Thanks to visit codestin.com
Credit goes to github.com

Skip to content

Releases: hellige/au

"Standardize" printing of NaN, etc.

12 Jun 15:55

Choose a tag to compare

When printing JSON, au now emits NaN, Infinity and -Infinity rather than nan, inf, and -inf. (None of this is standard JSON, but these seem to be more widely accepted.) There is no change to the binary format.

Bug fixes, build updates

03 Jun 16:02

Choose a tag to compare

This release fixes a bug in the behavior of tail -f and improves the handling of time/date prefix patterns from command line. There were also small build updates for newer compilers and distros.

Added ztail command

26 Apr 21:01

Choose a tag to compare

In this release we have the ability to tail a gzipped file, provided an index is availble. A new command, au ztail, forces au to treat the input stream as compressed (and error if no index is available), while the existing tail command will detect this automatically.

Bug fix: improve handling of large records

12 Feb 01:24

Choose a tag to compare

This release improves the handling of very large records during binary search (grep -o).

new features

07 Jan 21:18

Choose a tag to compare

This release contains various new features in the au command-line tool, mostly related to the grep command:

  • au will now detect gzipped files and automatically handle them correctly. zcat/zgrep are now deprecated.
  • au grep can now handle plain ASCII log files with timestamps at the start of (some) lines. See au grep -l.
  • au grep gained several other new arguments, including -g and -F, and the bisect functionality was improved.
  • au grep will now accept timestamps without a date (e.g., 12:34:56.100) and attempt to infer the date from the input stream.

grep now supports json files!

29 Dec 23:19

Choose a tag to compare

The au grep command now supports normal JSON files as well as au-encoded files. au will attempt to auto-detect whether the input stream is JSON. Gzipped JSON files are supported as well, provided you build an index first using au zindex (just like with au-encoded files).

The ability to binary search large gzipped JSON log files can be very convenient:

[~]% ls -lh big.json
-rw-r--r-- 1 mhellige mhellige 22G Dec 29 17:00 big.json
[~]% time au grep -c -o estdEventTime 2021-12-17T08:00:00 big.json
1263
au grep -c -o estdEventTime 2021-12-17T08:00:00 big.json  0.12s user 0.02s system 99% cpu 0.141 total

Not bad for a 22G JSON file!

Improve handling of large records

28 Dec 22:24

Choose a tag to compare

Previously, grep could fail upon encountering large records when the underlying stream was non-seekable (e.g., when grepping stdin or a non-indexed gzipped file). This release fixes that problem.

Fix performance regression

16 Nov 23:25

Choose a tag to compare

This release fixes a performance regression in zgrep -o that was introduced in the last release.

Added zcat command

28 Oct 20:56
519d414

Choose a tag to compare

This release adds the au zcat command as a convenience.

Performance improvements

27 Aug 18:08
47e55fd

Choose a tag to compare

Performance improvements, fixes for recent compilers.