Changeset - c5bed2ab87a7
[Not reviewed]
stable
0 1 0
Mads Kiilerich (mads) - 13 months ago 2025-02-18 13:02:17
mads@kiilerich.com
Grafted from: 98693b531d90
vcs: Always return node content as bytes

is_binary would fail to find nul bytes in unicode strings.
1 file changed with 2 insertions and 0 deletions:
0 comments (0 inline, 0 general)
kallithea/lib/vcs/nodes.py
Show inline comments
 
@@ -175,192 +175,194 @@ class Node(object):
 

	
 
    def is_root(self):
 
        """
 
        Returns ``True`` if node is a root node and ``False`` otherwise.
 
        """
 
        return self.kind == NodeKind.DIR and self.path == ''
 

	
 
    def is_submodule(self):
 
        """
 
        Returns ``True`` if node's kind is ``NodeKind.SUBMODULE``, ``False``
 
        otherwise.
 
        """
 
        return self.kind == NodeKind.SUBMODULE
 

	
 
    @LazyProperty
 
    def added(self):
 
        return self.state is NodeState.ADDED
 

	
 
    @LazyProperty
 
    def changed(self):
 
        return self.state is NodeState.CHANGED
 

	
 
    @LazyProperty
 
    def not_changed(self):
 
        return self.state is NodeState.NOT_CHANGED
 

	
 
    @LazyProperty
 
    def removed(self):
 
        return self.state is NodeState.REMOVED
 

	
 

	
 
class FileNode(Node):
 
    """
 
    Class representing file nodes.
 

	
 
    :attribute: path: path to the node, relative to repository's root
 
    :attribute: content: if given arbitrary sets content of the file
 
    :attribute: changeset: if given, first time content is accessed, callback
 
    :attribute: mode: octal stat mode for a node. Default is 0100644.
 
    """
 

	
 
    def __init__(self, path, content=None, changeset=None, mode=None):
 
        """
 
        Only one of ``content`` and ``changeset`` may be given. Passing both
 
        would raise ``NodeError`` exception.
 

	
 
        :param path: relative path to the node
 
        :param content: content may be passed to constructor
 
        :param changeset: if given, will use it to lazily fetch content
 
        :param mode: octal representation of ST_MODE (i.e. 0100644)
 
        """
 

	
 
        if content and changeset:
 
            raise NodeError("Cannot use both content and changeset")
 
        super(FileNode, self).__init__(path, kind=NodeKind.FILE)
 
        self.changeset = changeset
 
        if not isinstance(content, bytes) and content is not None:
 
            # File content is one thing that inherently must be bytes ... but
 
            # VCS module tries to be "user friendly" and support unicode ...
 
            content = safe_bytes(content)
 
        self._content = content
 
        self._mode = mode or 0o100644
 

	
 
    def __eq__(self, other):
 
        eq = super(FileNode, self).__eq__(other)
 
        if eq is not None:
 
            return eq
 
        return self.content == other.content
 

	
 
    def __lt__(self, other):
 
        lt = super(FileNode, self).__lt__(other)
 
        if lt is not None:
 
            return lt
 
        return self.content < other.content
 

	
 
    @LazyProperty
 
    def mode(self):
 
        """
 
        Returns lazily mode of the FileNode. If ``changeset`` is not set, would
 
        use value given at initialization or 0100644 (default).
 
        """
 
        if self.changeset:
 
            mode = self.changeset.get_file_mode(self.path)
 
        else:
 
            mode = self._mode
 
        return mode
 

	
 
    @property
 
    def content(self):
 
        """
 
        Returns lazily byte content of the FileNode.
 
        """
 
        if self.changeset:
 
            content = self.changeset.get_file_content(self.path)
 
        else:
 
            content = self._content
 
        if content is not None:
 
            content = safe_bytes(content)
 
        return content
 

	
 
    @LazyProperty
 
    def size(self):
 
        if self.changeset:
 
            return self.changeset.get_file_size(self.path)
 
        raise NodeError("Cannot retrieve size of the file without related "
 
            "changeset attribute")
 

	
 
    @LazyProperty
 
    def message(self):
 
        if self.changeset:
 
            return self.last_changeset.message
 
        raise NodeError("Cannot retrieve message of the file without related "
 
            "changeset attribute")
 

	
 
    @LazyProperty
 
    def last_changeset(self):
 
        if self.changeset:
 
            return self.changeset.get_file_changeset(self.path)
 
        raise NodeError("Cannot retrieve last changeset of the file without "
 
            "related changeset attribute")
 

	
 
    def get_mimetype(self):
 
        """
 
        Mimetype is calculated based on the file's content.
 
        """
 

	
 
        mtype, encoding = mimetypes.guess_type(self.name)
 

	
 
        if mtype is None:
 
            if self.is_binary:
 
                mtype = 'application/octet-stream'
 
                encoding = None
 
            else:
 
                mtype = 'text/plain'
 
                encoding = None
 

	
 
                # try with pygments
 
                try:
 
                    mt = lexers.get_lexer_for_filename(self.name).mimetypes
 
                except lexers.ClassNotFound:
 
                    mt = None
 

	
 
                if mt:
 
                    mtype = mt[0]
 

	
 
        return mtype, encoding
 

	
 
    @LazyProperty
 
    def mimetype(self):
 
        """
 
        Wrapper around full mimetype info. It returns only type of fetched
 
        mimetype without the encoding part. use get_mimetype function to fetch
 
        full set of (type,encoding)
 
        """
 
        return self.get_mimetype()[0]
 

	
 
    @LazyProperty
 
    def mimetype_main(self):
 
        return self.mimetype.split('/')[0]
 

	
 
    @LazyProperty
 
    def lexer(self):
 
        """
 
        Returns pygment's lexer class. Would try to guess lexer taking file's
 
        content, name and mimetype.
 
        """
 
        try:
 
            lexer = lexers.guess_lexer_for_filename(self.name, safe_str(self.content), stripnl=False)
 
        except lexers.ClassNotFound:
 
            lexer = lexers.TextLexer(stripnl=False)
 
        # returns first alias
 
        return lexer
 

	
 
    @LazyProperty
 
    def lexer_alias(self):
 
        """
 
        Returns first alias of the lexer guessed for this file.
 
        """
 
        return self.lexer.aliases[0]
 

	
 
    @LazyProperty
 
    def history(self):
 
        """
 
        Returns a list of changeset for this file in which the file was changed
 
        """
 
        if self.changeset is None:
 
            raise NodeError('Unable to get changeset for this FileNode')
 
        return self.changeset.get_file_history(self.path)
 

	
 
    @LazyProperty
 
    def annotate(self):
 
        """
 
        Returns a list of three element tuples with lineno,changeset and line
 
        """
0 comments (0 inline, 0 general)