Backup for process name and state

[Jimmy-653]

Retrieve process name and state from stat string if status map doesn't contain them already.
(I checked for Uid and Gid too, but it seems they are not contained in the string)

[dbwiddis]

I left a commit to fix the checkstyle issues CI was complaining about, and changed name and state not to conflict with the class variables. You need to rework the parsing logic, though. Filenames (and thus executables) can have spaces in the name.

[coveralls]

Coverage remained the same at 87.975% when pulling 946967c on J-Jimmy:upstream/1364_backup_process_name_state into e23ebd3 on oshi:master.

[Jimmy-653]

I can respect spaces in the name if I extract it like this.
However to get the state, I do need to split by whitespaces....There just is nothing else. (At least on my testsystem)

[dbwiddis]

Looking good. A few more requests and it'll be good to go!

[dbwiddis]

shouldn't this be nameStart + 1 to exclude the leading parenthesis and nameEnd (which is exclusive) to exclude the ending parenthesis?

[Jimmy-653]

I did output the original name from the status map and the name I extract here, the indexes are right as they are (on Ubuntu 20.04 LTS with Java 1.8_265)

[dbwiddis]

Ah, I see what you're doing. In this line you're capturing the full string with both parentheses, but then in the next line you're starting the substring from 1, and going to length-1 (which, since it's exclusive, has length-2 as last character), effectively removing the two parenthesis. If you just use start+1, end on this line, you don't need a second line.

[Jimmy-653]

Oh yeah, thats still in because I had the string split at first, obviously can be put together. I just woke up so I could do it in half an hour.

[dbwiddis]

I can make these fixes and merge if you'd like, but wanted to at least give feedback why.

[dbwiddis]

Scrutinzer says using the character '(' is faster than "(". I'm not sure it matters but why not? :)

[dbwiddis]

But benchmarks show only a tiny difference so I don't care that much. Still, if you're making other changes, the character version will make the CI tests happy and be a few unnoticed nanoseconds faster.

[dbwiddis]

Ah, I see what you're doing. In this line you're capturing the full string with both parentheses, but then in the next line you're starting the substring from 1, and going to length-1 (which, since it's exclusive, has length-2 as last character), effectively removing the two parenthesis. If you just use start+1, end on this line, you don't need a second line.

[dbwiddis]

Looks great! Merging.