Add ISLA worksheet generator

[fluffy-critter]

After enabling the isla_worksheet feature, the sample page will show a button to generate an ISLA worksheet, which provides an Excel spreadsheet suitable for providing ISLA data for the sample.

[silby]

These are the metadata fields I think we need for the most-routine ISLAs:

• DNA extraction date

• Cells used in extraction (10^6)

• DNA concentration of extraction (ng/µL)

• Eluted volume of extraction (µL)

• Input to PCR per replicate (µL)

• PCR round 1 date

• PCR round 2 date

• PCR round 3 date

• Notes

We should also include human-readable sample information in the output, as well as the sample ID. A link into Viroverse wouldn't go amiss. You could put $patient $tissue from $sample_date in a single cell or you could add patient, tissue type, sample date, sample name as prefilled, locked key-values.

Did we already go back and forth with Sheila or anyone about asking for the three-char prefix in our UI and have the PCR IDs filled in with defaults? People will probably otherwise drag-fill, which is fine.

Code thoughts when you're ready for final review.

[fluffy-critter]

I applied all of the refactoring suggestions and also improved the formatting. I also experimented with merging cells to make the formatting look a bit less grody in the metadata section but it only looked weirder instead.

You can re-review now if you want but I'm planning on adding in the automatic filling-in of known data as well. Also shouldn't there be metadata fields for pcr_primer and dna_reaction_kit? I had those before but I removed them since they weren't in the list on your first comment.

[silby]

Also shouldn't there be metadata fields for pcr_primer and dna_reaction_kit? I had those before but I removed them since they weren't in the list on your first comment.

The PCR primers used in each round of ISLA, as I understand it, are usually the same every time. My initial plan was to hardcode these in the importer. If we opt not to do that, note that there are 3 PCR rounds in the ISLA protocol with distinct primers. The extraction kits and PCR enzymes used also sounded like they don't really vary. These would be harder to capture correctly from spreadsheet input than primers because the namespace is much less standardized.