---
title: "Lint and Debug Workflow"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{Lint and Debug Workflow}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r, include = FALSE}
knitr::opts_chunk$set(collapse = TRUE, comment = "#>")
```

## User goal

Linting helps instructors detect common authoring problems before conversion. This is useful in teaching teams where several people maintain source lessons.

## Minimal example

```{r eval=FALSE}
library(tutorizeR)

lint <- lint_source("lesson.qmd", strict = FALSE)
print(lint)
```

## Strict conversion

```{r eval=FALSE}
library(tutorizeR)

report <- tutorize(
  input = "lesson.qmd",
  output_dir = tempdir(),
  assessment = "both",
  lint_strict = TRUE,
  overwrite = TRUE
)
```

## Realistic installed example

```{r eval=FALSE}
library(tutorizeR)

example_dir <- system.file("examples", "example_course_module", package = "tutorizeR")
source_file <- file.path(example_dir, "lesson-source.qmd")
question_bank <- load_question_bank(file.path(example_dir, "question-bank"))

lint <- lint_source(
  input = source_file,
  question_bank = question_bank,
  strict = FALSE
)

print(lint)
```

## Conversion reports

```{r eval=FALSE}
write_tutorize_report(
  report,
  file.path(tempdir(), "conversion-report.json"),
  format = "json"
)
```

Reports are useful in continuous integration and course release workflows because they record the generated output path, number of exercises, MCQs, warnings, and lint summary.

## Limits

Linting can detect structural issues, but it cannot judge whether a lesson is pedagogically effective. Generated reports should support instructor review, not replace it.

## Reproducibility checklist

```{r eval=FALSE}
stopifnot(inherits(lint, "tutorize_lint_report"))
```