Strict validation of input samplesheet.csv

[jannikseidelQBiC]

Following feature would be beneficial in the future:

• The headers of the samplesheet should be able to be validated in strict mode, meaning if they are misspelled or missing the pipelines should throw an error instead of a warning
• This behavior should be able to be turned on or off

The same feature would be nice for entries in the samplesheet, where required fields are missing.

e.g. if following samplesheet would be present

sample,short_reads_fastq_1,short_reads_fastq_2,long_reads_fastq_2
test,,path/to/first_fastq.gz,

and should be formatted like this

sample,short_reads_fastq_1,short_reads_fastq_2,long_reads_fastq_1
test,path/to/first_fastq.gz,,

and either short_reads_fastq_1 or long_reads_fastq_1are required fields

The pipeline should fail and not run empty with warnings. Also a meaningful error should be thrown in such a scenario.

It would be best this could be turned on or off in a schema_input.json.

[nvnieuwk]

The best solution for this is to create a new configuration option (validation.strictHeaders for example). This can then be turned on to throw errors instead of warnings for unknown headers.