diff options
author | Yusuke Tanaka <yusuktan@maguro.dev> | 2024-09-18 13:35:48 +0900 |
---|---|---|
committer | GitHub <noreply@github.com> | 2024-09-17 21:35:48 -0700 |
commit | d5c00ef50e6519fccde54a577e038f0ebb1282e9 (patch) | |
tree | 9429c5f09c6969fd8236041f48b354a9b0841e1f /cli | |
parent | 37315917625179063cb5653e2edd4ee0e5de99c5 (diff) |
feat(cli): evaluate code snippets in JSDoc and markdown (#25220)
This commit lets `deno test --doc` command actually evaluate code snippets in
JSDoc and markdown files.
## How it works
1. Extract code snippets from JSDoc or code fences
2. Convert them into pseudo files by wrapping them in `Deno.test(...)`
3. Register the pseudo files as in-memory files
4. Run type-check and evaluation
We apply some magic at the step 2 - let's say we have the following file named
`mod.ts` as an input:
````ts
/**
* ```ts
* import { assertEquals } from "jsr:@std/assert/equals";
*
* assertEquals(add(1, 2), 3);
* ```
*/
export function add(a: number, b: number) {
return a + b;
}
````
This is virtually transformed into:
```ts
import { assertEquals } from "jsr:@std/assert/equals";
import { add } from "files:///path/to/mod.ts";
Deno.test("mod.ts$2-7.ts", async () => {
assertEquals(add(1, 2), 3);
});
```
Note that a new import statement is inserted here to make `add` function
available. In a nutshell, all items exported from `mod.ts` become available in
the generated pseudo file with this automatic import insertion.
The intention behind this design is that, from library user's standpoint, it
should be very obvious that this `add` function is what this example code is
attached to. Also, if there is an explicit import statement like
`import { add } from "./mod.ts"`, this import path `./mod.ts` is not helpful for
doc readers because they will need to import it in a different way.
The automatic import insertion has some edge cases, in particular where there is
a local variable in a snippet with the same name as one of the exported items.
This case is addressed by employing swc's scope analysis (see test cases for
more details).
## "type-checking only" mode stays around
This change will likely impact a lot of existing doc tests in the ecosystem
because some doc tests rely on the fact that they are not evaluated - some cause
side effects if executed, some throw errors at runtime although they do pass the
type check, etc. To help those tests gradually transition to the ones runnable
with the new `deno test --doc`, we will keep providing the ability to run
type-checking only via `deno check --doc`. Additionally there is a `--doc-only`
option added to the `check` subcommand too, which is useful when you want to
type-check on code snippets in markdown files, as normal `deno check` command
doesn't accept markdown.
## Demo
https://github.com/user-attachments/assets/47e9af73-d16e-472d-b09e-1853b9e8f5ce
---
Closes #4716
Diffstat (limited to 'cli')
-rw-r--r-- | cli/args/flags.rs | 68 | ||||
-rw-r--r-- | cli/lsp/testing/execution.rs | 13 | ||||
-rw-r--r-- | cli/main.rs | 7 | ||||
-rw-r--r-- | cli/tools/check.rs | 46 | ||||
-rw-r--r-- | cli/tools/test/mod.rs | 336 | ||||
-rw-r--r-- | cli/util/extract.rs | 1410 | ||||
-rw-r--r-- | cli/util/mod.rs | 1 |
7 files changed, 1604 insertions, 277 deletions
diff --git a/cli/args/flags.rs b/cli/args/flags.rs index f832c2a62..92336a0a1 100644 --- a/cli/args/flags.rs +++ b/cli/args/flags.rs @@ -109,6 +109,8 @@ pub struct CacheFlags { #[derive(Clone, Debug, Eq, PartialEq)] pub struct CheckFlags { pub files: Vec<String>, + pub doc: bool, + pub doc_only: bool, } #[derive(Clone, Debug, Eq, PartialEq)] @@ -1695,6 +1697,19 @@ Unless --reload is specified, this command will not re-download already cached d .hide(true) ) .arg( + Arg::new("doc") + .long("doc") + .help("Type-check code blocks in JSDoc as well as actual code") + .action(ArgAction::SetTrue) + ) + .arg( + Arg::new("doc-only") + .long("doc-only") + .help("Type-check code blocks in JSDoc and Markdown only") + .action(ArgAction::SetTrue) + .conflicts_with("doc") + ) + .arg( Arg::new("file") .num_args(1..) .required_unless_present("help") @@ -2789,7 +2804,7 @@ or <c>**/__tests__/**</>: .arg( Arg::new("doc") .long("doc") - .help("Type-check code blocks in JSDoc and Markdown") + .help("Evaluate code blocks in JSDoc and Markdown") .action(ArgAction::SetTrue) .help_heading(TEST_HEADING), ) @@ -4121,7 +4136,11 @@ fn check_parse( if matches.get_flag("all") || matches.get_flag("remote") { flags.type_check_mode = TypeCheckMode::All; } - flags.subcommand = DenoSubcommand::Check(CheckFlags { files }); + flags.subcommand = DenoSubcommand::Check(CheckFlags { + files, + doc: matches.get_flag("doc"), + doc_only: matches.get_flag("doc-only"), + }); Ok(()) } @@ -6862,12 +6881,55 @@ mod tests { Flags { subcommand: DenoSubcommand::Check(CheckFlags { files: svec!["script.ts"], + doc: false, + doc_only: false, + }), + type_check_mode: TypeCheckMode::Local, + ..Flags::default() + } + ); + + let r = flags_from_vec(svec!["deno", "check", "--doc", "script.ts"]); + assert_eq!( + r.unwrap(), + Flags { + subcommand: DenoSubcommand::Check(CheckFlags { + files: svec!["script.ts"], + doc: true, + doc_only: false, }), type_check_mode: TypeCheckMode::Local, ..Flags::default() } ); + let r = flags_from_vec(svec!["deno", "check", "--doc-only", "markdown.md"]); + assert_eq!( + r.unwrap(), + Flags { + subcommand: DenoSubcommand::Check(CheckFlags { + files: svec!["markdown.md"], + doc: false, + doc_only: true, + }), + type_check_mode: TypeCheckMode::Local, + ..Flags::default() + } + ); + + // `--doc` and `--doc-only` are mutually exclusive + let r = flags_from_vec(svec![ + "deno", + "check", + "--doc", + "--doc-only", + "script.ts" + ]); + assert_eq!( + r.unwrap_err().kind(), + clap::error::ErrorKind::ArgumentConflict + ); + for all_flag in ["--remote", "--all"] { let r = flags_from_vec(svec!["deno", "check", all_flag, "script.ts"]); assert_eq!( @@ -6875,6 +6937,8 @@ mod tests { Flags { subcommand: DenoSubcommand::Check(CheckFlags { files: svec!["script.ts"], + doc: false, + doc_only: false, }), type_check_mode: TypeCheckMode::All, ..Flags::default() diff --git a/cli/lsp/testing/execution.rs b/cli/lsp/testing/execution.rs index a8cea8dd4..55d81f342 100644 --- a/cli/lsp/testing/execution.rs +++ b/cli/lsp/testing/execution.rs @@ -234,16 +234,9 @@ impl TestRun { &cli_options.permissions_options(), )?; let main_graph_container = factory.main_module_graph_container().await?; - test::check_specifiers( - factory.file_fetcher()?, - main_graph_container, - self - .queue - .iter() - .map(|s| (s.clone(), test::TestMode::Executable)) - .collect(), - ) - .await?; + main_graph_container + .check_specifiers(&self.queue.iter().cloned().collect::<Vec<_>>()) + .await?; let (concurrent_jobs, fail_fast) = if let DenoSubcommand::Test(test_flags) = cli_options.sub_command() { diff --git a/cli/main.rs b/cli/main.rs index 10d9ead4e..a7e1c8342 100644 --- a/cli/main.rs +++ b/cli/main.rs @@ -121,12 +121,7 @@ async fn run_subcommand(flags: Arc<Flags>) -> Result<i32, AnyError> { tools::installer::install_from_entrypoints(flags, &cache_flags.files).await }), DenoSubcommand::Check(check_flags) => spawn_subcommand(async move { - let factory = CliFactory::from_flags(flags); - let main_graph_container = - factory.main_module_graph_container().await?; - main_graph_container - .load_and_type_check_files(&check_flags.files) - .await + tools::check::check(flags, check_flags).await }), DenoSubcommand::Clean => spawn_subcommand(async move { tools::clean::clean() diff --git a/cli/tools/check.rs b/cli/tools/check.rs index d50af5230..9c464fa16 100644 --- a/cli/tools/check.rs +++ b/cli/tools/check.rs @@ -15,7 +15,9 @@ use once_cell::sync::Lazy; use regex::Regex; use crate::args::check_warn_tsconfig; +use crate::args::CheckFlags; use crate::args::CliOptions; +use crate::args::Flags; use crate::args::TsConfig; use crate::args::TsConfigType; use crate::args::TsTypeLib; @@ -24,13 +26,57 @@ use crate::cache::CacheDBHash; use crate::cache::Caches; use crate::cache::FastInsecureHasher; use crate::cache::TypeCheckCache; +use crate::factory::CliFactory; use crate::graph_util::BuildFastCheckGraphOptions; use crate::graph_util::ModuleGraphBuilder; use crate::npm::CliNpmResolver; use crate::tsc; use crate::tsc::Diagnostics; +use crate::util::extract; use crate::util::path::to_percent_decoded_str; +pub async fn check( + flags: Arc<Flags>, + check_flags: CheckFlags, +) -> Result<(), AnyError> { + let factory = CliFactory::from_flags(flags); + + let main_graph_container = factory.main_module_graph_container().await?; + + let specifiers = + main_graph_container.collect_specifiers(&check_flags.files)?; + if specifiers.is_empty() { + log::warn!("{} No matching files found.", colors::yellow("Warning")); + } + + let specifiers_for_typecheck = if check_flags.doc || check_flags.doc_only { + let file_fetcher = factory.file_fetcher()?; + + let mut specifiers_for_typecheck = if check_flags.doc { + specifiers.clone() + } else { + vec![] + }; + + for s in specifiers { + let file = file_fetcher.fetch_bypass_permissions(&s).await?; + let snippet_files = extract::extract_snippet_files(file)?; + for snippet_file in snippet_files { + specifiers_for_typecheck.push(snippet_file.specifier.clone()); + file_fetcher.insert_memory_files(snippet_file); + } + } + + specifiers_for_typecheck + } else { + specifiers + }; + + main_graph_container + .check_specifiers(&specifiers_for_typecheck) + .await +} + /// Options for performing a check of a module graph. Note that the decision to /// emit or not is determined by the `ts_config` settings. pub struct CheckOptions { diff --git a/cli/tools/test/mod.rs b/cli/tools/test/mod.rs index 63382ffc6..d043ffcba 100644 --- a/cli/tools/test/mod.rs +++ b/cli/tools/test/mod.rs @@ -9,21 +9,18 @@ use crate::display; use crate::factory::CliFactory; use crate::file_fetcher::File; use crate::file_fetcher::FileFetcher; -use crate::graph_container::MainModuleGraphContainer; use crate::graph_util::has_graph_root_local_dependent_changed; use crate::ops; +use crate::util::extract::extract_doc_tests; use crate::util::file_watcher; use crate::util::fs::collect_specifiers; use crate::util::path::get_extension; use crate::util::path::is_script_ext; -use crate::util::path::mapped_specifier_for_tsc; use crate::util::path::matches_pattern_or_exact_path; use crate::worker::CliMainWorkerFactory; use crate::worker::CoverageCollector; -use deno_ast::swc::common::comments::CommentKind; use deno_ast::MediaType; -use deno_ast::SourceRangedForSpanned; use deno_config::glob::FilePatterns; use deno_config::glob::WalkEntry; use deno_core::anyhow; @@ -151,6 +148,20 @@ pub enum TestMode { Both, } +impl TestMode { + /// Returns `true` if the test mode indicates that code snippet extraction is + /// needed. + fn needs_test_extraction(&self) -> bool { + matches!(self, Self::Documentation | Self::Both) + } + + /// Returns `true` if the test mode indicates that the test should be + /// type-checked and run. + fn needs_test_run(&self) -> bool { + matches!(self, Self::Executable | Self::Both) + } +} + #[derive(Clone, Debug, Default)] pub struct TestFilter { pub substring: Option<String>, @@ -1174,233 +1185,6 @@ async fn wait_for_activity_to_stabilize( }) } -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| { - block.get(1)?; - - 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, - _ => 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 = ModuleSpecifier::parse(&format!( - "{}${}-{}", - specifier, - file_line_index + line_offset + 1, - file_line_index + line_offset + line_count + 1, - )) - .unwrap(); - let file_specifier = - mapped_specifier_for_tsc(&file_specifier, file_media_type) - .map(|s| ModuleSpecifier::parse(&s).unwrap()) - .unwrap_or(file_specifier); - - Some(File { - specifier: file_specifier, - maybe_headers: None, - source: file_source.into_bytes().into(), - }) - }) - .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.clone(), - text: source, - media_type, - capture_tokens: false, - maybe_syntax: None, - scope_analysis: false, - })?; - let comments = parsed_source.comments().get_vec(); - let blocks_regex = lazy_regex::regex!(r"```([^\r\n]*)\r?\n([\S\s]*?)```"); - let lines_regex = lazy_regex::regex!(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_lazy().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 = - lazy_regex::regex!(r"(?s)<!--.*?-->|```([^\r\n]*)\r?\n([\S\s]*?)```"); - let lines_regex = lazy_regex::regex!(r"(?:\# ?)?(.*)"); - - extract_files_from_regex_blocks( - specifier, - source, - media_type, - /* file line index */ 0, - blocks_regex, - lines_regex, - ) -} - -async fn fetch_inline_files( - file_fetcher: &FileFetcher, - specifiers: Vec<ModuleSpecifier>, -) -> Result<Vec<File>, AnyError> { - let mut files = Vec::new(); - for specifier in specifiers { - let file = file_fetcher - .fetch_bypass_permissions(&specifier) - .await? - .into_text_decoded()?; - - 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, - file.media_type, - ) - }; - - files.extend(inline_files?); - } - - Ok(files) -} - -/// Type check a collection of module and document specifiers. -pub async fn check_specifiers( - file_fetcher: &FileFetcher, - main_graph_container: &Arc<MainModuleGraphContainer>, - specifiers: Vec<(ModuleSpecifier, TestMode)>, -) -> Result<(), AnyError> { - let inline_files = fetch_inline_files( - file_fetcher, - specifiers - .iter() - .filter_map(|(specifier, mode)| { - if *mode != TestMode::Executable { - Some(specifier.clone()) - } else { - None - } - }) - .collect(), - ) - .await?; - - let mut module_specifiers = specifiers - .into_iter() - .filter_map(|(specifier, mode)| { - if mode != TestMode::Documentation { - Some(specifier) - } else { - None - } - }) - .collect::<Vec<_>>(); - - if !inline_files.is_empty() { - module_specifiers - .extend(inline_files.iter().map(|file| file.specifier.clone())); - - for file in inline_files { - file_fetcher.insert_memory_files(file); - } - } - - main_graph_container - .check_specifiers(&module_specifiers) - .await?; - - Ok(()) -} - static HAS_TEST_RUN_SIGINT_HANDLER: AtomicBool = AtomicBool::new(false); /// Test a collection of specifiers with test modes concurrently. @@ -1788,14 +1572,19 @@ pub async fn run_tests( return Err(generic_error("No test modules found")); } + let doc_tests = get_doc_tests(&specifiers_with_mode, file_fetcher).await?; + let specifiers_for_typecheck_and_test = + get_target_specifiers(specifiers_with_mode, &doc_tests); + for doc_test in doc_tests { + file_fetcher.insert_memory_files(doc_test); + } + let main_graph_container = factory.main_module_graph_container().await?; - check_specifiers( - file_fetcher, - main_graph_container, - specifiers_with_mode.clone(), - ) - .await?; + // Typecheck + main_graph_container + .check_specifiers(&specifiers_for_typecheck_and_test) + .await?; if workspace_test_options.no_run { return Ok(()); @@ -1804,17 +1593,12 @@ pub async fn run_tests( let worker_factory = Arc::new(factory.create_cli_main_worker_factory().await?); + // Run tests test_specifiers( worker_factory, &permissions, permission_desc_parser, - specifiers_with_mode - .into_iter() - .filter_map(|(s, m)| match m { - TestMode::Documentation => None, - _ => Some(s), - }) - .collect(), + specifiers_for_typecheck_and_test, TestSpecifiersOptions { cwd: Url::from_directory_path(cli_options.initial_cwd()).map_err( |_| { @@ -1949,8 +1733,6 @@ pub async fn run_tests_with_watch( test_modules.clone() }; - let worker_factory = - Arc::new(factory.create_cli_main_worker_factory().await?); let specifiers_with_mode = fetch_specifiers_with_test_mode( &cli_options, file_fetcher, @@ -1962,30 +1744,34 @@ pub async fn run_tests_with_watch( .filter(|(specifier, _)| test_modules_to_reload.contains(specifier)) .collect::<Vec<(ModuleSpecifier, TestMode)>>(); + let doc_tests = + get_doc_tests(&specifiers_with_mode, file_fetcher).await?; + let specifiers_for_typecheck_and_test = + get_target_specifiers(specifiers_with_mode, &doc_tests); + for doc_test in doc_tests { + file_fetcher.insert_memory_files(doc_test); + } + let main_graph_container = factory.main_module_graph_container().await?; - check_specifiers( - file_fetcher, - main_graph_container, - specifiers_with_mode.clone(), - ) - .await?; + + // Typecheck + main_graph_container + .check_specifiers(&specifiers_for_typecheck_and_test) + .await?; if workspace_test_options.no_run { return Ok(()); } + let worker_factory = + Arc::new(factory.create_cli_main_worker_factory().await?); + test_specifiers( worker_factory, &permissions, permission_desc_parser, - specifiers_with_mode - .into_iter() - .filter_map(|(s, m)| match m { - TestMode::Documentation => None, - _ => Some(s), - }) - .collect(), + specifiers_for_typecheck_and_test, TestSpecifiersOptions { cwd: Url::from_directory_path(cli_options.initial_cwd()).map_err( |_| { @@ -2020,6 +1806,38 @@ pub async fn run_tests_with_watch( Ok(()) } +/// Extracts doc tests from files specified by the given specifiers. +async fn get_doc_tests( + specifiers_with_mode: &[(Url, TestMode)], + file_fetcher: &FileFetcher, +) -> Result<Vec<File>, AnyError> { + let specifiers_needing_extraction = specifiers_with_mode + .iter() + .filter(|(_, mode)| mode.needs_test_extraction()) + .map(|(s, _)| s); + + let mut doc_tests = Vec::new(); + for s in specifiers_needing_extraction { + let file = file_fetcher.fetch_bypass_permissions(s).await?; + doc_tests.extend(extract_doc_tests(file)?); + } + + Ok(doc_tests) +} + +/// Get a list of specifiers that we need to perform typecheck and run tests on. +/// The result includes "pseudo specifiers" for doc tests. +fn get_target_specifiers( + specifiers_with_mode: Vec<(Url, TestMode)>, + doc_tests: &[File], +) -> Vec<Url> { + specifiers_with_mode + .into_iter() + .filter_map(|(s, mode)| mode.needs_test_run().then_some(s)) + .chain(doc_tests.iter().map(|d| d.specifier.clone())) + .collect() +} + /// Tracks failures for the `--fail-fast` argument in /// order to tell when to stop running tests. #[derive(Clone, Default)] diff --git a/cli/util/extract.rs b/cli/util/extract.rs new file mode 100644 index 000000000..e27a79347 --- /dev/null +++ b/cli/util/extract.rs @@ -0,0 +1,1410 @@ +// Copyright 2018-2024 the Deno authors. All rights reserved. MIT license. + +use deno_ast::swc::ast; +use deno_ast::swc::atoms::Atom; +use deno_ast::swc::common::collections::AHashSet; +use deno_ast::swc::common::comments::CommentKind; +use deno_ast::swc::common::DUMMY_SP; +use deno_ast::swc::utils as swc_utils; +use deno_ast::swc::visit::as_folder; +use deno_ast::swc::visit::FoldWith as _; +use deno_ast::swc::visit::Visit; +use deno_ast::swc::visit::VisitMut; +use deno_ast::swc::visit::VisitWith as _; +use deno_ast::MediaType; +use deno_ast::SourceRangedForSpanned as _; +use deno_core::error::AnyError; +use deno_core::ModuleSpecifier; +use regex::Regex; +use std::collections::BTreeSet; +use std::fmt::Write as _; +use std::sync::Arc; + +use crate::file_fetcher::File; +use crate::util::path::mapped_specifier_for_tsc; + +/// Extracts doc tests from a given file, transforms them into pseudo test +/// files by wrapping the content of the doc tests in a `Deno.test` call, and +/// returns a list of the pseudo test files. +/// +/// The difference from [`extract_snippet_files`] is that this function wraps +/// extracted code snippets in a `Deno.test` call. +pub fn extract_doc_tests(file: File) -> Result<Vec<File>, AnyError> { + extract_inner(file, WrapKind::DenoTest) +} + +/// Extracts code snippets from a given file and returns a list of the extracted +/// files. +/// +/// The difference from [`extract_doc_tests`] is that this function does *not* +/// wrap extracted code snippets in a `Deno.test` call. +pub fn extract_snippet_files(file: File) -> Result<Vec<File>, AnyError> { + extract_inner(file, WrapKind::NoWrap) +} + +#[derive(Clone, Copy)] +enum WrapKind { + DenoTest, + NoWrap, +} + +fn extract_inner( + file: File, + wrap_kind: WrapKind, +) -> Result<Vec<File>, AnyError> { + let file = file.into_text_decoded()?; + + let exports = match deno_ast::parse_program(deno_ast::ParseParams { + specifier: file.specifier.clone(), + text: file.source.clone(), + media_type: file.media_type, + capture_tokens: false, + scope_analysis: false, + maybe_syntax: None, + }) { + Ok(parsed) => { + let mut c = ExportCollector::default(); + c.visit_program(parsed.program_ref()); + c + } + Err(_) => ExportCollector::default(), + }; + + let extracted_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, + )? + }; + + extracted_files + .into_iter() + .map(|extracted_file| { + generate_pseudo_file(extracted_file, &file.specifier, &exports, wrap_kind) + }) + .collect::<Result<_, _>>() +} + +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 = + lazy_regex::regex!(r"(?s)<!--.*?-->|```([^\r\n]*)\r?\n([\S\s]*?)```"); + let lines_regex = lazy_regex::regex!(r"(?:\# ?)?(.*)"); + + extract_files_from_regex_blocks( + specifier, + source, + media_type, + /* file line index */ 0, + blocks_regex, + lines_regex, + ) +} + +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.clone(), + text: source, + media_type, + capture_tokens: false, + maybe_syntax: None, + scope_analysis: false, + })?; + let comments = parsed_source.comments().get_vec(); + let blocks_regex = lazy_regex::regex!(r"```([^\r\n]*)\r?\n([\S\s]*?)```"); + let lines_regex = lazy_regex::regex!(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_lazy().line_index(comment.start()), + blocks_regex, + lines_regex, + ) + }) + .flatten() + .collect(); + + Ok(files) +} + +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| { + block.get(1)?; + + 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, + _ => 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 = ModuleSpecifier::parse(&format!( + "{}${}-{}", + specifier, + file_line_index + line_offset + 1, + file_line_index + line_offset + line_count + 1, + )) + .unwrap(); + let file_specifier = + mapped_specifier_for_tsc(&file_specifier, file_media_type) + .map(|s| ModuleSpecifier::parse(&s).unwrap()) + .unwrap_or(file_specifier); + + Some(File { + specifier: file_specifier, + maybe_headers: None, + source: file_source.into_bytes().into(), + }) + }) + .collect(); + + Ok(files) +} + +#[derive(Default)] +struct ExportCollector { + named_exports: BTreeSet<Atom>, + default_export: Option<Atom>, +} + +impl ExportCollector { + fn to_import_specifiers( + &self, + symbols_to_exclude: &AHashSet<Atom>, + ) -> Vec<ast::ImportSpecifier> { + let mut import_specifiers = vec![]; + + if let Some(default_export) = &self.default_export { + if !symbols_to_exclude.contains(default_export) { + import_specifiers.push(ast::ImportSpecifier::Default( + ast::ImportDefaultSpecifier { + span: DUMMY_SP, + local: ast::Ident { + span: DUMMY_SP, + ctxt: Default::default(), + sym: default_export.clone(), + optional: false, + }, + }, + )); + } + } + + for named_export in &self.named_exports { + if symbols_to_exclude.contains(named_export) { + continue; + } + + import_specifiers.push(ast::ImportSpecifier::Named( + ast::ImportNamedSpecifier { + span: DUMMY_SP, + local: ast::Ident { + span: DUMMY_SP, + ctxt: Default::default(), + sym: named_export.clone(), + optional: false, + }, + imported: None, + is_type_only: false, + }, + )); + } + + import_specifiers + } +} + +impl Visit for ExportCollector { + fn visit_ts_module_decl(&mut self, ts_module_decl: &ast::TsModuleDecl) { + if ts_module_decl.declare { + return; + } + + ts_module_decl.visit_children_with(self); + } + + fn visit_export_decl(&mut self, export_decl: &ast::ExportDecl) { + match &export_decl.decl { + ast::Decl::Class(class) => { + self.named_exports.insert(class.ident.sym.clone()); + } + ast::Decl::Fn(func) => { + self.named_exports.insert(func.ident.sym.clone()); + } + ast::Decl::Var(var) => { + for var_decl in &var.decls { + let atoms = extract_sym_from_pat(&var_decl.name); + self.named_exports.extend(atoms); + } + } + ast::Decl::TsEnum(ts_enum) => { + self.named_exports.insert(ts_enum.id.sym.clone()); + } + ast::Decl::TsModule(ts_module) => { + if ts_module.declare { + return; + } + + match &ts_module.id { + ast::TsModuleName::Ident(ident) => { + self.named_exports.insert(ident.sym.clone()); + } + ast::TsModuleName::Str(s) => { + self.named_exports.insert(s.value.clone()); + } + } + } + ast::Decl::TsTypeAlias(ts_type_alias) => { + self.named_exports.insert(ts_type_alias.id.sym.clone()); + } + ast::Decl::TsInterface(ts_interface) => { + self.named_exports.insert(ts_interface.id.sym.clone()); + } + ast::Decl::Using(_) => {} + } + } + + fn visit_export_default_decl( + &mut self, + export_default_decl: &ast::ExportDefaultDecl, + ) { + match &export_default_decl.decl { + ast::DefaultDecl::Class(class) => { + if let Some(ident) = &class.ident { + self.default_export = Some(ident.sym.clone()); + } + } + ast::DefaultDecl::Fn(func) => { + if let Some(ident) = &func.ident { + self.default_export = Some(ident.sym.clone()); + } + } + ast::DefaultDecl::TsInterfaceDecl(_) => {} + } + } + + fn visit_export_named_specifier( + &mut self, + export_named_specifier: &ast::ExportNamedSpecifier, + ) { + fn get_atom(export_name: &ast::ModuleExportName) -> Atom { + match export_name { + ast::ModuleExportName::Ident(ident) => ident.sym.clone(), + ast::ModuleExportName::Str(s) => s.value.clone(), + } + } + + match &export_named_specifier.exported { + Some(exported) => { + self.named_exports.insert(get_atom(exported)); + } + None => { + self + .named_exports + .insert(get_atom(&export_named_specifier.orig)); + } + } + } + + fn visit_named_export(&mut self, named_export: &ast::NamedExport) { + // ExportCollector does not handle re-exports + if named_export.src.is_some() { + return; + } + + named_export.visit_children_with(self); + } +} + +fn extract_sym_from_pat(pat: &ast::Pat) -> Vec<Atom> { + fn rec(pat: &ast::Pat, atoms: &mut Vec<Atom>) { + match pat { + ast::Pat::Ident(binding_ident) => { + atoms.push(binding_ident.sym.clone()); + } + ast::Pat::Array(array_pat) => { + for elem in array_pat.elems.iter().flatten() { + rec(elem, atoms); + } + } + ast::Pat::Rest(rest_pat) => { + rec(&rest_pat.arg, atoms); + } + ast::Pat::Object(object_pat) => { + for prop in &object_pat.props { + match prop { + ast::ObjectPatProp::Assign(assign_pat_prop) => { + atoms.push(assign_pat_prop.key.sym.clone()); + } + ast::ObjectPatProp::KeyValue(key_value_pat_prop) => { + rec(&key_value_pat_prop.value, atoms); + } + ast::ObjectPatProp::Rest(rest_pat) => { + rec(&rest_pat.arg, atoms); + } + } + } + } + ast::Pat::Assign(assign_pat) => { + rec(&assign_pat.left, atoms); + } + ast::Pat::Invalid(_) | ast::Pat::Expr(_) => {} + } + } + + let mut atoms = vec![]; + rec(pat, &mut atoms); + atoms +} + +/// Generates a "pseudo" file from a given file by applying the following +/// transformations: +/// +/// 1. Injects `import` statements for expoted items from the base file +/// 2. If `wrap_kind` is [`WrapKind::DenoTest`], wraps the content of the file +/// in a `Deno.test` call. +/// +/// For example, given a file that looks like: +/// +/// ```ts +/// import { assertEquals } from "@std/assert/equals"; +/// +/// assertEquals(increment(1), 2); +/// ``` +/// +/// and the base file (from which the above snippet was extracted): +/// +/// ```ts +/// export function increment(n: number): number { +/// return n + 1; +/// } +/// +/// export const SOME_CONST = "HELLO"; +/// ``` +/// +/// The generated pseudo test file would look like (if `wrap_in_deno_test` is enabled): +/// +/// ```ts +/// import { assertEquals } from "@std/assert/equals"; +/// import { increment, SOME_CONST } from "./base.ts"; +/// +/// Deno.test("./base.ts$1-3.ts", async () => { +/// assertEquals(increment(1), 2); +/// }); +/// ``` +/// +/// # Edge case - duplicate identifier +/// +/// If a given file imports, say, `doSomething` from an external module while +/// the base file exports `doSomething` as well, the generated pseudo test file +/// would end up having two duplciate imports for `doSomething`, causing the +/// duplicate identifier error. +/// +/// To avoid this issue, when a given file imports `doSomething`, this takes +/// precedence over the automatic import injection for the base file's +/// `doSomething`. So the generated pseudo test file would look like: +/// +/// ```ts +/// import { assertEquals } from "@std/assert/equals"; +/// import { doSomething } from "./some_external_module.ts"; +/// +/// Deno.test("./base.ts$1-3.ts", async () => { +/// assertEquals(doSomething(1), 2); +/// }); +/// ``` +fn generate_pseudo_file( + file: File, + base_file_specifier: &ModuleSpecifier, + exports: &ExportCollector, + wrap_kind: WrapKind, +) -> Result<File, AnyError> { + let file = file.into_text_decoded()?; + + let parsed = deno_ast::parse_program(deno_ast::ParseParams { + specifier: file.specifier.clone(), + text: file.source, + media_type: file.media_type, + capture_tokens: false, + scope_analysis: true, + maybe_syntax: None, + })?; + + let top_level_atoms = swc_utils::collect_decls_with_ctxt::<Atom, _>( + parsed.program_ref(), + parsed.top_level_context(), + ); + + let transformed = + parsed + .program_ref() + .clone() + .fold_with(&mut as_folder(Transform { + specifier: &file.specifier, + base_file_specifier, + exports_from_base: exports, + atoms_to_be_excluded_from_import: top_level_atoms, + wrap_kind, + })); + + let source = deno_ast::swc::codegen::to_code(&transformed); + + log::debug!("{}:\n{}", file.specifier, source); + + Ok(File { + specifier: file.specifier, + maybe_headers: None, + source: source.into_bytes().into(), + }) +} + +struct Transform<'a> { + specifier: &'a ModuleSpecifier, + base_file_specifier: &'a ModuleSpecifier, + exports_from_base: &'a ExportCollector, + atoms_to_be_excluded_from_import: AHashSet<Atom>, + wrap_kind: WrapKind, +} + +impl<'a> VisitMut for Transform<'a> { + fn visit_mut_program(&mut self, node: &mut ast::Program) { + let new_module_items = match node { + ast::Program::Module(module) => { + let mut module_decls = vec![]; + let mut stmts = vec![]; + + for item in &module.body { + match item { + ast::ModuleItem::ModuleDecl(decl) => { + module_decls.push(decl.clone()); + } + ast::ModuleItem::Stmt(stmt) => { + stmts.push(stmt.clone()); + } + } + } + + let mut transformed_items = vec![]; + transformed_items + .extend(module_decls.into_iter().map(ast::ModuleItem::ModuleDecl)); + let import_specifiers = self + .exports_from_base + .to_import_specifiers(&self.atoms_to_be_excluded_from_import); + if !import_specifiers.is_empty() { + transformed_items.push(ast::ModuleItem::ModuleDecl( + ast::ModuleDecl::Import(ast::ImportDecl { + span: DUMMY_SP, + specifiers: import_specifiers, + src: Box::new(ast::Str { + span: DUMMY_SP, + value: self.base_file_specifier.to_string().into(), + raw: None, + }), + type_only: false, + with: None, + phase: ast::ImportPhase::Evaluation, + }), + )); + } + match self.wrap_kind { + WrapKind::DenoTest => { + transformed_items.push(ast::ModuleItem::Stmt(wrap_in_deno_test( + stmts, + self.specifier.to_string().into(), + ))); + } + WrapKind::NoWrap => { + transformed_items + .extend(stmts.into_iter().map(ast::ModuleItem::Stmt)); + } + } + + transformed_items + } + ast::Program::Script(script) => { + let mut transformed_items = vec![]; + + let import_specifiers = self + .exports_from_base + .to_import_specifiers(&self.atoms_to_be_excluded_from_import); + if !import_specifiers.is_empty() { + transformed_items.push(ast::ModuleItem::ModuleDecl( + ast::ModuleDecl::Import(ast::ImportDecl { + span: DUMMY_SP, + specifiers: import_specifiers, + src: Box::new(ast::Str { + span: DUMMY_SP, + value: self.base_file_specifier.to_string().into(), + raw: None, + }), + type_only: false, + with: None, + phase: ast::ImportPhase::Evaluation, + }), + )); + } + + match self.wrap_kind { + WrapKind::DenoTest => { + transformed_items.push(ast::ModuleItem::Stmt(wrap_in_deno_test( + script.body.clone(), + self.specifier.to_string().into(), + ))); + } + WrapKind::NoWrap => { + transformed_items.extend( + script.body.clone().into_iter().map(ast::ModuleItem::Stmt), + ); + } + } + + transformed_items + } + }; + + *node = ast::Program::Module(ast::Module { + span: DUMMY_SP, + body: new_module_items, + shebang: None, + }); + } +} + +fn wrap_in_deno_test(stmts: Vec<ast::Stmt>, test_name: Atom) -> ast::Stmt { + ast::Stmt::Expr(ast::ExprStmt { + span: DUMMY_SP, + expr: Box::new(ast::Expr::Call(ast::CallExpr { + span: DUMMY_SP, + callee: ast::Callee::Expr(Box::new(ast::Expr::Member(ast::MemberExpr { + span: DUMMY_SP, + obj: Box::new(ast::Expr::Ident(ast::Ident { + span: DUMMY_SP, + sym: "Deno".into(), + optional: false, + ..Default::default() + })), + prop: ast::MemberProp::Ident(ast::IdentName { + span: DUMMY_SP, + sym: "test".into(), + }), + }))), + args: vec![ + ast::ExprOrSpread { + spread: None, + expr: Box::new(ast::Expr::Lit(ast::Lit::Str(ast::Str { + span: DUMMY_SP, + value: test_name, + raw: None, + }))), + }, + ast::ExprOrSpread { + spread: None, + expr: Box::new(ast::Expr::Arrow(ast::ArrowExpr { + span: DUMMY_SP, + params: vec![], + body: Box::new(ast::BlockStmtOrExpr::BlockStmt(ast::BlockStmt { + span: DUMMY_SP, + stmts, + ..Default::default() + })), + is_async: true, + is_generator: false, + type_params: None, + return_type: None, + ..Default::default() + })), + }, + ], + type_args: None, + ..Default::default() + })), + }) +} + +#[cfg(test)] +mod tests { + use super::*; + use crate::file_fetcher::TextDecodedFile; + use deno_ast::swc::atoms::Atom; + use pretty_assertions::assert_eq; + + #[test] + fn test_extract_doc_tests() { + struct Input { + source: &'static str, + specifier: &'static str, + } + struct Expected { + source: &'static str, + specifier: &'static str, + media_type: MediaType, + } + struct Test { + input: Input, + expected: Vec<Expected>, + } + + let tests = [ + Test { + input: Input { + source: r#""#, + specifier: "file:///main.ts", + }, + expected: vec![], + }, + Test { + input: Input { + source: r#" +/** + * ```ts + * import { assertEquals } from "@std/assert/equal"; + * + * assertEquals(add(1, 2), 3); + * ``` + */ +export function add(a: number, b: number): number { + return a + b; +} +"#, + specifier: "file:///main.ts", + }, + expected: vec![Expected { + source: r#"import { assertEquals } from "@std/assert/equal"; +import { add } from "file:///main.ts"; +Deno.test("file:///main.ts$3-8.ts", async ()=>{ + assertEquals(add(1, 2), 3); +}); +"#, + specifier: "file:///main.ts$3-8.ts", + media_type: MediaType::TypeScript, + }], + }, + Test { + input: Input { + source: r#" +/** + * ```ts + * foo(); + * ``` + */ +export function foo() {} + +export default class Bar {} +"#, + specifier: "file:///main.ts", + }, + expected: vec![Expected { + source: r#"import Bar, { foo } from "file:///main.ts"; +Deno.test("file:///main.ts$3-6.ts", async ()=>{ + foo(); +}); +"#, + specifier: "file:///main.ts$3-6.ts", + media_type: MediaType::TypeScript, + }], + }, + Test { + input: Input { + source: r#" +/** + * ```ts + * const input = { a: 42 } satisfies Args; + * foo(input); + * ``` + */ +export function foo(args: Args) {} + +export type Args = { a: number }; +"#, + specifier: "file:///main.ts", + }, + expected: vec![Expected { + source: r#"import { Args, foo } from "file:///main.ts"; +Deno.test("file:///main.ts$3-7.ts", async ()=>{ + const input = { + a: 42 + } satisfies Args; + foo(input); +}); +"#, + specifier: "file:///main.ts$3-7.ts", + media_type: MediaType::TypeScript, + }], + }, + Test { + input: Input { + source: r#" +/** + * This is a module-level doc. + * + * ```ts + * foo(); + * ``` + * + * @module doc + */ +"#, + specifier: "file:///main.ts", + }, + expected: vec![Expected { + source: r#"Deno.test("file:///main.ts$5-8.ts", async ()=>{ + foo(); +}); +"#, + specifier: "file:///main.ts$5-8.ts", + media_type: MediaType::TypeScript, + }], + }, + Test { + input: Input { + source: r#" +/** + * This is a module-level doc. + * + * ```js + * const cls = new MyClass(); + * ``` + * + * @module doc + */ + +/** + * ```ts + * foo(); + * ``` + */ +export function foo() {} + +export default class MyClass {} + +export * from "./other.ts"; +"#, + specifier: "file:///main.ts", + }, + expected: vec![ + Expected { + source: r#"import MyClass, { foo } from "file:///main.ts"; +Deno.test("file:///main.ts$5-8.js", async ()=>{ + const cls = new MyClass(); +}); +"#, + specifier: "file:///main.ts$5-8.js", + media_type: MediaType::JavaScript, + }, + Expected { + source: r#"import MyClass, { foo } from "file:///main.ts"; +Deno.test("file:///main.ts$13-16.ts", async ()=>{ + foo(); +}); +"#, + specifier: "file:///main.ts$13-16.ts", + media_type: MediaType::TypeScript, + }, + ], + }, + // Avoid duplicate imports + Test { + input: Input { + source: r#" +/** + * ```ts + * import { DUPLICATE1 } from "./other1.ts"; + * import * as DUPLICATE2 from "./other2.js"; + * import { foo as DUPLICATE3 } from "./other3.tsx"; + * + * foo(); + * ``` + */ +export function foo() {} + +export const DUPLICATE1 = "dup1"; +const DUPLICATE2 = "dup2"; +export default DUPLICATE2; +const DUPLICATE3 = "dup3"; +export { DUPLICATE3 }; +"#, + specifier: "file:///main.ts", + }, + expected: vec![Expected { + source: r#"import { DUPLICATE1 } from "./other1.ts"; +import * as DUPLICATE2 from "./other2.js"; +import { foo as DUPLICATE3 } from "./other3.tsx"; +import { foo } from "file:///main.ts"; +Deno.test("file:///main.ts$3-10.ts", async ()=>{ + foo(); +}); +"#, + specifier: "file:///main.ts$3-10.ts", + media_type: MediaType::TypeScript, + }], + }, + // duplication of imported identifier and local identifier is fine + Test { + input: Input { + source: r#" +/** + * ```ts + * const foo = createFoo(); + * foo(); + * ``` + */ +export function createFoo() { + return () => "created foo"; +} + +export const foo = () => "foo"; +"#, + specifier: "file:///main.ts", + }, + expected: vec![Expected { + source: r#"import { createFoo } from "file:///main.ts"; +Deno.test("file:///main.ts$3-7.ts", async ()=>{ + const foo = createFoo(); + foo(); +}); +"#, + specifier: "file:///main.ts$3-7.ts", + media_type: MediaType::TypeScript, + }], + }, + // example code has an exported item `foo` - because `export` must be at + // the top level, `foo` is "hoisted" to the top level instead of being + // wrapped in `Deno.test`. + Test { + input: Input { + source: r#" +/** + * ```ts + * doSomething(); + * export const foo = 42; + * ``` + */ +export function doSomething() {} +"#, + specifier: "file:///main.ts", + }, + expected: vec![Expected { + source: r#"export const foo = 42; +import { doSomething } from "file:///main.ts"; +Deno.test("file:///main.ts$3-7.ts", async ()=>{ + doSomething(); +}); +"#, + specifier: "file:///main.ts$3-7.ts", + media_type: MediaType::TypeScript, + }], + }, + Test { + input: Input { + source: r#" +# Header + +This is a *markdown*. + +```js +import { assertEquals } from "@std/assert/equal"; +import { add } from "jsr:@deno/non-existent"; + +assertEquals(add(1, 2), 3); +``` +"#, + specifier: "file:///README.md", + }, + expected: vec![Expected { + source: r#"import { assertEquals } from "@std/assert/equal"; +import { add } from "jsr:@deno/non-existent"; +Deno.test("file:///README.md$6-12.js", async ()=>{ + assertEquals(add(1, 2), 3); +}); +"#, + specifier: "file:///README.md$6-12.js", + media_type: MediaType::JavaScript, + }], + }, + ]; + + for test in tests { + let file = File { + specifier: ModuleSpecifier::parse(test.input.specifier).unwrap(), + maybe_headers: None, + source: test.input.source.as_bytes().into(), + }; + let got_decoded = extract_doc_tests(file) + .unwrap() + .into_iter() + .map(|f| f.into_text_decoded().unwrap()) + .collect::<Vec<_>>(); + let expected = test + .expected + .iter() + .map(|e| TextDecodedFile { + specifier: ModuleSpecifier::parse(e.specifier).unwrap(), + media_type: e.media_type, + source: e.source.into(), + }) + .collect::<Vec<_>>(); + assert_eq!(got_decoded, expected); + } + } + + #[test] + fn test_extract_snippet_files() { + struct Input { + source: &'static str, + specifier: &'static str, + } + struct Expected { + source: &'static str, + specifier: &'static str, + media_type: MediaType, + } + struct Test { + input: Input, + expected: Vec<Expected>, + } + + let tests = [ + Test { + input: Input { + source: r#""#, + specifier: "file:///main.ts", + }, + expected: vec![], + }, + Test { + input: Input { + source: r#" +/** + * ```ts + * import { assertEquals } from "@std/assert/equals"; + * + * assertEquals(add(1, 2), 3); + * ``` + */ +export function add(a: number, b: number): number { + return a + b; +} +"#, + specifier: "file:///main.ts", + }, + expected: vec![Expected { + source: r#"import { assertEquals } from "@std/assert/equals"; +import { add } from "file:///main.ts"; +assertEquals(add(1, 2), 3); +"#, + specifier: "file:///main.ts$3-8.ts", + media_type: MediaType::TypeScript, + }], + }, + Test { + input: Input { + source: r#" +/** + * ```ts + * import { assertEquals } from "@std/assert/equals"; + * import { DUPLICATE } from "./other.ts"; + * + * assertEquals(add(1, 2), 3); + * ``` + */ +export function add(a: number, b: number): number { + return a + b; +} + +export const DUPLICATE = "dup"; +"#, + specifier: "file:///main.ts", + }, + expected: vec![Expected { + source: r#"import { assertEquals } from "@std/assert/equals"; +import { DUPLICATE } from "./other.ts"; +import { add } from "file:///main.ts"; +assertEquals(add(1, 2), 3); +"#, + specifier: "file:///main.ts$3-9.ts", + media_type: MediaType::TypeScript, + }], + }, + // duplication of imported identifier and local identifier is fine, since + // we wrap the snippet in a block. + // This would be a problem if the local one is declared with `var`, as + // `var` is not block scoped but function scoped. For now we don't handle + // this case assuming that `var` is not used in modern code. + Test { + input: Input { + source: r#" + /** + * ```ts + * const foo = createFoo(); + * foo(); + * ``` + */ + export function createFoo() { + return () => "created foo"; + } + + export const foo = () => "foo"; + "#, + specifier: "file:///main.ts", + }, + expected: vec![Expected { + source: r#"import { createFoo } from "file:///main.ts"; +const foo = createFoo(); +foo(); +"#, + specifier: "file:///main.ts$3-7.ts", + media_type: MediaType::TypeScript, + }], + }, + Test { + input: Input { + source: r#" +# Header + +This is a *markdown*. + +```js +import { assertEquals } from "@std/assert/equal"; +import { add } from "jsr:@deno/non-existent"; + +assertEquals(add(1, 2), 3); +``` +"#, + specifier: "file:///README.md", + }, + expected: vec![Expected { + source: r#"import { assertEquals } from "@std/assert/equal"; +import { add } from "jsr:@deno/non-existent"; +assertEquals(add(1, 2), 3); +"#, + specifier: "file:///README.md$6-12.js", + media_type: MediaType::JavaScript, + }], + }, + ]; + + for test in tests { + let file = File { + specifier: ModuleSpecifier::parse(test.input.specifier).unwrap(), + maybe_headers: None, + source: test.input.source.as_bytes().into(), + }; + let got_decoded = extract_snippet_files(file) + .unwrap() + .into_iter() + .map(|f| f.into_text_decoded().unwrap()) + .collect::<Vec<_>>(); + let expected = test + .expected + .iter() + .map(|e| TextDecodedFile { + specifier: ModuleSpecifier::parse(e.specifier).unwrap(), + media_type: e.media_type, + source: e.source.into(), + }) + .collect::<Vec<_>>(); + assert_eq!(got_decoded, expected); + } + } + + #[test] + fn test_export_collector() { + fn helper(input: &'static str) -> ExportCollector { + let mut collector = ExportCollector::default(); + let parsed = deno_ast::parse_module(deno_ast::ParseParams { + specifier: deno_ast::ModuleSpecifier::parse("file:///main.ts").unwrap(), + text: input.into(), + media_type: deno_ast::MediaType::TypeScript, + capture_tokens: false, + scope_analysis: false, + maybe_syntax: None, + }) + .unwrap(); + + collector.visit_program(parsed.program_ref()); + collector + } + + struct Test { + input: &'static str, + named_expected: BTreeSet<Atom>, + default_expected: Option<Atom>, + } + + macro_rules! atom_set { + ($( $x:expr ),*) => { + [$( Atom::from($x) ),*].into_iter().collect::<BTreeSet<_>>() + }; + } + + let tests = [ + Test { + input: r#"export const foo = 42;"#, + named_expected: atom_set!("foo"), + default_expected: None, + }, + Test { + input: r#"export let foo = 42;"#, + named_expected: atom_set!("foo"), + default_expected: None, + }, + Test { + input: r#"export var foo = 42;"#, + named_expected: atom_set!("foo"), + default_expected: None, + }, + Test { + input: r#"export const foo = () => {};"#, + named_expected: atom_set!("foo"), + default_expected: None, + }, + Test { + input: r#"export function foo() {}"#, + named_expected: atom_set!("foo"), + default_expected: None, + }, + Test { + input: r#"export class Foo {}"#, + named_expected: atom_set!("Foo"), + default_expected: None, + }, + Test { + input: r#"export enum Foo {}"#, + named_expected: atom_set!("Foo"), + default_expected: None, + }, + Test { + input: r#"export module Foo {}"#, + named_expected: atom_set!("Foo"), + default_expected: None, + }, + Test { + input: r#"export module "foo" {}"#, + named_expected: atom_set!("foo"), + default_expected: None, + }, + Test { + input: r#"export namespace Foo {}"#, + named_expected: atom_set!("Foo"), + default_expected: None, + }, + Test { + input: r#"export type Foo = string;"#, + named_expected: atom_set!("Foo"), + default_expected: None, + }, + Test { + input: r#"export interface Foo {};"#, + named_expected: atom_set!("Foo"), + default_expected: None, + }, + Test { + input: r#"export let name1, name2;"#, + named_expected: atom_set!("name1", "name2"), + default_expected: None, + }, + Test { + input: r#"export const name1 = 1, name2 = 2;"#, + named_expected: atom_set!("name1", "name2"), + default_expected: None, + }, + Test { + input: r#"export function* generatorFunc() {}"#, + named_expected: atom_set!("generatorFunc"), + default_expected: None, + }, + Test { + input: r#"export const { name1, name2: bar } = obj;"#, + named_expected: atom_set!("name1", "bar"), + default_expected: None, + }, + Test { + input: r#"export const [name1, name2] = arr;"#, + named_expected: atom_set!("name1", "name2"), + default_expected: None, + }, + Test { + input: r#"export const { name1 = 42 } = arr;"#, + named_expected: atom_set!("name1"), + default_expected: None, + }, + Test { + input: r#"export default function foo() {}"#, + named_expected: atom_set!(), + default_expected: Some("foo".into()), + }, + Test { + input: r#"export { foo, bar as barAlias };"#, + named_expected: atom_set!("foo", "barAlias"), + default_expected: None, + }, + Test { + input: r#" +export default class Foo {} +export let value1 = 42; +const value2 = "Hello"; +const value3 = "World"; +export { value2 }; +"#, + named_expected: atom_set!("value1", "value2"), + default_expected: Some("Foo".into()), + }, + // overloaded function + Test { + input: r#" +export function foo(a: number): boolean; +export function foo(a: boolean): string; +export function foo(a: number | boolean): boolean | string { + return typeof a === "number" ? true : "hello"; +} +"#, + named_expected: atom_set!("foo"), + default_expected: None, + }, + // The collector deliberately does not handle re-exports, because from + // doc reader's perspective, an example code would become hard to follow + // if it uses re-exported items (as opposed to normal, non-re-exported + // items that would look verbose if an example code explicitly imports + // them). + Test { + input: r#" +export * from "./module1.ts"; +export * as name1 from "./module2.ts"; +export { name2, name3 as N3 } from "./module3.js"; +export { default } from "./module4.ts"; +export { default as myDefault } from "./module5.ts"; +"#, + named_expected: atom_set!(), + default_expected: None, + }, + Test { + input: r#" +export namespace Foo { + export type MyType = string; + export const myValue = 42; + export function myFunc(): boolean; +} +"#, + named_expected: atom_set!("Foo"), + default_expected: None, + }, + Test { + input: r#" +declare namespace Foo { + export type MyType = string; + export const myValue = 42; + export function myFunc(): boolean; +} +"#, + named_expected: atom_set!(), + default_expected: None, + }, + Test { + input: r#" +declare module Foo { + export type MyType = string; + export const myValue = 42; + export function myFunc(): boolean; +} +"#, + named_expected: atom_set!(), + default_expected: None, + }, + Test { + input: r#" +declare global { + export type MyType = string; + export const myValue = 42; + export function myFunc(): boolean; +} +"#, + named_expected: atom_set!(), + default_expected: None, + }, + ]; + + for test in tests { + let got = helper(test.input); + assert_eq!(got.named_exports, test.named_expected); + assert_eq!(got.default_export, test.default_expected); + } + } +} diff --git a/cli/util/mod.rs b/cli/util/mod.rs index b9071c496..e59b09d2c 100644 --- a/cli/util/mod.rs +++ b/cli/util/mod.rs @@ -7,6 +7,7 @@ pub mod console; pub mod diff; pub mod display; pub mod draw_thread; +pub mod extract; pub mod file_watcher; pub mod fs; pub mod logger; |