Fix even more print page links. (#963)

This commit is contained in:
Eric Huss 2019-07-01 08:52:25 -07:00 committed by Dylan DPC
parent 4b569edadd
commit 228e99ba11
5 changed files with 222 additions and 36 deletions

View File

@ -33,12 +33,10 @@ impl HtmlHandlebars {
let content = ch.content.clone(); let content = ch.content.clone();
let content = utils::render_markdown(&content, ctx.html_config.curly_quotes); let content = utils::render_markdown(&content, ctx.html_config.curly_quotes);
let string_path = ch.path.parent().unwrap().display().to_string(); let fixed_content = utils::render_markdown_with_path(
let fixed_content = utils::render_markdown_with_base(
&ch.content, &ch.content,
ctx.html_config.curly_quotes, ctx.html_config.curly_quotes,
&string_path, Some(&ch.path),
); );
print_content.push_str(&fixed_content); print_content.push_str(&fixed_content);

View File

@ -8,6 +8,8 @@ use regex::Regex;
use pulldown_cmark::{html, CowStr, Event, Options, Parser, Tag}; use pulldown_cmark::{html, CowStr, Event, Options, Parser, Tag};
use std::borrow::Cow; use std::borrow::Cow;
use std::fmt::Write;
use std::path::Path;
pub use self::string::take_lines; pub use self::string::take_lines;
@ -65,20 +67,47 @@ pub fn id_from_content(content: &str) -> String {
normalize_id(trimmed) normalize_id(trimmed)
} }
fn adjust_links<'a>(event: Event<'a>, with_base: &str) -> Event<'a> { /// Fix links to the correct location.
///
/// This adjusts links, such as turning `.md` extensions to `.html`.
///
/// `path` is the path to the page being rendered relative to the root of the
/// book. This is used for the `print.html` page so that links on the print
/// page go to the original location. Normal page rendering sets `path` to
/// None. Ideally, print page links would link to anchors on the print page,
/// but that is very difficult.
fn adjust_links<'a>(event: Event<'a>, path: Option<&Path>) -> Event<'a> {
lazy_static! { lazy_static! {
static ref SCHEME_LINK: Regex = Regex::new(r"^[a-z][a-z0-9+.-]*:").unwrap(); static ref SCHEME_LINK: Regex = Regex::new(r"^[a-z][a-z0-9+.-]*:").unwrap();
static ref MD_LINK: Regex = Regex::new(r"(?P<link>.*)\.md(?P<anchor>#.*)?").unwrap(); static ref MD_LINK: Regex = Regex::new(r"(?P<link>.*)\.md(?P<anchor>#.*)?").unwrap();
} }
fn fix<'a>(dest: CowStr<'a>, base: &str) -> CowStr<'a> { fn fix<'a>(dest: CowStr<'a>, path: Option<&Path>) -> CowStr<'a> {
if dest.starts_with('#') {
// Fragment-only link.
if let Some(path) = path {
let mut base = path.display().to_string();
if base.ends_with(".md") {
base.replace_range(base.len() - 3.., ".html");
}
return format!("{}{}", base, dest).into();
} else {
return dest;
}
}
// Don't modify links with schemes like `https`. // Don't modify links with schemes like `https`.
if !SCHEME_LINK.is_match(&dest) { if !SCHEME_LINK.is_match(&dest) {
// This is a relative link, adjust it as necessary. // This is a relative link, adjust it as necessary.
let mut fixed_link = String::new(); let mut fixed_link = String::new();
if !base.is_empty() { if let Some(path) = path {
fixed_link.push_str(base); let base = path
fixed_link.push_str("/"); .parent()
.expect("path can't be empty")
.to_str()
.expect("utf-8 paths only");
if !base.is_empty() {
write!(fixed_link, "{}/", base).unwrap();
}
} }
if let Some(caps) = MD_LINK.captures(&dest) { if let Some(caps) = MD_LINK.captures(&dest) {
@ -95,20 +124,45 @@ fn adjust_links<'a>(event: Event<'a>, with_base: &str) -> Event<'a> {
dest dest
} }
fn fix_html<'a>(html: CowStr<'a>, path: Option<&Path>) -> CowStr<'a> {
// This is a terrible hack, but should be reasonably reliable. Nobody
// should ever parse a tag with a regex. However, there isn't anything
// in Rust that I know of that is suitable for handling partial html
// fragments like those generated by pulldown_cmark.
//
// There are dozens of HTML tags/attributes that contain paths, so
// feel free to add more tags if desired; these are the only ones I
// care about right now.
lazy_static! {
static ref HTML_LINK: Regex =
Regex::new(r#"(<(?:a|img) [^>]*?(?:src|href)=")([^"]+?)""#).unwrap();
}
HTML_LINK
.replace_all(&html, |caps: &regex::Captures<'_>| {
let fixed = fix(caps[2].into(), path);
format!("{}{}\"", &caps[1], fixed)
})
.into_owned()
.into()
}
match event { match event {
Event::Start(Tag::Link(link_type, dest, title)) => { Event::Start(Tag::Link(link_type, dest, title)) => {
Event::Start(Tag::Link(link_type, fix(dest, with_base), title)) Event::Start(Tag::Link(link_type, fix(dest, path), title))
} }
Event::Start(Tag::Image(link_type, dest, title)) => { Event::Start(Tag::Image(link_type, dest, title)) => {
Event::Start(Tag::Image(link_type, fix(dest, with_base), title)) Event::Start(Tag::Image(link_type, fix(dest, path), title))
} }
Event::Html(html) => Event::Html(fix_html(html, path)),
Event::InlineHtml(html) => Event::InlineHtml(fix_html(html, path)),
_ => event, _ => event,
} }
} }
/// Wrapper around the pulldown-cmark parser for rendering markdown to HTML. /// Wrapper around the pulldown-cmark parser for rendering markdown to HTML.
pub fn render_markdown(text: &str, curly_quotes: bool) -> String { pub fn render_markdown(text: &str, curly_quotes: bool) -> String {
render_markdown_with_base(text, curly_quotes, "") render_markdown_with_path(text, curly_quotes, None)
} }
pub fn new_cmark_parser(text: &str) -> Parser<'_> { pub fn new_cmark_parser(text: &str) -> Parser<'_> {
@ -120,13 +174,13 @@ pub fn new_cmark_parser(text: &str) -> Parser<'_> {
Parser::new_ext(text, opts) Parser::new_ext(text, opts)
} }
pub fn render_markdown_with_base(text: &str, curly_quotes: bool, base: &str) -> String { pub fn render_markdown_with_path(text: &str, curly_quotes: bool, path: Option<&Path>) -> String {
let mut s = String::with_capacity(text.len() * 3 / 2); let mut s = String::with_capacity(text.len() * 3 / 2);
let p = new_cmark_parser(text); let p = new_cmark_parser(text);
let mut converter = EventQuoteConverter::new(curly_quotes); let mut converter = EventQuoteConverter::new(curly_quotes);
let events = p let events = p
.map(clean_codeblock_headers) .map(clean_codeblock_headers)
.map(|event| adjust_links(event, base)) .map(|event| adjust_links(event, path))
.map(|event| converter.convert(event)); .map(|event| converter.convert(event));
html::push_html(&mut s, events); html::push_html(&mut s, events);

View File

@ -3,6 +3,14 @@
When we link to [the first section](../first/nested.md), it should work on When we link to [the first section](../first/nested.md), it should work on
both the print page and the non-print page. both the print page and the non-print page.
A [fragment link](#some-section) should work.
Link [outside](../../std/foo/bar.html). Link [outside](../../std/foo/bar.html).
![Some image](../images/picture.png) ![Some image](../images/picture.png)
<a href="../first/markdown.md">HTML Link</a>
<img src="../images/picture.png" alt="raw html">
## Some section

View File

@ -124,6 +124,9 @@ fn check_correct_relative_links_in_print_page() {
r##"<a href="second/../first/nested.html">the first section</a>,"##, r##"<a href="second/../first/nested.html">the first section</a>,"##,
r##"<a href="second/../../std/foo/bar.html">outside</a>"##, r##"<a href="second/../../std/foo/bar.html">outside</a>"##,
r##"<img src="second/../images/picture.png" alt="Some image" />"##, r##"<img src="second/../images/picture.png" alt="Some image" />"##,
r##"<a href="second/nested.html#some-section">fragment link</a>"##,
r##"<a href="second/../first/markdown.html">HTML Link</a>"##,
r##"<img src="second/../images/picture.png" alt="raw html">"##,
], ],
); );
} }

View File

@ -15,6 +15,7 @@
"first/markdown.html#tasklisks", "first/markdown.html#tasklisks",
"second.html#second-chapter", "second.html#second-chapter",
"second/nested.html#testing-relative-links-for-the-print-page", "second/nested.html#testing-relative-links-for-the-print-page",
"second/nested.html#some-section",
"conclusion.html#conclusion" "conclusion.html#conclusion"
], ],
"index": { "index": {
@ -51,11 +52,16 @@
"title": 2 "title": 2
}, },
"14": { "14": {
"body": 13, "body": 18,
"breadcrumbs": 7, "breadcrumbs": 7,
"title": 5 "title": 5
}, },
"15": { "15": {
"body": 0,
"breadcrumbs": 3,
"title": 1
},
"16": {
"body": 3, "body": 3,
"breadcrumbs": 1, "breadcrumbs": 1,
"title": 1 "title": 1
@ -139,15 +145,21 @@
"title": "Second Chapter" "title": "Second Chapter"
}, },
"14": { "14": {
"body": "When we link to the first section , it should work on both the print page and the non-print page. Link outside . Some image", "body": "When we link to the first section , it should work on both the print page and the non-print page. A fragment link should work. Link outside . Some image HTML Link",
"breadcrumbs": "Second Chapter » Testing relative links for the print page", "breadcrumbs": "Second Chapter » Testing relative links for the print page",
"id": "14", "id": "14",
"title": "Testing relative links for the print page" "title": "Testing relative links for the print page"
}, },
"15": { "15": {
"body": "",
"breadcrumbs": "Second Chapter » Some section",
"id": "15",
"title": "Some section"
},
"16": {
"body": "I put &lt;HTML&gt; in here!", "body": "I put &lt;HTML&gt; in here!",
"breadcrumbs": "Conclusion", "breadcrumbs": "Conclusion",
"id": "15", "id": "16",
"title": "Conclusion" "title": "Conclusion"
}, },
"2": { "2": {
@ -199,7 +211,7 @@
"title": "Tables" "title": "Tables"
} }
}, },
"length": 16, "length": 17,
"save": true "save": true
}, },
"fields": [ "fields": [
@ -499,7 +511,7 @@
"s": { "s": {
"df": 2, "df": 2,
"docs": { "docs": {
"15": { "16": {
"tf": 1.0 "tf": 1.0
}, },
"7": { "7": {
@ -701,6 +713,38 @@
} }
} }
} }
},
"r": {
"a": {
"df": 0,
"docs": {},
"g": {
"df": 0,
"docs": {},
"m": {
"df": 0,
"docs": {},
"e": {
"df": 0,
"docs": {},
"n": {
"df": 0,
"docs": {},
"t": {
"df": 1,
"docs": {
"14": {
"tf": 1.0
}
}
}
}
}
}
}
},
"df": 0,
"docs": {}
} }
}, },
"g": { "g": {
@ -746,7 +790,7 @@
"0": { "0": {
"tf": 1.0 "tf": 1.0
}, },
"15": { "16": {
"tf": 1.0 "tf": 1.0
} }
} }
@ -784,6 +828,22 @@
}, },
"df": 0, "df": 0,
"docs": {} "docs": {}
},
"t": {
"df": 0,
"docs": {},
"m": {
"df": 0,
"docs": {},
"l": {
"df": 1,
"docs": {
"14": {
"tf": 1.0
}
}
}
}
} }
}, },
"i": { "i": {
@ -968,7 +1028,7 @@
"df": 1, "df": 1,
"docs": { "docs": {
"14": { "14": {
"tf": 1.7320508075688772 "tf": 2.23606797749979
} }
} }
} }
@ -1021,7 +1081,7 @@
"t": { "t": {
"df": 1, "df": 1,
"docs": { "docs": {
"15": { "16": {
"tf": 1.0 "tf": 1.0
} }
} }
@ -1382,7 +1442,7 @@
"t": { "t": {
"df": 1, "df": 1,
"docs": { "docs": {
"15": { "16": {
"tf": 1.0 "tf": 1.0
} }
} }
@ -1505,11 +1565,14 @@
"df": 0, "df": 0,
"docs": {}, "docs": {},
"n": { "n": {
"df": 3, "df": 4,
"docs": { "docs": {
"14": { "14": {
"tf": 1.0 "tf": 1.0
}, },
"15": {
"tf": 1.0
},
"3": { "3": {
"tf": 1.0 "tf": 1.0
}, },
@ -1793,7 +1856,7 @@
"df": 1, "df": 1,
"docs": { "docs": {
"14": { "14": {
"tf": 1.0 "tf": 1.4142135623730951
} }
} }
}, },
@ -2051,7 +2114,7 @@
"df": 0, "df": 0,
"docs": {}, "docs": {},
"r": { "r": {
"df": 12, "df": 13,
"docs": { "docs": {
"10": { "10": {
"tf": 1.0 "tf": 1.0
@ -2068,6 +2131,9 @@
"14": { "14": {
"tf": 1.0 "tf": 1.0
}, },
"15": {
"tf": 1.0
},
"2": { "2": {
"tf": 1.4142135623730951 "tf": 1.4142135623730951
}, },
@ -2129,7 +2195,7 @@
"s": { "s": {
"df": 2, "df": 2,
"docs": { "docs": {
"15": { "16": {
"tf": 1.4142135623730951 "tf": 1.4142135623730951
}, },
"7": { "7": {
@ -2355,6 +2421,38 @@
} }
} }
} }
},
"r": {
"a": {
"df": 0,
"docs": {},
"g": {
"df": 0,
"docs": {},
"m": {
"df": 0,
"docs": {},
"e": {
"df": 0,
"docs": {},
"n": {
"df": 0,
"docs": {},
"t": {
"df": 1,
"docs": {
"14": {
"tf": 1.0
}
}
}
}
}
}
}
},
"df": 0,
"docs": {}
} }
}, },
"g": { "g": {
@ -2400,7 +2498,7 @@
"0": { "0": {
"tf": 1.0 "tf": 1.0
}, },
"15": { "16": {
"tf": 1.0 "tf": 1.0
} }
} }
@ -2438,6 +2536,22 @@
}, },
"df": 0, "df": 0,
"docs": {} "docs": {}
},
"t": {
"df": 0,
"docs": {},
"m": {
"df": 0,
"docs": {},
"l": {
"df": 1,
"docs": {
"14": {
"tf": 1.0
}
}
}
}
} }
}, },
"i": { "i": {
@ -2622,7 +2736,7 @@
"df": 1, "df": 1,
"docs": { "docs": {
"14": { "14": {
"tf": 2.0 "tf": 2.449489742783178
} }
} }
} }
@ -2675,7 +2789,7 @@
"t": { "t": {
"df": 1, "df": 1,
"docs": { "docs": {
"15": { "16": {
"tf": 1.0 "tf": 1.0
} }
} }
@ -3036,7 +3150,7 @@
"t": { "t": {
"df": 1, "df": 1,
"docs": { "docs": {
"15": { "16": {
"tf": 1.0 "tf": 1.0
} }
} }
@ -3135,7 +3249,7 @@
"docs": {}, "docs": {},
"n": { "n": {
"d": { "d": {
"df": 3, "df": 4,
"docs": { "docs": {
"13": { "13": {
"tf": 1.4142135623730951 "tf": 1.4142135623730951
@ -3143,6 +3257,9 @@
"14": { "14": {
"tf": 1.0 "tf": 1.0
}, },
"15": {
"tf": 1.0
},
"7": { "7": {
"tf": 1.0 "tf": 1.0
} }
@ -3162,11 +3279,14 @@
"df": 0, "df": 0,
"docs": {}, "docs": {},
"n": { "n": {
"df": 3, "df": 4,
"docs": { "docs": {
"14": { "14": {
"tf": 1.0 "tf": 1.0
}, },
"15": {
"tf": 1.4142135623730951
},
"3": { "3": {
"tf": 1.4142135623730951 "tf": 1.4142135623730951
}, },
@ -3450,7 +3570,7 @@
"df": 1, "df": 1,
"docs": { "docs": {
"14": { "14": {
"tf": 1.0 "tf": 1.4142135623730951
} }
} }
}, },
@ -3546,7 +3666,7 @@
"s": { "s": {
"df": 1, "df": 1,
"docs": { "docs": {
"15": { "16": {
"tf": 1.0 "tf": 1.0
} }
} }
@ -3862,8 +3982,11 @@
"df": 0, "df": 0,
"docs": {}, "docs": {},
"n": { "n": {
"df": 2, "df": 3,
"docs": { "docs": {
"15": {
"tf": 1.0
},
"3": { "3": {
"tf": 1.0 "tf": 1.0
}, },