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
|
import dataclasses
import glob
import os
import os.path
import re
import sys
import urllib.parse
import bs4
@dataclasses.dataclass
class Link:
href: str
parsed_href: urllib.parse.ParseResult
line_no: int
col_no: int
def main(args):
if len(args) != 2:
print('Usage: check-rellinks dir')
sys.exit(1)
os.chdir(args[1])
filenames = glob.glob('**/*.html', recursive=True)
targets = {}
rellinks = {}
for filename in filenames:
with open(filename) as f:
soup = bs4.BeautifulSoup(f, 'html.parser')
targets[filename] = [e['id'] for e in soup.find_all(id=True)]
links = [Link(href=e['href'], parsed_href=urllib.parse.urlparse(e['href']),
line_no=e.sourceline, col_no=e.sourcepos+1) # e.sourcepos is 0-based
for e in soup.find_all('a', href=True)]
rellinks[filename] = [link for link in links if link.parsed_href.scheme == '']
def check(path, fragment):
return path in targets and (fragment == '' or fragment in targets[path])
has_broken = False
for srcfile in rellinks:
# Old release notes may refer to docs that have since been removed, so
# don't check them.
if srcfile.endswith('-release-notes.html'):
continue
for link in rellinks[srcfile]:
# Special-case the relative link to the feed; all the other links
# are to HTML files.
if link.href == 'feed.atom':
continue
def report_that_link(does_what):
print(f'{srcfile}:{link.line_no}:{link.col_no}: {link.href} {does_what}')
nonlocal has_broken
has_broken = True
# Check that the path part has slashes in the expected places.
# Slashes are significant for resolving relative links; relative
# links with missing or superfluous slashes can still lead the
# browser to the correct file, but will break the relative links on
# the target page.
#
# For example, a relative link in ref/str.html to ./re.html/ will
# navigate to the expected ref/re.html, but when on that page, a
# relative link from ref/re.html to ./math.html will resolve
# incorrectly to ref/re.html/math.html due to the trailing slash.
#
# These paths are less harmful in production since the HTTP server
# of https://elv.sh knows how to redirect them to the correct URL,
# but this still results in unnecessary roundtrips with the server
# and is best avoided altogether.
path = link.parsed_href.path
# Duplicate slashes like ..//ref/
if '//' in path:
report_that_link('has duplicate slashes')
continue
# Links relative to the root like /ref are valid on the website but
# breaks local previewing.
if path.startswith('/'):
report_that_link('is relative to the root and breaks local previewing')
continue
if path.endswith('/'):
# Paths ending with / should be directory links. Check that it's
# not a file link with a superfluous trailing slash, like
# ./re.html/.
if path.endswith('.html/'):
report_that_link('has a superfluous trailing slash')
continue
# Paths not ending with / should be either empty or a file link.
# Check that it's not a directory link with a missing trailing
# slash, like ../ref.
elif path != '' and not path.endswith('.html'):
report_that_link('lacks a trailing slash')
continue
# Now check if the link target is valid.
if path == '':
dstfile = srcfile
else:
if path.endswith('/'):
path += 'index.html'
dstfile = os.path.normpath(os.path.join(os.path.dirname(srcfile), path))
if dstfile not in targets:
report_that_link(f'links to non-existing page {dstfile}')
continue
fragment = link.parsed_href.fragment
if fragment != '' and fragment not in targets[dstfile]:
report_that_link(f'links to non-existing target {fragment} in {dstfile}')
if has_broken:
sys.exit(1)
if __name__ == '__main__':
main(sys.argv)
|
