Add the m option (from grep). #93
Conversation
|
This looks reasonable, but I want to test it offline first and it will be a week or two before I can do that. Thanks for this (in advance)! |
There was a problem hiding this comment.
This entire hunk is plagiarized from the GNU grep manual and cannot be used. Not just because of the blatant copyright violation, but also because it does not accurately describe the added functionality.
There was a problem hiding this comment.
Yes, it was included by accident I think, sorry for that m(_ _)m
I had decided to not include any change to the manual page, but it must have slip through somehow.
Also, I don't get why there are so many changes...
Sorry again, as I said below, you can safely ignore this PR.
| @@ -652,6 +666,10 @@ Copyright (c) 2001-2009 Ville Laurikari <vl@iki.fi>.\n")); | |||
| if (show_help) | |||
| tre_agrep_usage(0); | |||
|
|
|||
| /* Fail without reading if count number is 0 */ | |||
| if (max_count == 0) | |||
| return EXIT_FAILURE; | |||
There was a problem hiding this comment.
We don't use EXIT_SUCCESS or EXIT_FAILURE elsewhere, so we shouldn't use them here.
There was a problem hiding this comment.
Then would return 1; be more appropriate?
| /* exit after max_count match(es) */ | ||
| if((max_count > 0)&&(count > (max_count-1))) { | ||
| if(count_matches) break;/* to get the result */ | ||
| else exit(0);/* to leave the stream "as is" */ |
There was a problem hiding this comment.
Exiting here produces the wrong result when multiple inputs are provided and at least one of them has max_count or more matching lines. The correct thing to do is break in all cases. Moreover, we're reading from a file descriptor, not a stream, and since we read a variable amount of data (but at least 10 kB) at a time, there is no guarantee that the input will be left in any particular state when we reach the match limit.
There was a problem hiding this comment.
Yes, I added the ugly else exit(0); as a quick fix for my use case.
There were some trouble either when piping or using multiple files, but I can't recall what problems exactly, sorry m(_ _)m
This behaves similarly to the equivalent option in GNU and BSD grep, i.e. stops processing each input as soon as the specified number of matches is reached. Fixes laurikari#93.
|
Bonjour @dag-erling, Thank you very much for your review. Anyway, your PR is surely more thoughtful so you can ignore this one! Have a nice day, |
|
merged #110 instead |
I wanted to stop
agrepafter N matches, so I add a-moption as close as possible to that of the GNUgrep.I'm sharing this in the hope that it will be useful to someone else.
Note: When using
-mcombined with the-Boption,agrepwill look for the N-best candidates.