2 Check lesson files and their contents.
9 from argparse import ArgumentParser
11 from util import (Reporter, read_markdown, load_yaml, check_unwanted_files,
16 # Where to look for source Markdown files.
17 SOURCE_DIRS = ['', '_episodes', '_extras']
19 # Where to look for source Rmd files.
20 SOURCE_RMD_DIRS = ['_episodes_rmd']
22 # Required files: each entry is ('path': YAML_required).
23 # FIXME: We do not yet validate whether any files have the required
24 # YAML headers, but should in the future.
25 # The '%' is replaced with the source directory path for checking.
26 # Episodes are handled specially, and extra files in '_extras' are also handled
27 # specially. This list must include all the Markdown files listed in the
28 # 'bin/initialize' script.
30 'CODE_OF_CONDUCT.md': True,
31 'CONTRIBUTING.md': False,
34 os.path.join('_extras', 'discuss.md'): True,
35 os.path.join('_extras', 'guide.md'): True,
41 # Episode filename pattern.
42 P_EPISODE_FILENAME = re.compile(r'(\d\d)-[-\w]+.md$')
44 # Pattern to match lines ending with whitespace.
45 P_TRAILING_WHITESPACE = re.compile(r'\s+$')
47 # Pattern to match figure references in HTML.
48 P_FIGURE_REFS = re.compile(r'<img[^>]+src="([^"]+)"[^>]*>')
50 # Pattern to match internally-defined Markdown links.
51 P_INTERNAL_LINK_REF = re.compile(r'\[([^\]]+)\]\[([^\]]+)\]')
53 # Pattern to match reference links (to resolve internally-defined references).
54 P_INTERNAL_LINK_DEF = re.compile(r'^\[([^\]]+)\]:\s*(.+)')
56 # Pattern to match {% include ... %} statements
57 P_INTERNAL_INCLUDE_LINK = re.compile(r'^{% include ([^ ]*) %}$')
59 # What kinds of blockquotes are allowed?
75 # What kinds of code fragments are allowed?
76 # Below we allow all 'language-*' code blocks
84 # What fields are required in teaching episode metadata?
85 TEACHING_METADATA_FIELDS = {
94 # What fields are required in break episode metadata?
95 BREAK_METADATA_FIELDS = {
101 # How long are lines allowed to be?
102 # Please keep this in sync with .editorconfig!
110 args.reporter = Reporter()
111 life_cycle = check_config(args.reporter, args.source_dir)
112 # pre-alpha lessons should report without error
113 if life_cycle == "pre-alpha":
114 args.permissive = True
115 check_source_rmd(args.reporter, args.source_dir, args.parser)
116 args.references = read_references(args.reporter, args.reference_path)
118 docs = read_all_markdown(args.source_dir, args.parser)
119 check_fileset(args.source_dir, args.reporter, list(docs.keys()))
120 check_unwanted_files(args.source_dir, args.reporter)
121 for filename in list(docs.keys()):
122 checker = create_checker(args, filename, docs[filename])
125 args.reporter.report()
126 if args.reporter.messages and not args.permissive:
131 """Parse command-line arguments."""
133 parser = ArgumentParser(description="""Check episode files in a lesson.""")
134 parser.add_argument('-l', '--linelen',
138 help='Check line lengths')
139 parser.add_argument('-p', '--parser',
142 help='path to Markdown parser')
143 parser.add_argument('-r', '--references',
145 dest='reference_path',
146 help='path to Markdown file of external references')
147 parser.add_argument('-s', '--source',
150 help='source directory')
151 parser.add_argument('-w', '--whitespace',
154 dest='trailing_whitespace',
155 help='Check for trailing whitespace')
156 parser.add_argument('--permissive',
160 help='Do not raise an error even if issues are detected')
162 args, extras = parser.parse_known_args()
163 require(args.parser is not None,
164 'Path to Markdown parser not provided')
166 'Unexpected trailing command-line arguments "{0}"'.format(extras))
171 def check_config(reporter, source_dir):
172 """Check configuration file."""
174 config_file = os.path.join(source_dir, '_config.yml')
175 config = load_yaml(config_file)
176 reporter.check_field(config_file, 'configuration',
177 config, 'kind', 'lesson')
178 reporter.check_field(config_file, 'configuration',
179 config, 'carpentry', ('swc', 'dc', 'lc', 'cp'))
180 reporter.check_field(config_file, 'configuration', config, 'title')
181 reporter.check_field(config_file, 'configuration', config, 'email')
184 {'values': {'root': '.', 'layout': 'page'}},
185 {'values': {'root': '..', 'layout': 'episode'}, 'scope': {'type': 'episodes', 'path': ''}},
186 {'values': {'root': '..', 'layout': 'page'}, 'scope': {'type': 'extras', 'path': ''}}
188 reporter.check(defaults in config.get('defaults', []),
190 '"root" not set to "." in configuration')
191 return config['life_cycle']
193 def check_source_rmd(reporter, source_dir, parser):
194 """Check that Rmd episode files include `source: Rmd`"""
196 episode_rmd_dir = [os.path.join(source_dir, d) for d in SOURCE_RMD_DIRS]
197 episode_rmd_files = [os.path.join(d, '*.Rmd') for d in episode_rmd_dir]
199 for pat in episode_rmd_files:
200 for f in glob.glob(pat):
201 data = read_markdown(parser, f)
202 dy = data['metadata']
204 reporter.check_field(f, 'episode_rmd',
207 def read_references(reporter, ref_path):
208 """Read shared file of reference links, returning dictionary of valid references
209 {symbolic_name : URL}
213 raise Warning("No filename has been provided.")
218 with open(ref_path, 'r', encoding='utf-8') as reader:
219 for (num, line) in enumerate(reader, 1):
221 if P_INTERNAL_INCLUDE_LINK.search(line): continue
223 m = P_INTERNAL_LINK_DEF.search(line)
225 message = '{}: {} not a valid reference: {}'
226 require(m, message.format(ref_path, num, line.rstrip()))
231 message = 'Empty reference at {0}:{1}'
232 require(name, message.format(ref_path, num))
234 unique_name = name not in result
235 unique_url = url not in urls_seen
237 reporter.check(unique_name,
239 'Duplicate reference name {0} at line {1}',
242 reporter.check(unique_url,
244 'Duplicate definition of URL {0} at line {1}',
253 def read_all_markdown(source_dir, parser):
254 """Read source files, returning
255 {path : {'metadata':yaml, 'metadata_len':N, 'text':text, 'lines':[(i, line, len)], 'doc':doc}}
258 all_dirs = [os.path.join(source_dir, d) for d in SOURCE_DIRS]
259 all_patterns = [os.path.join(d, '*.md') for d in all_dirs]
261 for pat in all_patterns:
262 for filename in glob.glob(pat):
263 data = read_markdown(parser, filename)
265 result[filename] = data
269 def check_fileset(source_dir, reporter, filenames_present):
270 """Are all required files present? Are extraneous files present?"""
272 # Check files with predictable names.
273 required = [os.path.join(source_dir, p) for p in REQUIRED_FILES]
274 missing = set(required) - set(filenames_present)
276 reporter.add(None, 'Missing required file {0}', m)
278 # Check episode files' names.
280 for filename in filenames_present:
281 if '_episodes' not in filename:
284 # split path to check episode name
285 base_name = os.path.basename(filename)
286 m = P_EPISODE_FILENAME.search(base_name)
288 seen.append(m.group(1))
291 None, 'Episode {0} has badly-formatted filename', filename)
293 # Check for duplicate episode numbers.
294 reporter.check(len(seen) == len(set(seen)),
296 'Duplicate episode numbers {0} vs {1}',
297 sorted(seen), sorted(set(seen)))
299 # Check that numbers are consecutive.
300 seen = sorted([int(s) for s in seen])
302 for i in range(len(seen) - 1):
303 clean = clean and ((seen[i+1] - seen[i]) == 1)
304 reporter.check(clean,
306 'Missing or non-consecutive episode numbers {0}',
310 def create_checker(args, filename, info):
311 """Create appropriate checker for file."""
313 for (pat, cls) in CHECKERS:
314 if pat.search(filename):
315 return cls(args, filename, **info)
316 return NotImplemented
319 """Base class for checking Markdown files."""
321 def __init__(self, args, filename, metadata, metadata_len, text, lines, doc):
322 """Cache arguments for checking."""
325 self.reporter = self.args.reporter # for convenience
326 self.filename = filename
327 self.metadata = metadata
328 self.metadata_len = metadata_len
338 self.check_metadata()
339 self.check_line_lengths()
340 self.check_trailing_whitespace()
341 self.check_blockquote_classes()
342 self.check_codeblock_classes()
343 self.check_defined_link_references()
345 def check_metadata(self):
346 """Check the YAML metadata."""
348 self.reporter.check(self.metadata is not None,
350 'Missing metadata entirely')
352 if self.metadata and (self.layout is not None):
353 self.reporter.check_field(
354 self.filename, 'metadata', self.metadata, 'layout', self.layout)
356 def check_line_lengths(self):
357 """Check the raw text of the lesson body."""
359 if self.args.line_lengths:
360 over = [i for (i, l, n) in self.lines if (
361 n > MAX_LINE_LEN) and (not l.startswith('!'))]
362 self.reporter.check(not over,
364 'Line(s) too long: {0}',
365 ', '.join([str(i) for i in over]))
367 def check_trailing_whitespace(self):
368 """Check for whitespace at the ends of lines."""
370 if self.args.trailing_whitespace:
372 i for (i, l, n) in self.lines if P_TRAILING_WHITESPACE.match(l)]
373 self.reporter.check(not trailing,
375 'Line(s) end with whitespace: {0}',
376 ', '.join([str(i) for i in trailing]))
378 def check_blockquote_classes(self):
379 """Check that all blockquotes have known classes."""
381 for node in self.find_all(self.doc, {'type': 'blockquote'}):
382 cls = self.get_val(node, 'attr', 'class')
383 self.reporter.check(cls in KNOWN_BLOCKQUOTES,
384 (self.filename, self.get_loc(node)),
385 'Unknown or missing blockquote type {0}',
388 def check_codeblock_classes(self):
389 """Check that all code blocks have known classes."""
391 for node in self.find_all(self.doc, {'type': 'codeblock'}):
392 cls = self.get_val(node, 'attr', 'class')
393 self.reporter.check(cls in KNOWN_CODEBLOCKS or cls.startswith('language-'),
394 (self.filename, self.get_loc(node)),
395 'Unknown or missing code block type {0}',
398 def check_defined_link_references(self):
399 """Check that defined links resolve in the file.
401 Internally-defined links match the pattern [text][label].
405 for node in self.find_all(self.doc, {'type': 'text'}):
406 for match in P_INTERNAL_LINK_REF.findall(node['value']):
409 if link not in self.args.references:
410 result.add('"{0}"=>"{1}"'.format(text, link))
411 self.reporter.check(not result,
413 'Internally-defined links may be missing definitions: {0}',
414 ', '.join(sorted(result)))
416 def find_all(self, node, pattern, accum=None):
417 """Find all matches for a pattern."""
419 assert isinstance(pattern, dict), 'Patterns must be dictionaries'
422 if self.match(node, pattern):
424 for child in node.get('children', []):
425 self.find_all(child, pattern, accum)
428 def match(self, node, pattern):
429 """Does this node match the given pattern?"""
435 if isinstance(val, str):
438 elif isinstance(val, dict):
439 if not self.match(node[key], val):
444 def get_val(node, *chain):
445 """Get value one or more levels down."""
448 for selector in chain:
449 curr = curr.get(selector, None)
454 def get_loc(self, node):
455 """Convenience method to get node's line number."""
457 result = self.get_val(node, 'options', 'location')
458 if self.metadata_len is not None:
459 result += self.metadata_len
463 class CheckNonJekyll(CheckBase):
464 """Check a file that isn't translated by Jekyll."""
466 def check_metadata(self):
467 self.reporter.check(self.metadata is None,
469 'Unexpected metadata')
472 class CheckIndex(CheckBase):
473 """Check the main index page."""
475 def __init__(self, args, filename, metadata, metadata_len, text, lines, doc):
476 super().__init__(args, filename, metadata, metadata_len, text, lines, doc)
477 self.layout = 'lesson'
479 def check_metadata(self):
480 super().check_metadata()
481 self.reporter.check(self.metadata.get('root', '') == '.',
483 'Root not set to "."')
486 class CheckEpisode(CheckBase):
487 """Check an episode page."""
490 """Run extra tests."""
493 self.check_reference_inclusion()
495 def check_metadata(self):
496 super().check_metadata()
498 if 'layout' in self.metadata:
499 if self.metadata['layout'] == 'break':
500 self.check_metadata_fields(BREAK_METADATA_FIELDS)
502 self.reporter.add(self.filename,
503 'Unknown episode layout "{0}"',
504 self.metadata['layout'])
506 self.check_metadata_fields(TEACHING_METADATA_FIELDS)
508 def check_metadata_fields(self, expected):
509 """Check metadata fields."""
510 for (name, type_) in expected:
511 if name not in self.metadata:
512 self.reporter.add(self.filename,
513 'Missing metadata field {0}',
515 elif not isinstance(self.metadata[name], type_):
516 self.reporter.add(self.filename,
517 '"{0}" has wrong type in metadata ({1} instead of {2})',
518 name, type(self.metadata[name]), type_)
520 def check_reference_inclusion(self):
521 """Check that links file has been included."""
523 if not self.args.reference_path:
526 for (i, last_line, line_len) in reversed(self.lines):
531 'No non-empty lines in {0}'.format(self.filename))
533 include_filename = os.path.split(self.args.reference_path)[-1]
534 if include_filename not in last_line:
535 self.reporter.add(self.filename,
536 'episode does not include "{0}"',
540 class CheckReference(CheckBase):
541 """Check the reference page."""
543 def __init__(self, args, filename, metadata, metadata_len, text, lines, doc):
544 super().__init__(args, filename, metadata, metadata_len, text, lines, doc)
545 self.layout = 'reference'
548 class CheckGeneric(CheckBase):
549 """Check a generic page."""
551 def __init__(self, args, filename, metadata, metadata_len, text, lines, doc):
552 super().__init__(args, filename, metadata, metadata_len, text, lines, doc)
556 (re.compile(r'CONTRIBUTING\.md'), CheckNonJekyll),
557 (re.compile(r'README\.md'), CheckNonJekyll),
558 (re.compile(r'index\.md'), CheckIndex),
559 (re.compile(r'reference\.md'), CheckReference),
560 (re.compile(os.path.join('_episodes', '*\.md')), CheckEpisode),
561 (re.compile(r'.*\.md'), CheckGeneric)
565 if __name__ == '__main__':