857 lines
27 KiB
Rust
Raw Normal View History

#[macro_use]
extern crate lazy_static;
extern crate wasm_bindgen_cli_support as cli;
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
use std::env;
use std::fs::{self, File};
use std::thread;
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
use std::io::{self, Read, Write};
use std::path::{Path, PathBuf};
use std::process::{Command, Stdio, Child, ChildStdin};
use std::net::TcpStream;
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
use std::sync::atomic::*;
use std::sync::{Once, ONCE_INIT, Mutex};
use std::time::{Duration, Instant};
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
static CNT: AtomicUsize = ATOMIC_USIZE_INIT;
thread_local!(static IDX: usize = CNT.fetch_add(1, Ordering::SeqCst));
Create the `web-sys` crate mechanically from WebIDL (#409) * Create a new `web-sys` crate This will eventually contain all the WebIDL-generated bindings to Web APIs. * ci: Test the new `web-sys` crate in CI * web-sys: Add a small README * web-sys: Vendor all the WebIDL files from mozilla-central * backend: Add a pass to remove AST items that use undefined imports This is necessary for the WebIDL frontend, which can't translate many WebIDL constructs into equivalent wasm-bindgen AST things yet. It lets us make incremental progress: we can generate bindings to methods we can support right now even though there might be methods on the same interface that we can't support yet. * webidl: Add a bunch of missing semicolons * webidl: Make parsing private It was only `pub` so that we could test it, but we ended up moving towards integration tests rather than unit tests that assert particular ASTs are parsed from WebIDL files. * webidl: Remove uses of undefined import types * test-project-builder: Build projects in "very verbose" mode This helps for debugging failing WebIDL-related tests. * test-project-builder: Add more profiling timers * test-project-builder: Detect when webpack-dev-server fails Instead of going into an infinite loop, detect when webpack-dev-server fails to start up and early exit the test. * webidl: Specify version for dev-dependency on wasm-bindgen-backend Instead of only a relative path. * guide: Add section about contributing to `web-sys` * WIP enable Event.webidl Still need to fix and finish the test. * Update expected webidl output * Start out a test's status as incomplete That way if we don't fill it in the error message doesn't look quite so bizarre * Fix onerror function in headless mode Otherwise we don't see any output! * Fix package.json/node_modules handling in project generation Make sure these are looked up in the git project root rather than the crate root * Avoid logging body text This was meant for debugging and is otherwise pretty noisy * Fix a relative path * More expected test fixes * Fix a typo * test-project-builder: Allow asynchronous tests * webidl: Convert [Unforgeable] attributes into `#[wasm_bindgen(structural)]` Fixes #432 * test-project-builder: Print generated WebIDL bindings for debugging purposes Helps debug bad WebIDL bindings generation inside tests. * When we can't find a descriptor, say which one can't be found This helps when debugging things that need to become structural. * web-sys: Test bindings for Event * ci: Use `--manifest-path dir` instead of `cd dir && ...` * web-sys: Just move .webidl files isntead of symlinking to enable them * tests: Polyfill Array.prototype.values for older browsers in CI * test-project-builder: Don't panic on poisoned headless test mutex We only use it to serialize headless tests so that we don't try to bind the port concurrently. Its OK to run another headless test if an earlier one panicked. * JsValue: Add {is,as}_{object,function} methods Allows dynamically casting values to `js::Object` and `js::Function`. * tidy: Fix whitespace and missing semicolons * Allow for dynamic feature detection of methods If we create bindings to a method that doesn't exist in this implementation, then it shouldn't fail until if/when we actually try and invoke that missing method. * tests: Do feature detection in Array.prototype.values test * Add JsValue::{is_string, as_js_string} methods And document all the cast/convert/check methods for js value. * eslint: allow backtick string literals * Only generate a fallback import function for non-structural imports
2018-07-09 16:35:25 -07:00
#[derive(Clone)]
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
pub struct Project {
files: Vec<(String, String)>,
debug: bool,
node: bool,
nodejs_experimental_modules: bool,
no_std: bool,
serde: bool,
rlib: bool,
webpack: bool,
node_args: Vec<String>,
deps: Vec<String>,
headless: bool,
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
}
pub fn project() -> Project {
let dir = Path::new(env!("CARGO_MANIFEST_DIR")).join("../..");
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
let mut lockfile = String::new();
fs::File::open(&dir.join("Cargo.lock"))
.unwrap()
.read_to_string(&mut lockfile)
.unwrap();
Project {
debug: true,
no_std: false,
node: true,
nodejs_experimental_modules: true,
webpack: false,
serde: false,
rlib: false,
headless: false,
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
deps: Vec::new(),
node_args: Vec::new(),
files: vec![
("Cargo.lock".to_string(), lockfile),
],
}
}
fn root() -> PathBuf {
let idx = IDX.with(|x| *x);
let mut me = env::current_exe().unwrap();
me.pop(); // chop off exe name
me.pop(); // chop off `deps`
me.pop(); // chop off `debug` / `release`
me.push("generated-tests");
me.push(&format!("test{}", idx));
return me;
}
fn assert_bigint_support() -> Option<&'static str> {
static BIGINT_SUPPORED: AtomicUsize = ATOMIC_USIZE_INIT;
static INIT: Once = ONCE_INIT;
INIT.call_once(|| {
let mut cmd = Command::new("node");
cmd.arg("-e").arg("BigInt");
cmd.stdin(Stdio::null())
.stdout(Stdio::piped())
.stderr(Stdio::piped());
if cmd.status().unwrap().success() {
BIGINT_SUPPORED.store(1, Ordering::SeqCst);
return;
}
cmd.arg("--harmony-bigint");
if cmd.status().unwrap().success() {
BIGINT_SUPPORED.store(2, Ordering::SeqCst);
return;
}
});
match BIGINT_SUPPORED.load(Ordering::SeqCst) {
1 => return None,
2 => return Some("--harmony-bigint"),
_ => panic!(
"the version of node.js that is installed for these tests \
does not support `BigInt`, you may wish to try installing \
node 10 to fix this"
),
}
}
impl Project {
/// Add a new file with the specified contents to this project, the `name`
/// can have slahes for files in subdirectories.
pub fn file(&mut self, name: &str, contents: &str) -> &mut Project {
self.files.push((name.to_string(), contents.to_string()));
self
}
/// Enable debug mode in wasm-bindgen for this test
pub fn debug(&mut self, debug: bool) -> &mut Project {
self.debug = debug;
self
}
/// Depend on `wasm-bindgen` without the `std` feature enabled.
pub fn no_std(&mut self, no_std: bool) -> &mut Project {
self.no_std = no_std;
self
}
/// Depend on the `serde` feature of `wasm-bindgen`
pub fn serde(&mut self, serde: bool) -> &mut Project {
self.serde = serde;
self
}
/// Generate an rlib instead of a cdylib in the generated Cargo project
pub fn rlib(&mut self, rlib: bool) -> &mut Project {
self.rlib = rlib;
self
}
/// Depend on a crate from crates.io, like serde.
pub fn depend(&mut self, dep: &str) -> &mut Project {
self.deps.push(dep.to_string());
self
}
/// Enables or disables node.js experimental modules output
pub fn nodejs_experimental_modules(&mut self, node: bool) -> &mut Project {
self.nodejs_experimental_modules = node;
self
}
/// Enables or disables the usage of webpack for this project
pub fn webpack(&mut self, webpack: bool) -> &mut Project {
self.webpack = webpack;
self
}
/// Add a path dependency to the generated project
pub fn add_local_dependency(&mut self, name: &str, path: &str) -> &mut Project {
self.deps
.push(format!("{} = {{ path = '{}' }}", name, path));
self
}
/// Returns the crate name that will be used for the generated crate, this
/// name changes between test runs and is generated at runtime.
pub fn crate_name(&self) -> String {
format!("test{}", IDX.with(|x| *x))
}
/// Flag this project as requiring bigint support in Node
pub fn requires_bigint(&mut self) -> &mut Project {
if let Some(arg) = assert_bigint_support() {
self.node_args.push(arg.to_string());
}
self
}
/// This test requires a headless web browser
pub fn headless(&mut self, headless: bool) -> &mut Project {
self.headless = headless;
self
}
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
/// Write this project to the filesystem, ensuring all files are ready to
/// go.
pub fn build(&mut self) -> (PathBuf, PathBuf) {
if self.headless {
self.webpack = true;
}
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
if self.webpack {
self.node = false;
self.nodejs_experimental_modules = false;
}
self.ensure_webpack_config();
self.ensure_test_entry();
if self.headless {
self.ensure_index_html();
self.ensure_run_headless_js();
}
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
let webidl_modules = self.generate_webidl_bindings();
self.generate_js_entry(webidl_modules);
let mut manifest = format!(
r#"
[package]
name = "test{}"
version = "0.0.1"
authors = []
[workspace]
[lib]
"#,
IDX.with(|x| *x)
);
if !self.rlib {
manifest.push_str("crate-type = [\"cdylib\"]\n");
}
manifest.push_str("[build-dependencies]\n");
manifest.push_str("wasm-bindgen-webidl = { path = '");
manifest.push_str(env!("CARGO_MANIFEST_DIR"));
manifest.push_str("/../webidl' }\n");
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
manifest.push_str("[dependencies]\n");
for dep in self.deps.iter() {
manifest.push_str(dep);
manifest.push_str("\n");
}
manifest.push_str("wasm-bindgen = { path = '");
manifest.push_str(env!("CARGO_MANIFEST_DIR"));
manifest.push_str("/../..'");
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
if self.no_std {
manifest.push_str(", default-features = false");
}
if self.serde {
manifest.push_str(", features = ['serde-serialize']");
}
manifest.push_str(" }\n");
self.files.push(("Cargo.toml".to_string(), manifest));
let root = root();
drop(fs::remove_dir_all(&root));
for &(ref file, ref contents) in self.files.iter() {
let mut dst = root.join(file);
if self.nodejs_experimental_modules &&
dst.extension().and_then(|s| s.to_str()) == Some("js")
{
dst = dst.with_extension("mjs");
}
if dst.extension().and_then(|s| s.to_str()) == Some("ts") &&
!self.webpack
{
panic!("webpack needs to be enabled to use typescript");
}
fs::create_dir_all(dst.parent().unwrap()).unwrap();
fs::File::create(&dst)
.unwrap()
.write_all(contents.as_ref())
.unwrap();
}
let target_dir = root.parent().unwrap() // chop off test name
.parent().unwrap(); // chop off `generated-tests`
(root.clone(), target_dir.to_path_buf())
}
fn ensure_webpack_config(&mut self) {
if !self.webpack {
return
}
let needs_typescript = self.files.iter().any(|t| t.0.ends_with(".ts"));
let mut rules = String::new();
let mut extensions = format!("'.js', '.wasm'");
if needs_typescript {
rules.push_str("
{
test: /.ts$/,
use: 'ts-loader',
exclude: /node_modules/,
}
");
extensions.push_str(", '.ts'");
}
let target = if self.headless { "web" } else { "node" };
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
self.files.push((
"webpack.config.js".to_string(),
format!(r#"
const path = require('path');
const fs = require('fs');
let nodeModules = {{}};
// Webpack bundles the modules from node_modules.
// For node target, we will not have `fs` module
// inside the `node_modules` folder.
// This reads the directories in `node_modules`
// and give that to externals and webpack ignores
// to bundle the modules listed as external.
if ('{2}' == 'node') {{
fs.readdirSync('node_modules')
.filter(module => module !== '.bin')
.forEach(mod => {{
// External however,expects browser environment.
// To make it work in `node` target we
// prefix commonjs here.
nodeModules[mod] = 'commonjs ' + mod;
}});
}}
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
module.exports = {{
entry: './run.js',
mode: "development",
devtool: "source-map",
module: {{
rules: [{}]
}},
resolve: {{
extensions: [{}]
}},
output: {{
filename: 'bundle.js',
path: path.resolve(__dirname, '.')
}},
target: '{}',
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
externals: nodeModules
}};
"#, rules, extensions, target)
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
));
if needs_typescript {
self.files.push((
"tsconfig.json".to_string(),
r#"
{
"compilerOptions": {
"noEmitOnError": true,
"noImplicitAny": true,
"noImplicitThis": true,
"noUnusedParameters": true,
"noUnusedLocals": true,
"noImplicitReturns": true,
"strictFunctionTypes": true,
"strictNullChecks": true,
"alwaysStrict": true,
"strict": true,
"target": "es5",
"lib": ["es2015"]
}
}
"#.to_string(),
));
}
}
fn ensure_test_entry(&mut self) {
if !self
.files
.iter()
.any(|(path, _)| path == "test.ts" || path == "test.js")
{
self.files.push((
"test.js".to_string(),
r#"export {test} from './out';"#.to_string(),
));
}
}
fn ensure_index_html(&mut self) {
self.file(
"index.html",
r#"
<!DOCTYPE html>
<html>
<body>
<div id="error"></div>
<div id="logs"></div>
<div id="status">incomplete</div>
<script src="bundle.js"></script>
</body>
</html>
"#,
);
}
fn ensure_run_headless_js(&mut self) {
self.file("run-headless.js", include_str!("run-headless.js"));
}
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
fn generate_webidl_bindings(&mut self) -> Vec<PathBuf> {
let mut res = Vec::new();
let mut origpaths = Vec::new();
for (path, _) in &self.files {
let path = Path::new(&path);
let extension = path.extension().map(|x| x.to_str().unwrap());
if extension != Some("webidl") {
continue;
}
res.push(path.with_extension("rs"));
origpaths.push(path.to_owned());
}
if res.is_empty() {
return res;
}
let mut buildrs = r#"
extern crate wasm_bindgen_webidl;
use wasm_bindgen_webidl::compile_file;
use std::env;
use std::fs::{self, File};
use std::io::Write;
use std::path::Path;
fn main() {
let dest = env::var("OUT_DIR").unwrap();
"#.to_string();
for (path, origpath) in res.iter().zip(origpaths.iter()) {
buildrs.push_str(&format!(
r#"
fs::create_dir_all("{}").unwrap();
Create the `web-sys` crate mechanically from WebIDL (#409) * Create a new `web-sys` crate This will eventually contain all the WebIDL-generated bindings to Web APIs. * ci: Test the new `web-sys` crate in CI * web-sys: Add a small README * web-sys: Vendor all the WebIDL files from mozilla-central * backend: Add a pass to remove AST items that use undefined imports This is necessary for the WebIDL frontend, which can't translate many WebIDL constructs into equivalent wasm-bindgen AST things yet. It lets us make incremental progress: we can generate bindings to methods we can support right now even though there might be methods on the same interface that we can't support yet. * webidl: Add a bunch of missing semicolons * webidl: Make parsing private It was only `pub` so that we could test it, but we ended up moving towards integration tests rather than unit tests that assert particular ASTs are parsed from WebIDL files. * webidl: Remove uses of undefined import types * test-project-builder: Build projects in "very verbose" mode This helps for debugging failing WebIDL-related tests. * test-project-builder: Add more profiling timers * test-project-builder: Detect when webpack-dev-server fails Instead of going into an infinite loop, detect when webpack-dev-server fails to start up and early exit the test. * webidl: Specify version for dev-dependency on wasm-bindgen-backend Instead of only a relative path. * guide: Add section about contributing to `web-sys` * WIP enable Event.webidl Still need to fix and finish the test. * Update expected webidl output * Start out a test's status as incomplete That way if we don't fill it in the error message doesn't look quite so bizarre * Fix onerror function in headless mode Otherwise we don't see any output! * Fix package.json/node_modules handling in project generation Make sure these are looked up in the git project root rather than the crate root * Avoid logging body text This was meant for debugging and is otherwise pretty noisy * Fix a relative path * More expected test fixes * Fix a typo * test-project-builder: Allow asynchronous tests * webidl: Convert [Unforgeable] attributes into `#[wasm_bindgen(structural)]` Fixes #432 * test-project-builder: Print generated WebIDL bindings for debugging purposes Helps debug bad WebIDL bindings generation inside tests. * When we can't find a descriptor, say which one can't be found This helps when debugging things that need to become structural. * web-sys: Test bindings for Event * ci: Use `--manifest-path dir` instead of `cd dir && ...` * web-sys: Just move .webidl files isntead of symlinking to enable them * tests: Polyfill Array.prototype.values for older browsers in CI * test-project-builder: Don't panic on poisoned headless test mutex We only use it to serialize headless tests so that we don't try to bind the port concurrently. Its OK to run another headless test if an earlier one panicked. * JsValue: Add {is,as}_{object,function} methods Allows dynamically casting values to `js::Object` and `js::Function`. * tidy: Fix whitespace and missing semicolons * Allow for dynamic feature detection of methods If we create bindings to a method that doesn't exist in this implementation, then it shouldn't fail until if/when we actually try and invoke that missing method. * tests: Do feature detection in Array.prototype.values test * Add JsValue::{is_string, as_js_string} methods And document all the cast/convert/check methods for js value. * eslint: allow backtick string literals * Only generate a fallback import function for non-structural imports
2018-07-09 16:35:25 -07:00
let bindings = compile_file(Path::new("{}")).expect("should compile OK");
println!("generated WebIDL bindings = '''\n{{}}\n'''", bindings);
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
File::create(&Path::new(&dest).join("{}"))
.unwrap()
Create the `web-sys` crate mechanically from WebIDL (#409) * Create a new `web-sys` crate This will eventually contain all the WebIDL-generated bindings to Web APIs. * ci: Test the new `web-sys` crate in CI * web-sys: Add a small README * web-sys: Vendor all the WebIDL files from mozilla-central * backend: Add a pass to remove AST items that use undefined imports This is necessary for the WebIDL frontend, which can't translate many WebIDL constructs into equivalent wasm-bindgen AST things yet. It lets us make incremental progress: we can generate bindings to methods we can support right now even though there might be methods on the same interface that we can't support yet. * webidl: Add a bunch of missing semicolons * webidl: Make parsing private It was only `pub` so that we could test it, but we ended up moving towards integration tests rather than unit tests that assert particular ASTs are parsed from WebIDL files. * webidl: Remove uses of undefined import types * test-project-builder: Build projects in "very verbose" mode This helps for debugging failing WebIDL-related tests. * test-project-builder: Add more profiling timers * test-project-builder: Detect when webpack-dev-server fails Instead of going into an infinite loop, detect when webpack-dev-server fails to start up and early exit the test. * webidl: Specify version for dev-dependency on wasm-bindgen-backend Instead of only a relative path. * guide: Add section about contributing to `web-sys` * WIP enable Event.webidl Still need to fix and finish the test. * Update expected webidl output * Start out a test's status as incomplete That way if we don't fill it in the error message doesn't look quite so bizarre * Fix onerror function in headless mode Otherwise we don't see any output! * Fix package.json/node_modules handling in project generation Make sure these are looked up in the git project root rather than the crate root * Avoid logging body text This was meant for debugging and is otherwise pretty noisy * Fix a relative path * More expected test fixes * Fix a typo * test-project-builder: Allow asynchronous tests * webidl: Convert [Unforgeable] attributes into `#[wasm_bindgen(structural)]` Fixes #432 * test-project-builder: Print generated WebIDL bindings for debugging purposes Helps debug bad WebIDL bindings generation inside tests. * When we can't find a descriptor, say which one can't be found This helps when debugging things that need to become structural. * web-sys: Test bindings for Event * ci: Use `--manifest-path dir` instead of `cd dir && ...` * web-sys: Just move .webidl files isntead of symlinking to enable them * tests: Polyfill Array.prototype.values for older browsers in CI * test-project-builder: Don't panic on poisoned headless test mutex We only use it to serialize headless tests so that we don't try to bind the port concurrently. Its OK to run another headless test if an earlier one panicked. * JsValue: Add {is,as}_{object,function} methods Allows dynamically casting values to `js::Object` and `js::Function`. * tidy: Fix whitespace and missing semicolons * Allow for dynamic feature detection of methods If we create bindings to a method that doesn't exist in this implementation, then it shouldn't fail until if/when we actually try and invoke that missing method. * tests: Do feature detection in Array.prototype.values test * Add JsValue::{is_string, as_js_string} methods And document all the cast/convert/check methods for js value. * eslint: allow backtick string literals * Only generate a fallback import function for non-structural imports
2018-07-09 16:35:25 -07:00
.write_all(bindings.as_bytes())
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
.unwrap();
"#,
path.parent().unwrap().to_str().unwrap(),
origpath.to_str().unwrap(),
Create the `web-sys` crate mechanically from WebIDL (#409) * Create a new `web-sys` crate This will eventually contain all the WebIDL-generated bindings to Web APIs. * ci: Test the new `web-sys` crate in CI * web-sys: Add a small README * web-sys: Vendor all the WebIDL files from mozilla-central * backend: Add a pass to remove AST items that use undefined imports This is necessary for the WebIDL frontend, which can't translate many WebIDL constructs into equivalent wasm-bindgen AST things yet. It lets us make incremental progress: we can generate bindings to methods we can support right now even though there might be methods on the same interface that we can't support yet. * webidl: Add a bunch of missing semicolons * webidl: Make parsing private It was only `pub` so that we could test it, but we ended up moving towards integration tests rather than unit tests that assert particular ASTs are parsed from WebIDL files. * webidl: Remove uses of undefined import types * test-project-builder: Build projects in "very verbose" mode This helps for debugging failing WebIDL-related tests. * test-project-builder: Add more profiling timers * test-project-builder: Detect when webpack-dev-server fails Instead of going into an infinite loop, detect when webpack-dev-server fails to start up and early exit the test. * webidl: Specify version for dev-dependency on wasm-bindgen-backend Instead of only a relative path. * guide: Add section about contributing to `web-sys` * WIP enable Event.webidl Still need to fix and finish the test. * Update expected webidl output * Start out a test's status as incomplete That way if we don't fill it in the error message doesn't look quite so bizarre * Fix onerror function in headless mode Otherwise we don't see any output! * Fix package.json/node_modules handling in project generation Make sure these are looked up in the git project root rather than the crate root * Avoid logging body text This was meant for debugging and is otherwise pretty noisy * Fix a relative path * More expected test fixes * Fix a typo * test-project-builder: Allow asynchronous tests * webidl: Convert [Unforgeable] attributes into `#[wasm_bindgen(structural)]` Fixes #432 * test-project-builder: Print generated WebIDL bindings for debugging purposes Helps debug bad WebIDL bindings generation inside tests. * When we can't find a descriptor, say which one can't be found This helps when debugging things that need to become structural. * web-sys: Test bindings for Event * ci: Use `--manifest-path dir` instead of `cd dir && ...` * web-sys: Just move .webidl files isntead of symlinking to enable them * tests: Polyfill Array.prototype.values for older browsers in CI * test-project-builder: Don't panic on poisoned headless test mutex We only use it to serialize headless tests so that we don't try to bind the port concurrently. Its OK to run another headless test if an earlier one panicked. * JsValue: Add {is,as}_{object,function} methods Allows dynamically casting values to `js::Object` and `js::Function`. * tidy: Fix whitespace and missing semicolons * Allow for dynamic feature detection of methods If we create bindings to a method that doesn't exist in this implementation, then it shouldn't fail until if/when we actually try and invoke that missing method. * tests: Do feature detection in Array.prototype.values test * Add JsValue::{is_string, as_js_string} methods And document all the cast/convert/check methods for js value. * eslint: allow backtick string literals * Only generate a fallback import function for non-structural imports
2018-07-09 16:35:25 -07:00
path.to_str().unwrap(),
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
));
self.files.push((
Path::new("src").join(path).to_str().unwrap().to_string(),
format!(
r#"include!(concat!(env!("OUT_DIR"), "/{}"));"#,
path.display()
),
));
}
buildrs.push('}');
self.files.push(("build.rs".to_string(), buildrs));
res
}
fn generate_js_entry(&mut self, modules: Vec<PathBuf>) {
let mut runjs = String::new();
let esm_imports = self.webpack || !self.node || self.nodejs_experimental_modules;
if self.headless {
runjs.push_str(
r#"
window.document.body.innerHTML += "\nTEST_START\n";
console.log = function(...args) {
const logs = document.getElementById('logs');
for (let msg of args) {
logs.innerHTML += `${msg}<br/>\n`;
}
};
"#,
);
} else if esm_imports {
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
runjs.push_str("import * as process from 'process';\n");
} else {
runjs.push_str("const process = require('process');\n");
}
runjs.push_str("
Create the `web-sys` crate mechanically from WebIDL (#409) * Create a new `web-sys` crate This will eventually contain all the WebIDL-generated bindings to Web APIs. * ci: Test the new `web-sys` crate in CI * web-sys: Add a small README * web-sys: Vendor all the WebIDL files from mozilla-central * backend: Add a pass to remove AST items that use undefined imports This is necessary for the WebIDL frontend, which can't translate many WebIDL constructs into equivalent wasm-bindgen AST things yet. It lets us make incremental progress: we can generate bindings to methods we can support right now even though there might be methods on the same interface that we can't support yet. * webidl: Add a bunch of missing semicolons * webidl: Make parsing private It was only `pub` so that we could test it, but we ended up moving towards integration tests rather than unit tests that assert particular ASTs are parsed from WebIDL files. * webidl: Remove uses of undefined import types * test-project-builder: Build projects in "very verbose" mode This helps for debugging failing WebIDL-related tests. * test-project-builder: Add more profiling timers * test-project-builder: Detect when webpack-dev-server fails Instead of going into an infinite loop, detect when webpack-dev-server fails to start up and early exit the test. * webidl: Specify version for dev-dependency on wasm-bindgen-backend Instead of only a relative path. * guide: Add section about contributing to `web-sys` * WIP enable Event.webidl Still need to fix and finish the test. * Update expected webidl output * Start out a test's status as incomplete That way if we don't fill it in the error message doesn't look quite so bizarre * Fix onerror function in headless mode Otherwise we don't see any output! * Fix package.json/node_modules handling in project generation Make sure these are looked up in the git project root rather than the crate root * Avoid logging body text This was meant for debugging and is otherwise pretty noisy * Fix a relative path * More expected test fixes * Fix a typo * test-project-builder: Allow asynchronous tests * webidl: Convert [Unforgeable] attributes into `#[wasm_bindgen(structural)]` Fixes #432 * test-project-builder: Print generated WebIDL bindings for debugging purposes Helps debug bad WebIDL bindings generation inside tests. * When we can't find a descriptor, say which one can't be found This helps when debugging things that need to become structural. * web-sys: Test bindings for Event * ci: Use `--manifest-path dir` instead of `cd dir && ...` * web-sys: Just move .webidl files isntead of symlinking to enable them * tests: Polyfill Array.prototype.values for older browsers in CI * test-project-builder: Don't panic on poisoned headless test mutex We only use it to serialize headless tests so that we don't try to bind the port concurrently. Its OK to run another headless test if an earlier one panicked. * JsValue: Add {is,as}_{object,function} methods Allows dynamically casting values to `js::Object` and `js::Function`. * tidy: Fix whitespace and missing semicolons * Allow for dynamic feature detection of methods If we create bindings to a method that doesn't exist in this implementation, then it shouldn't fail until if/when we actually try and invoke that missing method. * tests: Do feature detection in Array.prototype.values test * Add JsValue::{is_string, as_js_string} methods And document all the cast/convert/check methods for js value. * eslint: allow backtick string literals * Only generate a fallback import function for non-structural imports
2018-07-09 16:35:25 -07:00
async function run(test, wasm) {
await test.test();
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
if (wasm.assertStackEmpty)
wasm.assertStackEmpty();
if (wasm.assertSlabEmpty)
wasm.assertSlabEmpty();
}
");
if self.headless {
runjs.push_str("
function onerror(e) {
const errors = document.getElementById('error');
let content = `exception: ${e.message}\\nstack: ${e.stack}`;
errors.innerHTML = `<pre>${content}</pre>`;
}
");
} else {
runjs.push_str("
function onerror(error) {
console.error(error);
process.exit(1);
}
");
}
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
if esm_imports {
runjs.push_str("console.log('importing modules...');\n");
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
runjs.push_str("const modules = [];\n");
for module in modules.iter() {
runjs.push_str(&format!("modules.push(import('./{}'))",
module.with_extension("").display()));
}
let import_wasm = if self.debug {
"import('./out')"
} else {
"new Promise((a, b) => a({}))"
};
runjs.push_str(&format!("
Promise.all(modules)
.then(results => {{
results.map(module => Object.assign(global, module));
return Promise.all([import('./test'), {}])
}})
.then(result => run(result[0], result[1]))
", import_wasm));
if self.headless {
runjs.push_str(".then(() => {
document.getElementById('status').innerHTML = 'good';
})");
}
runjs.push_str(".catch(onerror)\n");
if self.headless {
runjs.push_str("
.finally(() => {
window.document.body.innerHTML += \"\\nTEST_DONE\";
})
");
}
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
} else {
assert!(!self.debug);
assert!(modules.is_empty());
runjs.push_str("
const test = require('./test');
Create the `web-sys` crate mechanically from WebIDL (#409) * Create a new `web-sys` crate This will eventually contain all the WebIDL-generated bindings to Web APIs. * ci: Test the new `web-sys` crate in CI * web-sys: Add a small README * web-sys: Vendor all the WebIDL files from mozilla-central * backend: Add a pass to remove AST items that use undefined imports This is necessary for the WebIDL frontend, which can't translate many WebIDL constructs into equivalent wasm-bindgen AST things yet. It lets us make incremental progress: we can generate bindings to methods we can support right now even though there might be methods on the same interface that we can't support yet. * webidl: Add a bunch of missing semicolons * webidl: Make parsing private It was only `pub` so that we could test it, but we ended up moving towards integration tests rather than unit tests that assert particular ASTs are parsed from WebIDL files. * webidl: Remove uses of undefined import types * test-project-builder: Build projects in "very verbose" mode This helps for debugging failing WebIDL-related tests. * test-project-builder: Add more profiling timers * test-project-builder: Detect when webpack-dev-server fails Instead of going into an infinite loop, detect when webpack-dev-server fails to start up and early exit the test. * webidl: Specify version for dev-dependency on wasm-bindgen-backend Instead of only a relative path. * guide: Add section about contributing to `web-sys` * WIP enable Event.webidl Still need to fix and finish the test. * Update expected webidl output * Start out a test's status as incomplete That way if we don't fill it in the error message doesn't look quite so bizarre * Fix onerror function in headless mode Otherwise we don't see any output! * Fix package.json/node_modules handling in project generation Make sure these are looked up in the git project root rather than the crate root * Avoid logging body text This was meant for debugging and is otherwise pretty noisy * Fix a relative path * More expected test fixes * Fix a typo * test-project-builder: Allow asynchronous tests * webidl: Convert [Unforgeable] attributes into `#[wasm_bindgen(structural)]` Fixes #432 * test-project-builder: Print generated WebIDL bindings for debugging purposes Helps debug bad WebIDL bindings generation inside tests. * When we can't find a descriptor, say which one can't be found This helps when debugging things that need to become structural. * web-sys: Test bindings for Event * ci: Use `--manifest-path dir` instead of `cd dir && ...` * web-sys: Just move .webidl files isntead of symlinking to enable them * tests: Polyfill Array.prototype.values for older browsers in CI * test-project-builder: Don't panic on poisoned headless test mutex We only use it to serialize headless tests so that we don't try to bind the port concurrently. Its OK to run another headless test if an earlier one panicked. * JsValue: Add {is,as}_{object,function} methods Allows dynamically casting values to `js::Object` and `js::Function`. * tidy: Fix whitespace and missing semicolons * Allow for dynamic feature detection of methods If we create bindings to a method that doesn't exist in this implementation, then it shouldn't fail until if/when we actually try and invoke that missing method. * tests: Do feature detection in Array.prototype.values test * Add JsValue::{is_string, as_js_string} methods And document all the cast/convert/check methods for js value. * eslint: allow backtick string literals * Only generate a fallback import function for non-structural imports
2018-07-09 16:35:25 -07:00
(async function () {
try {
await run(test, {});
} catch (e) {
onerror(e);
}
}());
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
");
}
self.files.push(("run.js".to_string(), runjs));
}
/// Build the Cargo project for the wasm target, returning the root of the
/// project and the target directory where output is located.
pub fn cargo_build(&mut self) -> (PathBuf, PathBuf) {
let (root, target_dir) = self.build();
let mut cmd = Command::new("cargo");
cmd.arg("build")
Create the `web-sys` crate mechanically from WebIDL (#409) * Create a new `web-sys` crate This will eventually contain all the WebIDL-generated bindings to Web APIs. * ci: Test the new `web-sys` crate in CI * web-sys: Add a small README * web-sys: Vendor all the WebIDL files from mozilla-central * backend: Add a pass to remove AST items that use undefined imports This is necessary for the WebIDL frontend, which can't translate many WebIDL constructs into equivalent wasm-bindgen AST things yet. It lets us make incremental progress: we can generate bindings to methods we can support right now even though there might be methods on the same interface that we can't support yet. * webidl: Add a bunch of missing semicolons * webidl: Make parsing private It was only `pub` so that we could test it, but we ended up moving towards integration tests rather than unit tests that assert particular ASTs are parsed from WebIDL files. * webidl: Remove uses of undefined import types * test-project-builder: Build projects in "very verbose" mode This helps for debugging failing WebIDL-related tests. * test-project-builder: Add more profiling timers * test-project-builder: Detect when webpack-dev-server fails Instead of going into an infinite loop, detect when webpack-dev-server fails to start up and early exit the test. * webidl: Specify version for dev-dependency on wasm-bindgen-backend Instead of only a relative path. * guide: Add section about contributing to `web-sys` * WIP enable Event.webidl Still need to fix and finish the test. * Update expected webidl output * Start out a test's status as incomplete That way if we don't fill it in the error message doesn't look quite so bizarre * Fix onerror function in headless mode Otherwise we don't see any output! * Fix package.json/node_modules handling in project generation Make sure these are looked up in the git project root rather than the crate root * Avoid logging body text This was meant for debugging and is otherwise pretty noisy * Fix a relative path * More expected test fixes * Fix a typo * test-project-builder: Allow asynchronous tests * webidl: Convert [Unforgeable] attributes into `#[wasm_bindgen(structural)]` Fixes #432 * test-project-builder: Print generated WebIDL bindings for debugging purposes Helps debug bad WebIDL bindings generation inside tests. * When we can't find a descriptor, say which one can't be found This helps when debugging things that need to become structural. * web-sys: Test bindings for Event * ci: Use `--manifest-path dir` instead of `cd dir && ...` * web-sys: Just move .webidl files isntead of symlinking to enable them * tests: Polyfill Array.prototype.values for older browsers in CI * test-project-builder: Don't panic on poisoned headless test mutex We only use it to serialize headless tests so that we don't try to bind the port concurrently. Its OK to run another headless test if an earlier one panicked. * JsValue: Add {is,as}_{object,function} methods Allows dynamically casting values to `js::Object` and `js::Function`. * tidy: Fix whitespace and missing semicolons * Allow for dynamic feature detection of methods If we create bindings to a method that doesn't exist in this implementation, then it shouldn't fail until if/when we actually try and invoke that missing method. * tests: Do feature detection in Array.prototype.values test * Add JsValue::{is_string, as_js_string} methods And document all the cast/convert/check methods for js value. * eslint: allow backtick string literals * Only generate a fallback import function for non-structural imports
2018-07-09 16:35:25 -07:00
.arg("-vv")
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
.arg("--target")
.arg("wasm32-unknown-unknown")
.current_dir(&root)
.env("CARGO_TARGET_DIR", &target_dir)
// Catch any warnings in generated code because we don't want any
.env("RUSTFLAGS", "-Dwarnings");
run(&mut cmd, "cargo");
(root, target_dir)
}
/// Generate wasm-bindgen bindings for the compiled artifacts of this
/// project, returning the root of the project as well as the target
/// directory where output was generated.
pub fn gen_bindings(&mut self) -> (PathBuf, PathBuf) {
let (root, target_dir) = self.cargo_build();
let idx = IDX.with(|x| *x);
let out = target_dir.join(&format!("wasm32-unknown-unknown/debug/test{}.wasm", idx));
let as_a_module = root.join("out.wasm");
fs::hard_link(&out, &as_a_module).unwrap();
let _x = wrap_step("running wasm-bindgen");
let res = cli::Bindgen::new()
.input_path(&as_a_module)
.typescript(self.webpack)
.debug(self.debug)
.nodejs(self.node)
.nodejs_experimental_modules(self.nodejs_experimental_modules)
.generate(&root);
if let Err(e) = res {
for e in e.causes() {
println!("- {}", e);
}
panic!("failed");
}
(root, target_dir)
}
/// Execute this project's `run.js`, ensuring that everything runs through
/// node or a browser correctly
pub fn test(&mut self) {
let (root, _target_dir) = self.gen_bindings();
if !self.webpack {
let mut cmd = Command::new("node");
cmd.args(&self.node_args);
if self.nodejs_experimental_modules {
cmd.arg("--experimental-modules").arg("run.mjs");
} else {
cmd.arg("run.js");
}
cmd.current_dir(&root);
run(&mut cmd, "node");
return
}
// Generate typescript bindings for the wasm module
{
let _x = wrap_step("running wasm2es6js");
let mut wasm = Vec::new();
File::open(root.join("out_bg.wasm"))
.unwrap()
.read_to_end(&mut wasm)
.unwrap();
let obj = cli::wasm2es6js::Config::new()
.base64(true)
.generate(&wasm)
.expect("failed to convert wasm to js");
File::create(root.join("out_bg.d.ts"))
.unwrap()
.write_all(obj.typescript().as_bytes())
.unwrap();
}
// move files from the root into each test, it looks like this may be
// needed for webpack to work well when invoked concurrently.
Create the `web-sys` crate mechanically from WebIDL (#409) * Create a new `web-sys` crate This will eventually contain all the WebIDL-generated bindings to Web APIs. * ci: Test the new `web-sys` crate in CI * web-sys: Add a small README * web-sys: Vendor all the WebIDL files from mozilla-central * backend: Add a pass to remove AST items that use undefined imports This is necessary for the WebIDL frontend, which can't translate many WebIDL constructs into equivalent wasm-bindgen AST things yet. It lets us make incremental progress: we can generate bindings to methods we can support right now even though there might be methods on the same interface that we can't support yet. * webidl: Add a bunch of missing semicolons * webidl: Make parsing private It was only `pub` so that we could test it, but we ended up moving towards integration tests rather than unit tests that assert particular ASTs are parsed from WebIDL files. * webidl: Remove uses of undefined import types * test-project-builder: Build projects in "very verbose" mode This helps for debugging failing WebIDL-related tests. * test-project-builder: Add more profiling timers * test-project-builder: Detect when webpack-dev-server fails Instead of going into an infinite loop, detect when webpack-dev-server fails to start up and early exit the test. * webidl: Specify version for dev-dependency on wasm-bindgen-backend Instead of only a relative path. * guide: Add section about contributing to `web-sys` * WIP enable Event.webidl Still need to fix and finish the test. * Update expected webidl output * Start out a test's status as incomplete That way if we don't fill it in the error message doesn't look quite so bizarre * Fix onerror function in headless mode Otherwise we don't see any output! * Fix package.json/node_modules handling in project generation Make sure these are looked up in the git project root rather than the crate root * Avoid logging body text This was meant for debugging and is otherwise pretty noisy * Fix a relative path * More expected test fixes * Fix a typo * test-project-builder: Allow asynchronous tests * webidl: Convert [Unforgeable] attributes into `#[wasm_bindgen(structural)]` Fixes #432 * test-project-builder: Print generated WebIDL bindings for debugging purposes Helps debug bad WebIDL bindings generation inside tests. * When we can't find a descriptor, say which one can't be found This helps when debugging things that need to become structural. * web-sys: Test bindings for Event * ci: Use `--manifest-path dir` instead of `cd dir && ...` * web-sys: Just move .webidl files isntead of symlinking to enable them * tests: Polyfill Array.prototype.values for older browsers in CI * test-project-builder: Don't panic on poisoned headless test mutex We only use it to serialize headless tests so that we don't try to bind the port concurrently. Its OK to run another headless test if an earlier one panicked. * JsValue: Add {is,as}_{object,function} methods Allows dynamically casting values to `js::Object` and `js::Function`. * tidy: Fix whitespace and missing semicolons * Allow for dynamic feature detection of methods If we create bindings to a method that doesn't exist in this implementation, then it shouldn't fail until if/when we actually try and invoke that missing method. * tests: Do feature detection in Array.prototype.values test * Add JsValue::{is_string, as_js_string} methods And document all the cast/convert/check methods for js value. * eslint: allow backtick string literals * Only generate a fallback import function for non-structural imports
2018-07-09 16:35:25 -07:00
let mut cwd = PathBuf::from(env!("CARGO_MANIFEST_DIR"));
cwd.pop(); // chop off test-project-builder
cwd.pop(); // chop off crates
fs::copy(cwd.join("package.json"), root.join("package.json")).unwrap();
let modules = cwd.join("node_modules");
if !modules.exists() {
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
panic!("\n\nfailed to find `node_modules`, have you run `npm install` yet?\n\n");
}
Create the `web-sys` crate mechanically from WebIDL (#409) * Create a new `web-sys` crate This will eventually contain all the WebIDL-generated bindings to Web APIs. * ci: Test the new `web-sys` crate in CI * web-sys: Add a small README * web-sys: Vendor all the WebIDL files from mozilla-central * backend: Add a pass to remove AST items that use undefined imports This is necessary for the WebIDL frontend, which can't translate many WebIDL constructs into equivalent wasm-bindgen AST things yet. It lets us make incremental progress: we can generate bindings to methods we can support right now even though there might be methods on the same interface that we can't support yet. * webidl: Add a bunch of missing semicolons * webidl: Make parsing private It was only `pub` so that we could test it, but we ended up moving towards integration tests rather than unit tests that assert particular ASTs are parsed from WebIDL files. * webidl: Remove uses of undefined import types * test-project-builder: Build projects in "very verbose" mode This helps for debugging failing WebIDL-related tests. * test-project-builder: Add more profiling timers * test-project-builder: Detect when webpack-dev-server fails Instead of going into an infinite loop, detect when webpack-dev-server fails to start up and early exit the test. * webidl: Specify version for dev-dependency on wasm-bindgen-backend Instead of only a relative path. * guide: Add section about contributing to `web-sys` * WIP enable Event.webidl Still need to fix and finish the test. * Update expected webidl output * Start out a test's status as incomplete That way if we don't fill it in the error message doesn't look quite so bizarre * Fix onerror function in headless mode Otherwise we don't see any output! * Fix package.json/node_modules handling in project generation Make sure these are looked up in the git project root rather than the crate root * Avoid logging body text This was meant for debugging and is otherwise pretty noisy * Fix a relative path * More expected test fixes * Fix a typo * test-project-builder: Allow asynchronous tests * webidl: Convert [Unforgeable] attributes into `#[wasm_bindgen(structural)]` Fixes #432 * test-project-builder: Print generated WebIDL bindings for debugging purposes Helps debug bad WebIDL bindings generation inside tests. * When we can't find a descriptor, say which one can't be found This helps when debugging things that need to become structural. * web-sys: Test bindings for Event * ci: Use `--manifest-path dir` instead of `cd dir && ...` * web-sys: Just move .webidl files isntead of symlinking to enable them * tests: Polyfill Array.prototype.values for older browsers in CI * test-project-builder: Don't panic on poisoned headless test mutex We only use it to serialize headless tests so that we don't try to bind the port concurrently. Its OK to run another headless test if an earlier one panicked. * JsValue: Add {is,as}_{object,function} methods Allows dynamically casting values to `js::Object` and `js::Function`. * tidy: Fix whitespace and missing semicolons * Allow for dynamic feature detection of methods If we create bindings to a method that doesn't exist in this implementation, then it shouldn't fail until if/when we actually try and invoke that missing method. * tests: Do feature detection in Array.prototype.values test * Add JsValue::{is_string, as_js_string} methods And document all the cast/convert/check methods for js value. * eslint: allow backtick string literals * Only generate a fallback import function for non-structural imports
2018-07-09 16:35:25 -07:00
symlink_dir(&modules, &root.join("node_modules")).unwrap();
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
if self.headless {
return self.test_headless(&root)
}
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
// Execute webpack to generate a bundle
let mut cmd = self.npm();
cmd.arg("run").arg("run-webpack").current_dir(&root);
run(&mut cmd, "npm");
let mut cmd = Command::new("node");
cmd.args(&self.node_args);
cmd.arg(root.join("bundle.js")).current_dir(&root);
run(&mut cmd, "node");
}
fn npm(&self) -> Command {
if cfg!(windows) {
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
let mut c = Command::new("cmd");
c.arg("/c");
c.arg("npm");
c
} else {
Command::new("npm")
}
}
fn test_headless(&mut self, root: &Path) {
// Serialize all headless tests since they require starting
// webpack-dev-server on the same port.
lazy_static! {
static ref MUTEX: Mutex<()> = Mutex::new(());
}
Create the `web-sys` crate mechanically from WebIDL (#409) * Create a new `web-sys` crate This will eventually contain all the WebIDL-generated bindings to Web APIs. * ci: Test the new `web-sys` crate in CI * web-sys: Add a small README * web-sys: Vendor all the WebIDL files from mozilla-central * backend: Add a pass to remove AST items that use undefined imports This is necessary for the WebIDL frontend, which can't translate many WebIDL constructs into equivalent wasm-bindgen AST things yet. It lets us make incremental progress: we can generate bindings to methods we can support right now even though there might be methods on the same interface that we can't support yet. * webidl: Add a bunch of missing semicolons * webidl: Make parsing private It was only `pub` so that we could test it, but we ended up moving towards integration tests rather than unit tests that assert particular ASTs are parsed from WebIDL files. * webidl: Remove uses of undefined import types * test-project-builder: Build projects in "very verbose" mode This helps for debugging failing WebIDL-related tests. * test-project-builder: Add more profiling timers * test-project-builder: Detect when webpack-dev-server fails Instead of going into an infinite loop, detect when webpack-dev-server fails to start up and early exit the test. * webidl: Specify version for dev-dependency on wasm-bindgen-backend Instead of only a relative path. * guide: Add section about contributing to `web-sys` * WIP enable Event.webidl Still need to fix and finish the test. * Update expected webidl output * Start out a test's status as incomplete That way if we don't fill it in the error message doesn't look quite so bizarre * Fix onerror function in headless mode Otherwise we don't see any output! * Fix package.json/node_modules handling in project generation Make sure these are looked up in the git project root rather than the crate root * Avoid logging body text This was meant for debugging and is otherwise pretty noisy * Fix a relative path * More expected test fixes * Fix a typo * test-project-builder: Allow asynchronous tests * webidl: Convert [Unforgeable] attributes into `#[wasm_bindgen(structural)]` Fixes #432 * test-project-builder: Print generated WebIDL bindings for debugging purposes Helps debug bad WebIDL bindings generation inside tests. * When we can't find a descriptor, say which one can't be found This helps when debugging things that need to become structural. * web-sys: Test bindings for Event * ci: Use `--manifest-path dir` instead of `cd dir && ...` * web-sys: Just move .webidl files isntead of symlinking to enable them * tests: Polyfill Array.prototype.values for older browsers in CI * test-project-builder: Don't panic on poisoned headless test mutex We only use it to serialize headless tests so that we don't try to bind the port concurrently. Its OK to run another headless test if an earlier one panicked. * JsValue: Add {is,as}_{object,function} methods Allows dynamically casting values to `js::Object` and `js::Function`. * tidy: Fix whitespace and missing semicolons * Allow for dynamic feature detection of methods If we create bindings to a method that doesn't exist in this implementation, then it shouldn't fail until if/when we actually try and invoke that missing method. * tests: Do feature detection in Array.prototype.values test * Add JsValue::{is_string, as_js_string} methods And document all the cast/convert/check methods for js value. * eslint: allow backtick string literals * Only generate a fallback import function for non-structural imports
2018-07-09 16:35:25 -07:00
let _lock = {
let _x = wrap_step("waiting on headless test lock");
// Don't panic on a poisoned mutex, since we only use it to
// serialize servers.
MUTEX.lock().unwrap_or_else(|poisoned| poisoned.into_inner())
};
let mut cmd = self.npm();
cmd.arg("run")
.arg("run-webpack-dev-server")
.arg("--")
.arg("--quiet")
.arg("--watch-stdin")
.current_dir(&root);
Create the `web-sys` crate mechanically from WebIDL (#409) * Create a new `web-sys` crate This will eventually contain all the WebIDL-generated bindings to Web APIs. * ci: Test the new `web-sys` crate in CI * web-sys: Add a small README * web-sys: Vendor all the WebIDL files from mozilla-central * backend: Add a pass to remove AST items that use undefined imports This is necessary for the WebIDL frontend, which can't translate many WebIDL constructs into equivalent wasm-bindgen AST things yet. It lets us make incremental progress: we can generate bindings to methods we can support right now even though there might be methods on the same interface that we can't support yet. * webidl: Add a bunch of missing semicolons * webidl: Make parsing private It was only `pub` so that we could test it, but we ended up moving towards integration tests rather than unit tests that assert particular ASTs are parsed from WebIDL files. * webidl: Remove uses of undefined import types * test-project-builder: Build projects in "very verbose" mode This helps for debugging failing WebIDL-related tests. * test-project-builder: Add more profiling timers * test-project-builder: Detect when webpack-dev-server fails Instead of going into an infinite loop, detect when webpack-dev-server fails to start up and early exit the test. * webidl: Specify version for dev-dependency on wasm-bindgen-backend Instead of only a relative path. * guide: Add section about contributing to `web-sys` * WIP enable Event.webidl Still need to fix and finish the test. * Update expected webidl output * Start out a test's status as incomplete That way if we don't fill it in the error message doesn't look quite so bizarre * Fix onerror function in headless mode Otherwise we don't see any output! * Fix package.json/node_modules handling in project generation Make sure these are looked up in the git project root rather than the crate root * Avoid logging body text This was meant for debugging and is otherwise pretty noisy * Fix a relative path * More expected test fixes * Fix a typo * test-project-builder: Allow asynchronous tests * webidl: Convert [Unforgeable] attributes into `#[wasm_bindgen(structural)]` Fixes #432 * test-project-builder: Print generated WebIDL bindings for debugging purposes Helps debug bad WebIDL bindings generation inside tests. * When we can't find a descriptor, say which one can't be found This helps when debugging things that need to become structural. * web-sys: Test bindings for Event * ci: Use `--manifest-path dir` instead of `cd dir && ...` * web-sys: Just move .webidl files isntead of symlinking to enable them * tests: Polyfill Array.prototype.values for older browsers in CI * test-project-builder: Don't panic on poisoned headless test mutex We only use it to serialize headless tests so that we don't try to bind the port concurrently. Its OK to run another headless test if an earlier one panicked. * JsValue: Add {is,as}_{object,function} methods Allows dynamically casting values to `js::Object` and `js::Function`. * tidy: Fix whitespace and missing semicolons * Allow for dynamic feature detection of methods If we create bindings to a method that doesn't exist in this implementation, then it shouldn't fail until if/when we actually try and invoke that missing method. * tests: Do feature detection in Array.prototype.values test * Add JsValue::{is_string, as_js_string} methods And document all the cast/convert/check methods for js value. * eslint: allow backtick string literals * Only generate a fallback import function for non-structural imports
2018-07-09 16:35:25 -07:00
let mut server = run_in_background(&mut cmd, "webpack-dev-server".into());
// wait for webpack-dev-server to come online and bind its port
Create the `web-sys` crate mechanically from WebIDL (#409) * Create a new `web-sys` crate This will eventually contain all the WebIDL-generated bindings to Web APIs. * ci: Test the new `web-sys` crate in CI * web-sys: Add a small README * web-sys: Vendor all the WebIDL files from mozilla-central * backend: Add a pass to remove AST items that use undefined imports This is necessary for the WebIDL frontend, which can't translate many WebIDL constructs into equivalent wasm-bindgen AST things yet. It lets us make incremental progress: we can generate bindings to methods we can support right now even though there might be methods on the same interface that we can't support yet. * webidl: Add a bunch of missing semicolons * webidl: Make parsing private It was only `pub` so that we could test it, but we ended up moving towards integration tests rather than unit tests that assert particular ASTs are parsed from WebIDL files. * webidl: Remove uses of undefined import types * test-project-builder: Build projects in "very verbose" mode This helps for debugging failing WebIDL-related tests. * test-project-builder: Add more profiling timers * test-project-builder: Detect when webpack-dev-server fails Instead of going into an infinite loop, detect when webpack-dev-server fails to start up and early exit the test. * webidl: Specify version for dev-dependency on wasm-bindgen-backend Instead of only a relative path. * guide: Add section about contributing to `web-sys` * WIP enable Event.webidl Still need to fix and finish the test. * Update expected webidl output * Start out a test's status as incomplete That way if we don't fill it in the error message doesn't look quite so bizarre * Fix onerror function in headless mode Otherwise we don't see any output! * Fix package.json/node_modules handling in project generation Make sure these are looked up in the git project root rather than the crate root * Avoid logging body text This was meant for debugging and is otherwise pretty noisy * Fix a relative path * More expected test fixes * Fix a typo * test-project-builder: Allow asynchronous tests * webidl: Convert [Unforgeable] attributes into `#[wasm_bindgen(structural)]` Fixes #432 * test-project-builder: Print generated WebIDL bindings for debugging purposes Helps debug bad WebIDL bindings generation inside tests. * When we can't find a descriptor, say which one can't be found This helps when debugging things that need to become structural. * web-sys: Test bindings for Event * ci: Use `--manifest-path dir` instead of `cd dir && ...` * web-sys: Just move .webidl files isntead of symlinking to enable them * tests: Polyfill Array.prototype.values for older browsers in CI * test-project-builder: Don't panic on poisoned headless test mutex We only use it to serialize headless tests so that we don't try to bind the port concurrently. Its OK to run another headless test if an earlier one panicked. * JsValue: Add {is,as}_{object,function} methods Allows dynamically casting values to `js::Object` and `js::Function`. * tidy: Fix whitespace and missing semicolons * Allow for dynamic feature detection of methods If we create bindings to a method that doesn't exist in this implementation, then it shouldn't fail until if/when we actually try and invoke that missing method. * tests: Do feature detection in Array.prototype.values test * Add JsValue::{is_string, as_js_string} methods And document all the cast/convert/check methods for js value. * eslint: allow backtick string literals * Only generate a fallback import function for non-structural imports
2018-07-09 16:35:25 -07:00
{
let _x = wrap_step("waiting for webpack-dev-server");
loop {
if TcpStream::connect("127.0.0.1:8080").is_ok() {
break;
}
if server.exited() {
panic!("webpack-dev-server exited during headless test initialization")
}
thread::sleep(Duration::from_millis(100));
}
}
let path = env::var_os("PATH").unwrap_or_default();
let mut path = env::split_paths(&path).collect::<Vec<_>>();
path.push(root.join("node_modules/geckodriver"));
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
Create the `web-sys` crate mechanically from WebIDL (#409) * Create a new `web-sys` crate This will eventually contain all the WebIDL-generated bindings to Web APIs. * ci: Test the new `web-sys` crate in CI * web-sys: Add a small README * web-sys: Vendor all the WebIDL files from mozilla-central * backend: Add a pass to remove AST items that use undefined imports This is necessary for the WebIDL frontend, which can't translate many WebIDL constructs into equivalent wasm-bindgen AST things yet. It lets us make incremental progress: we can generate bindings to methods we can support right now even though there might be methods on the same interface that we can't support yet. * webidl: Add a bunch of missing semicolons * webidl: Make parsing private It was only `pub` so that we could test it, but we ended up moving towards integration tests rather than unit tests that assert particular ASTs are parsed from WebIDL files. * webidl: Remove uses of undefined import types * test-project-builder: Build projects in "very verbose" mode This helps for debugging failing WebIDL-related tests. * test-project-builder: Add more profiling timers * test-project-builder: Detect when webpack-dev-server fails Instead of going into an infinite loop, detect when webpack-dev-server fails to start up and early exit the test. * webidl: Specify version for dev-dependency on wasm-bindgen-backend Instead of only a relative path. * guide: Add section about contributing to `web-sys` * WIP enable Event.webidl Still need to fix and finish the test. * Update expected webidl output * Start out a test's status as incomplete That way if we don't fill it in the error message doesn't look quite so bizarre * Fix onerror function in headless mode Otherwise we don't see any output! * Fix package.json/node_modules handling in project generation Make sure these are looked up in the git project root rather than the crate root * Avoid logging body text This was meant for debugging and is otherwise pretty noisy * Fix a relative path * More expected test fixes * Fix a typo * test-project-builder: Allow asynchronous tests * webidl: Convert [Unforgeable] attributes into `#[wasm_bindgen(structural)]` Fixes #432 * test-project-builder: Print generated WebIDL bindings for debugging purposes Helps debug bad WebIDL bindings generation inside tests. * When we can't find a descriptor, say which one can't be found This helps when debugging things that need to become structural. * web-sys: Test bindings for Event * ci: Use `--manifest-path dir` instead of `cd dir && ...` * web-sys: Just move .webidl files isntead of symlinking to enable them * tests: Polyfill Array.prototype.values for older browsers in CI * test-project-builder: Don't panic on poisoned headless test mutex We only use it to serialize headless tests so that we don't try to bind the port concurrently. Its OK to run another headless test if an earlier one panicked. * JsValue: Add {is,as}_{object,function} methods Allows dynamically casting values to `js::Object` and `js::Function`. * tidy: Fix whitespace and missing semicolons * Allow for dynamic feature detection of methods If we create bindings to a method that doesn't exist in this implementation, then it shouldn't fail until if/when we actually try and invoke that missing method. * tests: Do feature detection in Array.prototype.values test * Add JsValue::{is_string, as_js_string} methods And document all the cast/convert/check methods for js value. * eslint: allow backtick string literals * Only generate a fallback import function for non-structural imports
2018-07-09 16:35:25 -07:00
let _x = wrap_step("running headless test");
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
let mut cmd = Command::new("node");
cmd.args(&self.node_args)
.arg(root.join("run-headless.js"))
.current_dir(&root)
.env("PATH", env::join_paths(&path).unwrap());
Speed up Travis by running Webpack in fewer tests (#381) * Reorganize Travis configuration * Add a `JOB` env var descriptor to all matrix entries. Not used anywhere but is useful when viewing the whole build on Travis's web interface. * Reorganize where builds are located, moving slow builds first and fast ones last. * Change checking the CLI builds from `cargo build` to `cargo check` * Use YAML references to reduce some duplication * Print some more timing statistics for each test * Extract `Project` helper in tests to a module This'll help make it a bit more extensible over time. At the same time the methods are also slightly reorganized to read more clearly from top to bottom. * Migrate all tests away from Webpack Wepback can take a significant amount of time to execute and when it's multiplied by hundreds of tests that adds up really quickly! After investigating Node's `--experimental-modules` option it looks like it's suitable for our use so this switches all tests to using JS files (moving away from TypeScript as well) with `--experimental-modules` with Node. Tests will be selectively re-enabled with webpack and node.js specific output (that doesn't require `--experimental-modules`), coming in later commits. * Restore the node test for node.js output Ensures it's workable as-is * Only generate typescript with webpack * Only read wasm files for webpack * Skip package.json/node_modules for now * Only generate webpack config if needed * Start a dedicated test module for typescript Will hopefully verify the generated Typescript compiles OK. * Remove unneeded `node` method * Fixup some rebase conflicts * Don't run asmjs example on travis * Fixup generator tests * Attempt to fix windows * Comment windows fix * More test fixes * More exclusions * More test fixes * Relax eslint regex Catch mjs modules as well * Fix eslint * Speed up travis on examples slightly
2018-07-04 22:37:09 -05:00
run(&mut cmd, "node");
}
/// Reads JS generated by `wasm-bindgen` to a string.
pub fn read_js(&self) -> String {
let path = root().join(if self.nodejs_experimental_modules {
"out.mjs"
} else {
"out.js"
});
println!("js, {:?}", &path);
fs::read_to_string(path).expect("Unable to read js")
}
}
#[cfg(unix)]
fn symlink_dir(a: &Path, b: &Path) -> io::Result<()> {
use std::os::unix::fs::symlink;
symlink(a, b)
}
#[cfg(windows)]
fn symlink_dir(a: &Path, b: &Path) -> io::Result<()> {
use std::os::windows::fs::symlink_dir;
symlink_dir(a, b)
}
fn wrap_step(desc: &str) -> WrapStep {
println!("···················································");
println!("{}", desc);
WrapStep { start: Instant::now() }
}
struct WrapStep { start: Instant }
impl Drop for WrapStep {
fn drop(&mut self) {
let dur = self.start.elapsed();
println!(
"dur: {}.{:03}s",
dur.as_secs(),
dur.subsec_nanos() / 1_000_000
);
}
}
pub fn run(cmd: &mut Command, program: &str) {
let _x = wrap_step(&format!("running {:?}", cmd));
let output = match cmd.output() {
Ok(output) => output,
Err(err) => panic!("failed to spawn `{}`: {}", program, err),
};
println!("exit: {}", output.status);
if output.stdout.len() > 0 {
println!("stdout ---\n{}", String::from_utf8_lossy(&output.stdout));
}
if output.stderr.len() > 0 {
println!("stderr ---\n{}", String::from_utf8_lossy(&output.stderr));
}
assert!(output.status.success());
}
struct BackgroundChild {
name: String,
child: Child,
stdin: Option<ChildStdin>,
stdout: Option<thread::JoinHandle<io::Result<String>>>,
stderr: Option<thread::JoinHandle<io::Result<String>>>,
}
Create the `web-sys` crate mechanically from WebIDL (#409) * Create a new `web-sys` crate This will eventually contain all the WebIDL-generated bindings to Web APIs. * ci: Test the new `web-sys` crate in CI * web-sys: Add a small README * web-sys: Vendor all the WebIDL files from mozilla-central * backend: Add a pass to remove AST items that use undefined imports This is necessary for the WebIDL frontend, which can't translate many WebIDL constructs into equivalent wasm-bindgen AST things yet. It lets us make incremental progress: we can generate bindings to methods we can support right now even though there might be methods on the same interface that we can't support yet. * webidl: Add a bunch of missing semicolons * webidl: Make parsing private It was only `pub` so that we could test it, but we ended up moving towards integration tests rather than unit tests that assert particular ASTs are parsed from WebIDL files. * webidl: Remove uses of undefined import types * test-project-builder: Build projects in "very verbose" mode This helps for debugging failing WebIDL-related tests. * test-project-builder: Add more profiling timers * test-project-builder: Detect when webpack-dev-server fails Instead of going into an infinite loop, detect when webpack-dev-server fails to start up and early exit the test. * webidl: Specify version for dev-dependency on wasm-bindgen-backend Instead of only a relative path. * guide: Add section about contributing to `web-sys` * WIP enable Event.webidl Still need to fix and finish the test. * Update expected webidl output * Start out a test's status as incomplete That way if we don't fill it in the error message doesn't look quite so bizarre * Fix onerror function in headless mode Otherwise we don't see any output! * Fix package.json/node_modules handling in project generation Make sure these are looked up in the git project root rather than the crate root * Avoid logging body text This was meant for debugging and is otherwise pretty noisy * Fix a relative path * More expected test fixes * Fix a typo * test-project-builder: Allow asynchronous tests * webidl: Convert [Unforgeable] attributes into `#[wasm_bindgen(structural)]` Fixes #432 * test-project-builder: Print generated WebIDL bindings for debugging purposes Helps debug bad WebIDL bindings generation inside tests. * When we can't find a descriptor, say which one can't be found This helps when debugging things that need to become structural. * web-sys: Test bindings for Event * ci: Use `--manifest-path dir` instead of `cd dir && ...` * web-sys: Just move .webidl files isntead of symlinking to enable them * tests: Polyfill Array.prototype.values for older browsers in CI * test-project-builder: Don't panic on poisoned headless test mutex We only use it to serialize headless tests so that we don't try to bind the port concurrently. Its OK to run another headless test if an earlier one panicked. * JsValue: Add {is,as}_{object,function} methods Allows dynamically casting values to `js::Object` and `js::Function`. * tidy: Fix whitespace and missing semicolons * Allow for dynamic feature detection of methods If we create bindings to a method that doesn't exist in this implementation, then it shouldn't fail until if/when we actually try and invoke that missing method. * tests: Do feature detection in Array.prototype.values test * Add JsValue::{is_string, as_js_string} methods And document all the cast/convert/check methods for js value. * eslint: allow backtick string literals * Only generate a fallback import function for non-structural imports
2018-07-09 16:35:25 -07:00
impl BackgroundChild {
pub fn exited(&mut self) -> bool {
self.child.try_wait().expect("should try_wait OK").is_some()
}
}
impl Drop for BackgroundChild {
fn drop(&mut self) {
drop(self.stdin.take());
let status = self.child.wait().expect("failed to wait on child");
let stdout = self
.stdout
.take()
.unwrap()
.join()
.unwrap()
.expect("failed to read stdout");
let stderr = self
.stderr
.take()
.unwrap()
.join()
.unwrap()
.expect("failed to read stderr");
println!("···················································");
println!("background {}", self.name);
println!("status: {}", status);
println!("stdout ---\n{}", stdout);
println!("stderr ---\n{}", stderr);
}
}
fn run_in_background(cmd: &mut Command, name: String) -> BackgroundChild {
cmd.stdout(Stdio::piped());
cmd.stderr(Stdio::piped());
cmd.stdin(Stdio::piped());
let mut child = cmd.spawn().expect(&format!("should spawn {} OK", name));
let mut stdout = child.stdout.take().unwrap();
let mut stderr = child.stderr.take().unwrap();
let stdin = child.stdin.take().unwrap();
let stdout = thread::spawn(move || {
let mut t = String::new();
stdout.read_to_string(&mut t)?;
Ok(t)
});
let stderr = thread::spawn(move || {
let mut t = String::new();
stderr.read_to_string(&mut t)?;
Ok(t)
});
BackgroundChild {
name,
child,
stdout: Some(stdout),
stderr: Some(stderr),
stdin: Some(stdin),
}
}