---
title: "Logging to File"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{Logging to File}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  eval = FALSE  # Set to TRUE when running interactively
)
```

### Passing arguments to `callr::r_bg()`

`{bakerrr}` uses `callr::r_bg()` under the hood. All `r_bg()` arguments can be passed via the `bg_args` argument in `bakerrr()`

#### Setup a sample function and its args_list

```{r}
# Example: Data processing pipeline
process_data <- function(dataset_id, filter_col, threshold) {
  # Input validation
  if (is.null(dataset_id) || !is.character(dataset_id)) {
    stop("Invalid or missing dataset_id: must be character and non-null.")
  }
  if (!filter_col %in% c("score", "value", "rating")) {
    stop("filter_col must be one of 'score', 'value', or 'rating'.")
  }
  if (!is.numeric(threshold) || threshold <= 0) {
    stop("threshold must be positive numeric.")
  }

  # Simulate loading data
  set.seed(as.integer(sub("DS_", "", dataset_id))) # Reproducible per job
  n <- sample(100:500, 1)
  data <- tibble::tibble(
    id = seq_len(n),
    score = round(runif(n, 0, 100)),
    value = round(rnorm(n, 50, 15)),
    rating = sample(1:10, n, replace = TRUE)
  )

  # Apply filter
  filtered <- dplyr::filter(data, data[[filter_col]] >= threshold)

  Sys.sleep(runif(1, 0.5, 2))  # Simulate processing time

  list(
    dataset_id = dataset_id,
    total_records = n,
    filtered_records = nrow(filtered),
    filter_applied = paste(filter_col, ">=", threshold),
    timestamp = Sys.time(),
    preview = head(filtered, 3)
  )
}

# Generate robust argument sets
set.seed(42)
args_list <- purrr::map(1:12, ~ list(
  dataset_id = paste0("DS_", .x),
  filter_col = sample(c("score", "value", "rating"), 1),
  threshold = sample(50:95, 1)
))

# Add intentional error cases
args_list[[5]] <- list(
  dataset_id = NULL,         # error: dataset_id missing
  filter_col = "score",
  threshold = 75
)
args_list[[7]] <- list(      # error: invalid filter_col
  dataset_id = "DS_7",
  filter_col = "unknown_col",
  threshold = 80
)

args_list[[9]] <- list(      # error: invalid threshold
  dataset_id = "DS_9",
  filter_col = "score",
  threshold = -15
)

```

#### Write outputs and errors to file while running jobs

When initiating the bakerrr with `bakerrr::bakerrr()`, pass any `callr::r_bg()` arguments to `bg_args` argument as below:

```{r}
stirr_job <- bakerrr::bakerrr(
  fun = process_data,
  args_list = args_list,
  n_daemons = min(4, length(args_list)),

  # Note how parameters are passed to callr::r_bg()
  bg_args = list(
    stdout = "bakerrr_out.log",
    stderr = "bakerrr_err.log"
  )
)

# Execute with real-time monitoring
stirr_job <- stirr_job |>
  bakerrr::run_jobs(wait_for_results = TRUE)

# Get results
stirr_job@results
```