without any filtering ability yet

plus refactor code a bit

- only show assumptions if there are errors related to parsing.

no valid sheets would result in app crashing, somade  e2e test for that and fixed it

e.g. if pollen calendar looks drastically different, then test will fail.

for this example, the way to solve it would fix bug that was ruining calendar OR if it is an improvement to calendar, update the snapshots in `cypress/snapshots/base/<testfilename>/<snapshotname>`

followed the `readme` of the package to setup ☝️
https://github.com/cypress-visual-regression/cypress-visual-regression

- can be used for the pollen calendar on `/pollen` page that user sees too

why?
0 has boolean value false so formatting data function was getting rid of those values

fix 🛠️
only get rid of undefined values (as on those dates, no pollen value was recorded)

- noticeable when clicking the filter buttons, the multichart would keep refreshing even tho the data hasn't changed yet

when changing the date period using the date filter, the last bar or line points on the chart is offsetted by 50%.

why?
when changing the date period using the date filter, the timestamp generated for the date would have time midday.

fix 🧑‍⚕️🧑‍⚕️
changing the date timestamp to have time of 00:00:00 (which is either midnight at the start of day or midnight at the end, idk)

no more weird offset

using the stack property to stack the bars only not the lines.

this is great because we don't have to manually set the max and min for each axis. and when

setting suggested min and max because if there is pollen values are all at 0, it should be obvious that the chart can go higher than 0 but the value was low. SORRY soooper confusing explanation, definitely easier to show.

…ontal line

it looked really weird and was especially noticeable for the dailyTotal because the horizontal line was around 40 on the y-axis

fix 🛠️: chart.js connects the points based on the order they appear in the array, not based on their x-axis value (in this case the date of the pollen measurement)

so sorting the parsed pollen data based on the date fixed this issue 👌👌👌😃😃😃

it should look as it does right now (50% difference is tolerated). but if it is not there, the test will let us know 😃🧪👌

even if errors, show any preview of any parsed data

update the types