Fix heading links in nested pages

Plus fixing the whitespace chars not being replaced by hyphen.

Also expand tests for link creations, and add test for nested pages.

Fixes <https://github.com/azerupi/mdBook/issues/416>
Fixes <https://github.com/azerupi/mdBook/issues/417>
This commit is contained in:
Behnam Esfahbod 2017-09-01 16:54:57 -07:00
parent cef62ec42e
commit 956a5cc7fd
6 changed files with 82 additions and 45 deletions

View File

@ -25,6 +25,8 @@ mod execs {
} }
// TODO: Drop after error_chain is fixed
#[allow(unused_doc_comment)]
error_chain!{ error_chain!{
foreign_links { foreign_links {
Io(std::io::Error); Io(std::io::Error);

View File

@ -63,14 +63,14 @@ impl HtmlHandlebars {
debug!("[*]: Render template"); debug!("[*]: Render template");
let rendered = ctx.handlebars.render("index", &ctx.data)?; let rendered = ctx.handlebars.render("index", &ctx.data)?;
let filename = Path::new(&ch.path).with_extension("html"); let filepath = Path::new(&ch.path).with_extension("html");
let rendered = self.post_process(rendered, let rendered = self.post_process(rendered,
filename.file_name().unwrap().to_str().unwrap_or(""), filepath.to_str().unwrap_or(""),
ctx.book.get_html_config().get_playpen_config()); ctx.book.get_html_config().get_playpen_config());
// Write to file // Write to file
info!("[*] Creating {:?} ✓", filename.display()); info!("[*] Creating {:?} ✓", filepath.display());
ctx.book.write_file(filename, &rendered.into_bytes())?; ctx.book.write_file(filepath, &rendered.into_bytes())?;
if ctx.is_index { if ctx.is_index {
self.render_index(ctx.book, ch, &ctx.destination)?; self.render_index(ctx.book, ch, &ctx.destination)?;
@ -111,9 +111,9 @@ impl HtmlHandlebars {
Ok(()) Ok(())
} }
fn post_process(&self, rendered: String, filename: &str, playpen_config: &PlaypenConfig) -> String { fn post_process(&self, rendered: String, filepath: &str, playpen_config: &PlaypenConfig) -> String {
let rendered = build_header_links(&rendered, filename); let rendered = build_header_links(&rendered, &filepath);
let rendered = fix_anchor_links(&rendered, filename); let rendered = fix_anchor_links(&rendered, &filepath);
let rendered = fix_code_blocks(&rendered); let rendered = fix_code_blocks(&rendered);
let rendered = add_playpen_pre(&rendered, playpen_config); let rendered = add_playpen_pre(&rendered, playpen_config);
@ -412,7 +412,7 @@ fn make_data(book: &MDBook) -> Result<serde_json::Map<String, serde_json::Value>
/// Goes through the rendered HTML, making sure all header tags are wrapped in /// Goes through the rendered HTML, making sure all header tags are wrapped in
/// an anchor so people can link to sections directly. /// an anchor so people can link to sections directly.
fn build_header_links(html: &str, filename: &str) -> String { fn build_header_links(html: &str, filepath: &str) -> String {
let regex = Regex::new(r"<h(\d)>(.*?)</h\d>").unwrap(); let regex = Regex::new(r"<h(\d)>(.*?)</h\d>").unwrap();
let mut id_counter = HashMap::new(); let mut id_counter = HashMap::new();
@ -422,14 +422,14 @@ fn build_header_links(html: &str, filename: &str) -> String {
"Regex should ensure we only ever get numbers here", "Regex should ensure we only ever get numbers here",
); );
wrap_header_with_link(level, &caps[2], &mut id_counter, filename) wrap_header_with_link(level, &caps[2], &mut id_counter, filepath)
}) })
.into_owned() .into_owned()
} }
/// Wraps a single header tag with a link, making sure each tag gets its own /// Wraps a single header tag with a link, making sure each tag gets its own
/// unique ID by appending an auto-incremented number (if necessary). /// unique ID by appending an auto-incremented number (if necessary).
fn wrap_header_with_link(level: usize, content: &str, id_counter: &mut HashMap<String, usize>, filename: &str) fn wrap_header_with_link(level: usize, content: &str, id_counter: &mut HashMap<String, usize>, filepath: &str)
-> String { -> String {
let raw_id = id_from_content(content); let raw_id = id_from_content(content);
@ -443,11 +443,11 @@ fn wrap_header_with_link(level: usize, content: &str, id_counter: &mut HashMap<S
*id_count += 1; *id_count += 1;
format!( format!(
r#"<a class="header" href="{filename}#{id}" id="{id}"><h{level}>{text}</h{level}></a>"#, r##"<a class="header" href="{filepath}#{id}" id="{id}"><h{level}>{text}</h{level}></a>"##,
level = level, level = level,
id = id, id = id,
text = content, text = content,
filename = filename filepath = filepath
) )
} }
@ -457,7 +457,7 @@ fn id_from_content(content: &str) -> String {
let mut content = content.to_string(); let mut content = content.to_string();
// Skip any tags or html-encoded stuff // Skip any tags or html-encoded stuff
let repl_sub = vec![ static REPL_SUB: &[&str] = &[
"<em>", "<em>",
"</em>", "</em>",
"<code>", "<code>",
@ -470,27 +470,25 @@ fn id_from_content(content: &str) -> String {
"&#39;", "&#39;",
"&quot;", "&quot;",
]; ];
for sub in repl_sub { for sub in REPL_SUB {
content = content.replace(sub, ""); content = content.replace(sub, "");
} }
let mut id = String::new(); let mut id = String::new();
for c in content.chars() { for c in content.chars() {
if c.is_alphanumeric() || c == '-' || c == '_' { if c.is_alphanumeric() || c == '_' {
id.push(c.to_ascii_lowercase()); id.push(c.to_ascii_lowercase());
} else if c.is_whitespace() { } else if c.is_whitespace() {
id.push(c); id.push('-');
} }
} }
id id
} }
// anchors to the same page (href="#anchor") do not work because of // anchors to the same page (href="#anchor") do not work because of
// <base href="../"> pointing to the root folder. This function *fixes* // <base href="../"> pointing to the root folder. This function *fixes*
// that in a very inelegant way // that in a very inelegant way
fn fix_anchor_links(html: &str, filename: &str) -> String { fn fix_anchor_links(html: &str, filepath: &str) -> String {
let regex = Regex::new(r##"<a([^>]+)href="#([^"]+)"([^>]*)>"##).unwrap(); let regex = Regex::new(r##"<a([^>]+)href="#([^"]+)"([^>]*)>"##).unwrap();
regex regex
.replace_all(html, |caps: &Captures| { .replace_all(html, |caps: &Captures| {
@ -499,9 +497,9 @@ fn fix_anchor_links(html: &str, filename: &str) -> String {
let after = &caps[3]; let after = &caps[3];
format!( format!(
"<a{before}href=\"{filename}#{anchor}\"{after}>", "<a{before}href=\"{filepath}#{anchor}\"{after}>",
before = before, before = before,
filename = filename, filepath = filepath,
anchor = anchor, anchor = anchor,
after = after after = after
) )
@ -601,17 +599,39 @@ mod tests {
#[test] #[test]
fn original_build_header_links() { fn original_build_header_links() {
let inputs = vec![ let inputs = vec![
("blah blah <h1>Foo</h1>", r#"blah blah <a class="header" href="bar.rs#foo" id="foo"><h1>Foo</h1></a>"#), (
("<h1>Foo</h1>", r#"<a class="header" href="bar.rs#foo" id="foo"><h1>Foo</h1></a>"#), "blah blah <h1>Foo</h1>",
("<h3>Foo^bar</h3>", r#"<a class="header" href="bar.rs#foobar" id="foobar"><h3>Foo^bar</h3></a>"#), r##"blah blah <a class="header" href="./some_chapter/some_section.html#foo" id="foo"><h1>Foo</h1></a>"##,
("<h4></h4>", r#"<a class="header" href="bar.rs#" id=""><h4></h4></a>"#), ),
("<h4><em>Hï</em></h4>", r#"<a class="header" href="bar.rs#hï" id="hï"><h4><em>Hï</em></h4></a>"#), (
("<h1>Foo</h1><h3>Foo</h3>", "<h1>Foo</h1>",
r#"<a class="header" href="bar.rs#foo" id="foo"><h1>Foo</h1></a><a class="header" href="bar.rs#foo-1" id="foo-1"><h3>Foo</h3></a>"#), r##"<a class="header" href="./some_chapter/some_section.html#foo" id="foo"><h1>Foo</h1></a>"##,
),
(
"<h3>Foo^bar</h3>",
r##"<a class="header" href="./some_chapter/some_section.html#foobar" id="foobar"><h3>Foo^bar</h3></a>"##,
),
(
"<h4></h4>",
r##"<a class="header" href="./some_chapter/some_section.html#" id=""><h4></h4></a>"##
),
(
"<h4><em>Hï</em></h4>",
r##"<a class="header" href="./some_chapter/some_section.html#hï" id="hï"><h4><em>Hï</em></h4></a>"##
),
(
"<h1>Foo</h1><h3>Foo</h3>",
r##"<a class="header" href="./some_chapter/some_section.html#foo" id="foo"><h1>Foo</h1></a><a class="header" href="./some_chapter/some_section.html#foo-1" id="foo-1"><h3>Foo</h3></a>"##
),
]; ];
for (src, should_be) in inputs { for (src, should_be) in inputs {
let got = build_header_links(src, "bar.rs"); let filepath = "./some_chapter/some_section.html";
let got = build_header_links(&src, filepath);
assert_eq!(got, should_be);
// This is redundant for most cases
let got = fix_anchor_links(&got, filepath);
assert_eq!(got, should_be); assert_eq!(got, should_be);
} }
} }

View File

@ -1,3 +1,5 @@
# First Chapter # First Chapter
more text. more text.
## Some Section

View File

@ -5,3 +5,5 @@ This file has some testable code.
```rust ```rust
assert!($TEST_STATUS); assert!($TEST_STATUS);
``` ```
## Some Section

View File

@ -1,8 +1,5 @@
//! Helpers for tests which exercise the overall application, in particular //! Helpers for tests which exercise the overall application, in particular
//! the `MDBook` initialization and build/rendering process. //! the `MDBook` initialization and build/rendering process.
//!
//! This will create an entire book in a temporary directory using some
//! dummy contents from the `tests/dummy-book/` directory.
use std::path::Path; use std::path::Path;

View File

@ -38,11 +38,11 @@ fn make_sure_bottom_level_files_contain_links_to_chapters() {
let dest = temp.path().join("book"); let dest = temp.path().join("book");
let links = vec![ let links = vec![
"intro.html", r#"href="intro.html""#,
"first/index.html", r#"href="./first/index.html""#,
"first/nested.html", r#"href="./first/nested.html""#,
"second.html", r#"href="./second.html""#,
"conclusion.html", r#"href="./conclusion.html""#,
]; ];
let files_in_bottom_dir = vec!["index.html", "intro.html", "second.html", "conclusion.html"]; let files_in_bottom_dir = vec!["index.html", "intro.html", "second.html", "conclusion.html"];
@ -61,11 +61,11 @@ fn check_correct_cross_links_in_nested_dir() {
let first = temp.path().join("book").join("first"); let first = temp.path().join("book").join("first");
let links = vec![ let links = vec![
r#"<base href="../">"#, r#"<base href="../">"#,
"intro.html", r#"href="intro.html""#,
"first/index.html", r#"href="./first/index.html""#,
"first/nested.html", r#"href="./first/nested.html""#,
"second.html", r#"href="./second.html""#,
"conclusion.html", r#"href="./conclusion.html""#,
]; ];
let files_in_nested_dir = vec!["index.html", "nested.html"]; let files_in_nested_dir = vec!["index.html", "nested.html"];
@ -73,6 +73,20 @@ fn check_correct_cross_links_in_nested_dir() {
for filename in files_in_nested_dir { for filename in files_in_nested_dir {
assert_contains_strings(first.join(filename), &links); assert_contains_strings(first.join(filename), &links);
} }
assert_contains_strings(
first.join("index.html"),
&[
r##"href="./first/index.html#some-section" id="some-section""##
],
);
assert_contains_strings(
first.join("nested.html"),
&[
r##"href="./first/nested.html#some-section" id="some-section""##
],
);
} }
#[test] #[test]