library(ggplot2)
library(dplyr)
library(readr)
library(stringr)

# Usage: Rscript combined_fault_rates.r exp_abspath1 exp_abspath2 ... [faults_file]

args <- commandArgs(trailingOnly = TRUE)
if (length(args) < 1) {
  stop("Need at least 1 experiment")
}

csv_suffix <- if (grepl("\\.csv$", args[length(args)])) {
  args[length(args)]
} else {
  "faults.csv"
}
exp_args <- if (grepl("\\.csv$", args[length(args)])) {
  args[-length(args)]
} else {
  args
}

# Faults / instruction count, per experiment
rates <- data.frame()

for (arg in exp_args) {
  faults_file <- file.path(arg, csv_suffix)
  mnem_file <- file.path(arg, "mnemonics.csv")

  if (!file.exists(faults_file)) {
    warning(paste("Missing:", faults_file))
    next
  }
  if (!file.exists(mnem_file)) {
    warning(paste("Missing:", mnem_file))
    next
  }

  df <- read_csv(faults_file, col_types = cols())
  mdf <- read_csv(mnem_file, col_types = cols())

  total_faults <- df |>
    filter(resulttype != "OK_MARKER") |>
    summarise(faults = sum(faults, na.rm = TRUE)) |>
    pull(faults)

  total_instrs <- sum(mdf$count, na.rm = TRUE)

  if (is.na(total_instrs) || total_instrs == 0) {
    warning(paste("Zero instruction count for", arg))
    next
  }

  match <- str_match(
    basename(arg),
    "^\\d{2}-\\d{2}_\\d{2}-\\d{2}-\\d{2}_(.+?)_(c|aot|interp)_"
  )
  if (is.na(match[1, 1])) {
    base_name <- basename(arg)
    variant <- "unknown"
  } else {
    base_name <- match[1, 2]
    variant <- match[1, 3]
  }

  rates <- bind_rows(
    rates,
    data.frame(
      base_name = base_name,
      variant = variant,
      fault_rate = total_faults / total_instrs
    )
  )
}

if (nrow(rates) == 0) {
  stop("No data loaded")
}

# Order base_names by their max fault rate
base_order <- rates |>
  group_by(base_name) |>
  summarise(max_rate = max(fault_rate), .groups = "drop") |>
  arrange(desc(max_rate)) |>
  pull(base_name)

rates <- rates |>
  mutate(
    base_name = factor(base_name, levels = base_order),
    variant = factor(variant, levels = c("c", "aot", "interp", "unknown"))
  )

plot <- ggplot(
  rates,
  aes(x = base_name, y = fault_rate, fill = variant)
) +
  geom_col(position = position_dodge(preserve = "single")) +
  scale_y_log10() +
  labs(
    title = "Fault Rate per Instruction",
    x = "Experiment",
    y = "Faults / Instruction Count",
    fill = "Variant"
  ) +
  theme_minimal() +
  theme(
    axis.text.x = element_text(angle = 90, hjust = 1),
    plot.title = element_text(size = 14, face = "bold")
  )

suffix <- gsub("^faults|\\.csv$", "", csv_suffix)
outfile <- paste0("injections/fault_rates_per_instruction", suffix, ".svg")
ggsave(outfile, plot = plot, width = 12, height = 6)
print(paste("Saved", outfile))