diff options
Diffstat (limited to 'lib/git/objects/tree.py')
| -rw-r--r-- | lib/git/objects/tree.py | 170 |
1 files changed, 136 insertions, 34 deletions
diff --git a/lib/git/objects/tree.py b/lib/git/objects/tree.py index 273384a3..707cebaa 100644 --- a/lib/git/objects/tree.py +++ b/lib/git/objects/tree.py @@ -9,26 +9,57 @@ import blob import base class Tree(base.IndexObject): + """ + Tress represent a ordered list of Blobs and other Trees. Hence it can be + accessed like a list. + + Tree's will cache their contents after first retrieval to improve efficiency. + + ``Tree as a list``:: + + Access a specific blob using the + tree['filename'] notation. + + You may as well access by index + blob = tree[0] + + + """ type = "tree" - __slots__ = "_contents" + __slots__ = "_cache" def __init__(self, repo, id, mode=None, path=None): super(Tree, self).__init__(repo, id, mode, path) def _set_cache_(self, attr): - if attr == "_contents": - # Read the tree contents. - self._contents = {} - for line in self.repo.git.ls_tree(self.id).splitlines(): - obj = self.content__from_string(self.repo, line) - if obj is not None: - self._contents[obj.path] = obj + if attr == "_cache": + # Set the data when we need it + self._cache = self._get_tree_cache(self.repo, self.id) else: super(Tree, self)._set_cache_(attr) - @staticmethod - def content__from_string(repo, text): + @classmethod + def _get_tree_cache(cls, repo, treeish): + """ + Return + list(object_instance, ...) + + ``treeish`` + sha or ref identifying a tree + """ + out = list() + for line in repo.git.ls_tree(treeish).splitlines(): + obj = cls.content_from_string(repo, line) + if obj is not None: + out.append(obj) + # END if object was handled + # END for each line from ls-tree + return out + + + @classmethod + def content_from_string(cls, repo, text): """ Parse a content item and create the appropriate object @@ -40,6 +71,8 @@ class Tree(base.IndexObject): Returns ``git.Blob`` or ``git.Tree`` + + NOTE: Currently sub-modules are ignored ! """ try: mode, typ, id, path = text.expandtabs(1).split(" ", 3) @@ -51,6 +84,7 @@ class Tree(base.IndexObject): elif typ == "blob": return blob.Blob(repo, id, mode, path) elif typ == "commit": + # TODO: Return a submodule return None else: raise(TypeError, "Invalid type: %s" % typ) @@ -67,36 +101,104 @@ class Tree(base.IndexObject): <git.Blob "8b1e02c0fb554eed2ce2ef737a68bb369d7527df"> Returns - ``git.Blob`` or ``git.Tree`` or ``None`` if not found + ``git.Blob`` or ``git.Tree`` + + Raise + KeyError if given file or tree does not exist in tree """ - return self.get(file) + return self[file] def __repr__(self): return '<git.Tree "%s">' % self.id + + @classmethod + def _iter_recursive(cls, repo, tree, cur_depth, max_depth, predicate ): + + for obj in tree: + # adjust path to be complete + obj.path = os.path.join(tree.path, obj.path) + if not predicate(obj): + continue + yield obj + if obj.type == "tree" and ( max_depth < 0 or cur_depth+1 <= max_depth ): + for recursive_obj in cls._iter_recursive( repo, obj, cur_depth+1, max_depth, predicate ): + yield recursive_obj + # END for each recursive object + # END if we may enter recursion + # END for each object + + def traverse(self, max_depth=-1, predicate = lambda i: True): + """ + Returns + Iterator to traverse the tree recursively up to the given level. + The iterator returns Blob and Tree objects + + ``max_depth`` + + if -1, the whole tree will be traversed + if 0, only the first level will be traversed which is the same as + the default non-recursive iterator + + ``predicate`` + + If predicate(item) returns True, item will be returned by iterator + """ + return self._iter_recursive( self.repo, self, 0, max_depth, predicate ) + + @property + def trees(self): + """ + Returns + list(Tree, ...) list of trees directly below this tree + """ + return [ i for i in self if i.type == "tree" ] + + @property + def blobs(self): + """ + Returns + list(Blob, ...) list of blobs directly below this tree + """ + return [ i for i in self if i.type == "blob" ] - # Implement the basics of the dict protocol: - # directories/trees can be seen as object dicts. - def __getitem__(self, key): - return self._contents[key] + # List protocol + def __getslice__(self,i,j): + return self._cache[i:j] + def __iter__(self): - return iter(self._contents) - + return iter(self._cache) + def __len__(self): - return len(self._contents) - - def __contains__(self, key): - return key in self._contents - - def get(self, key): - return self._contents.get(key) - - def items(self): - return self._contents.items() - - def keys(self): - return self._contents.keys() - - def values(self): - return self._contents.values() + return len(self._cache) + + def __getitem__(self,item): + if isinstance(item, int): + return self._cache[item] + + if isinstance(item, basestring): + # compatability + for obj in self._cache: + if obj.path == item: + return obj + # END for each obj + raise KeyError( "Blob or Tree named %s not found" % item ) + # END index is basestring + + raise TypeError( "Invalid index type: %r" % item ) + + + def __contains__(self,item): + if isinstance(item, base.IndexObject): + return item in self._cache + + # compatability + for obj in self._cache: + if item == obj.path: + return True + # END for each item + return False + + def __reversed__(self): + return reversed(self._cache) |