Mismatch between source and test files in tibble

# For paged output
options(max.print = 1000)

library(tidyverse)

── Attaching packages ──────────────────────── tidyverse 1.3.0.9000 ──

✓ ggplot2 3.3.0            ✓ purrr   0.3.3.9000  
✓ tibble  2.99.99.9014     ✓ dplyr   0.8.5       
✓ tidyr   1.0.2            ✓ stringr 1.4.0       
✓ readr   1.3.1.9000       ✓ forcats 0.5.0       

── Conflicts ──────────────────────────────── tidyverse_conflicts() ──
x dplyr::filter() masks stats::filter()
x dplyr::lag()    masks stats::lag()

# for as.data.frame.coverage
requireNamespace("covr", quietly = TRUE)

# Created by covr-per-file.R in
# https://gist.github.com/krlmlr/30a3c07d5114cde39236b2c08d69c873
test_results <- readRDS("covr-results.rds")

covr_enframe <- function(x) {
  x$result %>%
    as.data.frame() %>%
    as_tibble() %>%
    transmute(
      source_file = filename,
      source_pos = paste0(
        source_file, ":",
        first_line, ":", first_column, "-",
        last_line, ":", last_column,
        if_else(is.na(functions), paste0("-", functions), "")
      ),
      value
    )
}

Check errors —-

test_results %>%
  filter(map_lgl(covr, ~ !is.null(.x$error))) %>%
  nrow()

[1] 0

Collect coverage per expression per test file —-

long <-
  test_results %>%
  mutate(covr_frames = map(covr, covr_enframe)) %>%
  select(-code, -covr) %>%
  rename(test_file = file) %>%
  unnest(covr_frames) %>%
  select(source_file, everything())

# Munge file names, don't care about native code for now
long_clean <-
  long %>%
  filter(grepl("^R/", source_file)) %>%
  mutate(source_file = gsub("^R/", "", source_file)) %>%
  mutate(test_file = gsub("^test-", "", test_file))

covr_per_test_file <-
  long_clean %>%
  group_by(source_file, test_file) %>%
  summarize(coverage = mean(value != 0)) %>%
  ungroup() %>%
  arrange(source_file, -coverage) %>%
  group_by(source_file) %>%
  mutate(rank = row_number()) %>%
  ungroup()

best <-
  covr_per_test_file %>%
  filter(rank == 1) %>%
  select(-rank)

All tests where the best coverage is obtained from a test file that’s named differently from the source file.

best %>%
  filter(test_file != source_file) %>%
  filter(!grepl("^utils-", source_file)) %>%
  arrange(-coverage)

Which source files are covered by tests in other files? —-

long_clean_rank <-
  long_clean %>%
  left_join(covr_per_test_file %>% select(source_file, test_file, rank))

Joining, by = c("source_file", "test_file")

best_hit <-
  long_clean_rank %>%
  filter(value > 0, source_file == test_file) %>%
  select(-value, -rank, -test_file)

other_hit <-
  long_clean_rank %>%
  filter(value > 0, source_file != test_file)

List locations that were missed by the same-file test but hit by the other test

anti_join(other_hit, best_hit, by = c("source_file", "source_pos")) %>%
  group_by(source_pos) %>%
  summarize(test_file = paste0(head(test_file, 3), collapse = ", ")) %>%
  ungroup()