Dodanie biblioteki vstorage (pierwszy commit zmierzający do uliniowienia historii...
[redakcja.git] / lib / vstorage / __init__.py
diff --git a/lib/vstorage/__init__.py b/lib/vstorage/__init__.py
new file mode 100644 (file)
index 0000000..88005da
--- /dev/null
@@ -0,0 +1,399 @@
+# -*- coding: utf-8 -*-
+import os
+import tempfile
+import datetime
+import mimetypes
+import urllib
+
+import sys
+
+# Note: we have to set these before importing Mercurial
+os.environ['HGENCODING'] = 'utf-8'
+os.environ['HGMERGE'] = "internal:merge"
+
+import mercurial.hg
+import mercurial.ui
+import mercurial.revlog
+import mercurial.util
+
+
+def urlquote(url, safe='/'):
+    """Quotes URL 
+    
+    >>> urlquote(u'Za\u017c\xf3\u0142\u0107 g\u0119\u015bl\u0105 ja\u017a\u0144')
+    'Za%C5%BC%C3%B3%C5%82%C4%87_g%C4%99%C5%9Bl%C4%85_ja%C5%BA%C5%84'
+    """
+    return urllib.quote(url.replace(' ', '_').encode('utf-8', 'ignore'), safe)
+    
+def urlunquote(url):
+    """Unqotes URL 
+    
+    # >>> urlunquote('Za%C5%BC%C3%B3%C5%82%C4%87_g%C4%99%C5%9Bl%C4%85_ja%C5%BA%C5%84')
+    # u'Za\u017c\xf3\u0142\u0107 g\u0119\u015bl\u0105 ja\u017a\u0144'
+    """
+    return unicode(urllib.unquote(url), 'utf-8', 'ignore').replace('_', ' ')
+
+def find_repo_path(path):
+    """Go up the directory tree looking for a Mercurial repository (a directory containing a .hg subdirectory)."""
+    while not os.path.isdir(os.path.join(path, ".hg")):
+        old_path, path = path, os.path.dirname(path)
+        if path == old_path:
+            return None
+    return path
+
+def locked_repo(func):
+    """A decorator for locking the repository when calling a method."""
+
+    def new_func(self, *args, **kwargs):
+        """Wrap the original function in locks."""
+
+        wlock = self.repo.wlock()
+        lock = self.repo.lock()
+        try:
+            func(self, *args, **kwargs)
+        finally:
+            lock.release()
+            wlock.release()
+
+    return new_func
+
+
+class DocumentNotFound(Exception):
+    pass
+
+
+class VersionedStorage(object):
+    """
+    Provides means of storing text pages and keeping track of their
+    change history, using Mercurial repository as the storage method.
+    """
+
+    def __init__(self, path, charset=None):
+        """
+        Takes the path to the directory where the pages are to be kept.
+        If the directory doen't exist, it will be created. If it's inside
+        a Mercurial repository, that repository will be used, otherwise
+        a new repository will be created in it.
+        """
+
+        self.charset = charset or 'utf-8'
+        self.path = path
+        if not os.path.exists(self.path):
+            os.makedirs(self.path)
+        self.repo_path = find_repo_path(self.path)
+        try:
+            self.ui = mercurial.ui.ui(report_untrusted=False,
+                                      interactive=False, quiet=True)
+        except TypeError:
+            # Mercurial 1.3 changed the way we setup the ui object.
+            self.ui = mercurial.ui.ui()
+            self.ui.quiet = True
+            self.ui._report_untrusted = False
+            self.ui.setconfig('ui', 'interactive', False)
+        if self.repo_path is None:
+            self.repo_path = self.path
+            create = True
+        else:
+            create = False
+        self.repo_prefix = self.path[len(self.repo_path):].strip('/')
+        self.repo = mercurial.hg.repository(self.ui, self.repo_path,
+                                            create=create)
+
+    def reopen(self):
+        """Close and reopen the repo, to make sure we are up to date."""
+
+        self.repo = mercurial.hg.repository(self.ui, self.repo_path)
+
+    def _file_path(self, title):
+        return os.path.join(self.path, urlquote(title, safe=''))
+
+    def _title_to_file(self, title):
+        return os.path.join(self.repo_prefix, urlquote(title, safe=''))
+
+    def _file_to_title(self, filename):
+        assert filename.startswith(self.repo_prefix)
+        name = filename[len(self.repo_prefix):].strip('/')
+        return urlunquote(name)
+
+    def __contains__(self, title):
+        return os.path.exists(self._file_path(title))
+
+    def __iter__(self):
+        return self.all_pages()
+
+    def merge_changes(self, changectx, repo_file, text, user, parent):
+        """Commits and merges conflicting changes in the repository."""
+        tip_node = changectx.node()
+        filectx = changectx[repo_file].filectx(parent)
+        parent_node = filectx.changectx().node()
+
+        self.repo.dirstate.setparents(parent_node)
+        node = self._commit([repo_file], text, user)
+        
+        partial = lambda filename: repo_file == filename
+        
+        # If p1 is equal to p2, there is no work to do. Even the dirstate is correct.
+        p1, p2 = self.repo[None].parents()[0], self.repo[tip_node]
+        if p1 == p2:
+            return text
+        
+        try:
+            unresolved = mercurial.merge.update(self.repo, tip_node, True, False, partial)
+        except mercurial.util.Abort:
+            raise
+            unresolved = 1, 1, 1, 1
+
+        self.repo.dirstate.setparents(tip_node, node)
+        # Mercurial 1.1 and later need updating the merge state
+        try:
+            mercurial.merge.mergestate(self.repo).mark(repo_file, "r")
+        except (AttributeError, KeyError):
+            pass
+        return u'merge of edit conflict'
+
+    @locked_repo
+    def save_file(self, title, file_name, author=u'', comment=u'', parent=None):
+        """Save an existing file as specified page."""
+
+        user = author.encode('utf-8') or u'anon'.encode('utf-8')
+        text = comment.encode('utf-8') or u'comment'.encode('utf-8')
+        repo_file = self._title_to_file(title)
+        file_path = self._file_path(title)
+        mercurial.util.rename(file_name, file_path)
+        changectx = self._changectx()
+        try:
+            filectx_tip = changectx[repo_file]
+            current_page_rev = filectx_tip.filerev()
+        except mercurial.revlog.LookupError:
+            self.repo.add([repo_file])
+            current_page_rev = -1
+        if parent is not None and current_page_rev != parent:
+            msg = self.merge_changes(changectx, repo_file, text, user, parent)
+            user = '<wiki>'
+            text = msg.encode('utf-8')
+        self._commit([repo_file], text, user)
+
+
+    def _commit(self, files, text, user):
+        try:
+            return self.repo.commit(files=files, text=text, user=user,
+                                    force=True, empty_ok=True)
+        except TypeError:
+            # Mercurial 1.3 doesn't accept empty_ok or files parameter
+            match = mercurial.match.exact(self.repo_path, '', list(files))
+            return self.repo.commit(match=match, text=text, user=user,
+                                    force=True)
+
+
+    def save_data(self, title, data, author=u'', comment=u'', parent=None):
+        """Save data as specified page."""
+
+        try:
+            temp_path = tempfile.mkdtemp(dir=self.path)
+            file_path = os.path.join(temp_path, 'saved')
+            f = open(file_path, "wb")
+            f.write(data)
+            f.close()
+            self.save_file(title, file_path, author, comment, parent)
+        finally:
+            try:
+                os.unlink(file_path)
+            except OSError:
+                pass
+            try:
+                os.rmdir(temp_path)
+            except OSError:
+                pass
+
+    def save_text(self, title, text, author=u'', comment=u'', parent=None):
+        """Save text as specified page, encoded to charset."""
+
+        data = text.encode(self.charset)
+        self.save_data(title, data, author, comment, parent)
+
+    def page_text(self, title):
+        """Read unicode text of a page."""
+
+        data = self.open_page(title).read()
+        text = unicode(data, self.charset, 'replace')
+        return text
+
+    def page_lines(self, page):
+        for data in page:
+            yield unicode(data, self.charset, 'replace')
+
+    @locked_repo
+    def delete_page(self, title, author=u'', comment=u''):
+        user = author.encode('utf-8') or 'anon'
+        text = comment.encode('utf-8') or 'deleted'
+        repo_file = self._title_to_file(title)
+        file_path = self._file_path(title)
+        try:
+            os.unlink(file_path)
+        except OSError:
+            pass
+        self.repo.remove([repo_file])
+        self._commit([repo_file], text, user)
+
+    def open_page(self, title):
+        try:
+            return open(self._file_path(title), "rb")
+        except IOError:
+            raise DocumentNotFound()
+
+    def page_file_meta(self, title):
+        """Get page's inode number, size and last modification time."""
+
+        try:
+            (st_mode, st_ino, st_dev, st_nlink, st_uid, st_gid, st_size,
+             st_atime, st_mtime, st_ctime) = os.stat(self._file_path(title))
+        except OSError:
+            return 0, 0, 0
+        return st_ino, st_size, st_mtime
+
+    def page_meta(self, title):
+        """Get page's revision, date, last editor and his edit comment."""
+
+        filectx_tip = self._find_filectx(title)
+        if filectx_tip is None:
+            raise DocumentNotFound()
+            #return -1, None, u'', u''
+        rev = filectx_tip.filerev()
+        filectx = filectx_tip.filectx(rev)
+        date = datetime.datetime.fromtimestamp(filectx.date()[0])
+        author = unicode(filectx.user(), "utf-8",
+                         'replace').split('<')[0].strip()
+        comment = unicode(filectx.description(), "utf-8", 'replace')
+        return rev, date, author, comment
+
+    def repo_revision(self):
+        return self._changectx().rev()
+
+    def page_mime(self, title):
+        """
+        Guess page's mime type ased on corresponding file name.
+        Default ot text/x-wiki for files without an extension.
+
+        # >>> page_mime('something.txt')
+        # 'text/plain'
+        # >>> page_mime('SomePage')
+        # 'text/x-wiki'
+        # >>> page_mime(u'ąęśUnicodePage')
+        # 'text/x-wiki'
+        # >>> page_mime('image.png')
+        # 'image/png'
+        # >>> page_mime('style.css')
+        # 'text/css'
+        # >>> page_mime('archive.tar.gz')
+        # 'archive/gzip'
+        """
+
+        addr = self._file_path(title)
+        mime, encoding = mimetypes.guess_type(addr, strict=False)
+        if encoding:
+            mime = 'archive/%s' % encoding
+        if mime is None:
+            mime = 'text/x-wiki'
+        return mime
+
+    def _changectx(self):
+        """Get the changectx of the tip."""
+        try:
+            # This is for Mercurial 1.0
+            return self.repo.changectx()
+        except TypeError:
+            # Mercurial 1.3 (and possibly earlier) needs an argument
+            return self.repo.changectx('tip')
+
+    def _find_filectx(self, title):
+        """Find the last revision in which the file existed."""
+
+        repo_file = self._title_to_file(title)
+        changectx = self._changectx()
+        stack = [changectx]
+        while repo_file not in changectx:
+            if not stack:
+                return None
+            changectx = stack.pop()
+            for parent in changectx.parents():
+                if parent != changectx:
+                    stack.append(parent)
+        return changectx[repo_file]
+
+    def page_history(self, title):
+        """Iterate over the page's history."""
+
+        filectx_tip = self._find_filectx(title)
+        if filectx_tip is None:
+            return
+        maxrev = filectx_tip.filerev()
+        minrev = 0
+        for rev in range(maxrev, minrev-1, -1):
+            filectx = filectx_tip.filectx(rev)
+            date = datetime.datetime.fromtimestamp(filectx.date()[0])
+            author = unicode(filectx.user(), "utf-8",
+                             'replace').split('<')[0].strip()
+            comment = unicode(filectx.description(), "utf-8", 'replace')
+            yield rev, date, author, comment
+
+    def page_revision(self, title, rev):
+        """Get unicode contents of specified revision of the page."""
+
+        filectx_tip = self._find_filectx(title)
+        if filectx_tip is None:
+            raise DocumentNotFound()
+        try:
+            data = filectx_tip.filectx(rev).data()
+        except IndexError:
+            raise DocumentNotFound()
+        return data
+
+    def revision_text(self, title, rev):
+        data = self.page_revision(title, rev)
+        text = unicode(data, self.charset, 'replace')
+        return text
+
+    def history(self):
+        """Iterate over the history of entire wiki."""
+
+        changectx = self._changectx()
+        maxrev = changectx.rev()
+        minrev = 0
+        for wiki_rev in range(maxrev, minrev-1, -1):
+            change = self.repo.changectx(wiki_rev)
+            date = datetime.datetime.fromtimestamp(change.date()[0])
+            author = unicode(change.user(), "utf-8",
+                             'replace').split('<')[0].strip()
+            comment = unicode(change.description(), "utf-8", 'replace')
+            for repo_file in change.files():
+                if repo_file.startswith(self.repo_prefix):
+                    title = self._file_to_title(repo_file)
+                    try:
+                        rev = change[repo_file].filerev()
+                    except mercurial.revlog.LookupError:
+                        rev = -1
+                    yield title, rev, date, author, comment
+
+    def all_pages(self):
+        """Iterate over the titles of all pages in the wiki."""
+
+        for filename in os.listdir(self.path):
+            if (os.path.isfile(os.path.join(self.path, filename))
+                and not filename.startswith('.')):
+                yield urlunquote(filename)
+
+    def changed_since(self, rev):
+        """Return all pages that changed since specified repository revision."""
+
+        try:
+            last = self.repo.lookup(int(rev))
+        except IndexError:
+            for page in self.all_pages():
+                yield page
+                return
+        current = self.repo.lookup('tip')
+        status = self.repo.status(current, last)
+        modified, added, removed, deleted, unknown, ignored, clean = status
+        for filename in modified+added+removed+deleted:
+            if filename.startswith(self.repo_prefix):
+                yield self._file_to_title(filename)