Added tests for multiline text support

.github/workflows/tests.yaml

@@ -20,6 +20,7 @@ jobs:
       GITHUB_PAT: ${{ secrets.PAT_GITHUB }}
       GH_ACTION: "ENABLED"
       CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
+      ANIMINT2_TEST_GHPAGES: ${{ github.event_name == 'push' && (github.ref == 'refs/heads/master' || github.ref == 'refs/heads/main') }}
     steps:
       - uses: actions/checkout@v3
       - name: install and update texlive
@@ -48,14 +49,18 @@ jobs:
         run: |
           npm install v8-to-istanbul
           echo "Node modules installed"
+
+      - name: Install xvfb (JS_coverage)
+        if: matrix.test-suite == 'JS_coverage'
+        run: /usr/bin/sudo DEBIAN_FRONTEND=noninteractive apt-get install -y xvfb
 
       - name: run tests
         run: |
           if [ "$TEST_SUITE" == "CRAN" ]; then
             bash build.sh
           elif [ "$TEST_SUITE" == "JS_coverage" ]; then
             echo "Running testthat with JS coverage collection..."
-            Rscript -e "source('tests/testthat.R', chdir = TRUE)"
+            xvfb-run -a Rscript -e "source('tests/testthat.R', chdir = TRUE)"
           else
             echo "Running testthat with R coverage collection..."
             Rscript -e 'covr::codecov(quiet = FALSE, type = "tests", test_files = "tests/testthat.R", flags = "r")'

DESCRIPTION

@@ -281,6 +281,7 @@ Collate:
     'uu_zzz.r'
     'z_animint.R'
     'z_animintHelpers.R'
+    'z_multiline.R'
     'z_facets.R'
     'z_geoms.R'
     'z_helperFunctions.R'

NEWS.md

@@ -14,6 +14,11 @@
 
 - `update_axes`: Fixed issue #273 where axis tick text font-size was inconsistent between plots with and without `update_axes`. Previously, plots using `theme_animint(update_axes="x")` would lose `theme(axis.text = element_text(size=...))` styling after axis updates.
 
+# Changes in version 2025.11.7 (PR#261)
+
+- Fixed multiline text spacing: plot titles no longer overlap with plot area, and X/Y axis label spacing is now consistent.
+- Fixed axis titles to scale correctly with `theme(text=element_text(size=X))` (issue #64).
+- 
 # Changes in version 2025.12.4 (PR#272)
 
 - Download status table now shows `files`, `disk`, and `rows` columns in "downloaded / total" format. Row counts display with comma separators for readability. Disk sizes show KiB or MiB units (using binary 1024 divisor, consistent with `man du`). Chunk sizes are calculated in R using `file.size()` and exported via plot.json.
@@ -34,6 +39,10 @@
 
 - `geom_text(vjust!=0)` warning mentions vjust support in `geom_label_aligned()`.
 
+# Changes in version 2025.10.21 (PR#221)
+
+- Multi-line text support: `\n` now works in plot titles, axis titles, legend titles, and `geom_text()` labels. Created `R/z_multiline.R` with helper to convert `\n` to `<br/>` during R compilation. JavaScript renderer converts `<br/>` to SVG `<tspan>` elements for proper multi-line display.
+
 # Changes in version 2025.10.17 (PR#255)
 
 - `getCommonChunk()` uses default group=1 (previously 1:N which was slower).
@@ -263,7 +272,7 @@
 
 # Changes in 2022.5.25
 
-- Add ability to rotate geom_text labels, following ggplot2's semantics of rotation direction. 
+- Add ability to rotate geom_text labels, following ggplot2\'s semantics of rotation direction. 
 
 # Changes in 2022.5.24
 
@@ -328,4 +337,4 @@
 
 # Changes in 2017.08.24
 
-- DSL: clickSelects/showSelected are now specified as parameters rather than aesthetics.
+- DSL: clickSelects/showSelected are now specified as parameters rather than aesthetics.

R/z_animint.R

@@ -130,7 +130,7 @@ parsePlot <- function(meta, plot, plot.name){
   for (xy in c("x", "y")) {
     s <- function(tmp) sprintf(tmp, xy)
     # one axis name per plot (ie, a xtitle/ytitle is shared across panels)
-    plot.info[[s("%stitle")]] <- if(is.blank(s("axis.title.%s"))){
+    axis_title_raw <- if(is.blank(s("axis.title.%s"))){
       ""
     } else {
       scale.i <- which(plot$scales$find(xy))
@@ -143,6 +143,10 @@ parsePlot <- function(meta, plot, plot.name){
         lab.or.null
       }
     }
+    # Convert newlines to <br/> for multi-line axis titles (Issue #221)
+    plot.info[[s("%stitle")]] <- convertNewlinesToBreaks(axis_title_raw)
+    ## axis title size.
+    plot.info[[s("%stitle_size")]] <- getTextSize(s("axis.title.%s"), theme.pars)
     ## panel text size.
     plot.info[[s("strip_text_%ssize")]] <- getTextSize(
       s("strip.text.%s"), theme.pars)
@@ -187,12 +191,13 @@ parsePlot <- function(meta, plot, plot.name){
   # grab the unique axis labels (makes rendering simpler)
   plot.info <- getUniqueAxisLabels(plot.info)
 
-  # grab plot title if present
-  plot.info$title <- if(is(theme.pars$plot.title, "blank")){
+  # grab plot title if present and convert newlines (Issue #221)
+  plot_title_raw <- if(is(theme.pars$plot.title, "blank")){
     ""
   }else{
     plot$labels$title
   }
+  plot.info$title <- convertNewlinesToBreaks(plot_title_raw)
   plot.info$title_size <- getTextSize("plot.title", theme.pars)
 
   ## Set plot width and height from animint.* options if they are
@@ -839,6 +844,8 @@ getLegendList <- function(plistextra){
     }
   }
   legend.list <- lapply(gdefs, getLegend)
+  # Use the 'class' field from getLegend output for legend key names (Issue #221)
+  # This ensures JSON keys don't contain newlines or other special characters
   names(legend.list) <- sapply(legend.list, function(i) i$class)
   ## Add a flag to specify whether or not there is both a color and a
   ## fill legend to display. If so, we need to draw the interior of

R/z_animintHelpers.R

@@ -702,6 +702,8 @@ getLegend <- function(mb){
     names(data) <- paste0(geom, names(data))# aesthetics by geom
     names(data) <- gsub(paste0(geom, "."), "", names(data), fixed=TRUE) # label isn't geom-specific
     data$label <- paste(data$label) # otherwise it is AsIs.
+    # Convert newlines to <br/> for multi-line legend labels (Issue #221)
+    data$label <- convertNewlinesToBreaks(data$label)
     data
   }
   dataframes <- mapply(function(i, j) cleanData(i$data, mb$key, j, i$params),
@@ -726,10 +728,16 @@ getLegend <- function(mb){
   if(guidetype=="none"){
     NULL
   }else{
+    # Convert newlines to <br/> for multi-line legend title (Issue #221)
+    legend_title <- convertNewlinesToBreaks(mb$title)
+    # For the 'class' field (used as JSON key), sanitize newlines to spaces
+    # to avoid JSON parsing issues with control characters (Issue #221)
+    safe_title <- gsub("\n", " ", mb$title, fixed = TRUE)
+    legend_class <- if(mb$is.discrete) mb$selector else safe_title
     list(guide = guidetype,
          geoms = unlist(mb$geom.legend.list),
-         title = mb$title,
-         class = if(mb$is.discrete)mb$selector else mb$title,
+         title = legend_title,
+         class = legend_class,
          selector = mb$selector,
          is_discrete= mb$is.discrete,
          legend_type = mb$legend_type,
@@ -913,6 +921,10 @@ split_recursive <- function(x, vars){
 ##' @author Toby Dylan Hocking
 saveChunks <- function(x, meta){
   if(is.data.frame(x)){
+    # Convert newlines to <br/> in label column for multi-line text (Issue #221)
+    if("label" %in% names(x)){
+      x$label <- convertNewlinesToBreaks(x$label)
+    }
     this.i <- meta$chunk.i
     csv.name <- sprintf("%s_chunk%d.tsv", meta$g$classed, this.i)
     ## Some geoms should be split into separate groups if there are NAs.

R/z_multiline.R

@@ -0,0 +1,4 @@
+convertNewlinesToBreaks <- function(text) {
+  gsub("\n", "<br/>", text, fixed = TRUE)
+}
+