Modify merge in 'update.commit.interactions'

The merge now always keeps all commit interactions, even if there is no commit data for them (for example if the commits were made by deleted users). Also introduced a warning if that happens. Case is tested with new part in 'test-data.R' Signed-off-by: Leo Sendelbach <[email protected]>
se-sic · Feb 27, 2024 · 8148614 · 8148614
1 parent 63f2737
commit 8148614
Show file tree

Hide file tree

Showing 2 changed files with 43 additions and 5 deletions.
diff --git a/tests/test-data.R b/tests/test-data.R
@@ -540,5 +540,36 @@ test_that("Compare two ProjectData Objects with commit.interactions", {
     commit.data[["hash"]][[5]] = 1
     proj.data.one$set.commits(commit.data)
 
-    expect_false(isTRUE(all.equal(proj.data.one$get.commit.interactions(), proj.data.two$get.commit.interactions())))
+    expect_false(isTRUE(all.equal(proj.data.one$get.commit.interactions(),
+                                  proj.data.two$get.commit.interactions())))
+
+    ## set commit list of one project data to empty and test that last
+    ## two rows of result data frame are empty
+    proj.data.two$set.commits(create.empty.commits.list())
+
+    ## create empty data frame of correct size
+    commit.interactions.data.expected = data.frame(matrix(nrow = 4, ncol = 8))
+    ## assure that the correct type is used
+    for(i in seq_len(8)) {
+        commit.interactions.data.expected[[i]] = as.character(commit.interactions.data.expected[[i]])
+    }
+    ## set everything except for authors as expected
+    colnames(commit.interactions.data.expected) = c("commit.hash", "base.hash", "func", "file",
+                                                    "base.func", "base.file", "base.author",
+                                                    "interacting.author")
+    commit.interactions.data.expected[["commit.hash"]] =
+                                                        c("0a1a5c523d835459c42f33e863623138555e2526",
+                                                        "418d1dc4929ad1df251d2aeb833dd45757b04a6f",
+                                                        "5a5ec9675e98187e1e92561e1888aa6f04faa338",
+                                                        "d01921773fae4bed8186b0aa411d6a2f7a6626e6")
+    commit.interactions.data.expected[["base.hash"]] =
+                                                      c("3a0ed78458b3976243db6829f63eba3eead26774",
+                                                        "0a1a5c523d835459c42f33e863623138555e2526",
+                                                        "72c8dd25d3dd6d18f46e2b26a5f5b1e2e8dc28d0",
+                                                        "0a1a5c523d835459c42f33e863623138555e2526")
+    commit.interactions.data.expected[["func"]] = c("test2.c", "test2.c", "test.c", "test2.c")
+    commit.interactions.data.expected[["file"]] = c("test2.c", "test2.c", "test.c", "test2.c")
+    commit.interactions.data.expected[["base.func"]] = c("test2.c", "test2.c", "test.c", "test2.c")
+    commit.interactions.data.expected[["base.file"]] = c("test2.c", "test2.c", "test.c", "test2.c")
+    expect_equal(proj.data.two$get.commit.interactions(), commit.interactions.data.expected)
 })
diff --git a/util-data.R b/util-data.R
@@ -420,11 +420,11 @@ ProjectData = R6::R6Class("ProjectData",
 
             ## remove existing columns named 'base.author' and 'interaction.author'
             indices.to.remove = which("base.author" == colnames(private$commit.interactions))
-            if (length(indices.to.remove)>0) {
+            if (length(indices.to.remove) > 0) {
                 private$commit.interactions = private$commit.interactions[, -indices.to.remove]
             }
             indices.to.remove = which("interacting.author" == colnames(private$commit.interactions))
-            if (length(indices.to.remove)>0) {
+            if (length(indices.to.remove) > 0) {
                 private$commit.interactions = private$commit.interactions[, -indices.to.remove]
             }
 
@@ -435,17 +435,24 @@ ProjectData = R6::R6Class("ProjectData",
 
             ## merge commit interactions with commits and change colnames to avoid duplicates
             commit.interaction.data = merge(private$commit.interactions, commit.data.subset,
-                                            by.x = "base.hash", by.y = "hash")
+                                            by.x = "base.hash", by.y = "hash", all.x = TRUE)
 
             author.index = match("author.name", colnames(commit.interaction.data))
             colnames(commit.interaction.data)[[author.index]] = "base.author"
 
             commit.interaction.data = merge(commit.interaction.data, commit.data.subset,
-                                            by.x = "commit.hash", by.y = "hash")
+                                            by.x = "commit.hash", by.y = "hash", all.x = TRUE)
 
             author.index = match("author.name", colnames(commit.interaction.data))
             colnames(commit.interaction.data)[[author.index]] = "interacting.author"
 
+            ## warning if we have interactions without authors
+            if (anyNA(commit.interaction.data[["base.author"]]) ||
+                anyNA(commit.interaction.data[["interacting.author"]])) {
+                logging::logwarn("There are authors in the commit-interactions that are not in the commit data!
+                                  This results in the commit-interactions having empty entries.
+                                  To clean up these entries, call cleanup.commit.interactions.")
+            }
             private$commit.interactions = commit.interaction.data
 
         },