Expand error message for `col_select`

[peterdesmet]

A col_select error currently warns if a column could not be found (👍):

Can't find column(s) animal-id in schema.

The user might not necessarily know what columns are available, so it would be helpful to show those:

Can't find column(s) animal-id, vis in field names.
ℹ Field names: event-id, visible, timestamp, location-long, location-lat, bar:barometric-pressure, external-temperature, gps:dop, gps:satellite-count, gps-time-to-fix, ground-speed, heading, height-above-msl, import-marked-outlier, location-error-numerical, manually-marked-outlier, vertical-error-numerical, sensor-type, individual-taxon-canonical-name, tag-local-identifier, individual-local-identifier, study-name

I also rephrased the first sentence, to mention the word "field names". I think it is less important to mention "schema".

[PietrH]

I discovered something interesting, R will output something different depending on if you specify the unicode symbol "information source" by it's hex code "\u2139" or by typing it directly (or pasting it) "ℹ️".

These display differently, and R doesn't consider them the same:

"\u2139"
#> [1] "ℹ"

"ℹ️"
#> [1] "ℹ️"

"ℹ️" == "\u2139"
#> [1] FALSE

^{Created on 2023-04-04 with reprex v2.0.2}

This is because they have a different variation form specified, the first in text representation mode, the second using variation selector sequence 16 (ref)

The symbol we are using is even an example on wikipedia: https://en.wikipedia.org/wiki/Emoji#Emoji_versus_text_presentation

So both characters are the same, but different.

---

Anyway, we should pick if we want to use text or emoji presentation, I don't know what the fallback is when the emoji is unavailable, I suspect text mode.

---

I actually really like how the emoji displays. Using text representation is probably more defensive. Any opinions?

Emoji

Text Mode

[peterdesmet]

I think we should use text mode. I've encountered R CMD Check issues regarding "not ascii" when I relied on the emoji symbol.