Merge pull request #438 from kivikakk/broken-link-callback-cleanup

broken reflink callback updates & big cleanup.

Author: kivikakk

README.md

@@ -193,7 +193,7 @@ fn replace_text(document: &str, orig_string: &str, replacement: &str) -> String
     for node in root.descendants() {
         if let NodeValue::Text(ref mut text) = node.data.borrow_mut().value {
             // If the node is a text node, perform the string replacement.
-            *text = text.replace(orig_string, replacement)
+            *text = text.replace(orig_string, replacement);
         }
     }
 

fuzz/fuzz_targets/all_options.rs

@@ -3,9 +3,10 @@
 use libfuzzer_sys::fuzz_target;
 
 use comrak::{
-    markdown_to_html, ExtensionOptions, Options, ParseOptions,
-    RenderOptions, ListStyleType,
+    markdown_to_html, BrokenLinkReference, ExtensionOptions, ListStyleType, Options, ParseOptions,
+    RenderOptions, ResolvedReference,
 };
+use std::sync::{Arc, Mutex};
 
 fuzz_target!(|s: &str| {
     let mut extension = ExtensionOptions::default();
@@ -18,22 +19,29 @@ fuzz_target!(|s: &str| {
     extension.header_ids = Some("user-content-".to_string());
     extension.footnotes = true;
     extension.description_lists = true;
+    extension.front_matter_delimiter = Some("---".to_string());
     extension.multiline_block_quotes = true;
     extension.math_dollars = true;
     extension.math_code = true;
-    extension.front_matter_delimiter = Some("---".to_string());
     extension.shortcodes = true;
     extension.wikilinks_title_after_pipe = true;
     extension.wikilinks_title_before_pipe = true;
     extension.underline = true;
     extension.spoiler = true;
     extension.greentext = true;
-    
+
     let mut parse = ParseOptions::default();
     parse.smart = true;
     parse.default_info_string = Some("rust".to_string());
     parse.relaxed_tasklist_matching = true;
     parse.relaxed_autolinks = true;
+    let mut cb = |link_ref: BrokenLinkReference| {
+        Some(ResolvedReference {
+            url: link_ref.normalized.to_string(),
+            title: link_ref.original.to_string(),
+        })
+    };
+    parse.broken_link_callback = Some(Arc::new(Mutex::new(&mut cb)));
 
     let mut render = RenderOptions::default();
     render.hardbreaks = true;
@@ -47,9 +55,15 @@ fuzz_target!(|s: &str| {
     render.escaped_char_spans = true;
     render.ignore_setext = true;
     render.ignore_empty_links = true;
+    render.gfm_quirks = true;
+    render.prefer_fenced = true;
 
     markdown_to_html(
         s,
-        &Options { extension, parse, render },
+        &Options {
+            extension,
+            parse,
+            render,
+        },
     );
 });

fuzz/fuzz_targets/cli_default.rs

@@ -2,9 +2,7 @@
 
 use libfuzzer_sys::fuzz_target;
 
-use comrak::{
-    markdown_to_html_with_plugins, plugins::syntect::SyntectAdapter, Plugins,
-};
+use comrak::{markdown_to_html_with_plugins, plugins::syntect::SyntectAdapter, Plugins};
 
 // Note that we end up fuzzing Syntect here.
 

fuzz/fuzz_targets/quadratic.rs

@@ -2,9 +2,8 @@
 #![feature(int_roundings)]
 #![no_main]
 use comrak::{
-    markdown_to_html, markdown_to_commonmark, markdown_to_commonmark_xml,
-    ExtensionOptions, Options, ParseOptions,
-    RenderOptions, ListStyleType,
+    markdown_to_commonmark, markdown_to_commonmark_xml, markdown_to_html, ExtensionOptions,
+    ListStyleType, Options, ParseOptions, RenderOptions,
 };
 use libfuzzer_sys::arbitrary::{self, Arbitrary};
 use libfuzzer_sys::fuzz_target;
@@ -297,18 +296,10 @@ fn fuzz_one_input(input: &Input, num_bytes: usize) -> (usize, Duration, f64) {
     let duration_per_byte = duration.as_secs_f64() / (byte_length as f64);
 
     if DEBUG {
-        println!(
-            "do_one: {} bytes, duration = {:?}",
-            byte_length,
-            duration
-        );
+        println!("do_one: {} bytes, duration = {:?}", byte_length, duration);
     }
 
-    (
-        byte_length,
-        duration,
-        duration_per_byte
-    )
+    (byte_length, duration, duration_per_byte)
 }
 
 /// The maximum number of steps to run in the main fuzzing loop below.

src/cm.rs

@@ -47,9 +47,9 @@ pub fn format_document_with_plugins<'a>(
     Ok(())
 }
 
-struct CommonMarkFormatter<'a, 'o> {
+struct CommonMarkFormatter<'a, 'o, 'c> {
     node: &'a AstNode<'a>,
-    options: &'o Options,
+    options: &'o Options<'c>,
     v: Vec<u8>,
     prefix: Vec<u8>,
     column: usize,
@@ -72,7 +72,7 @@ enum Escaping {
     Title,
 }
 
-impl<'a, 'o> Write for CommonMarkFormatter<'a, 'o> {
+impl<'a, 'o, 'c> Write for CommonMarkFormatter<'a, 'o, 'c> {
     fn write(&mut self, buf: &[u8]) -> std::io::Result<usize> {
         self.output(buf, false, Escaping::Literal);
         Ok(buf.len())
@@ -83,8 +83,8 @@ impl<'a, 'o> Write for CommonMarkFormatter<'a, 'o> {
     }
 }
 
-impl<'a, 'o> CommonMarkFormatter<'a, 'o> {
-    fn new(node: &'a AstNode<'a>, options: &'o Options) -> Self {
+impl<'a, 'o, 'c> CommonMarkFormatter<'a, 'o, 'c> {
+    fn new(node: &'a AstNode<'a>, options: &'o Options<'c>) -> Self {
         CommonMarkFormatter {
             node,
             options,

src/html.rs

@@ -63,25 +63,24 @@ impl<'w> Write for WriteWithLast<'w> {
     }
 }
 
-/// Converts header Strings to canonical, unique, but still human-readable, anchors.
+/// Converts header strings to canonical, unique, but still human-readable,
+/// anchors.
 ///
-/// To guarantee uniqueness, an anchorizer keeps track of the anchors
-/// it has returned.  So, for example, to parse several MarkDown
-/// files, use a new anchorizer per file.
+/// To guarantee uniqueness, an anchorizer keeps track of the anchors it has
+/// returned; use one per output file.
 ///
 /// ## Example
 ///
 /// ```
-/// use comrak::Anchorizer;
-///
+/// # use comrak::Anchorizer;
 /// let mut anchorizer = Anchorizer::new();
-///
 /// // First "stuff" is unsuffixed.
 /// assert_eq!("stuff".to_string(), anchorizer.anchorize("Stuff".to_string()));
 /// // Second "stuff" has "-1" appended to make it unique.
 /// assert_eq!("stuff-1".to_string(), anchorizer.anchorize("Stuff".to_string()));
 /// ```
 #[derive(Debug, Default)]
+#[doc(hidden)]
 pub struct Anchorizer(HashSet<String>);
 
 impl Anchorizer {
@@ -96,12 +95,9 @@ impl Anchorizer {
     /// resultant anchor unique.
     ///
     /// ```
-    /// use comrak::Anchorizer;
-    ///
+    /// # use comrak::Anchorizer;
     /// let mut anchorizer = Anchorizer::new();
-    ///
     /// let source = "Ticks aren't in";
-    ///
     /// assert_eq!("ticks-arent-in".to_string(), anchorizer.anchorize(source.to_string()));
     /// ```
     pub fn anchorize(&mut self, header: String) -> String {
@@ -130,9 +126,9 @@ impl Anchorizer {
     }
 }
 
-struct HtmlFormatter<'o> {
+struct HtmlFormatter<'o, 'c> {
     output: &'o mut WriteWithLast<'o>,
-    options: &'o Options,
+    options: &'o Options<'c>,
     anchorizer: Anchorizer,
     footnote_ix: u32,
     written_footnote_ix: u32,
@@ -365,8 +361,12 @@ where
     Ok(())
 }
 
-impl<'o> HtmlFormatter<'o> {
-    fn new(options: &'o Options, output: &'o mut WriteWithLast<'o>, plugins: &'o Plugins) -> Self {
+impl<'o, 'c: 'o> HtmlFormatter<'o, 'c> {
+    fn new(
+        options: &'o Options<'c>,
+        output: &'o mut WriteWithLast<'o>,
+        plugins: &'o Plugins,
+    ) -> Self {
         HtmlFormatter {
             options,
             output,

src/lib.rs

@@ -1,8 +1,7 @@
 //! A 100% [CommonMark](http://commonmark.org/) and [GFM](https://github.github.com/gfm/)
-//! compatible Markdown parser.  Source repository is at <https://github.com/kivikakk/comrak>.
+//! compatible Markdown parser.
 //!
-//! The design is based on [cmark-gfm](https://github.com/github/cmark-gfm), so
-//! familiarity with that will help.
+//! Source repository and detailed `README` is at <https://github.com/kivikakk/comrak>.
 //!
 //! You can use `comrak::markdown_to_html` directly:
 //!
@@ -20,41 +19,28 @@
 //! use comrak::nodes::{AstNode, NodeValue};
 //!
 //! # fn main() {
-//! // The returned nodes are created in the supplied Arena, and are bound by its lifetime.
 //! let arena = Arena::new();
 //!
 //! let root = parse_document(
 //!     &arena,
-//!     "This is my input.\n\n1. Also my input.\n2. Certainly my input.\n",
+//!     "This is my input.\n\n1. Also [my](#) input.\n2. Certainly *my* input.\n",
 //!     &Options::default());
 //!
-//! fn iter_nodes<'a, F>(node: &'a AstNode<'a>, f: &F)
-//!     where F : Fn(&'a AstNode<'a>) {
-//!     f(node);
-//!     for c in node.children() {
-//!         iter_nodes(c, f);
+//! for node in root.descendants() {
+//!     if let NodeValue::Text(ref mut text) = node.data.borrow_mut().value {
+//!         *text = text.replace("my", "your");
 //!     }
 //! }
 //!
-//! iter_nodes(root, &|node| {
-//!     match &mut node.data.borrow_mut().value {
-//!         &mut NodeValue::Text(ref mut text) => {
-//!             let orig = std::mem::replace(text, String::new());
-//!             *text = orig.replace("my", "your");
-//!         }
-//!         _ => (),
-//!     }
-//! });
-//!
 //! let mut html = vec![];
 //! format_html(root, &Options::default(), &mut html).unwrap();
 //!
 //! assert_eq!(
 //!     String::from_utf8(html).unwrap(),
 //!     "<p>This is your input.</p>\n\
 //!      <ol>\n\
-//!      <li>Also your input.</li>\n\
-//!      <li>Certainly your input.</li>\n\
+//!      <li>Also <a href=\"#\">your</a> input.</li>\n\
+//!      <li>Certainly <em>your</em> input.</li>\n\
 //!      </ol>\n");
 //! # }
 //! ```
@@ -98,11 +84,15 @@ pub use cm::format_document as format_commonmark;
 pub use cm::format_document_with_plugins as format_commonmark_with_plugins;
 pub use html::format_document as format_html;
 pub use html::format_document_with_plugins as format_html_with_plugins;
+#[doc(inline)]
 pub use html::Anchorizer;
+#[allow(deprecated)]
+pub use parser::parse_document_with_broken_link_callback;
 pub use parser::{
-    parse_document, parse_document_with_broken_link_callback, ExtensionOptions,
+    parse_document, BrokenLinkCallback, BrokenLinkReference, ExtensionOptions,
     ExtensionOptionsBuilder, ListStyleType, Options, ParseOptions, ParseOptionsBuilder, Plugins,
     PluginsBuilder, RenderOptions, RenderOptionsBuilder, RenderPlugins, RenderPluginsBuilder,
+    ResolvedReference,
 };
 pub use typed_arena::Arena;
 pub use xml::format_document as format_xml;
@@ -111,9 +101,9 @@ pub use xml::format_document_with_plugins as format_xml_with_plugins;
 /// Legacy naming of [`ExtensionOptions`]
 pub type ComrakExtensionOptions = ExtensionOptions;
 /// Legacy naming of [`Options`]
-pub type ComrakOptions = Options;
+pub type ComrakOptions<'c> = Options<'c>;
 /// Legacy naming of [`ParseOptions`]
-pub type ComrakParseOptions = ParseOptions;
+pub type ComrakParseOptions<'c> = ParseOptions<'c>;
 /// Legacy naming of [`Plugins`]
 pub type ComrakPlugins<'a> = Plugins<'a>;
 /// Legacy naming of [`RenderOptions`]