deno_3p_lib 1.25.2

// Copyright 2018-2022 the Deno authors. All rights reserved. MIT license.

use crate::args::Flags;
use crate::args::TestFlags;
use crate::args::TypeCheckMode;
use crate::checksum;
use crate::colors;
use crate::create_main_worker;
use crate::display;
use crate::file_fetcher::File;
use crate::file_watcher;
use crate::file_watcher::ResolutionResult;
use crate::fs_util::collect_specifiers;
use crate::fs_util::is_supported_test_ext;
use crate::fs_util::is_supported_test_path;
use crate::fs_util::specifier_to_file_path;
use crate::graph_util::contains_specifier;
use crate::graph_util::graph_valid;
use crate::ops;
use crate::proc_state::ProcState;

use deno_ast::swc::common::comments::CommentKind;
use deno_ast::MediaType;
use deno_ast::SourceRangedForSpanned;
use deno_core::error::generic_error;
use deno_core::error::AnyError;
use deno_core::error::JsError;
use deno_core::futures::future;
use deno_core::futures::stream;
use deno_core::futures::FutureExt;
use deno_core::futures::StreamExt;
use deno_core::parking_lot::Mutex;
use deno_core::url::Url;
use deno_core::ModuleSpecifier;
use deno_graph::ModuleKind;
use deno_runtime::fmt_errors::format_js_error;
use deno_runtime::ops::io::Stdio;
use deno_runtime::ops::io::StdioPipe;
use deno_runtime::permissions::Permissions;
use deno_runtime::tokio_util::run_local;
use indexmap::IndexMap;
use log::Level;
use rand::rngs::SmallRng;
use rand::seq::SliceRandom;
use rand::SeedableRng;
use regex::Regex;
use serde::Deserialize;
use std::collections::BTreeMap;
use std::collections::HashSet;
use std::fmt::Write as _;
use std::io::Read;
use std::io::Write;
use std::num::NonZeroUsize;
use std::path::PathBuf;
use std::sync::Arc;
use std::time::Duration;
use std::time::Instant;
use tokio::sync::mpsc::unbounded_channel;
use tokio::sync::mpsc::UnboundedSender;

/// The test mode is used to determine how a specifier is to be tested.
#[derive(Debug, Clone, PartialEq)]
pub enum TestMode {
  /// Test as documentation, type-checking fenced code blocks.
  Documentation,
  /// Test as an executable module, loading the module into the isolate and running each test it
  /// defines.
  Executable,
  /// Test as both documentation and an executable module.
  Both,
}

#[derive(Clone, Debug, Default)]
pub struct TestFilter {
  pub substring: Option<String>,
  pub regex: Option<Regex>,
  pub include: Option<Vec<String>>,
  pub exclude: Vec<String>,
}

impl TestFilter {
  pub fn includes(&self, name: &String) -> bool {
    if let Some(substring) = &self.substring {
      if !name.contains(substring) {
        return false;
      }
    }
    if let Some(regex) = &self.regex {
      if !regex.is_match(name) {
        return false;
      }
    }
    if let Some(include) = &self.include {
      if !include.contains(name) {
        return false;
      }
    }
    if self.exclude.contains(name) {
      return false;
    }
    true
  }

  pub fn from_flag(flag: &Option<String>) -> Self {
    let mut substring = None;
    let mut regex = None;
    if let Some(flag) = flag {
      if flag.starts_with('/') && flag.ends_with('/') {
        let rs = flag.trim_start_matches('/').trim_end_matches('/');
        regex =
          Some(Regex::new(rs).unwrap_or_else(|_| Regex::new("$^").unwrap()));
      } else {
        substring = Some(flag.clone());
      }
    }
    Self {
      substring,
      regex,
      ..Default::default()
    }
  }
}

#[derive(Debug, Clone, PartialEq, Deserialize, Eq, Hash)]
#[serde(rename_all = "camelCase")]
pub struct TestLocation {
  pub file_name: String,
  pub line_number: u32,
  pub column_number: u32,
}

#[derive(Debug, Clone, PartialEq, Deserialize, Eq, Hash)]
#[serde(rename_all = "camelCase")]
pub struct TestDescription {
  pub id: usize,
  pub name: String,
  pub origin: String,
  pub location: TestLocation,
}

impl TestDescription {
  pub fn static_id(&self) -> String {
    checksum::gen(&[self.location.file_name.as_bytes(), self.name.as_bytes()])
  }
}

#[derive(Debug, Clone, PartialEq, Deserialize)]
#[serde(rename_all = "camelCase")]
pub enum TestOutput {
  String(String),
  Bytes(Vec<u8>),
}

#[derive(Debug, Clone, PartialEq, Deserialize)]
#[serde(rename_all = "camelCase")]
pub enum TestResult {
  Ok,
  Ignored,
  Failed(Box<JsError>),
  Cancelled,
}

#[derive(Debug, Clone, PartialEq, Deserialize)]
#[serde(rename_all = "camelCase")]
pub struct TestStepDescription {
  pub id: usize,
  pub name: String,
  pub origin: String,
  pub location: TestLocation,
  pub level: usize,
  pub parent_id: usize,
  pub root_id: usize,
  pub root_name: String,
}

impl TestStepDescription {
  pub fn static_id(&self) -> String {
    checksum::gen(&[
      self.location.file_name.as_bytes(),
      &self.level.to_be_bytes(),
      self.name.as_bytes(),
    ])
  }
}

#[derive(Debug, Clone, PartialEq, Deserialize)]
#[serde(rename_all = "camelCase")]
pub enum TestStepResult {
  Ok,
  Ignored,
  Failed(Option<Box<JsError>>),
  Pending(Option<Box<JsError>>),
}

impl TestStepResult {
  fn error(&self) -> Option<&JsError> {
    match self {
      TestStepResult::Failed(Some(error)) => Some(error),
      TestStepResult::Pending(Some(error)) => Some(error),
      _ => None,
    }
  }
}

#[derive(Debug, Clone, PartialEq, Deserialize)]
#[serde(rename_all = "camelCase")]
pub struct TestPlan {
  pub origin: String,
  pub total: usize,
  pub filtered_out: usize,
  pub used_only: bool,
}

#[derive(Debug, Clone, Deserialize)]
#[serde(rename_all = "camelCase")]
pub enum TestEvent {
  Register(TestDescription),
  Plan(TestPlan),
  Wait(usize),
  Output(Vec<u8>),
  Result(usize, TestResult, u64),
  UncaughtError(String, Box<JsError>),
  StepRegister(TestStepDescription),
  StepWait(usize),
  StepResult(usize, TestStepResult, u64),
}

#[derive(Debug, Clone, Deserialize)]
pub struct TestSummary {
  pub total: usize,
  pub passed: usize,
  pub failed: usize,
  pub ignored: usize,
  pub passed_steps: usize,
  pub failed_steps: usize,
  pub pending_steps: usize,
  pub ignored_steps: usize,
  pub filtered_out: usize,
  pub measured: usize,
  pub failures: Vec<(TestDescription, Box<JsError>)>,
  pub uncaught_errors: Vec<(String, Box<JsError>)>,
}

#[derive(Debug, Clone)]
struct TestSpecifierOptions {
  concurrent_jobs: NonZeroUsize,
  fail_fast: Option<NonZeroUsize>,
  filter: TestFilter,
}

impl TestSummary {
  pub fn new() -> TestSummary {
    TestSummary {
      total: 0,
      passed: 0,
      failed: 0,
      ignored: 0,
      passed_steps: 0,
      failed_steps: 0,
      pending_steps: 0,
      ignored_steps: 0,
      filtered_out: 0,
      measured: 0,
      failures: Vec::new(),
      uncaught_errors: Vec::new(),
    }
  }

  fn has_failed(&self) -> bool {
    self.failed > 0 || !self.failures.is_empty()
  }
}

struct PrettyTestReporter {
  parallel: bool,
  echo_output: bool,
  in_new_line: bool,
  last_wait_id: Option<usize>,
  cwd: Url,
  did_have_user_output: bool,
  started_tests: bool,
}

impl PrettyTestReporter {
  fn new(parallel: bool, echo_output: bool) -> PrettyTestReporter {
    PrettyTestReporter {
      parallel,
      echo_output,
      in_new_line: true,
      last_wait_id: None,
      cwd: Url::from_directory_path(std::env::current_dir().unwrap()).unwrap(),
      did_have_user_output: false,
      started_tests: false,
    }
  }

  fn force_report_wait(&mut self, description: &TestDescription) {
    if !self.in_new_line {
      println!();
    }
    if self.parallel {
      print!(
        "{}",
        colors::gray(format!(
          "{} => ",
          self.to_relative_path_or_remote_url(&description.origin)
        ))
      );
    }
    print!("{} ...", description.name);
    self.in_new_line = false;
    // flush for faster feedback when line buffered
    std::io::stdout().flush().unwrap();
    self.last_wait_id = Some(description.id);
  }

  fn to_relative_path_or_remote_url(&self, path_or_url: &str) -> String {
    let url = Url::parse(path_or_url).unwrap();
    if url.scheme() == "file" {
      if let Some(mut r) = self.cwd.make_relative(&url) {
        if !r.starts_with("../") {
          r = format!("./{}", r);
        }
        return r;
      }
    }
    path_or_url.to_string()
  }

  fn force_report_step_wait(&mut self, description: &TestStepDescription) {
    self.write_output_end();
    if !self.in_new_line {
      println!();
    }
    print!("{}{} ...", "  ".repeat(description.level), description.name);
    self.in_new_line = false;
    // flush for faster feedback when line buffered
    std::io::stdout().flush().unwrap();
    self.last_wait_id = Some(description.id);
  }

  fn force_report_step_result(
    &mut self,
    description: &TestStepDescription,
    result: &TestStepResult,
    elapsed: u64,
  ) {
    let status = match result {
      TestStepResult::Ok => colors::green("ok").to_string(),
      TestStepResult::Ignored => colors::yellow("ignored").to_string(),
      TestStepResult::Pending(_) => colors::gray("pending").to_string(),
      TestStepResult::Failed(_) => colors::red("FAILED").to_string(),
    };

    self.write_output_end();
    if self.in_new_line || self.last_wait_id != Some(description.id) {
      self.force_report_step_wait(description);
    }

    println!(
      " {} {}",
      status,
      colors::gray(format!("({})", display::human_elapsed(elapsed.into())))
    );

    if let Some(js_error) = result.error() {
      let err_string = format_test_error(js_error);
      let err_string = format!("{}: {}", colors::red_bold("error"), err_string);
      for line in err_string.lines() {
        println!("{}{}", "  ".repeat(description.level + 1), line);
      }
    }
    self.in_new_line = true;
  }

  fn write_output_end(&mut self) {
    if self.did_have_user_output {
      println!("{}", colors::gray("----- output end -----"));
      self.in_new_line = true;
      self.did_have_user_output = false;
    }
  }

  fn report_register(&mut self, _description: &TestDescription) {}

  fn report_plan(&mut self, plan: &TestPlan) {
    if self.parallel {
      return;
    }
    let inflection = if plan.total == 1 { "test" } else { "tests" };
    println!(
      "{}",
      colors::gray(format!(
        "running {} {} from {}",
        plan.total,
        inflection,
        self.to_relative_path_or_remote_url(&plan.origin)
      ))
    );
    self.in_new_line = true;
  }

  fn report_wait(&mut self, description: &TestDescription) {
    if !self.parallel {
      self.force_report_wait(description);
    }
    self.started_tests = true;
  }

  fn report_output(&mut self, output: &[u8]) {
    if !self.echo_output {
      return;
    }

    if !self.did_have_user_output && self.started_tests {
      self.did_have_user_output = true;
      if !self.in_new_line {
        println!();
      }
      println!("{}", colors::gray("------- output -------"));
      self.in_new_line = true;
    }

    // output everything to stdout in order to prevent
    // stdout and stderr racing
    std::io::stdout().write_all(output).unwrap();
  }

  fn report_result(
    &mut self,
    description: &TestDescription,
    result: &TestResult,
    elapsed: u64,
  ) {
    if self.parallel {
      self.force_report_wait(description);
    }

    self.write_output_end();
    if self.in_new_line || self.last_wait_id != Some(description.id) {
      self.force_report_wait(description);
    }

    let status = match result {
      TestResult::Ok => colors::green("ok").to_string(),
      TestResult::Ignored => colors::yellow("ignored").to_string(),
      TestResult::Failed(_) => colors::red("FAILED").to_string(),
      TestResult::Cancelled => colors::gray("cancelled").to_string(),
    };

    println!(
      " {} {}",
      status,
      colors::gray(format!("({})", display::human_elapsed(elapsed.into())))
    );
    self.in_new_line = true;
  }

  fn report_uncaught_error(&mut self, origin: &str, _error: &JsError) {
    if !self.in_new_line {
      println!();
    }
    println!(
      "Uncaught error from {} {}",
      self.to_relative_path_or_remote_url(origin),
      colors::red("FAILED")
    );
    self.in_new_line = true;
    self.did_have_user_output = false;
  }

  fn report_step_register(&mut self, _description: &TestStepDescription) {}

  fn report_step_wait(&mut self, description: &TestStepDescription) {
    if !self.parallel {
      self.force_report_step_wait(description);
    }
  }

  fn report_step_result(
    &mut self,
    description: &TestStepDescription,
    result: &TestStepResult,
    elapsed: u64,
    tests: &IndexMap<usize, TestDescription>,
    test_steps: &IndexMap<usize, TestStepDescription>,
  ) {
    if self.parallel {
      self.write_output_end();
      let root;
      let mut ancestor_names = vec![];
      let mut current_desc = description;
      loop {
        if let Some(step_desc) = test_steps.get(&current_desc.parent_id) {
          ancestor_names.push(&step_desc.name);
          current_desc = step_desc;
        } else {
          root = tests.get(&current_desc.parent_id).unwrap();
          break;
        }
      }
      ancestor_names.reverse();
      print!(
        "{}",
        colors::gray(format!(
          "{} =>",
          self.to_relative_path_or_remote_url(&description.origin)
        ))
      );
      print!(" {} ...", root.name);
      for name in ancestor_names {
        print!(" {} ...", name);
      }
      print!(" {} ...", description.name);
      self.in_new_line = false;
      self.last_wait_id = Some(description.id);
    }
    self.force_report_step_result(description, result, elapsed);
  }

  fn report_summary(&mut self, summary: &TestSummary, elapsed: &Duration) {
    if !summary.failures.is_empty() || !summary.uncaught_errors.is_empty() {
      #[allow(clippy::type_complexity)] // Type alias doesn't look better here
      let mut failures_by_origin: BTreeMap<
        String,
        (Vec<(&TestDescription, &JsError)>, Option<&JsError>),
      > = BTreeMap::default();
      let mut failure_titles = vec![];
      for (description, js_error) in &summary.failures {
        let (failures, _) = failures_by_origin
          .entry(description.origin.clone())
          .or_default();
        failures.push((description, js_error.as_ref()));
      }
      for (origin, js_error) in &summary.uncaught_errors {
        let (_, uncaught_error) =
          failures_by_origin.entry(origin.clone()).or_default();
        let _ = uncaught_error.insert(js_error.as_ref());
      }
      println!("\n{}\n", colors::white_bold_on_red(" ERRORS "));
      for (origin, (failures, uncaught_error)) in failures_by_origin {
        for (description, js_error) in failures {
          let failure_title = format!(
            "{} {}",
            &description.name,
            colors::gray(format!(
              "=> {}:{}:{}",
              self.to_relative_path_or_remote_url(
                &description.location.file_name
              ),
              description.location.line_number,
              description.location.column_number
            ))
          );
          println!("{}", &failure_title);
          println!(
            "{}: {}",
            colors::red_bold("error"),
            format_test_error(js_error)
          );
          println!();
          failure_titles.push(failure_title);
        }
        if let Some(js_error) = uncaught_error {
          let failure_title = format!(
            "{} (uncaught error)",
            self.to_relative_path_or_remote_url(&origin)
          );
          println!("{}", &failure_title);
          println!(
            "{}: {}",
            colors::red_bold("error"),
            format_test_error(js_error)
          );
          println!("This error was not caught from a test and caused the test runner to fail on the referenced module.");
          println!("It most likely originated from a dangling promise, event/timeout handler or top-level code.");
          println!();
          failure_titles.push(failure_title);
        }
      }
      println!("{}\n", colors::white_bold_on_red(" FAILURES "));
      for failure_title in failure_titles {
        println!("{}", failure_title);
      }
    }

    let status = if summary.has_failed() {
      colors::red("FAILED").to_string()
    } else {
      colors::green("ok").to_string()
    };

    let get_steps_text = |count: usize| -> String {
      if count == 0 {
        String::new()
      } else if count == 1 {
        " (1 step)".to_string()
      } else {
        format!(" ({} steps)", count)
      }
    };

    let mut summary_result = String::new();

    write!(
      summary_result,
      "{} passed{} | {} failed{}",
      summary.passed,
      get_steps_text(summary.passed_steps),
      summary.failed,
      get_steps_text(summary.failed_steps + summary.pending_steps),
    )
    .unwrap();

    let ignored_steps = get_steps_text(summary.ignored_steps);
    if summary.ignored > 0 || !ignored_steps.is_empty() {
      write!(
        summary_result,
        " | {} ignored{}",
        summary.ignored, ignored_steps
      )
      .unwrap()
    }

    if summary.measured > 0 {
      write!(summary_result, " | {} measured", summary.measured,).unwrap();
    }

    if summary.filtered_out > 0 {
      write!(summary_result, " | {} filtered out", summary.filtered_out)
        .unwrap()
    };

    println!(
      "\n{} | {} {}\n",
      status,
      summary_result,
      colors::gray(format!(
        "({})",
        display::human_elapsed(elapsed.as_millis())
      )),
    );
    self.in_new_line = true;
  }
}

fn abbreviate_test_error(js_error: &JsError) -> JsError {
  let mut js_error = js_error.clone();
  let frames = std::mem::take(&mut js_error.frames);

  // check if there are any stack frames coming from user code
  let should_filter = frames.iter().any(|f| {
    if let Some(file_name) = &f.file_name {
      !(file_name.starts_with("[deno:") || file_name.starts_with("deno:"))
    } else {
      true
    }
  });

  if should_filter {
    let mut frames = frames
      .into_iter()
      .rev()
      .skip_while(|f| {
        if let Some(file_name) = &f.file_name {
          file_name.starts_with("[deno:") || file_name.starts_with("deno:")
        } else {
          false
        }
      })
      .into_iter()
      .collect::<Vec<_>>();
    frames.reverse();
    js_error.frames = frames;
  } else {
    js_error.frames = frames;
  }

  js_error.cause = js_error
    .cause
    .as_ref()
    .map(|e| Box::new(abbreviate_test_error(e)));
  js_error.aggregated = js_error
    .aggregated
    .as_ref()
    .map(|es| es.iter().map(abbreviate_test_error).collect());
  js_error
}

// This function prettifies `JsError` and applies some changes specifically for
// test runner purposes:
//
// - filter out stack frames:
//   - if stack trace consists of mixed user and internal code, the frames
//     below the first user code frame are filtered out
//   - if stack trace consists only of internal code it is preserved as is
pub fn format_test_error(js_error: &JsError) -> String {
  let mut js_error = abbreviate_test_error(js_error);
  js_error.exception_message = js_error
    .exception_message
    .trim_start_matches("Uncaught ")
    .to_string();
  format_js_error(&js_error)
}

/// Test a single specifier as documentation containing test programs, an executable test module or
/// both.
async fn test_specifier(
  ps: ProcState,
  permissions: Permissions,
  specifier: ModuleSpecifier,
  mode: TestMode,
  sender: &TestEventSender,
  options: TestSpecifierOptions,
) -> Result<(), AnyError> {
  let mut worker = create_main_worker(
    &ps,
    specifier.clone(),
    permissions,
    vec![ops::testing::init(sender.clone(), options.filter.clone())],
    Stdio {
      stdin: StdioPipe::Inherit,
      stdout: StdioPipe::File(sender.stdout()),
      stderr: StdioPipe::File(sender.stderr()),
    },
  )
  .await?;

  worker.run_test_specifier(mode).await
}

fn extract_files_from_regex_blocks(
  specifier: &ModuleSpecifier,
  source: &str,
  media_type: MediaType,
  file_line_index: usize,
  blocks_regex: &Regex,
  lines_regex: &Regex,
) -> Result<Vec<File>, AnyError> {
  let files = blocks_regex
    .captures_iter(source)
    .filter_map(|block| {
      if block.get(1) == None {
        return None;
      }

      let maybe_attributes: Option<Vec<_>> = block
        .get(1)
        .map(|attributes| attributes.as_str().split(' ').collect());

      let file_media_type = if let Some(attributes) = maybe_attributes {
        if attributes.contains(&"ignore") {
          return None;
        }

        match attributes.first() {
          Some(&"js") => MediaType::JavaScript,
          Some(&"javascript") => MediaType::JavaScript,
          Some(&"mjs") => MediaType::Mjs,
          Some(&"cjs") => MediaType::Cjs,
          Some(&"jsx") => MediaType::Jsx,
          Some(&"ts") => MediaType::TypeScript,
          Some(&"typescript") => MediaType::TypeScript,
          Some(&"mts") => MediaType::Mts,
          Some(&"cts") => MediaType::Cts,
          Some(&"tsx") => MediaType::Tsx,
          Some(&"") => media_type,
          _ => MediaType::Unknown,
        }
      } else {
        media_type
      };

      if file_media_type == MediaType::Unknown {
        return None;
      }

      let line_offset = source[0..block.get(0).unwrap().start()]
        .chars()
        .filter(|c| *c == '\n')
        .count();

      let line_count = block.get(0).unwrap().as_str().split('\n').count();

      let body = block.get(2).unwrap();
      let text = body.as_str();

      // TODO(caspervonb) generate an inline source map
      let mut file_source = String::new();
      for line in lines_regex.captures_iter(text) {
        let text = line.get(1).unwrap();
        writeln!(file_source, "{}", text.as_str()).unwrap();
      }

      let file_specifier = deno_core::resolve_url_or_path(&format!(
        "{}${}-{}{}",
        specifier,
        file_line_index + line_offset + 1,
        file_line_index + line_offset + line_count + 1,
        file_media_type.as_ts_extension(),
      ))
      .unwrap();

      Some(File {
        local: file_specifier.to_file_path().unwrap(),
        maybe_types: None,
        media_type: file_media_type,
        source: file_source.into(),
        specifier: file_specifier,
        maybe_headers: None,
      })
    })
    .collect();

  Ok(files)
}

fn extract_files_from_source_comments(
  specifier: &ModuleSpecifier,
  source: Arc<str>,
  media_type: MediaType,
) -> Result<Vec<File>, AnyError> {
  let parsed_source = deno_ast::parse_module(deno_ast::ParseParams {
    specifier: specifier.as_str().to_string(),
    text_info: deno_ast::SourceTextInfo::new(source),
    media_type,
    capture_tokens: false,
    maybe_syntax: None,
    scope_analysis: false,
  })?;
  let comments = parsed_source.comments().get_vec();
  let blocks_regex = Regex::new(r"```([^\r\n]*)\r?\n([\S\s]*?)```")?;
  let lines_regex = Regex::new(r"(?:\* ?)(?:\# ?)?(.*)")?;

  let files = comments
    .iter()
    .filter(|comment| {
      if comment.kind != CommentKind::Block || !comment.text.starts_with('*') {
        return false;
      }

      true
    })
    .flat_map(|comment| {
      extract_files_from_regex_blocks(
        specifier,
        &comment.text,
        media_type,
        parsed_source.text_info().line_index(comment.start()),
        &blocks_regex,
        &lines_regex,
      )
    })
    .flatten()
    .collect();

  Ok(files)
}

fn extract_files_from_fenced_blocks(
  specifier: &ModuleSpecifier,
  source: &str,
  media_type: MediaType,
) -> Result<Vec<File>, AnyError> {
  // The pattern matches code blocks as well as anything in HTML comment syntax,
  // but it stores the latter without any capturing groups. This way, a simple
  // check can be done to see if a block is inside a comment (and skip typechecking)
  // or not by checking for the presence of capturing groups in the matches.
  let blocks_regex =
    Regex::new(r"(?s)<!--.*?-->|```([^\r\n]*)\r?\n([\S\s]*?)```")?;
  let lines_regex = Regex::new(r"(?:\# ?)?(.*)")?;

  extract_files_from_regex_blocks(
    specifier,
    source,
    media_type,
    /* file line index */ 0,
    &blocks_regex,
    &lines_regex,
  )
}

async fn fetch_inline_files(
  ps: ProcState,
  specifiers: Vec<ModuleSpecifier>,
) -> Result<Vec<File>, AnyError> {
  let mut files = Vec::new();
  for specifier in specifiers {
    let mut fetch_permissions = Permissions::allow_all();
    let file = ps
      .file_fetcher
      .fetch(&specifier, &mut fetch_permissions)
      .await?;

    let inline_files = if file.media_type == MediaType::Unknown {
      extract_files_from_fenced_blocks(
        &file.specifier,
        &file.source,
        file.media_type,
      )
    } else {
      extract_files_from_source_comments(
        &file.specifier,
        file.source.clone(),
        file.media_type,
      )
    };

    files.extend(inline_files?);
  }

  Ok(files)
}

/// Type check a collection of module and document specifiers.
pub async fn check_specifiers(
  ps: &ProcState,
  permissions: Permissions,
  specifiers: Vec<(ModuleSpecifier, TestMode)>,
) -> Result<(), AnyError> {
  let lib = ps.options.ts_type_lib_window();
  let inline_files = fetch_inline_files(
    ps.clone(),
    specifiers
      .iter()
      .filter_map(|(specifier, mode)| {
        if *mode != TestMode::Executable {
          Some(specifier.clone())
        } else {
          None
        }
      })
      .collect(),
  )
  .await?;

  if !inline_files.is_empty() {
    let specifiers = inline_files
      .iter()
      .map(|file| file.specifier.clone())
      .collect();

    for file in inline_files {
      ps.file_fetcher.insert_cached(file);
    }

    ps.prepare_module_load(
      specifiers,
      false,
      lib,
      Permissions::allow_all(),
      permissions.clone(),
      false,
    )
    .await?;
  }

  let module_specifiers = specifiers
    .iter()
    .filter_map(|(specifier, mode)| {
      if *mode != TestMode::Documentation {
        Some(specifier.clone())
      } else {
        None
      }
    })
    .collect();

  ps.prepare_module_load(
    module_specifiers,
    false,
    lib,
    Permissions::allow_all(),
    permissions,
    true,
  )
  .await?;

  Ok(())
}

/// Test a collection of specifiers with test modes concurrently.
async fn test_specifiers(
  ps: ProcState,
  permissions: Permissions,
  specifiers_with_mode: Vec<(ModuleSpecifier, TestMode)>,
  options: TestSpecifierOptions,
) -> Result<(), AnyError> {
  let log_level = ps.options.log_level();
  let specifiers_with_mode = if let Some(seed) = ps.options.shuffle_tests() {
    let mut rng = SmallRng::seed_from_u64(seed);
    let mut specifiers_with_mode = specifiers_with_mode.clone();
    specifiers_with_mode.sort_by_key(|(specifier, _)| specifier.clone());
    specifiers_with_mode.shuffle(&mut rng);
    specifiers_with_mode
  } else {
    specifiers_with_mode
  };

  let (sender, mut receiver) = unbounded_channel::<TestEvent>();
  let sender = TestEventSender::new(sender);
  let concurrent_jobs = options.concurrent_jobs;
  let fail_fast = options.fail_fast;

  let join_handles =
    specifiers_with_mode.iter().map(move |(specifier, mode)| {
      let ps = ps.clone();
      let permissions = permissions.clone();
      let specifier = specifier.clone();
      let mode = mode.clone();
      let mut sender = sender.clone();
      let options = options.clone();

      tokio::task::spawn_blocking(move || {
        let origin = specifier.to_string();
        let file_result = run_local(test_specifier(
          ps,
          permissions,
          specifier,
          mode,
          &sender,
          options,
        ));
        if let Err(error) = file_result {
          if error.is::<JsError>() {
            sender.send(TestEvent::UncaughtError(
              origin,
              Box::new(error.downcast::<JsError>().unwrap()),
            ))?;
          } else {
            return Err(error);
          }
        }
        Ok(())
      })
    });

  let join_stream = stream::iter(join_handles)
    .buffer_unordered(concurrent_jobs.get())
    .collect::<Vec<Result<Result<(), AnyError>, tokio::task::JoinError>>>();

  let mut reporter = Box::new(PrettyTestReporter::new(
    concurrent_jobs.get() > 1,
    log_level != Some(Level::Error),
  ));

  let handler = {
    tokio::task::spawn(async move {
      let earlier = Instant::now();
      let mut tests = IndexMap::new();
      let mut test_steps = IndexMap::new();
      let mut tests_with_result = HashSet::new();
      let mut summary = TestSummary::new();
      let mut used_only = false;

      while let Some(event) = receiver.recv().await {
        match event {
          TestEvent::Register(description) => {
            reporter.report_register(&description);
            tests.insert(description.id, description);
          }

          TestEvent::Plan(plan) => {
            summary.total += plan.total;
            summary.filtered_out += plan.filtered_out;

            if plan.used_only {
              used_only = true;
            }

            reporter.report_plan(&plan);
          }

          TestEvent::Wait(id) => {
            reporter.report_wait(tests.get(&id).unwrap());
          }

          TestEvent::Output(output) => {
            reporter.report_output(&output);
          }

          TestEvent::Result(id, result, elapsed) => {
            if tests_with_result.insert(id) {
              let description = tests.get(&id).unwrap().clone();
              match &result {
                TestResult::Ok => {
                  summary.passed += 1;
                }
                TestResult::Ignored => {
                  summary.ignored += 1;
                }
                TestResult::Failed(error) => {
                  summary.failed += 1;
                  summary.failures.push((description.clone(), error.clone()));
                }
                TestResult::Cancelled => {
                  unreachable!("should be handled in TestEvent::UncaughtError");
                }
              }
              reporter.report_result(&description, &result, elapsed);
            }
          }

          TestEvent::UncaughtError(origin, error) => {
            reporter.report_uncaught_error(&origin, &error);
            summary.failed += 1;
            summary.uncaught_errors.push((origin.clone(), error));
            for desc in tests.values() {
              if desc.origin == origin && tests_with_result.insert(desc.id) {
                summary.failed += 1;
                reporter.report_result(desc, &TestResult::Cancelled, 0);
              }
            }
          }

          TestEvent::StepRegister(description) => {
            reporter.report_step_register(&description);
            test_steps.insert(description.id, description);
          }

          TestEvent::StepWait(id) => {
            reporter.report_step_wait(test_steps.get(&id).unwrap());
          }

          TestEvent::StepResult(id, result, duration) => {
            match &result {
              TestStepResult::Ok => {
                summary.passed_steps += 1;
              }
              TestStepResult::Ignored => {
                summary.ignored_steps += 1;
              }
              TestStepResult::Failed(_) => {
                summary.failed_steps += 1;
              }
              TestStepResult::Pending(_) => {
                summary.pending_steps += 1;
              }
            }

            reporter.report_step_result(
              test_steps.get(&id).unwrap(),
              &result,
              duration,
              &tests,
              &test_steps,
            );
          }
        }

        if let Some(x) = fail_fast {
          if summary.failed >= x.get() {
            break;
          }
        }
      }

      let elapsed = Instant::now().duration_since(earlier);
      reporter.report_summary(&summary, &elapsed);

      if used_only {
        return Err(generic_error(
          "Test failed because the \"only\" option was used",
        ));
      }

      if summary.failed > 0 {
        return Err(generic_error("Test failed"));
      }

      Ok(())
    })
  };

  let (join_results, result) = future::join(join_stream, handler).await;

  // propagate any errors
  for join_result in join_results {
    join_result??;
  }

  result??;

  Ok(())
}

/// Collects specifiers marking them with the appropriate test mode while maintaining the natural
/// input order.
///
/// - Specifiers matching the `is_supported_test_ext` predicate are marked as
/// `TestMode::Documentation`.
/// - Specifiers matching the `is_supported_test_path` are marked as `TestMode::Executable`.
/// - Specifiers matching both predicates are marked as `TestMode::Both`
fn collect_specifiers_with_test_mode(
  include: Vec<String>,
  ignore: Vec<PathBuf>,
  include_inline: bool,
) -> Result<Vec<(ModuleSpecifier, TestMode)>, AnyError> {
  let module_specifiers =
    collect_specifiers(include.clone(), &ignore, is_supported_test_path)?;

  if include_inline {
    return collect_specifiers(include, &ignore, is_supported_test_ext).map(
      |specifiers| {
        specifiers
          .into_iter()
          .map(|specifier| {
            let mode = if module_specifiers.contains(&specifier) {
              TestMode::Both
            } else {
              TestMode::Documentation
            };

            (specifier, mode)
          })
          .collect()
      },
    );
  }

  let specifiers_with_mode = module_specifiers
    .into_iter()
    .map(|specifier| (specifier, TestMode::Executable))
    .collect();

  Ok(specifiers_with_mode)
}

/// Collects module and document specifiers with test modes via
/// `collect_specifiers_with_test_mode` which are then pre-fetched and adjusted
/// based on the media type.
///
/// Specifiers that do not have a known media type that can be executed as a
/// module are marked as `TestMode::Documentation`. Type definition files
/// cannot be run, and therefore need to be marked as `TestMode::Documentation`
/// as well.
async fn fetch_specifiers_with_test_mode(
  ps: &ProcState,
  include: Vec<String>,
  ignore: Vec<PathBuf>,
  include_inline: bool,
) -> Result<Vec<(ModuleSpecifier, TestMode)>, AnyError> {
  let maybe_test_config = ps.options.to_test_config()?;

  let mut include_files = include.clone();
  let mut exclude_files = ignore.clone();

  if let Some(test_config) = maybe_test_config.as_ref() {
    if include_files.is_empty() {
      include_files = test_config
        .files
        .include
        .iter()
        .map(|s| s.to_string())
        .collect::<Vec<_>>();
    }

    if exclude_files.is_empty() {
      exclude_files = test_config
        .files
        .exclude
        .iter()
        .filter_map(|s| specifier_to_file_path(s).ok())
        .collect::<Vec<_>>();
    }
  }

  if include_files.is_empty() {
    include_files.push(".".to_string());
  }

  let mut specifiers_with_mode = collect_specifiers_with_test_mode(
    include_files,
    exclude_files,
    include_inline,
  )?;
  for (specifier, mode) in &mut specifiers_with_mode {
    let file = ps
      .file_fetcher
      .fetch(specifier, &mut Permissions::allow_all())
      .await?;

    if file.media_type == MediaType::Unknown
      || file.media_type == MediaType::Dts
    {
      *mode = TestMode::Documentation
    }
  }

  Ok(specifiers_with_mode)
}

pub async fn run_tests(
  flags: Flags,
  test_flags: TestFlags,
) -> Result<(), AnyError> {
  let ps = ProcState::build(flags).await?;
  let permissions =
    Permissions::from_options(&ps.options.permissions_options())?;
  let specifiers_with_mode = fetch_specifiers_with_test_mode(
    &ps,
    test_flags.include,
    test_flags.ignore.clone(),
    test_flags.doc,
  )
  .await?;

  if !test_flags.allow_none && specifiers_with_mode.is_empty() {
    return Err(generic_error("No test modules found"));
  }

  check_specifiers(&ps, permissions.clone(), specifiers_with_mode.clone())
    .await?;

  if test_flags.no_run {
    return Ok(());
  }

  test_specifiers(
    ps,
    permissions,
    specifiers_with_mode,
    TestSpecifierOptions {
      concurrent_jobs: test_flags.concurrent_jobs,
      fail_fast: test_flags.fail_fast,
      filter: TestFilter::from_flag(&test_flags.filter),
    },
  )
  .await?;

  Ok(())
}

pub async fn run_tests_with_watch(
  flags: Flags,
  test_flags: TestFlags,
) -> Result<(), AnyError> {
  let ps = ProcState::build(flags).await?;
  let permissions =
    Permissions::from_options(&ps.options.permissions_options())?;

  let include = test_flags.include;
  let ignore = test_flags.ignore.clone();
  let paths_to_watch: Vec<_> = include.iter().map(PathBuf::from).collect();
  let no_check = ps.options.type_check_mode() == TypeCheckMode::None;

  let resolver = |changed: Option<Vec<PathBuf>>| {
    let paths_to_watch = paths_to_watch.clone();
    let paths_to_watch_clone = paths_to_watch.clone();

    let files_changed = changed.is_some();
    let include = include.clone();
    let ignore = ignore.clone();
    let ps = ps.clone();

    async move {
      let test_modules = if test_flags.doc {
        collect_specifiers(include.clone(), &ignore, is_supported_test_ext)
      } else {
        collect_specifiers(include.clone(), &ignore, is_supported_test_path)
      }?;

      let mut paths_to_watch = paths_to_watch_clone;
      let mut modules_to_reload = if files_changed {
        Vec::new()
      } else {
        test_modules
          .iter()
          .map(|url| (url.clone(), ModuleKind::Esm))
          .collect()
      };
      let graph = ps
        .create_graph(
          test_modules
            .iter()
            .map(|s| (s.clone(), ModuleKind::Esm))
            .collect(),
        )
        .await?;
      graph_valid(&graph, !no_check, ps.options.check_js())?;

      // TODO(@kitsonk) - This should be totally derivable from the graph.
      for specifier in test_modules {
        fn get_dependencies<'a>(
          graph: &'a deno_graph::ModuleGraph,
          maybe_module: Option<&'a deno_graph::Module>,
          // This needs to be accessible to skip getting dependencies if they're already there,
          // otherwise this will cause a stack overflow with circular dependencies
          output: &mut HashSet<&'a ModuleSpecifier>,
          no_check: bool,
        ) {
          if let Some(module) = maybe_module {
            for dep in module.dependencies.values() {
              if let Some(specifier) = &dep.get_code() {
                if !output.contains(specifier) {
                  output.insert(specifier);
                  get_dependencies(
                    graph,
                    graph.get(specifier),
                    output,
                    no_check,
                  );
                }
              }
              if !no_check {
                if let Some(specifier) = &dep.get_type() {
                  if !output.contains(specifier) {
                    output.insert(specifier);
                    get_dependencies(
                      graph,
                      graph.get(specifier),
                      output,
                      no_check,
                    );
                  }
                }
              }
            }
          }
        }

        // This test module and all it's dependencies
        let mut modules = HashSet::new();
        modules.insert(&specifier);
        get_dependencies(&graph, graph.get(&specifier), &mut modules, no_check);

        paths_to_watch.extend(
          modules
            .iter()
            .filter_map(|specifier| specifier.to_file_path().ok()),
        );

        if let Some(changed) = &changed {
          for path in changed.iter().filter_map(|path| {
            deno_core::resolve_url_or_path(&path.to_string_lossy()).ok()
          }) {
            if modules.contains(&&path) {
              modules_to_reload.push((specifier, ModuleKind::Esm));
              break;
            }
          }
        }
      }

      Ok((paths_to_watch, modules_to_reload))
    }
    .map(move |result| {
      if files_changed
        && matches!(result, Ok((_, ref modules)) if modules.is_empty())
      {
        ResolutionResult::Ignore
      } else {
        match result {
          Ok((paths_to_watch, modules_to_reload)) => {
            ResolutionResult::Restart {
              paths_to_watch,
              result: Ok(modules_to_reload),
            }
          }
          Err(e) => ResolutionResult::Restart {
            paths_to_watch,
            result: Err(e),
          },
        }
      }
    })
  };

  let cli_options = ps.options.clone();
  let operation = |modules_to_reload: Vec<(ModuleSpecifier, ModuleKind)>| {
    let filter = test_flags.filter.clone();
    let include = include.clone();
    let ignore = ignore.clone();
    let permissions = permissions.clone();
    let ps = ps.clone();

    async move {
      let specifiers_with_mode = fetch_specifiers_with_test_mode(
        &ps,
        include.clone(),
        ignore.clone(),
        test_flags.doc,
      )
      .await?
      .iter()
      .filter(|(specifier, _)| {
        contains_specifier(&modules_to_reload, specifier)
      })
      .cloned()
      .collect::<Vec<(ModuleSpecifier, TestMode)>>();

      check_specifiers(&ps, permissions.clone(), specifiers_with_mode.clone())
        .await?;

      if test_flags.no_run {
        return Ok(());
      }

      test_specifiers(
        ps,
        permissions.clone(),
        specifiers_with_mode,
        TestSpecifierOptions {
          concurrent_jobs: test_flags.concurrent_jobs,
          fail_fast: test_flags.fail_fast,
          filter: TestFilter::from_flag(&filter),
        },
      )
      .await?;

      Ok(())
    }
  };

  file_watcher::watch_func(
    resolver,
    operation,
    file_watcher::PrintConfig {
      job_name: "Test".to_string(),
      clear_screen: !cli_options.no_clear_screen(),
    },
  )
  .await?;

  Ok(())
}

#[derive(Clone)]
pub struct TestEventSender {
  sender: UnboundedSender<TestEvent>,
  stdout_writer: TestOutputPipe,
  stderr_writer: TestOutputPipe,
}

impl TestEventSender {
  pub fn new(sender: UnboundedSender<TestEvent>) -> Self {
    Self {
      stdout_writer: TestOutputPipe::new(sender.clone()),
      stderr_writer: TestOutputPipe::new(sender.clone()),
      sender,
    }
  }

  pub fn stdout(&self) -> std::fs::File {
    self.stdout_writer.as_file()
  }

  pub fn stderr(&self) -> std::fs::File {
    self.stderr_writer.as_file()
  }

  pub fn send(&mut self, message: TestEvent) -> Result<(), AnyError> {
    // for any event that finishes collecting output, we need to
    // ensure that the collected stdout and stderr pipes are flushed
    if matches!(
      message,
      TestEvent::Result(_, _, _)
        | TestEvent::StepWait(_)
        | TestEvent::StepResult(_, _, _)
    ) {
      self.flush_stdout_and_stderr();
    }

    self.sender.send(message)?;
    Ok(())
  }

  fn flush_stdout_and_stderr(&mut self) {
    self.stdout_writer.flush();
    self.stderr_writer.flush();
  }
}

// use a string that if it ends up in the output won't affect how things are displayed
const ZERO_WIDTH_SPACE: &str = "\u{200B}";

struct TestOutputPipe {
  writer: os_pipe::PipeWriter,
  state: Arc<Mutex<Option<std::sync::mpsc::Sender<()>>>>,
}

impl Clone for TestOutputPipe {
  fn clone(&self) -> Self {
    Self {
      writer: self.writer.try_clone().unwrap(),
      state: self.state.clone(),
    }
  }
}

impl TestOutputPipe {
  pub fn new(sender: UnboundedSender<TestEvent>) -> Self {
    let (reader, writer) = os_pipe::pipe().unwrap();
    let state = Arc::new(Mutex::new(None));

    start_output_redirect_thread(reader, sender, state.clone());

    Self { writer, state }
  }

  pub fn flush(&mut self) {
    // We want to wake up the other thread and have it respond back
    // that it's done clearing out its pipe before returning.
    let (sender, receiver) = std::sync::mpsc::channel();
    if let Some(sender) = self.state.lock().replace(sender) {
      let _ = sender.send(()); // just in case
    }
    // Bit of a hack to send a zero width space in order to wake
    // the thread up. It seems that sending zero bytes here does
    // not work on windows.
    self.writer.write_all(ZERO_WIDTH_SPACE.as_bytes()).unwrap();
    self.writer.flush().unwrap();
    // ignore the error as it might have been picked up and closed
    let _ = receiver.recv();
  }

  pub fn as_file(&self) -> std::fs::File {
    pipe_writer_to_file(self.writer.try_clone().unwrap())
  }
}

#[cfg(windows)]
fn pipe_writer_to_file(writer: os_pipe::PipeWriter) -> std::fs::File {
  use std::os::windows::prelude::FromRawHandle;
  use std::os::windows::prelude::IntoRawHandle;
  // SAFETY: Requires consuming ownership of the provided handle
  unsafe { std::fs::File::from_raw_handle(writer.into_raw_handle()) }
}

#[cfg(unix)]
fn pipe_writer_to_file(writer: os_pipe::PipeWriter) -> std::fs::File {
  use std::os::unix::io::FromRawFd;
  use std::os::unix::io::IntoRawFd;
  // SAFETY: Requires consuming ownership of the provided handle
  unsafe { std::fs::File::from_raw_fd(writer.into_raw_fd()) }
}

fn start_output_redirect_thread(
  mut pipe_reader: os_pipe::PipeReader,
  sender: UnboundedSender<TestEvent>,
  flush_state: Arc<Mutex<Option<std::sync::mpsc::Sender<()>>>>,
) {
  tokio::task::spawn_blocking(move || loop {
    let mut buffer = [0; 512];
    let size = match pipe_reader.read(&mut buffer) {
      Ok(0) | Err(_) => break,
      Ok(size) => size,
    };
    let oneshot_sender = flush_state.lock().take();
    let mut data = &buffer[0..size];
    if data.ends_with(ZERO_WIDTH_SPACE.as_bytes()) {
      data = &data[0..data.len() - ZERO_WIDTH_SPACE.len()];
    }

    if !data.is_empty()
      && sender
        .send(TestEvent::Output(buffer[0..size].to_vec()))
        .is_err()
    {
      break;
    }

    // Always respond back if this was set. Ideally we would also check to
    // ensure the pipe reader is empty before sending back this response.
    if let Some(sender) = oneshot_sender {
      let _ignore = sender.send(());
    }
  });
}