]>
git.ipfire.org Git - ipfire.org.git/blob - src/backend/wiki.py
7 import markdown
.extensions
8 import markdown
.preprocessors
15 from .decorators
import *
17 class Wiki(misc
.Object
):
18 def _get_pages(self
, query
, *args
):
19 res
= self
.db
.query(query
, *args
)
22 yield Page(self
.backend
, row
.id, data
=row
)
24 def _get_page(self
, query
, *args
):
25 res
= self
.db
.get(query
, *args
)
28 return Page(self
.backend
, res
.id, data
=res
)
31 return self
._get
_pages
("""
37 wiki ON current.id = wiki.id
39 current.deleted IS FALSE
44 def make_path(self
, page
, path
):
45 # Nothing to do for absolute links
46 if path
.startswith("/"):
49 # Relative links (one-level down)
50 elif path
.startswith("./"):
51 path
= os
.path
.join(page
, path
)
53 # All other relative links
55 p
= os
.path
.dirname(page
)
56 path
= os
.path
.join(p
, path
)
59 return os
.path
.normpath(path
)
61 def page_exists(self
, path
):
62 page
= self
.get_page(path
)
64 # Page must have been found and not deleted
65 return page
and not page
.was_deleted()
67 def get_page_title(self
, page
, default
=None):
68 doc
= self
.get_page(page
)
72 title
= os
.path
.basename(page
)
76 def get_page(self
, page
, revision
=None):
77 page
= Page
.sanitise_page_name(page
)
79 # Split the path into parts
80 parts
= page
.split("/")
82 # Check if this is an action
83 if any((part
.startswith("_") for part
in parts
)):
87 return self
._get
_page
("SELECT * FROM wiki WHERE page = %s \
88 AND timestamp = %s", page
, revision
)
90 return self
._get
_page
("SELECT * FROM wiki WHERE page = %s \
91 ORDER BY timestamp DESC LIMIT 1", page
)
93 def get_recent_changes(self
, account
, limit
=None):
94 pages
= self
._get
_pages
("SELECT * FROM wiki \
95 ORDER BY timestamp DESC")
98 if not page
.check_acl(account
):
107 def create_page(self
, page
, author
, content
, changes
=None, address
=None):
108 page
= Page
.sanitise_page_name(page
)
110 # Write page to the database
111 page
= self
._get
_page
("""
124 """, page
, author
.uid
, content
or None, changes
, address
,
127 # Store any linked files
128 page
._store
_linked
_files
()
130 # Send email to all watchers
131 page
._send
_watcher
_emails
(excludes
=[author
])
135 def delete_page(self
, page
, author
, **kwargs
):
136 # Do nothing if the page does not exist
137 if not self
.get_page(page
):
140 # Just creates a blank last version of the page
141 self
.create_page(page
, author
=author
, content
=None, **kwargs
)
143 def make_breadcrumbs(self
, url
):
144 # Split and strip all empty elements (double slashes)
145 parts
= list(e
for e
in url
.split("/") if e
)
148 for part
in ("/".join(parts
[:i
]) for i
in range(1, len(parts
))):
149 ret
.append(("/%s" % part
, self
.get_page_title(part
, os
.path
.basename(part
))))
153 def search(self
, query
, account
=None, limit
=None):
154 res
= self
._get
_pages
("""
158 wiki_search_index search_index
160 wiki ON search_index.wiki_id = wiki.id
162 search_index.document @@ websearch_to_tsquery('english', %s)
164 ts_rank(search_index.document, websearch_to_tsquery('english', %s)) DESC
170 # Skip any pages the user doesn't have permission for
171 if not page
.check_acl(account
):
174 # Return any other pages
177 # Break when we have found enough pages
178 if limit
and len(pages
) >= limit
:
185 Needs to be called after a page has been changed
187 self
.db
.execute("REFRESH MATERIALIZED VIEW wiki_search_index")
189 def get_watchlist(self
, account
):
190 pages
= self
._get
_pages
("""
197 wiki ON wiki_current.id = wiki.id
203 wiki_watchlist watchlist
205 pages ON watchlist.page = pages.page
215 def check_acl(self
, page
, account
):
216 res
= self
.db
.query("""
222 %s ILIKE (path || '%%')
230 # Access not permitted when user is not logged in
234 # If user is in a matching group, we grant permission
235 for group
in row
.groups
:
236 if account
.is_member_of_group(group
):
239 # Otherwise access is not permitted
242 # If no ACLs are found, we permit access
247 def _get_files(self
, query
, *args
):
248 res
= self
.db
.query(query
, *args
)
251 yield File(self
.backend
, row
.id, data
=row
)
253 def _get_file(self
, query
, *args
):
254 res
= self
.db
.get(query
, *args
)
257 return File(self
.backend
, res
.id, data
=res
)
259 def get_files(self
, path
):
260 files
= self
._get
_files
("""
275 def get_file_by_path(self
, path
, revision
=None):
276 path
, filename
= os
.path
.dirname(path
), os
.path
.basename(path
)
279 # Fetch a specific revision
280 return self
._get
_file
("""
294 """, path
, filename
, revision
,
297 # Fetch latest version
298 return self
._get
_file
("""
312 def get_file_by_path_and_filename(self
, path
, filename
):
313 return self
._get
_file
("""
327 def upload(self
, path
, filename
, data
, mimetype
, author
, address
):
328 # Replace any existing files
329 file = self
.get_file_by_path_and_filename(path
, filename
)
333 # Upload the blob first
334 blob
= self
.db
.get("""
342 SET data = EXCLUDED.data
347 # Create entry for file
348 return self
._get
_file
("""
360 %s, %s, %s, %s, %s, %s, %s
363 """, path
, filename
, author
.uid
, address
, mimetype
, blob
.id, len(data
),
366 def render(self
, path
, text
, **kwargs
):
367 return WikiRenderer(self
.backend
, path
, text
, **kwargs
)
370 class Page(misc
.Object
):
371 def init(self
, id, data
=None):
376 return "<%s %s %s>" % (self
.__class
__.__name
__, self
.page
, self
.timestamp
)
378 def __eq__(self
, other
):
379 if isinstance(other
, self
.__class
__):
380 return self
.id == other
.id
382 return NotImplemented
384 def __lt__(self
, other
):
385 if isinstance(other
, self
.__class
__):
386 if self
.page
== other
.page
:
387 return self
.timestamp
< other
.timestamp
389 return self
.page
< other
.page
391 return NotImplemented
394 def sanitise_page_name(page
):
398 # Make sure that the page name does NOT end with a /
399 if page
.endswith("/"):
402 # Make sure the page name starts with a /
403 if not page
.startswith("/"):
406 # Remove any double slashes
407 page
= page
.replace("//", "/")
413 return "/docs%s" % self
.page
417 return "https://www.ipfire.org%s" % self
.url
421 return self
.data
.page
425 return self
._title
or os
.path
.basename(self
.page
[1:])
429 if not self
.markdown
:
432 # Find first H1 headline in markdown
433 markdown
= self
.markdown
.splitlines()
435 m
= re
.match(r
"^#\s*(.*)( #)?$", markdown
[0])
441 if self
.data
.author_uid
:
442 return self
.backend
.accounts
.get_by_uid(self
.data
.author_uid
)
446 return self
.data
.markdown
or ""
452 # Strip off the first line if it contains a heading (as it will be shown separately)
453 for i
, line
in enumerate(self
.markdown
.splitlines()):
454 if i
== 0 and line
.startswith("#"):
459 renderer
= self
.backend
.wiki
.render(self
.page
, "\n".join(lines
), revision
=self
.timestamp
)
467 renderer
= self
.backend
.wiki
.render(self
.page
, self
.markdown
, revision
=self
.timestamp
)
469 return renderer
.files
471 def _store_linked_files(self
):
472 self
.db
.executemany("INSERT INTO wiki_linked_files(page_id, path) \
473 VALUES(%s, %s)", ((self
.id, file) for file in self
.files
))
477 return self
.data
.timestamp
479 def was_deleted(self
):
480 return not self
.markdown
483 def breadcrumbs(self
):
484 return self
.backend
.wiki
.make_breadcrumbs(self
.page
)
486 def is_latest_revision(self
):
487 return self
.get_latest_revision() == self
489 def get_latest_revision(self
):
490 revisions
= self
.get_revisions()
492 # Return first object
493 for rev
in revisions
:
496 def get_revisions(self
):
497 return self
.backend
.wiki
._get
_pages
("SELECT * FROM wiki \
498 WHERE page = %s ORDER BY timestamp DESC", self
.page
)
501 def previous_revision(self
):
502 return self
.backend
.wiki
._get
_page
("SELECT * FROM wiki \
503 WHERE page = %s AND timestamp < %s ORDER BY timestamp DESC \
504 LIMIT 1", self
.page
, self
.timestamp
)
508 return self
.data
.changes
512 def check_acl(self
, account
):
513 return self
.backend
.wiki
.check_acl(self
.page
, account
)
519 if self
.previous_revision
:
520 diff
= difflib
.unified_diff(
521 self
.previous_revision
.markdown
.splitlines(),
522 self
.markdown
.splitlines(),
525 return "\n".join(diff
)
529 res
= self
.db
.query("SELECT uid FROM wiki_watchlist \
530 WHERE page = %s", self
.page
)
533 # Search for account by UID and skip if none was found
534 account
= self
.backend
.accounts
.get_by_uid(row
.uid
)
541 def is_watched_by(self
, account
):
542 res
= self
.db
.get("SELECT 1 FROM wiki_watchlist \
543 WHERE page = %s AND uid = %s", self
.page
, account
.uid
)
550 def add_watcher(self
, account
):
551 if self
.is_watched_by(account
):
554 self
.db
.execute("INSERT INTO wiki_watchlist(page, uid) \
555 VALUES(%s, %s)", self
.page
, account
.uid
)
557 def remove_watcher(self
, account
):
558 self
.db
.execute("DELETE FROM wiki_watchlist \
559 WHERE page = %s AND uid = %s", self
.page
, account
.uid
)
561 def _send_watcher_emails(self
, excludes
=[]):
562 # Nothing to do if there was no previous revision
563 if not self
.previous_revision
:
566 for watcher
in self
.watchers
:
567 # Skip everyone who is excluded
568 if watcher
in excludes
:
569 logging
.debug("Excluding %s" % watcher
)
573 if not self
.backend
.wiki
.check_acl(self
.page
, watcher
):
574 logging
.debug("Watcher %s does not have permissions" % watcher
)
577 logging
.debug("Sending watcher email to %s" % watcher
)
580 self
.backend
.messages
.send_template("wiki/messages/page-changed",
581 account
=watcher
, page
=self
, priority
=-10)
583 def restore(self
, author
, address
, comment
=None):
584 changes
= "Restore to revision from %s" % self
.timestamp
.isoformat()
588 changes
= "%s: %s" % (changes
, comment
)
590 return self
.backend
.wiki
.create_page(self
.page
,
591 author
, self
.markdown
, changes
=changes
, address
=address
)
594 class File(misc
.Object
):
595 def init(self
, id, data
):
599 def __eq__(self
, other
):
600 if isinstance(other
, self
.__class
__):
601 return self
.id == other
.id
603 return NotImplemented
607 return "/docs%s" % os
.path
.join(self
.path
, self
.filename
)
611 return self
.data
.path
615 return self
.data
.filename
619 return self
.data
.mimetype
623 return self
.data
.size
627 if self
.data
.author_uid
:
628 return self
.backend
.accounts
.get_by_uid(self
.data
.author_uid
)
631 def created_at(self
):
632 return self
.data
.created_at
634 timestamp
= created_at
636 def delete(self
, author
=None):
637 if not self
.can_be_deleted():
638 raise RuntimeError("Cannot delete %s" % self
)
640 self
.db
.execute("UPDATE wiki_files SET deleted_at = NOW(), deleted_by = %s \
641 WHERE id = %s", author
.uid
if author
else None, self
.id)
643 def can_be_deleted(self
):
644 # Cannot be deleted if still in use
652 def deleted_at(self
):
653 return self
.data
.deleted_at
655 def get_latest_revision(self
):
656 revisions
= self
.get_revisions()
658 # Return first object
659 for rev
in revisions
:
662 def get_revisions(self
):
663 revisions
= self
.backend
.wiki
._get
_files
("SELECT * FROM wiki_files \
664 WHERE path = %s AND filename = %s ORDER BY created_at DESC", self
.path
, self
.filename
)
666 return list(revisions
)
669 return self
.mimetype
in ("application/pdf", "application/x-pdf")
672 return self
.mimetype
.startswith("image/")
674 def is_vector_image(self
):
675 return self
.mimetype
in ("image/svg+xml",)
677 def is_bitmap_image(self
):
678 return self
.is_image() and not self
.is_vector_image()
682 res
= self
.db
.get("SELECT data FROM wiki_blobs \
683 WHERE id = %s", self
.data
.blob_id
)
686 return bytes(res
.data
)
688 async def get_thumbnail(self
, size
):
689 assert self
.is_bitmap_image()
691 cache_key
= "-".join((
693 util
.normalize(self
.filename
),
694 self
.created_at
.isoformat(),
698 # Try to fetch the data from the cache
699 thumbnail
= await self
.backend
.cache
.get(cache_key
)
703 # Generate the thumbnail
704 thumbnail
= util
.generate_thumbnail(self
.blob
, size
)
706 # Put it into the cache for forever
707 await self
.backend
.cache
.set(cache_key
, thumbnail
)
714 Returns a list of all pages this file is linked by
716 pages
= self
.backend
.wiki
._get
_pages
("""
722 wiki_current ON wiki_linked_files.page_id = wiki_current.id
724 wiki ON wiki_linked_files.page_id = wiki.id
726 wiki_linked_files.path = %s
729 """, os
.path
.join(self
.path
, self
.filename
),
735 class WikiRenderer(misc
.Object
):
748 _links
= re
.compile(r
"<a href=\"(.*?
)\">(.*?
)</a
>")
751 _images = re.compile(r"<img
alt(?
:=\"(.*?
)\")? src
=\"(.*?
)\" (?
:title
=\"(.*?
)\" )?
/>")
753 def init(self, path, text, revision=None):
757 # Optionally, the revision of the rendered page
758 self.revision = revision
761 self.renderer = markdown.Markdown(
763 LinkedFilesExtractorExtension(),
764 PrettyLinksExtension(),
776 self.html = self._render()
778 def _render_link(self, m):
779 url, text = m.groups()
782 for schema in self.schemas:
783 if url.startswith(schema):
784 return """<a class="link
-external
" href="%s">%s</a>""" % \
790 if url.startswith("mailto
:"):
793 return """<a class="link
-external
" href="mailto
:%s">%s</a>""" % \
796 # Everything else must be an internal link
797 path = self.backend.wiki.make_path(self.path, url)
799 return """<a href="/docs
%s">%s</a>""" % \
800 (path, text or self.backend.wiki.get_page_title(path))
802 def _render_image(self, m):
803 alt_text, url, caption = m.groups()
805 # Compute a hash over the URL
806 h = hashlib.new("md5
")
807 h.update(url.encode())
811 <div class="columns
is-centered
">
812 <div class="column
is-8">
813 <figure class="image modal
-trigger
" data-target="%(id)s">
814 <img src="/docs
%(url)s?s
=640&
;%(args)s" alt="%(caption)s">
816 <figcaption class="figure
-caption
">%(caption)s</figcaption>
819 <div class="modal
is-large
" id="%(id)s">
820 <div class="modal
-background
"></div>
822 <div class="modal
-content
">
824 <img src="/docs
%(url)s?s
=1920&
;%(args)s" alt="%(caption)s"
828 <a class="button
is-small
" href="/docs
%(url)s?action
=detail
">
830 <i class="fa
-solid fa
-circle
-info
"></i>
835 <button class="modal
-close
is-large
" aria-label="close
"></button>
841 # Skip any absolute and external URLs
842 if url.startswith("https
://") or url.startswith("http
://"):
844 "caption
" : caption or "",
850 # Try to split query string
851 url, delimiter, qs = url.partition("?
")
853 # Parse query arguments
854 args = urllib.parse.parse_qs(qs)
856 # Build absolute path
857 url = self.backend.wiki.make_path(self.path, url)
860 file = self.backend.wiki.get_file_by_path(url, revision=self.revision)
861 if not file or not file.is_image():
862 return "<!-- Could
not find image
%s in %s -->" % (url, self.path)
864 # Remove any requested size
868 # Link the image that has been the current version at the time of the page edit
870 args["revision
"] = file.timestamp
873 "caption
" : caption or "",
876 "args
" : urllib.parse.urlencode(args),
880 logging.debug("Rendering
%s" % self.path)
883 text = self.renderer.convert(self.text)
886 text = self._links.sub(self._render_link, text)
888 # Postprocess images to <figure>
889 text = self._images.sub(self._render_image, text)
896 A list of all linked files that have been part of the rendered markup
900 for url in self.renderer.files:
901 # Skip external images
902 if url.startswith("https
://") or url.startswith("http
://"):
905 # Make the URL absolute
906 url = self.backend.wiki.make_path(self.path, url)
908 # Check if this is a file (it could also just be a page)
909 file = self.backend.wiki.get_file_by_path(url)
916 class PrettyLinksExtension(markdown.extensions.Extension):
917 def extendMarkdown(self, md):
918 # Create links to Bugzilla
919 md.preprocessors.register(BugzillaLinksPreprocessor(md), "bugzilla
", 10)
921 # Create links to CVE
922 md.preprocessors.register(CVELinksPreprocessor(md), "cve
", 10)
925 class BugzillaLinksPreprocessor(markdown.preprocessors.Preprocessor):
926 regex = re.compile(r"(?
:#(\d{5,}))", re.I)
928 def run(self
, lines
):
930 yield self
.regex
.sub(r
"[#\1](https://bugzilla.ipfire.org/show_bug.cgi?id=\1)", line
)
933 class CVELinksPreprocessor(markdown
.preprocessors
.Preprocessor
):
934 regex
= re
.compile(r
"(?:CVE)[\s\-](\d{4}\-\d+)")
936 def run(self
, lines
):
938 yield self
.regex
.sub(r
"[CVE-\1](https://cve.mitre.org/cgi-bin/cvename.cgi?name=\1)", line
)
941 class LinkedFilesExtractor(markdown
.treeprocessors
.Treeprocessor
):
943 Finds all Linked Files
948 # Find all images and store the URLs
949 for image
in root
.findall(".//img"):
950 src
= image
.get("src")
952 self
.md
.files
.append(src
)
955 for link
in root
.findall(".//a"):
956 href
= link
.get("href")
958 self
.md
.files
.append(href
)
961 class LinkedFilesExtractorExtension(markdown
.extensions
.Extension
):
962 def extendMarkdown(self
, md
):
963 md
.treeprocessors
.register(LinkedFilesExtractor(md
), "linked-files-extractor", 10)