1 # Copyright 2015 The Rust Project Developers. See the COPYRIGHT
2 # file at the top-level directory of this distribution and at
3 # http://rust-lang.org/COPYRIGHT.
5 # Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
6 # http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
7 # <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
8 # option. This file may not be copied, modified, or distributed
9 # except according to those terms.
12 htmldocck.py is a custom checker script for Rustdoc HTML outputs.
16 The principle is simple: This script receives a path to generated HTML
17 documentation and a "template" script, which has a series of check
18 commands like `@has` or `@matches`. Each command can be used to check if
19 some pattern is present or not present in the particular file or in
20 the particular node of HTML tree. In many cases, the template script
21 happens to be a source code given to rustdoc.
23 While it indeed is possible to test in smaller portions, it has been
24 hard to construct tests in this fashion and major rendering errors were
25 discovered much later. This script is designed for making the black-box
26 and regression testing of Rustdoc easy. This does not preclude the needs
27 for unit testing, but can be used to complement related tests by quickly
28 showing the expected renderings.
30 In order to avoid one-off dependencies for this task, this script uses
31 a reasonably working HTML parser and the existing XPath implementation
32 from Python 2's standard library. Hopefully we won't render
37 Commands start with an `@` followed by a command name (letters and
38 hyphens), and zero or more arguments separated by one or more whitespace
39 and optionally delimited with single or double quotes. The `@` mark
40 cannot be preceded by a non-whitespace character. Other lines (including
41 every text up to the first `@`) are ignored, but it is recommended to
42 avoid the use of `@` in the template file.
44 There are a number of supported commands:
46 * `@has PATH` checks for the existence of given file.
48 `PATH` is relative to the output directory. It can be given as `-`
49 which repeats the most recently used `PATH`.
51 * `@has PATH PATTERN` and `@matches PATH PATTERN` checks for
52 the occurrence of given `PATTERN` in the given file. Only one
53 occurrence of given pattern is enough.
55 For `@has`, `PATTERN` is a whitespace-normalized (every consecutive
56 whitespace being replaced by one single space character) string.
57 The entire file is also whitespace-normalized including newlines.
59 For `@matches`, `PATTERN` is a Python-supported regular expression.
60 The file remains intact but the regexp is matched with no `MULTILINE`
61 and `IGNORECASE` option. You can still use a prefix `(?m)` or `(?i)`
62 to override them, and `\A` and `\Z` for definitely matching
63 the beginning and end of the file.
65 (The same distinction goes to other variants of these commands.)
67 * `@has PATH XPATH PATTERN` and `@matches PATH XPATH PATTERN` checks for
68 the presence of given `XPATH` in the given HTML file, and also
69 the occurrence of given `PATTERN` in the matching node or attribute.
70 Only one occurrence of given pattern in the match is enough.
72 `PATH` should be a valid and well-formed HTML file. It does *not*
73 accept arbitrary HTML5; it should have matching open and close tags
74 and correct entity references at least.
76 `XPATH` is an XPath expression to match. This is fairly limited:
77 `tag`, `*`, `.`, `//`, `..`, `[@attr]`, `[@attr='value']`, `[tag]`,
78 `[POS]` (element located in given `POS`), `[last()-POS]`, `text()`
79 and `@attr` (both as the last segment) are supported. Some examples:
81 - `//pre` or `.//pre` matches any element with a name `pre`.
82 - `//a[@href]` matches any element with an `href` attribute.
83 - `//*[@class="impl"]//code` matches any element with a name `code`,
84 which is an ancestor of some element which `class` attr is `impl`.
85 - `//h1[@class="fqn"]/span[1]/a[last()]/@class` matches a value of
86 `class` attribute in the last `a` element (can be followed by more
87 elements that are not `a`) inside the first `span` in the `h1` with
88 a class of `fqn`. Note that there cannot be no additional elements
89 between them due to the use of `/` instead of `//`.
91 Do not try to use non-absolute paths, it won't work due to the flawed
92 ElementTree implementation. The script rejects them.
94 For the text matches (i.e. paths not ending with `@attr`), any
95 subelements are flattened into one string; this is handy for ignoring
96 highlights for example. If you want to simply check the presence of
97 given node or attribute, use an empty string (`""`) as a `PATTERN`.
99 * `@count PATH XPATH COUNT' checks for the occurrence of given XPath
100 in the given file. The number of occurrences must match the given count.
102 All conditions can be negated with `!`. `@!has foo/type.NoSuch.html`
103 checks if the given file does not exist, for example.
107 from __future__ import print_function
112 from collections import namedtuple
113 from HTMLParser import HTMLParser
114 from xml.etree import cElementTree as ET
116 # ⇤/⇥ are not in HTML 4 but are in HTML 5
117 from htmlentitydefs import entitydefs
118 entitydefs['larrb'] = u'\u21e4'
119 entitydefs['rarrb'] = u'\u21e5'
121 # "void elements" (no closing tag) from the HTML Standard section 12.1.2
122 VOID_ELEMENTS = set(['area', 'base', 'br', 'col', 'embed', 'hr', 'img', 'input', 'keygen',
123 'link', 'menuitem', 'meta', 'param', 'source', 'track', 'wbr'])
126 class CustomHTMLParser(HTMLParser):
127 """simplified HTML parser.
129 this is possible because we are dealing with very regular HTML from
130 rustdoc; we only have to deal with i) void elements and ii) empty
132 def __init__(self, target=None):
133 HTMLParser.__init__(self)
134 self.__builder = target or ET.TreeBuilder()
136 def handle_starttag(self, tag, attrs):
137 attrs = dict((k, v or '') for k, v in attrs)
138 self.__builder.start(tag, attrs)
139 if tag in VOID_ELEMENTS:
140 self.__builder.end(tag)
142 def handle_endtag(self, tag):
143 self.__builder.end(tag)
145 def handle_startendtag(self, tag, attrs):
146 attrs = dict((k, v or '') for k, v in attrs)
147 self.__builder.start(tag, attrs)
148 self.__builder.end(tag)
150 def handle_data(self, data):
151 self.__builder.data(data)
153 def handle_entityref(self, name):
154 self.__builder.data(entitydefs[name])
156 def handle_charref(self, name):
157 code = int(name[1:], 16) if name.startswith(('x', 'X')) else int(name, 10)
158 self.__builder.data(unichr(code).encode('utf-8'))
161 HTMLParser.close(self)
162 return self.__builder.close()
164 Command = namedtuple('Command', 'negated cmd args lineno context')
166 class FailedCheck(Exception):
169 class InvalidCheck(Exception):
172 def concat_multi_lines(f):
173 """returns a generator out of the file object, which
174 - removes `\\` then `\n` then a shared prefix with the previous line then
176 - keeps a line number (starting from 0) of the first line being
178 lastline = None # set to the last line when the last line has a backslash
181 for lineno, line in enumerate(f):
182 line = line.rstrip('\r\n')
184 # strip the common prefix from the current line if needed
185 if lastline is not None:
187 for i in xrange(min(len(line), len(lastline))):
188 if line[i] != lastline[i]:
191 line = line[maxprefix:].lstrip()
193 firstlineno = firstlineno or lineno
194 if line.endswith('\\'):
197 catenated += line[:-1]
199 yield firstlineno, catenated + line
204 if lastline is not None:
205 print_err(lineno, line, 'Trailing backslash at the end of the file')
207 LINE_PATTERN = re.compile(r'''
208 (?<=(?<!\S)@)(?P<negated>!?)
209 (?P<cmd>[A-Za-z]+(?:-[A-Za-z]+)*)
214 def get_commands(template):
215 with open(template, 'rUb') as f:
216 for lineno, line in concat_multi_lines(f):
217 m = LINE_PATTERN.search(line)
221 negated = (m.group('negated') == '!')
223 args = m.group('args')
224 if args and not args[:1].isspace():
225 print_err(lineno, line, 'Invalid template syntax')
227 args = shlex.split(args)
228 yield Command(negated=negated, cmd=cmd, args=args, lineno=lineno+1, context=line)
231 def _flatten(node, acc):
233 acc.append(node.text)
246 def normalize_xpath(path):
247 if path.startswith('//'):
248 return '.' + path # avoid warnings
249 elif path.startswith('.//'):
252 raise InvalidCheck('Non-absolute XPath is not supported due to implementation issues')
255 class CachedFiles(object):
256 def __init__(self, root):
260 self.last_path = None
262 def resolve_path(self, path):
264 path = os.path.normpath(path)
265 self.last_path = path
267 elif self.last_path is None:
268 raise InvalidCheck('Tried to use the previous path in the first command')
270 return self.last_path
272 def get_file(self, path):
273 path = self.resolve_path(path)
274 if path in self.files:
275 return self.files[path]
277 abspath = os.path.join(self.root, path)
278 if not(os.path.exists(abspath) and os.path.isfile(abspath)):
279 raise FailedCheck('File does not exist {!r}'.format(path))
281 with open(abspath) as f:
283 self.files[path] = data
286 def get_tree(self, path):
287 path = self.resolve_path(path)
288 if path in self.trees:
289 return self.trees[path]
291 abspath = os.path.join(self.root, path)
292 if not(os.path.exists(abspath) and os.path.isfile(abspath)):
293 raise FailedCheck('File does not exist {!r}'.format(path))
295 with open(abspath) as f:
297 tree = ET.parse(f, CustomHTMLParser())
298 except Exception as e:
299 raise RuntimeError('Cannot parse an HTML file {!r}: {}'.format(path, e))
300 self.trees[path] = tree
301 return self.trees[path]
304 def check_string(data, pat, regexp):
306 return True # special case a presence testing
308 return re.search(pat, data) is not None
310 data = ' '.join(data.split())
311 pat = ' '.join(pat.split())
315 def check_tree_attr(tree, path, attr, pat, regexp):
316 path = normalize_xpath(path)
318 for e in tree.findall(path):
320 value = e.attrib[attr]
324 ret = check_string(value, pat, regexp)
330 def check_tree_text(tree, path, pat, regexp):
331 path = normalize_xpath(path)
333 for e in tree.findall(path):
339 ret = check_string(value, pat, regexp)
345 def check_tree_count(tree, path, count):
346 path = normalize_xpath(path)
347 return len(tree.findall(path)) == count
350 print(*args, file=sys.stderr)
352 def print_err(lineno, context, err, message=None):
355 stderr("{}: {}".format(lineno, message or err))
357 stderr("\t{}".format(err))
360 stderr("\t{}".format(context))
364 def check_command(c, cache):
367 if c.cmd == 'has' or c.cmd == 'matches': # string test
368 regexp = (c.cmd == 'matches')
369 if len(c.args) == 1 and not regexp: # @has <path> = file existence
371 cache.get_file(c.args[0])
373 except FailedCheck as err:
376 elif len(c.args) == 2: # @has/matches <path> <pat> = string test
377 cerr = "`PATTERN` did not match"
378 ret = check_string(cache.get_file(c.args[0]), c.args[1], regexp)
379 elif len(c.args) == 3: # @has/matches <path> <pat> <match> = XML tree test
380 cerr = "`XPATH PATTERN` did not match"
381 tree = cache.get_tree(c.args[0])
382 pat, sep, attr = c.args[1].partition('/@')
384 tree = cache.get_tree(c.args[0])
385 ret = check_tree_attr(tree, pat, attr, c.args[2], regexp)
386 else: # normalized text
388 if pat.endswith('/text()'):
390 ret = check_tree_text(cache.get_tree(c.args[0]), pat, c.args[2], regexp)
392 raise InvalidCheck('Invalid number of @{} arguments'.format(c.cmd))
394 elif c.cmd == 'count': # count test
395 if len(c.args) == 3: # @count <path> <pat> <count> = count test
396 ret = check_tree_count(cache.get_tree(c.args[0]), c.args[1], int(c.args[2]))
398 raise InvalidCheck('Invalid number of @{} arguments'.format(c.cmd))
399 elif c.cmd == 'valid-html':
400 raise InvalidCheck('Unimplemented @valid-html')
402 elif c.cmd == 'valid-links':
403 raise InvalidCheck('Unimplemented @valid-links')
405 raise InvalidCheck('Unrecognized @{}'.format(c.cmd))
408 raise FailedCheck(cerr)
410 except FailedCheck as err:
411 message = '@{}{} check failed'.format('!' if c.negated else '', c.cmd)
412 print_err(c.lineno, c.context, err.message, message)
413 except InvalidCheck as err:
414 print_err(c.lineno, c.context, err.message)
416 def check(target, commands):
417 cache = CachedFiles(target)
419 check_command(c, cache)
421 if __name__ == '__main__':
422 if len(sys.argv) != 3:
423 stderr('Usage: {} <doc dir> <template>'.format(sys.argv[0]))
426 check(sys.argv[1], get_commands(sys.argv[2]))
428 stderr("\nEncountered {} errors".format(ERR_COUNT))