117 ensure fresh install works even if ifadv missing (#121)

* Remove ifadv from dependencies

It is not an official CRAN dependency

* Hot fix the problem with ifadv

* Hot fix the problem with ifadv without using devtools

* Improvements in get_ifadv

But still causing odd problems: tests go fine on testing, but fail on checking

* Greatly simplify the download process

For some reason, native R data files (.rda) are less comfortable for R than the good-old csv. I pushed a csv copy of the rda file to https://github.com/elpaco-escience/ifadv/tree/csv, and that's the one we are using now.

DESCRIPTION

@@ -31,8 +31,7 @@ Imports:
 Suggests: 
     rmarkdown,
     testthat (>= 3.0.0),
-    ifadv,
     pkgdown,
-    ggrepel
-Remotes: git::https://github.com/elpaco-escience/ifadv.git
+    ggrepel,
+    utils
 Config/testthat/edition: 3

NAMESPACE

@@ -6,6 +6,7 @@ export(add_lines)
 export(calculate_timing)
 export(geom_token)
 export(geom_turn)
+export(get_ifadv)
 export(init)
 export(plot_density)
 export(plot_quality)

R/auxs.R

@@ -0,0 +1,16 @@
+#' Get IFADV data
+#'
+#' IFA Dialog Video corpus data
+#' Available in the public repository:
+#' https://github.com/elpaco-escience/ifadv
+#'
+#' This function requires an internet connection.
+#'
+#' @param source (default = "https://raw.githubusercontent.com/elpaco-escience/ifadv/csv/data/ifadv.csv")
+#'
+#' @return A data frame containing the IFADV dataset
+#' @export
+#'
+get_ifadv <- function(source="https://raw.githubusercontent.com/elpaco-escience/ifadv/csv/data/ifadv.csv") {
+  return(utils::read.csv(source))
+}

tests/testthat/_snaps/report.md

@@ -1,7 +1,7 @@
 # summary reports are accurate
 
     Code
-      report_stats(data)
+      report_stats(testdata)
     Output
       ### Corpus-level statistics 
       

tests/testthat/test-plots.R

@@ -12,7 +12,7 @@
 # between new and reference image. They are clickable from the test menu.
 
 # Load the test data
-testdata <- ifadv::ifadv
+testdata <- get_ifadv()
 
 #' Auxiliary decorator
 #'

tests/testthat/test-report.R

@@ -1,10 +1,10 @@
-## set up the test environment
-data <- ifadv::ifadv
+## Load the test data
+testdata <- get_ifadv()
 
 test_that("summary reports are accurate", {
   expect_snapshot(
     report_stats(
-      data
+      testdata
     )
   )
 })

vignettes/workflows.Rmd

@@ -33,8 +33,8 @@ The snippet below initializes the talkr dataset using the ifadv data. For more i
 
 ```{r}
 
-data <- init(ifadv::ifadv) 
-
+data <- get_ifadv()
+data <- init(data)
 ```
 
 Essential to any `talkr` workflow is a minimal set of data fields. These are the following: