Retain resource attributes in write_package()

[peterdesmet]

Fix #191, fix #160.

• write_resource() no longer builds a resource from scratch for added resources, but uses the existing metadata (including custom properties).
• add_resource() will set some of the reserved properties like mediatype (to NULL), so they are at a logical position. These are then populated by write_package()
• The custom properties ... are now included BEFORE the schema.
• I also changed ifelse(condition, x, y) for if(condition) x else y, since it allows to return NULL in y 64e4f9d

Reprex for fixed #191:

library(magrittr)
library(frictionless)
my_package <-
  create_package() %>%
  add_resource(
    resource_name = "iris",
    data = iris,
    title = "Iris dataset", # Your additional metadata
    description = "The built-in dataset in R."
  )

# the package indeed includes the properties title and description
str(my_package)
#> List of 2
#>  $ resources:List of 1
#>   ..$ :List of 10
#>   .. ..$ name       : chr "iris"
#>   .. ..$ data       :'data.frame':   150 obs. of  5 variables:
#>   .. .. ..$ Sepal.Length: num [1:150] 5.1 4.9 4.7 4.6 5 5.4 4.6 5 4.4 4.9 ...
#>   .. .. ..$ Sepal.Width : num [1:150] 3.5 3 3.2 3.1 3.6 3.9 3.4 3.4 2.9 3.1 ...
#>   .. .. ..$ Petal.Length: num [1:150] 1.4 1.4 1.3 1.5 1.4 1.7 1.4 1.5 1.4 1.5 ...
#>   .. .. ..$ Petal.Width : num [1:150] 0.2 0.2 0.2 0.2 0.2 0.4 0.3 0.2 0.2 0.1 ...
#>   .. .. ..$ Species     : Factor w/ 3 levels "setosa","versicolor",..: 1 1 1 1 1 1 1 1 1 1 ...
#>   .. ..$ profile    : chr "tabular-data-resource"
#>   .. ..$ format     : NULL
#>   .. ..$ mediatype  : NULL
#>   .. ..$ encoding   : NULL
#>   .. ..$ dialect    : NULL
#>   .. ..$ title      : chr "Iris dataset"
#>   .. ..$ description: chr "The built-in dataset in R."
#>   .. ..$ schema     :List of 1
#>   .. .. ..$ fields:List of 5
#>   .. .. .. ..$ :List of 2
#>   .. .. .. .. ..$ name: chr "Sepal.Length"
#>   .. .. .. .. ..$ type: chr "number"
#>   .. .. .. ..$ :List of 2
#>   .. .. .. .. ..$ name: chr "Sepal.Width"
#>   .. .. .. .. ..$ type: chr "number"
#>   .. .. .. ..$ :List of 2
#>   .. .. .. .. ..$ name: chr "Petal.Length"
#>   .. .. .. .. ..$ type: chr "number"
#>   .. .. .. ..$ :List of 2
#>   .. .. .. .. ..$ name: chr "Petal.Width"
#>   .. .. .. .. ..$ type: chr "number"
#>   .. .. .. ..$ :List of 3
#>   .. .. .. .. ..$ name       : chr "Species"
#>   .. .. .. .. ..$ type       : chr "string"
#>   .. .. .. .. ..$ constraints:List of 1
#>   .. .. .. .. .. ..$ enum: chr [1:3] "setosa" "versicolor" "virginica"
#>  $ directory: chr "."
#>  - attr(*, "class")= chr [1:2] "datapackage" "list"

# writing the package
write_package(my_package, "irisdir")

# when reading the package again, title and description have disappeared FIXED
p <- read_package("irisdir/datapackage.json")
unclass(p)
#> $resources
#> $resources[[1]]
#> $resources[[1]]$name
#> [1] "iris"
#> 
#> $resources[[1]]$path
#> [1] "iris.csv"
#> 
#> $resources[[1]]$profile
#> [1] "tabular-data-resource"
#> 
#> $resources[[1]]$format
#> [1] "csv"
#> 
#> $resources[[1]]$mediatype
#> [1] "text/csv"
#> 
#> $resources[[1]]$encoding
#> [1] "utf-8"
#> 
#> $resources[[1]]$title
#> [1] "Iris dataset"
#> 
#> $resources[[1]]$description
#> [1] "The built-in dataset in R."
#> 
#> $resources[[1]]$schema
#> $resources[[1]]$schema$fields
#> $resources[[1]]$schema$fields[[1]]
#> $resources[[1]]$schema$fields[[1]]$name
#> [1] "Sepal.Length"
#> 
#> $resources[[1]]$schema$fields[[1]]$type
#> [1] "number"
#> 
#> 
#> $resources[[1]]$schema$fields[[2]]
#> $resources[[1]]$schema$fields[[2]]$name
#> [1] "Sepal.Width"
#> 
#> $resources[[1]]$schema$fields[[2]]$type
#> [1] "number"
#> 
#> 
#> $resources[[1]]$schema$fields[[3]]
#> $resources[[1]]$schema$fields[[3]]$name
#> [1] "Petal.Length"
#> 
#> $resources[[1]]$schema$fields[[3]]$type
#> [1] "number"
#> 
#> 
#> $resources[[1]]$schema$fields[[4]]
#> $resources[[1]]$schema$fields[[4]]$name
#> [1] "Petal.Width"
#> 
#> $resources[[1]]$schema$fields[[4]]$type
#> [1] "number"
#> 
#> 
#> $resources[[1]]$schema$fields[[5]]
#> $resources[[1]]$schema$fields[[5]]$name
#> [1] "Species"
#> 
#> $resources[[1]]$schema$fields[[5]]$type
#> [1] "string"
#> 
#> $resources[[1]]$schema$fields[[5]]$constraints
#> $resources[[1]]$schema$fields[[5]]$constraints$enum
#> [1] "setosa"     "versicolor" "virginica" 
#> 
#> 
#> 
#> 
#> 
#> 
#> 
#> $directory
#> [1] "irisdir"

^{Created on 2024-03-26 with reprex v2.1.0}

[damianooldoni]

Nice solutions to these two issues. And nothing against the use of standard if else on one line. It allows indeed to assign NULL to the output, and it's in my opinion slightly more readable than ifelse() or if_else() expressions. I will keep it in mind in my code writing.

Just two questions. Do you think worth to:

1. Add a small paragraph in the vignette where you show what is retained in the metadata by running write_package()?
2. Add a sentence with the limitations of write_package()? For example, it writes resources as CSVs only right now.

Just a reflection while reading the very clear Getting Started.

Tests, checks passed locally and documentation correctly compiled.

[peterdesmet]

Thanks for the review and suggestions Damiano! Here's why I won't update the documentation further:

• I show in the vignette how to add properties to a resource, but I don't find a good place to indicate that they are written, without breaking the flow of the text. I think it's a fair assumption of the user that they are written, without having to show it.
• The "limitations" (or features 😉 ) of write_package() are explained in the function documentation: it lists what happens for each type of resource. The vignette shows what files were written for my_package.

I'll merge this and get started on releasing 1.1.0.