lychee/lychee-bin/src/main.rs

286 lines
9 KiB
Rust
Raw Normal View History

//! `lychee` is a fast, asynchronous, resource-friendly link checker.
//! It is able to find broken hyperlinks and mail addresses inside Markdown,
//! HTML, `reStructuredText`, and any other format.
//!
//! The lychee binary is a wrapper around lychee-lib, which provides
//! convenience functions for calling lychee from the command-line.
//!
//! Run it inside a repository with a `README.md`:
//! ```
//! lychee
//! ```
//!
//! You can also specify various types of inputs:
//!
//! Check links on a website:
//!
//! ```sh
//! lychee https://endler.dev/
//! ```
//!
//! Check links in a remote file:
//! ```sh
//! lychee https://raw.githubusercontent.com/lycheeverse/lychee/master/README.md
//! ```
//!
//! Check links in local file(s):
//! ```sh
//! lychee README.md
//! lychee test.html info.txt
//! ```
//!
//! Check links in local files (by shell glob):
//! ```sh
//! lychee ~/projects/*/README.md
//! ```
//!
//! Check links in local files (lychee supports advanced globbing and `~` expansion):
//! ```sh
//! lychee "~/projects/big_project/**/README.*"
//! ```
//!
//! Ignore case when globbing and check result for each link:
//! ```sh
//! lychee --glob-ignore-case --verbose "~/projects/**/[r]eadme.*"
//! ```
Major refactor of codebase (#208) - The binary component and library component are separated as two packages in the same workspace. - `lychee` is the binary component, in `lychee-bin/*`. - `lychee-lib` is the library component, in `lychee-lib/*`. - Users can now install only the `lychee-lib`, instead of both components, that would require fewer dependencies and faster compilation. - Dependencies for each component are adjusted and updated. E.g., no CLI dependencies for `lychee-lib`. - CLI tests are only moved to `lychee`, as it has nothing to do with the library component. - `Status::Error` is refactored to contain dedicated error enum, `ErrorKind`. - The motivation is to delay the formatting of errors to strings. Note that `e.to_string()` is not necessarily cheap (though trivial in many cases). The formatting is no delayed until the error is needed to be displayed to users. So in some cases, if the error is never used, it means that it won't be formatted at all. - Replaced `regex` based matching with one of the following: - Simple string equality test in the case of 'false positivie'. - URL parsing based test, in the case of extracting repository and user name for GitHub links. - Either cases would be much more efficient than `regex` based matching. First, there's no need to construct a state machine for regex. Second, URL is already verified and parsed on its creation, and extracting its components is fairly cheap. Also, this removes the dependency on `lazy-static` in `lychee-lib`. - `types` module now has a sub-directory, and its components are now separated into their own modules (in that sub-directory). - `lychee-lib::test_utils` module is only compiled for tests. - `wiremock` is moved to `dev-dependency` as it's only needed for `test` modules. - Dependencies are listed in alphabetical order. - Imports are organized in the following fashion: - Imports from `std` - Imports from 3rd-party crates, and `lychee-lib`. - Imports from `crate::*` or `super::*`. - No glob import. - I followed suggestion from `cargo clippy`, with `clippy::all` and `clippy:pedantic`. Co-authored-by: Lucius Hu <lebensterben@users.noreply.github.com>
2021-04-14 23:24:11 +00:00
#![warn(clippy::all, clippy::pedantic)]
#![warn(
absolute_paths_not_starting_with_crate,
rustdoc::invalid_html_tags,
Major refactor of codebase (#208) - The binary component and library component are separated as two packages in the same workspace. - `lychee` is the binary component, in `lychee-bin/*`. - `lychee-lib` is the library component, in `lychee-lib/*`. - Users can now install only the `lychee-lib`, instead of both components, that would require fewer dependencies and faster compilation. - Dependencies for each component are adjusted and updated. E.g., no CLI dependencies for `lychee-lib`. - CLI tests are only moved to `lychee`, as it has nothing to do with the library component. - `Status::Error` is refactored to contain dedicated error enum, `ErrorKind`. - The motivation is to delay the formatting of errors to strings. Note that `e.to_string()` is not necessarily cheap (though trivial in many cases). The formatting is no delayed until the error is needed to be displayed to users. So in some cases, if the error is never used, it means that it won't be formatted at all. - Replaced `regex` based matching with one of the following: - Simple string equality test in the case of 'false positivie'. - URL parsing based test, in the case of extracting repository and user name for GitHub links. - Either cases would be much more efficient than `regex` based matching. First, there's no need to construct a state machine for regex. Second, URL is already verified and parsed on its creation, and extracting its components is fairly cheap. Also, this removes the dependency on `lazy-static` in `lychee-lib`. - `types` module now has a sub-directory, and its components are now separated into their own modules (in that sub-directory). - `lychee-lib::test_utils` module is only compiled for tests. - `wiremock` is moved to `dev-dependency` as it's only needed for `test` modules. - Dependencies are listed in alphabetical order. - Imports are organized in the following fashion: - Imports from `std` - Imports from 3rd-party crates, and `lychee-lib`. - Imports from `crate::*` or `super::*`. - No glob import. - I followed suggestion from `cargo clippy`, with `clippy::all` and `clippy:pedantic`. Co-authored-by: Lucius Hu <lebensterben@users.noreply.github.com>
2021-04-14 23:24:11 +00:00
missing_copy_implementations,
missing_debug_implementations,
semicolon_in_expressions_from_macros,
unreachable_pub,
unused_extern_crates,
variant_size_differences,
clippy::missing_const_for_fn
)]
#![deny(anonymous_parameters, macro_use_extern_crate, pointer_structural_match)]
#![deny(missing_docs)]
Major refactor of codebase (#208) - The binary component and library component are separated as two packages in the same workspace. - `lychee` is the binary component, in `lychee-bin/*`. - `lychee-lib` is the library component, in `lychee-lib/*`. - Users can now install only the `lychee-lib`, instead of both components, that would require fewer dependencies and faster compilation. - Dependencies for each component are adjusted and updated. E.g., no CLI dependencies for `lychee-lib`. - CLI tests are only moved to `lychee`, as it has nothing to do with the library component. - `Status::Error` is refactored to contain dedicated error enum, `ErrorKind`. - The motivation is to delay the formatting of errors to strings. Note that `e.to_string()` is not necessarily cheap (though trivial in many cases). The formatting is no delayed until the error is needed to be displayed to users. So in some cases, if the error is never used, it means that it won't be formatted at all. - Replaced `regex` based matching with one of the following: - Simple string equality test in the case of 'false positivie'. - URL parsing based test, in the case of extracting repository and user name for GitHub links. - Either cases would be much more efficient than `regex` based matching. First, there's no need to construct a state machine for regex. Second, URL is already verified and parsed on its creation, and extracting its components is fairly cheap. Also, this removes the dependency on `lazy-static` in `lychee-lib`. - `types` module now has a sub-directory, and its components are now separated into their own modules (in that sub-directory). - `lychee-lib::test_utils` module is only compiled for tests. - `wiremock` is moved to `dev-dependency` as it's only needed for `test` modules. - Dependencies are listed in alphabetical order. - Imports are organized in the following fashion: - Imports from `std` - Imports from 3rd-party crates, and `lychee-lib`. - Imports from `crate::*` or `super::*`. - No glob import. - I followed suggestion from `cargo clippy`, with `clippy::all` and `clippy:pedantic`. Co-authored-by: Lucius Hu <lebensterben@users.noreply.github.com>
2021-04-14 23:24:11 +00:00
use color::YELLOW;
Improve concurrency with streams (#330) * Move to from vec to streams Previously we collected all inputs in one vector before checking the links, which is not ideal. Especially when reading many inputs (e.g. by using a glob pattern), this could cause issues like running out of file handles. By moving to streams we avoid that scenario. This is also the first step towards improving performance for many inputs. To stay as close to the pre-stream behaviour, we want to stop processing as soon as an Err value appears in the stream. This is easiest when the stream is consumed in the main thread. Previously, the stream was consumed in a tokio task and the main thread waited for responses. Now, a tokio task waits for responses (and displays them/registers response stats) and the main thread sends links to the ClientPool. To ensure that the main thread waits for all responses to have arrived before finishing the ProgressBar and printing the stats, it waits for the show_results_task to finish. * Return collected links as Stream * Initialize ProgressBar without length because we can't know the amount of links without blocking * Handle stream results in main thread, not in task * Add basic directory support using jwalk * Add test for HTTP protocol file type (http://) * Remove deadpool (once again): Replaced with `futures::StreamExt::for_each_concurrent`. * Refactor main; fix tests * Move commands into separate submodule * Simplify input handling * Simplify collector * Remove unnecessary unwrap * Simplify main * cleanup check * clean up dump command * Handle requests in parallel * Fix formatting and lints Co-authored-by: Timo Freiberg <self@timofreiberg.com>
2021-12-01 17:25:11 +00:00
use lychee_lib::Collector;
Major refactor of codebase (#208) - The binary component and library component are separated as two packages in the same workspace. - `lychee` is the binary component, in `lychee-bin/*`. - `lychee-lib` is the library component, in `lychee-lib/*`. - Users can now install only the `lychee-lib`, instead of both components, that would require fewer dependencies and faster compilation. - Dependencies for each component are adjusted and updated. E.g., no CLI dependencies for `lychee-lib`. - CLI tests are only moved to `lychee`, as it has nothing to do with the library component. - `Status::Error` is refactored to contain dedicated error enum, `ErrorKind`. - The motivation is to delay the formatting of errors to strings. Note that `e.to_string()` is not necessarily cheap (though trivial in many cases). The formatting is no delayed until the error is needed to be displayed to users. So in some cases, if the error is never used, it means that it won't be formatted at all. - Replaced `regex` based matching with one of the following: - Simple string equality test in the case of 'false positivie'. - URL parsing based test, in the case of extracting repository and user name for GitHub links. - Either cases would be much more efficient than `regex` based matching. First, there's no need to construct a state machine for regex. Second, URL is already verified and parsed on its creation, and extracting its components is fairly cheap. Also, this removes the dependency on `lazy-static` in `lychee-lib`. - `types` module now has a sub-directory, and its components are now separated into their own modules (in that sub-directory). - `lychee-lib::test_utils` module is only compiled for tests. - `wiremock` is moved to `dev-dependency` as it's only needed for `test` modules. - Dependencies are listed in alphabetical order. - Imports are organized in the following fashion: - Imports from `std` - Imports from 3rd-party crates, and `lychee-lib`. - Imports from `crate::*` or `super::*`. - No glob import. - I followed suggestion from `cargo clippy`, with `clippy::all` and `clippy:pedantic`. Co-authored-by: Lucius Hu <lebensterben@users.noreply.github.com>
2021-04-14 23:24:11 +00:00
// required for apple silicon
use ring as _;
use anyhow::{Context, Error, Result};
Major refactor of codebase (#208) - The binary component and library component are separated as two packages in the same workspace. - `lychee` is the binary component, in `lychee-bin/*`. - `lychee-lib` is the library component, in `lychee-lib/*`. - Users can now install only the `lychee-lib`, instead of both components, that would require fewer dependencies and faster compilation. - Dependencies for each component are adjusted and updated. E.g., no CLI dependencies for `lychee-lib`. - CLI tests are only moved to `lychee`, as it has nothing to do with the library component. - `Status::Error` is refactored to contain dedicated error enum, `ErrorKind`. - The motivation is to delay the formatting of errors to strings. Note that `e.to_string()` is not necessarily cheap (though trivial in many cases). The formatting is no delayed until the error is needed to be displayed to users. So in some cases, if the error is never used, it means that it won't be formatted at all. - Replaced `regex` based matching with one of the following: - Simple string equality test in the case of 'false positivie'. - URL parsing based test, in the case of extracting repository and user name for GitHub links. - Either cases would be much more efficient than `regex` based matching. First, there's no need to construct a state machine for regex. Second, URL is already verified and parsed on its creation, and extracting its components is fairly cheap. Also, this removes the dependency on `lazy-static` in `lychee-lib`. - `types` module now has a sub-directory, and its components are now separated into their own modules (in that sub-directory). - `lychee-lib::test_utils` module is only compiled for tests. - `wiremock` is moved to `dev-dependency` as it's only needed for `test` modules. - Dependencies are listed in alphabetical order. - Imports are organized in the following fashion: - Imports from `std` - Imports from 3rd-party crates, and `lychee-lib`. - Imports from `crate::*` or `super::*`. - No glob import. - I followed suggestion from `cargo clippy`, with `clippy::all` and `clippy:pedantic`. Co-authored-by: Lucius Hu <lebensterben@users.noreply.github.com>
2021-04-14 23:24:11 +00:00
use openssl_sys as _; // required for vendored-openssl feature
Improve concurrency with streams (#330) * Move to from vec to streams Previously we collected all inputs in one vector before checking the links, which is not ideal. Especially when reading many inputs (e.g. by using a glob pattern), this could cause issues like running out of file handles. By moving to streams we avoid that scenario. This is also the first step towards improving performance for many inputs. To stay as close to the pre-stream behaviour, we want to stop processing as soon as an Err value appears in the stream. This is easiest when the stream is consumed in the main thread. Previously, the stream was consumed in a tokio task and the main thread waited for responses. Now, a tokio task waits for responses (and displays them/registers response stats) and the main thread sends links to the ClientPool. To ensure that the main thread waits for all responses to have arrived before finishing the ProgressBar and printing the stats, it waits for the show_results_task to finish. * Return collected links as Stream * Initialize ProgressBar without length because we can't know the amount of links without blocking * Handle stream results in main thread, not in task * Add basic directory support using jwalk * Add test for HTTP protocol file type (http://) * Remove deadpool (once again): Replaced with `futures::StreamExt::for_each_concurrent`. * Refactor main; fix tests * Move commands into separate submodule * Simplify input handling * Simplify collector * Remove unnecessary unwrap * Simplify main * cleanup check * clean up dump command * Handle requests in parallel * Fix formatting and lints Co-authored-by: Timo Freiberg <self@timofreiberg.com>
2021-12-01 17:25:11 +00:00
use ring as _;
use std::fs::{self, File};
use std::io::{self, BufRead, BufReader, ErrorKind, Write};
use std::sync::Arc;
2020-10-21 00:10:25 +00:00
use structopt::StructOpt;
2020-08-09 20:48:02 +00:00
mod cache;
Improve concurrency with streams (#330) * Move to from vec to streams Previously we collected all inputs in one vector before checking the links, which is not ideal. Especially when reading many inputs (e.g. by using a glob pattern), this could cause issues like running out of file handles. By moving to streams we avoid that scenario. This is also the first step towards improving performance for many inputs. To stay as close to the pre-stream behaviour, we want to stop processing as soon as an Err value appears in the stream. This is easiest when the stream is consumed in the main thread. Previously, the stream was consumed in a tokio task and the main thread waited for responses. Now, a tokio task waits for responses (and displays them/registers response stats) and the main thread sends links to the ClientPool. To ensure that the main thread waits for all responses to have arrived before finishing the ProgressBar and printing the stats, it waits for the show_results_task to finish. * Return collected links as Stream * Initialize ProgressBar without length because we can't know the amount of links without blocking * Handle stream results in main thread, not in task * Add basic directory support using jwalk * Add test for HTTP protocol file type (http://) * Remove deadpool (once again): Replaced with `futures::StreamExt::for_each_concurrent`. * Refactor main; fix tests * Move commands into separate submodule * Simplify input handling * Simplify collector * Remove unnecessary unwrap * Simplify main * cleanup check * clean up dump command * Handle requests in parallel * Fix formatting and lints Co-authored-by: Timo Freiberg <self@timofreiberg.com>
2021-12-01 17:25:11 +00:00
mod client;
mod color;
Improve concurrency with streams (#330) * Move to from vec to streams Previously we collected all inputs in one vector before checking the links, which is not ideal. Especially when reading many inputs (e.g. by using a glob pattern), this could cause issues like running out of file handles. By moving to streams we avoid that scenario. This is also the first step towards improving performance for many inputs. To stay as close to the pre-stream behaviour, we want to stop processing as soon as an Err value appears in the stream. This is easiest when the stream is consumed in the main thread. Previously, the stream was consumed in a tokio task and the main thread waited for responses. Now, a tokio task waits for responses (and displays them/registers response stats) and the main thread sends links to the ClientPool. To ensure that the main thread waits for all responses to have arrived before finishing the ProgressBar and printing the stats, it waits for the show_results_task to finish. * Return collected links as Stream * Initialize ProgressBar without length because we can't know the amount of links without blocking * Handle stream results in main thread, not in task * Add basic directory support using jwalk * Add test for HTTP protocol file type (http://) * Remove deadpool (once again): Replaced with `futures::StreamExt::for_each_concurrent`. * Refactor main; fix tests * Move commands into separate submodule * Simplify input handling * Simplify collector * Remove unnecessary unwrap * Simplify main * cleanup check * clean up dump command * Handle requests in parallel * Fix formatting and lints Co-authored-by: Timo Freiberg <self@timofreiberg.com>
2021-12-01 17:25:11 +00:00
mod commands;
2020-08-14 09:43:45 +00:00
mod options;
mod parse;
mod stats;
mod time;
mod writer;
Major refactor of codebase (#208) - The binary component and library component are separated as two packages in the same workspace. - `lychee` is the binary component, in `lychee-bin/*`. - `lychee-lib` is the library component, in `lychee-lib/*`. - Users can now install only the `lychee-lib`, instead of both components, that would require fewer dependencies and faster compilation. - Dependencies for each component are adjusted and updated. E.g., no CLI dependencies for `lychee-lib`. - CLI tests are only moved to `lychee`, as it has nothing to do with the library component. - `Status::Error` is refactored to contain dedicated error enum, `ErrorKind`. - The motivation is to delay the formatting of errors to strings. Note that `e.to_string()` is not necessarily cheap (though trivial in many cases). The formatting is no delayed until the error is needed to be displayed to users. So in some cases, if the error is never used, it means that it won't be formatted at all. - Replaced `regex` based matching with one of the following: - Simple string equality test in the case of 'false positivie'. - URL parsing based test, in the case of extracting repository and user name for GitHub links. - Either cases would be much more efficient than `regex` based matching. First, there's no need to construct a state machine for regex. Second, URL is already verified and parsed on its creation, and extracting its components is fairly cheap. Also, this removes the dependency on `lazy-static` in `lychee-lib`. - `types` module now has a sub-directory, and its components are now separated into their own modules (in that sub-directory). - `lychee-lib::test_utils` module is only compiled for tests. - `wiremock` is moved to `dev-dependency` as it's only needed for `test` modules. - Dependencies are listed in alphabetical order. - Imports are organized in the following fashion: - Imports from `std` - Imports from 3rd-party crates, and `lychee-lib`. - Imports from `crate::*` or `super::*`. - No glob import. - I followed suggestion from `cargo clippy`, with `clippy::all` and `clippy:pedantic`. Co-authored-by: Lucius Hu <lebensterben@users.noreply.github.com>
2021-04-14 23:24:11 +00:00
use crate::{
cache::{Cache, StoreExt},
color::color,
options::{Config, Format, LycheeOptions, LYCHEE_CACHE_FILE, LYCHEE_IGNORE_FILE},
stats::ResponseStats,
writer::StatsWriter,
Major refactor of codebase (#208) - The binary component and library component are separated as two packages in the same workspace. - `lychee` is the binary component, in `lychee-bin/*`. - `lychee-lib` is the library component, in `lychee-lib/*`. - Users can now install only the `lychee-lib`, instead of both components, that would require fewer dependencies and faster compilation. - Dependencies for each component are adjusted and updated. E.g., no CLI dependencies for `lychee-lib`. - CLI tests are only moved to `lychee`, as it has nothing to do with the library component. - `Status::Error` is refactored to contain dedicated error enum, `ErrorKind`. - The motivation is to delay the formatting of errors to strings. Note that `e.to_string()` is not necessarily cheap (though trivial in many cases). The formatting is no delayed until the error is needed to be displayed to users. So in some cases, if the error is never used, it means that it won't be formatted at all. - Replaced `regex` based matching with one of the following: - Simple string equality test in the case of 'false positivie'. - URL parsing based test, in the case of extracting repository and user name for GitHub links. - Either cases would be much more efficient than `regex` based matching. First, there's no need to construct a state machine for regex. Second, URL is already verified and parsed on its creation, and extracting its components is fairly cheap. Also, this removes the dependency on `lazy-static` in `lychee-lib`. - `types` module now has a sub-directory, and its components are now separated into their own modules (in that sub-directory). - `lychee-lib::test_utils` module is only compiled for tests. - `wiremock` is moved to `dev-dependency` as it's only needed for `test` modules. - Dependencies are listed in alphabetical order. - Imports are organized in the following fashion: - Imports from `std` - Imports from 3rd-party crates, and `lychee-lib`. - Imports from `crate::*` or `super::*`. - No glob import. - I followed suggestion from `cargo clippy`, with `clippy::all` and `clippy:pedantic`. Co-authored-by: Lucius Hu <lebensterben@users.noreply.github.com>
2021-04-14 23:24:11 +00:00
};
/// A C-like enum that can be cast to `i32` and used as process exit code.
enum ExitCode {
Success = 0,
// NOTE: exit code 1 is used for any `Result::Err` bubbled up to `main()` using the `?` operator.
// For now, 1 acts as a catch-all for everything non-link related (including config errors),
// until we find a way to structure the error code handling better.
#[allow(unused)]
UnexpectedFailure = 1,
LinkCheckFailure = 2,
}
2020-08-14 09:48:55 +00:00
fn main() -> Result<()> {
#[cfg(feature = "tokio-console")]
console_subscriber::init();
// std::process::exit doesn't guarantee that all destructors will be ran,
Improve concurrency with streams (#330) * Move to from vec to streams Previously we collected all inputs in one vector before checking the links, which is not ideal. Especially when reading many inputs (e.g. by using a glob pattern), this could cause issues like running out of file handles. By moving to streams we avoid that scenario. This is also the first step towards improving performance for many inputs. To stay as close to the pre-stream behaviour, we want to stop processing as soon as an Err value appears in the stream. This is easiest when the stream is consumed in the main thread. Previously, the stream was consumed in a tokio task and the main thread waited for responses. Now, a tokio task waits for responses (and displays them/registers response stats) and the main thread sends links to the ClientPool. To ensure that the main thread waits for all responses to have arrived before finishing the ProgressBar and printing the stats, it waits for the show_results_task to finish. * Return collected links as Stream * Initialize ProgressBar without length because we can't know the amount of links without blocking * Handle stream results in main thread, not in task * Add basic directory support using jwalk * Add test for HTTP protocol file type (http://) * Remove deadpool (once again): Replaced with `futures::StreamExt::for_each_concurrent`. * Refactor main; fix tests * Move commands into separate submodule * Simplify input handling * Simplify collector * Remove unnecessary unwrap * Simplify main * cleanup check * clean up dump command * Handle requests in parallel * Fix formatting and lints Co-authored-by: Timo Freiberg <self@timofreiberg.com>
2021-12-01 17:25:11 +00:00
// therefore we wrap "main" code in another function to ensure that.
// See: https://doc.rust-lang.org/stable/std/process/fn.exit.html
// Also see: https://www.youtube.com/watch?v=zQC8T71Y8e4
let exit_code = run_main()?;
std::process::exit(exit_code);
}
/// Read lines from file; ignore empty lines
fn read_lines(file: &File) -> Result<Vec<String>> {
let lines: Vec<_> = BufReader::new(file).lines().collect::<Result<_, _>>()?;
Ok(lines.into_iter().filter(|line| !line.is_empty()).collect())
}
Improve concurrency with streams (#330) * Move to from vec to streams Previously we collected all inputs in one vector before checking the links, which is not ideal. Especially when reading many inputs (e.g. by using a glob pattern), this could cause issues like running out of file handles. By moving to streams we avoid that scenario. This is also the first step towards improving performance for many inputs. To stay as close to the pre-stream behaviour, we want to stop processing as soon as an Err value appears in the stream. This is easiest when the stream is consumed in the main thread. Previously, the stream was consumed in a tokio task and the main thread waited for responses. Now, a tokio task waits for responses (and displays them/registers response stats) and the main thread sends links to the ClientPool. To ensure that the main thread waits for all responses to have arrived before finishing the ProgressBar and printing the stats, it waits for the show_results_task to finish. * Return collected links as Stream * Initialize ProgressBar without length because we can't know the amount of links without blocking * Handle stream results in main thread, not in task * Add basic directory support using jwalk * Add test for HTTP protocol file type (http://) * Remove deadpool (once again): Replaced with `futures::StreamExt::for_each_concurrent`. * Refactor main; fix tests * Move commands into separate submodule * Simplify input handling * Simplify collector * Remove unnecessary unwrap * Simplify main * cleanup check * clean up dump command * Handle requests in parallel * Fix formatting and lints Co-authored-by: Timo Freiberg <self@timofreiberg.com>
2021-12-01 17:25:11 +00:00
/// Merge all provided config options into one
/// This includes a potential config file, command-line- and environment variables
fn load_config() -> Result<LycheeOptions> {
let mut opts = LycheeOptions::from_args();
2020-08-14 09:48:55 +00:00
2020-10-21 00:10:25 +00:00
// Load a potentially existing config file and merge it into the config from the CLI
if let Some(c) = Config::load_from_file(&opts.config_file)? {
2021-09-03 00:18:58 +00:00
opts.config.merge(c);
}
2021-09-01 15:37:31 +00:00
if let Ok(lycheeignore) = File::open(LYCHEE_IGNORE_FILE) {
opts.config.exclude.append(&mut read_lines(&lycheeignore)?);
}
2021-09-01 15:37:31 +00:00
// Load excludes from file
for path in &opts.config.exclude_file {
2021-09-03 00:24:02 +00:00
let file = File::open(path)?;
opts.config.exclude.append(&mut read_lines(&file)?);
2021-09-01 15:37:31 +00:00
}
Improve concurrency with streams (#330) * Move to from vec to streams Previously we collected all inputs in one vector before checking the links, which is not ideal. Especially when reading many inputs (e.g. by using a glob pattern), this could cause issues like running out of file handles. By moving to streams we avoid that scenario. This is also the first step towards improving performance for many inputs. To stay as close to the pre-stream behaviour, we want to stop processing as soon as an Err value appears in the stream. This is easiest when the stream is consumed in the main thread. Previously, the stream was consumed in a tokio task and the main thread waited for responses. Now, a tokio task waits for responses (and displays them/registers response stats) and the main thread sends links to the ClientPool. To ensure that the main thread waits for all responses to have arrived before finishing the ProgressBar and printing the stats, it waits for the show_results_task to finish. * Return collected links as Stream * Initialize ProgressBar without length because we can't know the amount of links without blocking * Handle stream results in main thread, not in task * Add basic directory support using jwalk * Add test for HTTP protocol file type (http://) * Remove deadpool (once again): Replaced with `futures::StreamExt::for_each_concurrent`. * Refactor main; fix tests * Move commands into separate submodule * Simplify input handling * Simplify collector * Remove unnecessary unwrap * Simplify main * cleanup check * clean up dump command * Handle requests in parallel * Fix formatting and lints Co-authored-by: Timo Freiberg <self@timofreiberg.com>
2021-12-01 17:25:11 +00:00
Ok(opts)
}
2020-10-21 00:10:25 +00:00
#[must_use]
/// Load cache (if exists and is still valid)
/// This returns an `Option` as starting without a cache is a common scenario
/// and we silently discard errors on purpose
fn load_cache(cfg: &Config) -> Option<Cache> {
if !cfg.cache {
return None;
}
// Discard entire cache if it hasn't been updated since `max_cache_age`.
// This is an optimization, which avoids iterating over the file and
// checking the age of each entry.
match fs::metadata(LYCHEE_CACHE_FILE) {
Err(_e) => {
// No cache found; silently start with empty cache
return None;
}
Ok(metadata) => {
let modified = metadata.modified().ok()?;
let elapsed = modified.elapsed().ok()?;
if elapsed > cfg.max_cache_age {
eprintln!(
"Cache is too old (age: {}, max age: {}). Discarding",
humantime::format_duration(elapsed),
humantime::format_duration(cfg.max_cache_age)
);
return None;
}
}
}
let cache = Cache::load(LYCHEE_CACHE_FILE, cfg.max_cache_age.as_secs());
match cache {
Ok(cache) => Some(cache),
Err(e) => {
eprintln!("Error while loading cache: {e}. Continuing without.");
None
}
}
}
/// Set up runtime and call lychee entrypoint
Improve concurrency with streams (#330) * Move to from vec to streams Previously we collected all inputs in one vector before checking the links, which is not ideal. Especially when reading many inputs (e.g. by using a glob pattern), this could cause issues like running out of file handles. By moving to streams we avoid that scenario. This is also the first step towards improving performance for many inputs. To stay as close to the pre-stream behaviour, we want to stop processing as soon as an Err value appears in the stream. This is easiest when the stream is consumed in the main thread. Previously, the stream was consumed in a tokio task and the main thread waited for responses. Now, a tokio task waits for responses (and displays them/registers response stats) and the main thread sends links to the ClientPool. To ensure that the main thread waits for all responses to have arrived before finishing the ProgressBar and printing the stats, it waits for the show_results_task to finish. * Return collected links as Stream * Initialize ProgressBar without length because we can't know the amount of links without blocking * Handle stream results in main thread, not in task * Add basic directory support using jwalk * Add test for HTTP protocol file type (http://) * Remove deadpool (once again): Replaced with `futures::StreamExt::for_each_concurrent`. * Refactor main; fix tests * Move commands into separate submodule * Simplify input handling * Simplify collector * Remove unnecessary unwrap * Simplify main * cleanup check * clean up dump command * Handle requests in parallel * Fix formatting and lints Co-authored-by: Timo Freiberg <self@timofreiberg.com>
2021-12-01 17:25:11 +00:00
fn run_main() -> Result<i32> {
use std::process::exit;
Improve concurrency with streams (#330) * Move to from vec to streams Previously we collected all inputs in one vector before checking the links, which is not ideal. Especially when reading many inputs (e.g. by using a glob pattern), this could cause issues like running out of file handles. By moving to streams we avoid that scenario. This is also the first step towards improving performance for many inputs. To stay as close to the pre-stream behaviour, we want to stop processing as soon as an Err value appears in the stream. This is easiest when the stream is consumed in the main thread. Previously, the stream was consumed in a tokio task and the main thread waited for responses. Now, a tokio task waits for responses (and displays them/registers response stats) and the main thread sends links to the ClientPool. To ensure that the main thread waits for all responses to have arrived before finishing the ProgressBar and printing the stats, it waits for the show_results_task to finish. * Return collected links as Stream * Initialize ProgressBar without length because we can't know the amount of links without blocking * Handle stream results in main thread, not in task * Add basic directory support using jwalk * Add test for HTTP protocol file type (http://) * Remove deadpool (once again): Replaced with `futures::StreamExt::for_each_concurrent`. * Refactor main; fix tests * Move commands into separate submodule * Simplify input handling * Simplify collector * Remove unnecessary unwrap * Simplify main * cleanup check * clean up dump command * Handle requests in parallel * Fix formatting and lints Co-authored-by: Timo Freiberg <self@timofreiberg.com>
2021-12-01 17:25:11 +00:00
let opts = load_config()?;
let runtime = match opts.config.threads {
2020-08-14 09:48:55 +00:00
Some(threads) => {
2021-02-18 22:33:14 +00:00
// We define our own runtime instead of the `tokio::main` attribute
// since we want to make the number of threads configurable
tokio::runtime::Builder::new_multi_thread()
.worker_threads(threads)
2020-08-14 09:48:55 +00:00
.enable_all()
.build()?
}
None => tokio::runtime::Runtime::new()?,
};
match runtime.block_on(run(&opts)) {
Err(e) if Some(ErrorKind::BrokenPipe) == underlying_io_error_kind(&e) => {
exit(ExitCode::Success as i32);
}
res => res,
}
}
/// Check if the given error can be traced back to an `io::ErrorKind`
/// This is helpful for troubleshooting the root cause of an error.
/// Code is taken from the anyhow documentation.
fn underlying_io_error_kind(error: &Error) -> Option<io::ErrorKind> {
for cause in error.chain() {
if let Some(io_error) = cause.downcast_ref::<io::Error>() {
return Some(io_error.kind());
}
}
None
2020-08-14 09:48:55 +00:00
}
/// Run lychee on the given inputs
Improve concurrency with streams (#330) * Move to from vec to streams Previously we collected all inputs in one vector before checking the links, which is not ideal. Especially when reading many inputs (e.g. by using a glob pattern), this could cause issues like running out of file handles. By moving to streams we avoid that scenario. This is also the first step towards improving performance for many inputs. To stay as close to the pre-stream behaviour, we want to stop processing as soon as an Err value appears in the stream. This is easiest when the stream is consumed in the main thread. Previously, the stream was consumed in a tokio task and the main thread waited for responses. Now, a tokio task waits for responses (and displays them/registers response stats) and the main thread sends links to the ClientPool. To ensure that the main thread waits for all responses to have arrived before finishing the ProgressBar and printing the stats, it waits for the show_results_task to finish. * Return collected links as Stream * Initialize ProgressBar without length because we can't know the amount of links without blocking * Handle stream results in main thread, not in task * Add basic directory support using jwalk * Add test for HTTP protocol file type (http://) * Remove deadpool (once again): Replaced with `futures::StreamExt::for_each_concurrent`. * Refactor main; fix tests * Move commands into separate submodule * Simplify input handling * Simplify collector * Remove unnecessary unwrap * Simplify main * cleanup check * clean up dump command * Handle requests in parallel * Fix formatting and lints Co-authored-by: Timo Freiberg <self@timofreiberg.com>
2021-12-01 17:25:11 +00:00
async fn run(opts: &LycheeOptions) -> Result<i32> {
let inputs = opts.inputs();
let requests = Collector::new(opts.config.base.clone())
.skip_missing_inputs(opts.config.skip_missing)
.include_verbatim(opts.config.include_verbatim)
// File a bug if you rely on this envvar! It's going to go away eventually.
.use_html5ever(std::env::var("LYCHEE_USE_HTML5EVER").map_or(false, |x| x == "1"))
.collect_links(inputs)
.await;
Improve concurrency with streams (#330) * Move to from vec to streams Previously we collected all inputs in one vector before checking the links, which is not ideal. Especially when reading many inputs (e.g. by using a glob pattern), this could cause issues like running out of file handles. By moving to streams we avoid that scenario. This is also the first step towards improving performance for many inputs. To stay as close to the pre-stream behaviour, we want to stop processing as soon as an Err value appears in the stream. This is easiest when the stream is consumed in the main thread. Previously, the stream was consumed in a tokio task and the main thread waited for responses. Now, a tokio task waits for responses (and displays them/registers response stats) and the main thread sends links to the ClientPool. To ensure that the main thread waits for all responses to have arrived before finishing the ProgressBar and printing the stats, it waits for the show_results_task to finish. * Return collected links as Stream * Initialize ProgressBar without length because we can't know the amount of links without blocking * Handle stream results in main thread, not in task * Add basic directory support using jwalk * Add test for HTTP protocol file type (http://) * Remove deadpool (once again): Replaced with `futures::StreamExt::for_each_concurrent`. * Refactor main; fix tests * Move commands into separate submodule * Simplify input handling * Simplify collector * Remove unnecessary unwrap * Simplify main * cleanup check * clean up dump command * Handle requests in parallel * Fix formatting and lints Co-authored-by: Timo Freiberg <self@timofreiberg.com>
2021-12-01 17:25:11 +00:00
let client = client::create(&opts.config)?;
Improve concurrency with streams (#330) * Move to from vec to streams Previously we collected all inputs in one vector before checking the links, which is not ideal. Especially when reading many inputs (e.g. by using a glob pattern), this could cause issues like running out of file handles. By moving to streams we avoid that scenario. This is also the first step towards improving performance for many inputs. To stay as close to the pre-stream behaviour, we want to stop processing as soon as an Err value appears in the stream. This is easiest when the stream is consumed in the main thread. Previously, the stream was consumed in a tokio task and the main thread waited for responses. Now, a tokio task waits for responses (and displays them/registers response stats) and the main thread sends links to the ClientPool. To ensure that the main thread waits for all responses to have arrived before finishing the ProgressBar and printing the stats, it waits for the show_results_task to finish. * Return collected links as Stream * Initialize ProgressBar without length because we can't know the amount of links without blocking * Handle stream results in main thread, not in task * Add basic directory support using jwalk * Add test for HTTP protocol file type (http://) * Remove deadpool (once again): Replaced with `futures::StreamExt::for_each_concurrent`. * Refactor main; fix tests * Move commands into separate submodule * Simplify input handling * Simplify collector * Remove unnecessary unwrap * Simplify main * cleanup check * clean up dump command * Handle requests in parallel * Fix formatting and lints Co-authored-by: Timo Freiberg <self@timofreiberg.com>
2021-12-01 17:25:11 +00:00
let exit_code = if opts.config.dump {
commands::dump(client, requests, opts.config.verbose).await?
Major refactor of codebase (#208) - The binary component and library component are separated as two packages in the same workspace. - `lychee` is the binary component, in `lychee-bin/*`. - `lychee-lib` is the library component, in `lychee-lib/*`. - Users can now install only the `lychee-lib`, instead of both components, that would require fewer dependencies and faster compilation. - Dependencies for each component are adjusted and updated. E.g., no CLI dependencies for `lychee-lib`. - CLI tests are only moved to `lychee`, as it has nothing to do with the library component. - `Status::Error` is refactored to contain dedicated error enum, `ErrorKind`. - The motivation is to delay the formatting of errors to strings. Note that `e.to_string()` is not necessarily cheap (though trivial in many cases). The formatting is no delayed until the error is needed to be displayed to users. So in some cases, if the error is never used, it means that it won't be formatted at all. - Replaced `regex` based matching with one of the following: - Simple string equality test in the case of 'false positivie'. - URL parsing based test, in the case of extracting repository and user name for GitHub links. - Either cases would be much more efficient than `regex` based matching. First, there's no need to construct a state machine for regex. Second, URL is already verified and parsed on its creation, and extracting its components is fairly cheap. Also, this removes the dependency on `lazy-static` in `lychee-lib`. - `types` module now has a sub-directory, and its components are now separated into their own modules (in that sub-directory). - `lychee-lib::test_utils` module is only compiled for tests. - `wiremock` is moved to `dev-dependency` as it's only needed for `test` modules. - Dependencies are listed in alphabetical order. - Imports are organized in the following fashion: - Imports from `std` - Imports from 3rd-party crates, and `lychee-lib`. - Imports from `crate::*` or `super::*`. - No glob import. - I followed suggestion from `cargo clippy`, with `clippy::all` and `clippy:pedantic`. Co-authored-by: Lucius Hu <lebensterben@users.noreply.github.com>
2021-04-14 23:24:11 +00:00
} else {
let cache = load_cache(&opts.config).unwrap_or_default();
let cache = Arc::new(cache);
let (stats, cache, exit_code) =
commands::check(client, cache, requests, &opts.config).await?;
Improve concurrency with streams (#330) * Move to from vec to streams Previously we collected all inputs in one vector before checking the links, which is not ideal. Especially when reading many inputs (e.g. by using a glob pattern), this could cause issues like running out of file handles. By moving to streams we avoid that scenario. This is also the first step towards improving performance for many inputs. To stay as close to the pre-stream behaviour, we want to stop processing as soon as an Err value appears in the stream. This is easiest when the stream is consumed in the main thread. Previously, the stream was consumed in a tokio task and the main thread waited for responses. Now, a tokio task waits for responses (and displays them/registers response stats) and the main thread sends links to the ClientPool. To ensure that the main thread waits for all responses to have arrived before finishing the ProgressBar and printing the stats, it waits for the show_results_task to finish. * Return collected links as Stream * Initialize ProgressBar without length because we can't know the amount of links without blocking * Handle stream results in main thread, not in task * Add basic directory support using jwalk * Add test for HTTP protocol file type (http://) * Remove deadpool (once again): Replaced with `futures::StreamExt::for_each_concurrent`. * Refactor main; fix tests * Move commands into separate submodule * Simplify input handling * Simplify collector * Remove unnecessary unwrap * Simplify main * cleanup check * clean up dump command * Handle requests in parallel * Fix formatting and lints Co-authored-by: Timo Freiberg <self@timofreiberg.com>
2021-12-01 17:25:11 +00:00
write_stats(stats, &opts.config)?;
if opts.config.cache {
cache.store(LYCHEE_CACHE_FILE)?;
}
exit_code
2020-10-10 04:31:28 +00:00
};
Improve concurrency with streams (#330) * Move to from vec to streams Previously we collected all inputs in one vector before checking the links, which is not ideal. Especially when reading many inputs (e.g. by using a glob pattern), this could cause issues like running out of file handles. By moving to streams we avoid that scenario. This is also the first step towards improving performance for many inputs. To stay as close to the pre-stream behaviour, we want to stop processing as soon as an Err value appears in the stream. This is easiest when the stream is consumed in the main thread. Previously, the stream was consumed in a tokio task and the main thread waited for responses. Now, a tokio task waits for responses (and displays them/registers response stats) and the main thread sends links to the ClientPool. To ensure that the main thread waits for all responses to have arrived before finishing the ProgressBar and printing the stats, it waits for the show_results_task to finish. * Return collected links as Stream * Initialize ProgressBar without length because we can't know the amount of links without blocking * Handle stream results in main thread, not in task * Add basic directory support using jwalk * Add test for HTTP protocol file type (http://) * Remove deadpool (once again): Replaced with `futures::StreamExt::for_each_concurrent`. * Refactor main; fix tests * Move commands into separate submodule * Simplify input handling * Simplify collector * Remove unnecessary unwrap * Simplify main * cleanup check * clean up dump command * Handle requests in parallel * Fix formatting and lints Co-authored-by: Timo Freiberg <self@timofreiberg.com>
2021-12-01 17:25:11 +00:00
Ok(exit_code as i32)
}
2020-10-10 04:31:28 +00:00
Improve concurrency with streams (#330) * Move to from vec to streams Previously we collected all inputs in one vector before checking the links, which is not ideal. Especially when reading many inputs (e.g. by using a glob pattern), this could cause issues like running out of file handles. By moving to streams we avoid that scenario. This is also the first step towards improving performance for many inputs. To stay as close to the pre-stream behaviour, we want to stop processing as soon as an Err value appears in the stream. This is easiest when the stream is consumed in the main thread. Previously, the stream was consumed in a tokio task and the main thread waited for responses. Now, a tokio task waits for responses (and displays them/registers response stats) and the main thread sends links to the ClientPool. To ensure that the main thread waits for all responses to have arrived before finishing the ProgressBar and printing the stats, it waits for the show_results_task to finish. * Return collected links as Stream * Initialize ProgressBar without length because we can't know the amount of links without blocking * Handle stream results in main thread, not in task * Add basic directory support using jwalk * Add test for HTTP protocol file type (http://) * Remove deadpool (once again): Replaced with `futures::StreamExt::for_each_concurrent`. * Refactor main; fix tests * Move commands into separate submodule * Simplify input handling * Simplify collector * Remove unnecessary unwrap * Simplify main * cleanup check * clean up dump command * Handle requests in parallel * Fix formatting and lints Co-authored-by: Timo Freiberg <self@timofreiberg.com>
2021-12-01 17:25:11 +00:00
/// Write final statistics to stdout or to file
fn write_stats(stats: ResponseStats, cfg: &Config) -> Result<()> {
let writer: Box<dyn StatsWriter> = match cfg.format {
Format::Compact => Box::new(writer::Compact::new()),
Format::Detailed => Box::new(writer::Detailed::new()),
Format::Json => Box::new(writer::Json::new()),
Format::Markdown => Box::new(writer::Markdown::new()),
};
let is_empty = stats.is_empty();
let github_issues = stats
.fail_map
.values()
.flatten()
.any(|body| body.uri.domain() == Some("github.com"));
let formatted = writer.write(stats)?;
if let Some(output) = &cfg.output {
fs::write(output, formatted).context("Cannot write status output to file")?;
} else {
if cfg.verbose && !is_empty {
// separate summary from the verbose list of links above
writeln!(io::stdout())?;
}
// we assume that the formatted stats don't have a final newline
writeln!(io::stdout(), "{formatted}")?;
}
if github_issues && cfg.github_token.is_none() {
let mut f = io::stdout();
color!(f, YELLOW, "\u{1f4a1} There were issues with Github URLs. You could try setting a Github token and running lychee again.",)?;
}
Ok(())
}