The catalogue of field forms a project can carry, vendored from rfp's
rfp_form_types.csv. Non-spatial child tables are excluded — this is a
roster of map layers.
Value
A data.frame with columns: layer_key, form_type, label, description,
layer_name, geometry, symbol, color, label_expression. One row per spatial
form, ordered by layer_key.
Details
This is a separate table from gq_groups() rather than part of it, because
the two answer different questions. Forms are not baked into the QGIS
templates: rfp_qgs_form_add() injects them per project, and a project's
config selects which. So groups.csv carries the two forms the templates
ship, and this carries every form a project could ask for.
Folding the roster into groups.csv would make gq_template_layers() report
thirteen forms for a template that ships two — a project styled for layers it
never downloaded, which is the defect the composition guards exist to catch.
symbol and color are NA for a form rfp has registered but not styled.
That is a statement about upstream, not a gap here: gq does not invent
symbology for a form whose appearance nobody has decided.
Examples
forms <- gq_form_types()
nrow(forms)
#> [1] 13
# the key is derived from rfp's layer name, not from its type — these differ
forms[forms$form_type == "monitoring_fish_passage", c("form_type", "layer_key")]
#> form_type layer_key
#> 5 monitoring_fish_passage form_fish_passage_monitoring
# which forms rfp has styled, and which are still undecided
forms$layer_key[is.na(forms$color)]
#> [1] "form_edna" "form_fhap" "form_fish_sample" "form_monitoring"
# the subset the shipped templates actually carry
intersect(forms$layer_key, gq_group_layers("Forms")$layer_key)
#> [1] "form_fiss_site" "form_pscis"