Refactor MIB parser

[jorpic]

• [x] unit tests for cassava FromRecord
• [x] low-boilerplate parser generator
• [x] unit tests for conversions
• [x] applicative style in conversions

I added unit tests for each of the FromRecord instances but for VDF and LGF there is not enough data in ASCII_RTE folder vfd.dat contains single entry which appears to be invalid and lgf.dat is just empty.

Also added unit tests for calibration conversions. Not sure if they are adequate, please check.

Some parsers check input record length with (>=). E.g. can input for PCFentry really be longer than 23? https://github.com/oswald2/AURIS/blob/0712070ebb4ffd4c13b609629f23ad28820f670b/esa-mib/src/Data/MIB/PCF.hs#L95-L97 PCF also provides some defaults in case input is a bit shorter. https://github.com/oswald2/AURIS/blob/0712070ebb4ffd4c13b609629f23ad28820f670b/esa-mib/src/Data/MIB/PCF.hs#L145 Can the length be 21? In this case should parser overwrite fields 19 and 20 with defaults?

[oswald2]

The VDF entry is indeed not right, I corrected it on the master branch.

I added a single entry to LGF.dat, which is comletely analog to mcf.dat, just it doesn't do a polynomial calculation, but a logarithmic one.

For the PCF, I just checked with the documentation and:

• No, there are not more than 23 fields
• The last fields are optional because in older versions of the MIB these were not present and they were added later. In order to have a backwards compatibility, the parser should not fail and therefore the default values were added. In principal it should not happen, that 21 fields are present, it's either 19 or 23. Still, we have seen in practice that some MIBs sometimes leave the last fields empty (which means apply the default value), but still the TAB separators need to be present, so the number of fields is either 19 or 23.

Do you need further updates on this with these answers? Otherwise I would merge it.

[jorpic]

Thank you for the answers. No further information is needed. Please do not merge it yet, I'll rebase my changes on top of the updated master and will change tests accordingly.

[oswald2]

I pushed changes to master:

• corrected the vdf entry
• added 2 lgf entries