def hash_file(path, block_size, hash_func):
hash = hash_func()
with open(path, "rb") as stream:
while True:
# Read a block of character data.
data = stream.read(block_size)
if not data:
return encode(hash.hexdigest())
# Update hash.
hash.update(data)
def _product_hash_rec(root, resolve_root, resolve_links, hash_func,
block_size):
if os.path.islink(root) and not (resolve_root or resolve_links):
# Hash link _contents_.
return hash_string(path_utf8(os.readlink(root)), hash_func)
elif os.path.isfile(root):
# Hash file contents.
return hash_file(root, block_size, hash_func)
elif os.path.isdir(root):
# Create a fingerprint of the directory by computing the hash of (for each entry in the directory) the hash
# of the entry name, the type of entry (link, file, or directory), and the hash of the contents of the
# entry.
hash = hash_func()
for basename in sorted(os.listdir(root)):
hash.update(hash_string(path_utf8(basename), hash_func))
path = os.path.join(root, basename)
if os.path.islink(path) and not (resolve_root
or resolve_links):
hash.update(b"l")
elif os.path.isdir(path):
hash.update(b"d")
else:
hash.update(b"f")
hash.update(
_product_hash_rec(path, False, resolve_links, hash_func,
block_size))
return encode(hash.hexdigest())
else:
raise IOError(
"path does not refer to a regular file or directory: %s" %
root)
def hash_string(string, hash_func):
hash = hash_func()
hash.update(string)
return encode(hash.hexdigest())