]> git.ipfire.org Git - ipfire.org.git/blobdiff - src/backend/wiki.py
wiki.py: fixed breadcrumb generation
[ipfire.org.git] / src / backend / wiki.py
index cc5964d14a693e3633d3b65ebfbfb731958f0336..03368e8057e8f2521df2e78117571876d4580508 100644 (file)
@@ -3,6 +3,9 @@
 import difflib
 import hashlib
 import logging
+import markdown
+import markdown.extensions
+import markdown.preprocessors
 import os.path
 import re
 import urllib.parse
@@ -25,11 +28,17 @@ class Wiki(misc.Object):
                        return Page(self.backend, res.id, data=res)
 
        def __iter__(self):
-               return self._get_pages(
-                       "SELECT wiki.* FROM wiki_current current \
-                               LEFT JOIN wiki ON current.id = wiki.id \
-                               WHERE current.deleted IS FALSE \
-                               ORDER BY page",
+               return self._get_pages("""
+                       SELECT
+                               wiki.*
+                       FROM
+                               wiki_current current
+                       LEFT JOIN
+                               wiki ON current.id = wiki.id
+                       WHERE
+                               current.deleted IS FALSE
+                       ORDER BY page
+                       """,
                )
 
        def make_path(self, page, path):
@@ -99,8 +108,24 @@ class Wiki(misc.Object):
                page = Page.sanitise_page_name(page)
 
                # Write page to the database
-               page = self._get_page("INSERT INTO wiki(page, author_uid, markdown, changes, address) \
-                       VALUES(%s, %s, %s, %s, %s) RETURNING *", page, author.uid, content or None, changes, address)
+               page = self._get_page("""
+                       INSERT INTO
+                               wiki
+                       (
+                               page,
+                               author_uid,
+                               markdown,
+                               changes,
+                               address
+                       ) VALUES (
+                               %s, %s, %s, %s, %s
+                       )
+                       RETURNING *
+                       """, page, author.uid, content or None, changes, address,
+               )
+
+               # Store any linked files
+               page._store_linked_files()
 
                # Send email to all watchers
                page._send_watcher_emails(excludes=[author])
@@ -121,16 +146,24 @@ class Wiki(misc.Object):
 
                ret = []
                for part in ("/".join(parts[:i]) for i in range(1, len(parts))):
-                       ret.append(("/%s" % part, self.get_page_title(part, os.path.basename(part))))
+                       ret.append(("/docs/%s" % part, self.get_page_title(part, os.path.basename(part))))
 
                return ret
 
        def search(self, query, account=None, limit=None):
-               res = self._get_pages("SELECT wiki.* FROM wiki_search_index search_index \
-                       LEFT JOIN wiki ON search_index.wiki_id = wiki.id \
-                       WHERE search_index.document @@ websearch_to_tsquery('english', %s) \
-                               ORDER BY ts_rank(search_index.document, websearch_to_tsquery('english', %s)) DESC",
-                       query, query)
+               res = self._get_pages("""
+                       SELECT
+                               wiki.*
+                       FROM
+                               wiki_search_index search_index
+                       LEFT JOIN
+                               wiki ON search_index.wiki_id = wiki.id
+                       WHERE
+                               search_index.document @@ websearch_to_tsquery('english', %s)
+                       ORDER BY
+                               ts_rank(search_index.document, websearch_to_tsquery('english', %s)) DESC
+                       """, query, query,
+               )
 
                pages = []
                for page in res:
@@ -180,8 +213,18 @@ class Wiki(misc.Object):
        # ACL
 
        def check_acl(self, page, account):
-               res = self.db.query("SELECT * FROM wiki_acls \
-                       WHERE %s ILIKE (path || '%%') ORDER BY LENGTH(path) DESC LIMIT 1", page)
+               res = self.db.query("""
+                       SELECT
+                               *
+                       FROM
+                               wiki_acls
+                       WHERE
+                               %s ILIKE (path || '%%')
+                       ORDER BY
+                               LENGTH(path) DESC
+                       LIMIT 1
+                       """, page,
+               )
 
                for row in res:
                        # Access not permitted when user is not logged in
@@ -214,8 +257,18 @@ class Wiki(misc.Object):
                        return File(self.backend, res.id, data=res)
 
        def get_files(self, path):
-               files = self._get_files("SELECT * FROM wiki_files \
-                       WHERE path = %s AND deleted_at IS NULL ORDER BY filename", path)
+               files = self._get_files("""
+                       SELECT
+                               *
+                       FROM
+                               wiki_files
+                       WHERE
+                               path = %s
+                       AND
+                               deleted_at IS NULL
+                       ORDER BY filename
+                       """, path,
+               )
 
                return list(files)
 
@@ -224,19 +277,52 @@ class Wiki(misc.Object):
 
                if revision:
                        # Fetch a specific revision
-                       return self._get_file("SELECT * FROM wiki_files \
-                               WHERE path = %s AND filename = %s AND created_at <= %s \
-                               ORDER BY created_at DESC LIMIT 1", path, filename, revision)
+                       return self._get_file("""
+                               SELECT
+                                       *
+                               FROM
+                                       wiki_files
+                               WHERE
+                                       path = %s
+                               AND
+                                       filename = %s
+                               AND
+                                       created_at <= %s
+                               ORDER BY
+                                       created_at DESC
+                                       LIMIT 1
+                               """, path, filename, revision,
+                       )
 
                # Fetch latest version
-               return self._get_file("SELECT * FROM wiki_files \
-                       WHERE path = %s AND filename = %s AND deleted_at IS NULL",
-                       path, filename)
+               return self._get_file("""
+                       SELECT
+                               *
+                       FROM
+                               wiki_files
+                       WHERE
+                               path = %s
+                       AND
+                               filename = %s
+                       AND
+                               deleted_at IS NULL
+                       """, path, filename,
+               )
 
        def get_file_by_path_and_filename(self, path, filename):
-               return self._get_file("SELECT * FROM wiki_files \
-                       WHERE path = %s AND filename = %s AND deleted_at IS NULL",
-                       path, filename)
+               return self._get_file("""
+                       SELECT
+                               *
+                       FROM
+                               wiki_files
+                       WHERE
+                               path = %s
+                       AND
+                               filename = %s
+                       AND
+                               deleted_at IS NULL
+                       """, path, filename,
+               )
 
        def upload(self, path, filename, data, mimetype, author, address):
                # Replace any existing files
@@ -245,19 +331,40 @@ class Wiki(misc.Object):
                        file.delete(author)
 
                # Upload the blob first
-               blob = self.db.get("INSERT INTO wiki_blobs(data) VALUES(%s) \
-                       ON CONFLICT (digest(data, %s)) DO UPDATE SET data = EXCLUDED.data \
-                       RETURNING id", data, "MD5")
+               blob = self.db.get("""
+                       INSERT INTO
+                               wiki_blobs(data)
+                       VALUES
+                               (%s)
+                       ON CONFLICT
+                               (digest(data, %s))
+                       DO UPDATE
+                               SET data = EXCLUDED.data
+                       RETURNING id
+                       """, data, "MD5",
+               )
 
                # Create entry for file
-               return self._get_file("INSERT INTO wiki_files(path, filename, author_uid, address, \
-                       mimetype, blob_id, size) VALUES(%s,  %s, %s, %s, %s, %s, %s) RETURNING *", path,
-                       filename, author.uid, address, mimetype, blob.id, len(data))
-
-       def render(self, path, text):
-               r = WikiRenderer(self.backend, path)
+               return self._get_file("""
+                       INSERT INTO
+                               wiki_files
+                       (
+                               path,
+                               filename,
+                               author_uid,
+                               address,
+                               mimetype,
+                               blob_id,
+                               size
+                       ) VALUES (
+                               %s, %s, %s, %s, %s, %s, %s
+                       )
+                       RETURNING *
+                       """, path, filename, author.uid, address, mimetype, blob.id, len(data),
+               )
 
-               return r.render(text)
+       def render(self, path, text, **kwargs):
+               return WikiRenderer(self.backend, path, text, **kwargs)
 
 
 class Page(misc.Object):
@@ -283,6 +390,9 @@ class Page(misc.Object):
 
                return NotImplemented
 
+       def __hash__(self):
+               return hash(self.page)
+
        @staticmethod
        def sanitise_page_name(page):
                if not page:
@@ -349,7 +459,21 @@ class Page(misc.Object):
 
                        lines.append(line)
 
-               return self.backend.wiki.render(self.page, "\n".join(lines))
+               renderer = self.backend.wiki.render(self.page, "\n".join(lines), revision=self.timestamp)
+
+               return renderer.html
+
+       # Linked Files
+
+       @property
+       def files(self):
+               renderer = self.backend.wiki.render(self.page, self.markdown, revision=self.timestamp)
+
+               return renderer.files
+
+       def _store_linked_files(self):
+               self.db.executemany("INSERT INTO wiki_linked_files(page_id, path) \
+                       VALUES(%s, %s)", ((self.id, file) for file in self.files))
 
        @property
        def timestamp(self):
@@ -479,6 +603,8 @@ class File(misc.Object):
                if isinstance(other, self.__class__):
                        return self.id == other.id
 
+               return NotImplemented
+
        @property
        def url(self):
                return "/docs%s" % os.path.join(self.path, self.filename)
@@ -508,10 +634,23 @@ class File(misc.Object):
        def created_at(self):
                return self.data.created_at
 
+       timestamp = created_at
+
        def delete(self, author=None):
+               if not self.can_be_deleted():
+                       raise RuntimeError("Cannot delete %s" % self)
+
                self.db.execute("UPDATE wiki_files SET deleted_at = NOW(), deleted_by = %s \
                        WHERE id = %s", author.uid if author else None, self.id)
 
+       def can_be_deleted(self):
+               # Cannot be deleted if still in use
+               if self.pages:
+                       return False
+
+               # Can be deleted
+               return True
+
        @property
        def deleted_at(self):
                return self.data.deleted_at
@@ -549,13 +688,16 @@ class File(misc.Object):
                if res:
                        return bytes(res.data)
 
-       async def get_thumbnail(self, size):
+       async def get_thumbnail(self, size, format=None):
                assert self.is_bitmap_image()
 
-               cache_key = "-".join((
+               cache_key = ":".join((
+                       "wiki",
+                       "thumbnail",
                        self.path,
                        util.normalize(self.filename),
                        self.created_at.isoformat(),
+                       format or "N/A",
                        "%spx" % size,
                ))
 
@@ -565,13 +707,36 @@ class File(misc.Object):
                        return thumbnail
 
                # Generate the thumbnail
-               thumbnail = util.generate_thumbnail(self.blob, size)
+               thumbnail = util.generate_thumbnail(self.blob, size, format=format, quality=95)
 
                # Put it into the cache for forever
                await self.backend.cache.set(cache_key, thumbnail)
 
                return thumbnail
 
+       @property
+       def pages(self):
+               """
+                       Returns a list of all pages this file is linked by
+               """
+               pages = self.backend.wiki._get_pages("""
+                       SELECT
+                               wiki.*
+                       FROM
+                               wiki_linked_files
+                       JOIN
+                               wiki_current ON wiki_linked_files.page_id = wiki_current.id
+                       LEFT JOIN
+                               wiki ON wiki_linked_files.page_id = wiki.id
+                       WHERE
+                               wiki_linked_files.path = %s
+                       ORDER BY
+                               wiki.page
+                       """, os.path.join(self.path, self.filename),
+               )
+
+               return list(pages)
+
 
 class WikiRenderer(misc.Object):
        schemas = (
@@ -586,13 +751,35 @@ class WikiRenderer(misc.Object):
        )
 
        # Links
-       links = re.compile(r"<a href=\"(.*?)\">(.*?)</a>")
+       _links = re.compile(r"<a href=\"(.*?)\">(.*?)</a>")
 
        # Images
-       images = re.compile(r"<img alt(?:=\"(.*?)\")? src=\"(.*?)\" (?:title=\"(.*?)\" )?/>")
+       _images = re.compile(r"<img alt(?:=\"(.*?)\")? src=\"(.*?)\" (?:title=\"(.*?)\" )?/>")
 
-       def init(self, path):
+       def init(self, path, text, revision=None):
                self.path = path
+               self.text = text
+
+               # Optionally, the revision of the rendered page
+               self.revision = revision
+
+               # Markdown Renderer
+               self.renderer = markdown.Markdown(
+                       extensions=[
+                               LinkedFilesExtractorExtension(),
+                               PrettyLinksExtension(),
+                               "codehilite",
+                               "fenced_code",
+                               "footnotes",
+                               "nl2br",
+                               "sane_lists",
+                               "tables",
+                               "toc",
+                       ],
+               )
+
+               # Render!
+               self.html = self._render()
 
        def _render_link(self, m):
                url, text = m.groups()
@@ -630,7 +817,7 @@ class WikiRenderer(misc.Object):
                        <div class="columns is-centered">
                                <div class="column is-8">
                                        <figure class="image modal-trigger" data-target="%(id)s">
-                                               <img src="/docs%(url)s" alt="%(caption)s">
+                                               <img src="/docs%(url)s?s=960&amp;%(args)s" alt="%(caption)s">
 
                                                <figcaption class="figure-caption">%(caption)s</figcaption>
                                        </figure>
@@ -640,9 +827,15 @@ class WikiRenderer(misc.Object):
 
                                                <div class="modal-content">
                                                        <p class="image">
-                                                               <img src="/docs%(plain_url)s?s=1920" alt="%(caption)s"
+                                                               <img src="/docs%(url)s?s=2048&amp;%(args)s" alt="%(caption)s"
                                                                        loading="lazy">
                                                        </p>
+
+                                                       <a class="button is-small" href="/docs%(url)s?action=detail">
+                                                               <span class="icon">
+                                                                       <i class="fa-solid fa-circle-info"></i>
+                                                               </span>
+                                                       </a>
                                                </div>
 
                                                <button class="modal-close is-large" aria-label="close"></button>
@@ -651,54 +844,129 @@ class WikiRenderer(misc.Object):
                        </div>
                """
 
+               # Try to split query string
+               url, delimiter, qs = url.partition("?")
+
+               # Parse query arguments
+               args = urllib.parse.parse_qs(qs)
+
                # Skip any absolute and external URLs
-               if url.startswith("/") or url.startswith("https://") or url.startswith("http://"):
+               if url.startswith("https://") or url.startswith("http://"):
                        return html % {
                                "caption"   : caption or "",
                                "id"        : id,
-                               "plain_url" : url,
                                "url"       : url,
+                               "args"      : args,
                        }
 
-               # Try to split query string
-               url, delimiter, qs = url.partition("?")
-
-               # Parse query arguments
-               args = urllib.parse.parse_qs(qs)
-
                # Build absolute path
-               plain_url = url = self.backend.wiki.make_path(self.path, url)
+               url = self.backend.wiki.make_path(self.path, url)
 
                # Find image
-               file = self.backend.wiki.get_file_by_path(url)
+               file = self.backend.wiki.get_file_by_path(url, revision=self.revision)
                if not file or not file.is_image():
                        return "<!-- Could not find image %s in %s -->" % (url, self.path)
 
-               # Scale down the image if not already done
-               if not "s" in args:
-                       args["s"] = "920"
+               # Remove any requested size
+               if "s" in args:
+                       del args["s"]
 
-               # Append arguments to the URL
-               if args:
-                       url = "%s?%s" % (url, urllib.parse.urlencode(args))
+               # Link the image that has been the current version at the time of the page edit
+               if file:
+                       args["revision"] = file.timestamp
 
                return html % {
                        "caption"   : caption or "",
                        "id"        : id,
-                       "plain_url" : plain_url,
                        "url"       : url,
+                       "args"      : urllib.parse.urlencode(args),
                }
 
-       def render(self, text):
+       def _render(self):
                logging.debug("Rendering %s" % self.path)
 
-               # Borrow this from the blog
-               text = self.backend.blog._render_text(text, lang="markdown")
+               # Render...
+               text = self.renderer.convert(self.text)
 
                # Postprocess links
-               text = self.links.sub(self._render_link, text)
+               text = self._links.sub(self._render_link, text)
 
                # Postprocess images to <figure>
-               text = self.images.sub(self._render_image, text)
+               text = self._images.sub(self._render_image, text)
 
                return text
+
+       @lazy_property
+       def files(self):
+               """
+                       A list of all linked files that have been part of the rendered markup
+               """
+               files = []
+
+               for url in self.renderer.files:
+                       # Skip external images
+                       if url.startswith("https://") or url.startswith("http://"):
+                               continue
+
+                       # Make the URL absolute
+                       url = self.backend.wiki.make_path(self.path, url)
+
+                       # Check if this is a file (it could also just be a page)
+                       file = self.backend.wiki.get_file_by_path(url)
+                       if file:
+                               files.append(url)
+
+               return files
+
+
+class PrettyLinksExtension(markdown.extensions.Extension):
+       def extendMarkdown(self, md):
+               # Create links to Bugzilla
+               md.preprocessors.register(BugzillaLinksPreprocessor(md), "bugzilla", 10)
+
+               # Create links to CVE
+               md.preprocessors.register(CVELinksPreprocessor(md), "cve", 10)
+
+
+class BugzillaLinksPreprocessor(markdown.preprocessors.Preprocessor):
+       regex = re.compile(r"(?:#(\d{5,}))", re.I)
+
+       def run(self, lines):
+               for line in lines:
+                       yield self.regex.sub(r"[#\1](https://bugzilla.ipfire.org/show_bug.cgi?id=\1)", line)
+
+
+class CVELinksPreprocessor(markdown.preprocessors.Preprocessor):
+       regex = re.compile(r"(?:CVE)[\s\-](\d{4}\-\d+)")
+
+       def run(self, lines):
+               for line in lines:
+                       yield self.regex.sub(r"[CVE-\1](https://cve.mitre.org/cgi-bin/cvename.cgi?name=\1)", line)
+
+
+class LinkedFilesExtractor(markdown.treeprocessors.Treeprocessor):
+       """
+               Finds all Linked Files
+       """
+       def __init__(self, *args, **kwargs):
+               super().__init__(*args, **kwargs)
+
+               self.md.files = []
+
+       def run(self, root):
+               # Find all images and store the URLs
+               for image in root.findall(".//img"):
+                       src = image.get("src")
+
+                       self.md.files.append(src)
+
+               # Find all links
+               for link in root.findall(".//a"):
+                       href = link.get("href")
+
+                       self.md.files.append(href)
+
+
+class LinkedFilesExtractorExtension(markdown.extensions.Extension):
+    def extendMarkdown(self, md):
+        md.treeprocessors.register(LinkedFilesExtractor(md), "linked-files-extractor", 10)