Demonstrate the Bytelines iterator for logreduce

The goals of this iterator are:

• Split sub line to handle cmd output embedded as a long oneliner.
• Work with Read object, such as file decompressors or network endpoints.
• CPU and Memory efficient.

Here are some benchmark implementation:

• readline.py: Python readline
• readline.rs: Rust read_line
• byteslines.rs: Zero copy implementation using Bytes:

Without the sub line split:

• lines.rs: Rust lines
• byteslines-ns.rs

Using /bin/time and valgrind to measure performance with a 91MB journal.txt file:

$ wc journal.txt
  266400  6408707 90348877 journal.txt
$ /bin/time --format "Run in %e sec, Max RSS: %M KB" true
Run in 0.00 sec, Max RSS: 1988 KB
$ valgrind true |& grep "total heap usage"
   total heap usage: 0 allocs, 0 frees, 0 bytes allocated

Results:

Using Python 3.10.2 and rustc 1.52.1:

Implementation	Max RSS	Allocs	Frees	Bytes allocated	Run time
readline.py	7420 KB	1,814,409	1,810,434	475,434,838	0.33 sec
readline.rs	2260 KB	692,114	692,112	285,799,923	0.15 sec
byteslines.rs	2068 KB	24	22	265,577	0.12 sec
----------------	----------	-----------	-----------	-----------------	----------
lines.rs	2124 KB	277,396	277,394	104,598,961	0.05 sec
byteslines-ns	2148 KB	24	22	265,577	0.05 sec