry. Removed redundant category tags. |
view |
389 |
2012-03-29 19:32:35 |
44031 |
to previous |
ormaaj |
...and add some workarounds to make it generalize to not only loops. This might be long enough for an FAQ. |
view |
388 |
2012-03-28 00:07:32 |
42503 |
to previous |
ormaaj |
Brace expansion "portability" note. The Bash way isn't entirely "unfortunate". |
view |
387 |
2012-02-22 09:26:34 |
41881 |
to previous |
129 |
"[[ -f $i ]] || continue" changed to '[[ -f "$i" ]] || continue' |
view |
386 |
2011-11-27 02:05:53 |
41879 |
to previous |
184-175-28-167 |
|
view |
385 |
2011-11-08 16:16:20 |
41838 |
to previous |
GreyCat |
answer question in #1 |
view |
384 |
2011-11-08 12:53:12 |
41701 |
to previous |
129 |
|
view |
383 |
2011-09-22 21:51:35 |
41528 |
to previous |
cust |
zsh will accept such syntax too |
view |
382 |
2011-08-21 18:44:50 |
41482 |
to previous |
35 |
|
view |
381 |
2011-08-10 18:07:47 |
41476 |
to previous |
GreyCat |
perl's grep command has nothing to do with variables being lost because they were set in a subshell in a bash script |
view |