File: main.rs

package info (click to toggle)
rustc 1.46.0+dfsg1-1
  • links: PTS, VCS
  • area: main
  • in suites: bullseye, sid
  • size: 269,876 kB
  • sloc: xml: 133,832; sh: 16,876; ansic: 16,844; javascript: 6,779; python: 6,128; cpp: 4,555; makefile: 3,186; asm: 1,392; ruby: 68; awk: 12
file content (370 lines) | stat: -rw-r--r-- 13,066 bytes parent folder | download
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
//! Script to check the validity of `href` links in our HTML documentation.
//!
//! In the past we've been quite error prone to writing in broken links as most
//! of them are manually rather than automatically added. As files move over
//! time or apis change old links become stale or broken. The purpose of this
//! script is to check all relative links in our documentation to make sure they
//! actually point to a valid place.
//!
//! Currently this doesn't actually do any HTML parsing or anything fancy like
//! that, it just has a simple "regex" to search for `href` and `id` tags.
//! These values are then translated to file URLs if possible and then the
//! destination is asserted to exist.
//!
//! A few exceptions are allowed as there's known bugs in rustdoc, but this
//! should catch the majority of "broken link" cases.

use std::collections::hash_map::Entry;
use std::collections::{HashMap, HashSet};
use std::env;
use std::fs;
use std::path::{Component, Path, PathBuf};
use std::rc::Rc;

use crate::Redirect::*;

macro_rules! t {
    ($e:expr) => {
        match $e {
            Ok(e) => e,
            Err(e) => panic!("{} failed with {:?}", stringify!($e), e),
        }
    };
}

fn main() {
    let docs = env::args_os().nth(1).unwrap();
    let docs = env::current_dir().unwrap().join(docs);
    let mut errors = false;
    walk(&mut HashMap::new(), &docs, &docs, &mut errors);
    if errors {
        panic!("found some broken links");
    }
}

#[derive(Debug)]
pub enum LoadError {
    IOError(std::io::Error),
    BrokenRedirect(PathBuf, std::io::Error),
    IsRedirect,
}

enum Redirect {
    SkipRedirect,
    FromRedirect(bool),
}

struct FileEntry {
    source: Rc<String>,
    ids: HashSet<String>,
}

type Cache = HashMap<PathBuf, FileEntry>;

fn small_url_encode(s: &str) -> String {
    s.replace("<", "%3C")
        .replace(">", "%3E")
        .replace(" ", "%20")
        .replace("?", "%3F")
        .replace("'", "%27")
        .replace("&", "%26")
        .replace(",", "%2C")
        .replace(":", "%3A")
        .replace(";", "%3B")
        .replace("[", "%5B")
        .replace("]", "%5D")
        .replace("\"", "%22")
}

impl FileEntry {
    fn parse_ids(&mut self, file: &Path, contents: &str, errors: &mut bool) {
        if self.ids.is_empty() {
            with_attrs_in_source(contents, " id", |fragment, i, _| {
                let frag = fragment.trim_start_matches("#").to_owned();
                let encoded = small_url_encode(&frag);
                if !self.ids.insert(frag) {
                    *errors = true;
                    println!("{}:{}: id is not unique: `{}`", file.display(), i, fragment);
                }
                // Just in case, we also add the encoded id.
                self.ids.insert(encoded);
            });
        }
    }
}

fn walk(cache: &mut Cache, root: &Path, dir: &Path, errors: &mut bool) {
    for entry in t!(dir.read_dir()).map(|e| t!(e)) {
        let path = entry.path();
        let kind = t!(entry.file_type());
        if kind.is_dir() {
            walk(cache, root, &path, errors);
        } else {
            let pretty_path = check(cache, root, &path, errors);
            if let Some(pretty_path) = pretty_path {
                let entry = cache.get_mut(&pretty_path).unwrap();
                // we don't need the source anymore,
                // so drop to reduce memory-usage
                entry.source = Rc::new(String::new());
            }
        }
    }
}

fn check(cache: &mut Cache, root: &Path, file: &Path, errors: &mut bool) -> Option<PathBuf> {
    // Ignore non-HTML files.
    if file.extension().and_then(|s| s.to_str()) != Some("html") {
        return None;
    }

    // Unfortunately we're not 100% full of valid links today to we need a few
    // exceptions to get this past `make check` today.
    // FIXME(#32129)
    if file.ends_with("std/io/struct.IoSlice.html")
        || file.ends_with("std/string/struct.String.html")
    {
        return None;
    }
    // FIXME(#32553)
    if file.ends_with("alloc/string/struct.String.html") {
        return None;
    }
    // FIXME(#32130)
    if file.ends_with("alloc/collections/btree_map/struct.BTreeMap.html")
        || file.ends_with("alloc/collections/btree_set/struct.BTreeSet.html")
        || file.ends_with("std/collections/btree_map/struct.BTreeMap.html")
        || file.ends_with("std/collections/btree_set/struct.BTreeSet.html")
        || file.ends_with("std/collections/hash_map/struct.HashMap.html")
        || file.ends_with("std/collections/hash_set/struct.HashSet.html")
    {
        return None;
    }

    let res = load_file(cache, root, file, SkipRedirect);
    let (pretty_file, contents) = match res {
        Ok(res) => res,
        Err(_) => return None,
    };
    {
        cache.get_mut(&pretty_file).unwrap().parse_ids(&pretty_file, &contents, errors);
    }

    // Search for anything that's the regex 'href[ ]*=[ ]*".*?"'
    with_attrs_in_source(&contents, " href", |url, i, base| {
        // Ignore external URLs
        if url.starts_with("http:")
            || url.starts_with("https:")
            || url.starts_with("javascript:")
            || url.starts_with("ftp:")
            || url.starts_with("irc:")
            || url.starts_with("data:")
        {
            return;
        }
        // Ignore parent URLs, so that the package installation process can
        // provide a symbolic link later
        if url.starts_with("../") {
            return;
        }
        let mut parts = url.splitn(2, "#");
        let url = parts.next().unwrap();
        let fragment = parts.next();
        let mut parts = url.splitn(2, "?");
        let url = parts.next().unwrap();

        // Once we've plucked out the URL, parse it using our base url and
        // then try to extract a file path.
        let mut path = file.to_path_buf();
        if !base.is_empty() || !url.is_empty() {
            path.pop();
            for part in Path::new(base).join(url).components() {
                match part {
                    Component::Prefix(_) | Component::RootDir => {
                        // Avoid absolute paths as they make the docs not
                        // relocatable by making assumptions on where the docs
                        // are hosted relative to the site root.
                        *errors = true;
                        println!(
                            "{}:{}: absolute path - {}",
                            pretty_file.display(),
                            i + 1,
                            Path::new(base).join(url).display()
                        );
                        return;
                    }
                    Component::CurDir => {}
                    Component::ParentDir => {
                        path.pop();
                    }
                    Component::Normal(s) => {
                        path.push(s);
                    }
                }
            }
        }

        // Alright, if we've found a file name then this file had better
        // exist! If it doesn't then we register and print an error.
        if path.exists() {
            if path.is_dir() {
                // Links to directories show as directory listings when viewing
                // the docs offline so it's best to avoid them.
                *errors = true;
                let pretty_path = path.strip_prefix(root).unwrap_or(&path);
                println!(
                    "{}:{}: directory link - {}",
                    pretty_file.display(),
                    i + 1,
                    pretty_path.display()
                );
                return;
            }
            if let Some(extension) = path.extension() {
                // Ignore none HTML files.
                if extension != "html" {
                    return;
                }
            }
            let res = load_file(cache, root, &path, FromRedirect(false));
            let (pretty_path, contents) = match res {
                Ok(res) => res,
                Err(LoadError::IOError(err)) => {
                    panic!("error loading {}: {}", path.display(), err);
                }
                Err(LoadError::BrokenRedirect(target, _)) => {
                    *errors = true;
                    println!(
                        "{}:{}: broken redirect to {}",
                        pretty_file.display(),
                        i + 1,
                        target.display()
                    );
                    return;
                }
                Err(LoadError::IsRedirect) => unreachable!(),
            };

            if let Some(ref fragment) = fragment {
                // Fragments like `#1-6` are most likely line numbers to be
                // interpreted by javascript, so we're ignoring these
                if fragment.splitn(2, '-').all(|f| f.chars().all(|c| c.is_numeric())) {
                    return;
                }

                // These appear to be broken in mdbook right now?
                if fragment.starts_with("-") {
                    return;
                }

                let entry = &mut cache.get_mut(&pretty_path).unwrap();
                entry.parse_ids(&pretty_path, &contents, errors);

                if !entry.ids.contains(*fragment) {
                    *errors = true;
                    print!("{}:{}: broken link fragment ", pretty_file.display(), i + 1);
                    println!("`#{}` pointing to `{}`", fragment, pretty_path.display());
                };
            }
        } else {
            *errors = true;
            print!("{}:{}: broken link - ", pretty_file.display(), i + 1);
            let pretty_path = path.strip_prefix(root).unwrap_or(&path);
            println!("{}", pretty_path.display());
        }
    });
    Some(pretty_file)
}

fn load_file(
    cache: &mut Cache,
    root: &Path,
    file: &Path,
    redirect: Redirect,
) -> Result<(PathBuf, Rc<String>), LoadError> {
    let pretty_file = PathBuf::from(file.strip_prefix(root).unwrap_or(&file));

    let (maybe_redirect, contents) = match cache.entry(pretty_file.clone()) {
        Entry::Occupied(entry) => (None, entry.get().source.clone()),
        Entry::Vacant(entry) => {
            let contents = match fs::read_to_string(file) {
                Ok(s) => Rc::new(s),
                Err(err) => {
                    return Err(if let FromRedirect(true) = redirect {
                        LoadError::BrokenRedirect(file.to_path_buf(), err)
                    } else {
                        LoadError::IOError(err)
                    });
                }
            };

            let maybe = maybe_redirect(&contents);
            if maybe.is_some() {
                if let SkipRedirect = redirect {
                    return Err(LoadError::IsRedirect);
                }
            } else {
                entry.insert(FileEntry { source: contents.clone(), ids: HashSet::new() });
            }
            (maybe, contents)
        }
    };
    match maybe_redirect.map(|url| file.parent().unwrap().join(url)) {
        Some(redirect_file) => load_file(cache, root, &redirect_file, FromRedirect(true)),
        None => Ok((pretty_file, contents)),
    }
}

fn maybe_redirect(source: &str) -> Option<String> {
    const REDIRECT: &'static str = "<p>Redirecting to <a href=";

    let mut lines = source.lines();
    let redirect_line = lines.nth(6)?;

    redirect_line.find(REDIRECT).map(|i| {
        let rest = &redirect_line[(i + REDIRECT.len() + 1)..];
        let pos_quote = rest.find('"').unwrap();
        rest[..pos_quote].to_owned()
    })
}

fn with_attrs_in_source<F: FnMut(&str, usize, &str)>(contents: &str, attr: &str, mut f: F) {
    let mut base = "";
    for (i, mut line) in contents.lines().enumerate() {
        while let Some(j) = line.find(attr) {
            let rest = &line[j + attr.len()..];
            // The base tag should always be the first link in the document so
            // we can get away with using one pass.
            let is_base = line[..j].ends_with("<base");
            line = rest;
            let pos_equals = match rest.find("=") {
                Some(i) => i,
                None => continue,
            };
            if rest[..pos_equals].trim_start_matches(" ") != "" {
                continue;
            }

            let rest = &rest[pos_equals + 1..];

            let pos_quote = match rest.find(&['"', '\''][..]) {
                Some(i) => i,
                None => continue,
            };
            let quote_delim = rest.as_bytes()[pos_quote] as char;

            if rest[..pos_quote].trim_start_matches(" ") != "" {
                continue;
            }
            let rest = &rest[pos_quote + 1..];
            let url = match rest.find(quote_delim) {
                Some(i) => &rest[..i],
                None => continue,
            };
            if is_base {
                base = url;
                continue;
            }
            f(url, i, base)
        }
    }
}