From af61998e1f8f46b666594d3b1540c5a530282eb8 Mon Sep 17 00:00:00 2001 From: Rob Browning Date: Sat, 2 Dec 2017 13:31:29 -0600 Subject: [PATCH 1/1] Port web to vfs2 Signed-off-by: Rob Browning Tested-by: Rob Browning --- cmd/web-cmd.py | 155 +++++++++++++++++++++++++++++-------------------- 1 file changed, 92 insertions(+), 63 deletions(-) diff --git a/cmd/web-cmd.py b/cmd/web-cmd.py index 4add17c..ea09088 100755 --- a/cmd/web-cmd.py +++ b/cmd/web-cmd.py @@ -5,12 +5,15 @@ exec "$bup_python" "$0" ${1+"$@"} """ # end of bup preamble +from __future__ import print_function from collections import namedtuple import mimetypes, os, posixpath, signal, stat, sys, time, urllib, webbrowser -from bup import options, git, vfs -from bup.helpers import (chunkyreader, debug1, handle_ctrl_c, log, - resource_path, saved_errors) +from bup import options, git, vfs2 as vfs +from bup.helpers import (chunkyreader, debug1, format_filesize, handle_ctrl_c, + log, resource_path, saved_errors) +from bup.metadata import Metadata +from bup.repo import LocalRepo try: from tornado import gen @@ -23,9 +26,16 @@ except ImportError: sys.exit(1) +# FIXME: right now the way hidden files are handled causes every +# directory to be traversed twice. + handle_ctrl_c() +def http_date_from_utc_ns(utc_ns): + return time.strftime('%a, %d %b %Y %H:%M:%S', time.gmtime(utc_ns / 10**9)) + + def _compute_breadcrumbs(path, show_hidden=False): """Returns a list of breadcrumb objects for a path.""" breadcrumbs = [] @@ -41,50 +51,69 @@ def _compute_breadcrumbs(path, show_hidden=False): return breadcrumbs -def _contains_hidden_files(n): - """Return True if n contains files starting with a '.', False otherwise.""" - for sub in n: - name = sub.name - if len(name)>1 and name.startswith('.'): - return True +def _contains_hidden_files(repo, dir_item): + """Return true if the directory contains items with names other than + '.' and '..' that begin with '.' + """ + for name, item in vfs.contents(repo, dir_item, want_meta=False): + if name in ('.', '..'): + continue + if name.startswith('.'): + return True return False -def _compute_dir_contents(n, path, show_hidden=False): - """Given a vfs node, returns an iterator for display info of all subs.""" - url_append = "" - if show_hidden: - url_append = "?hidden=1" +def _dir_contents(repo, resolution, show_hidden=False): + """Yield the display information for the contents of dir_item.""" - if path != "/": - yield('..', '../' + url_append, '') - for sub in n: - display = sub.name - link = urllib.quote(sub.name) + url_query = '?hidden=1' if show_hidden else '' + def display_info(name, item, resolved_item, display_name=None): # link should be based on fully resolved type to avoid extra # HTTP redirect. - if stat.S_ISDIR(sub.try_resolve().mode): - link += "/" - - if not show_hidden and len(display)>1 and display.startswith('.'): - continue - - size = None - if stat.S_ISDIR(sub.mode): - display += '/' - elif stat.S_ISLNK(sub.mode): - display += '@' + if stat.S_ISDIR(vfs.item_mode(resolved_item)): + link = urllib.quote(name) + '/' else: - size = sub.size() - size = (opt.human_readable and format_filesize(size)) or size + link = urllib.quote(name) - yield (display, link + url_append, size) + size = vfs.item_size(repo, item) + if opt.human_readable: + display_size = format_filesize(size) + else: + display_size = size + + if not display_name: + mode = vfs.item_mode(item) + if stat.S_ISDIR(mode): + display_name = name + '/' + elif stat.S_ISLNK(mode): + display_name = name + '@' + else: + display_name = name + + return display_name, link + url_query, display_size + + dir_item = resolution[-1][1] + for name, item in vfs.contents(repo, dir_item): + if not show_hidden: + if (name not in ('.', '..')) and name.startswith('.'): + continue + if name == '.': + yield display_info(name, item, item, '.') + parent_item = resolution[-2][1] if len(resolution) > 1 else dir_item + yield display_info('..', parent_item, parent_item, '..') + continue + res = vfs.try_resolve(repo, name, parent=resolution, want_meta=False) + res_name, res_item = res[-1] + yield display_info(name, item, res_item) class BupRequestHandler(tornado.web.RequestHandler): + def initialize(self, repo=None): + self.repo = repo + def decode_argument(self, value, name=None): if name == 'path': return value @@ -98,30 +127,33 @@ class BupRequestHandler(tornado.web.RequestHandler): def _process_request(self, path): path = urllib.unquote(path) - print 'Handling request for %s' % path - try: - n = top.resolve(path) - except vfs.NoSuchFile: + print('Handling request for %s' % path) + # Set want_meta because dir metadata won't be fetched, and if + # it's not a dir, then we're going to want the metadata. + res = vfs.resolve(self.repo, path, want_meta=True) + leaf_name, leaf_item = res[-1] + if not leaf_item: self.send_error(404) return - f = None - if stat.S_ISDIR(n.mode): - self._list_directory(path, n) + mode = vfs.item_mode(leaf_item) + if stat.S_ISDIR(mode): + self._list_directory(path, res) else: - self._get_file(path, n) + self._get_file(self.repo, path, res) - def _list_directory(self, path, n): + def _list_directory(self, path, resolution): """Helper to produce a directory listing. Return value is either a file object, or None (indicating an error). In either case, the headers are sent. """ if not path.endswith('/') and len(path) > 0: - print 'Redirecting from %s to %s' % (path, path + '/') + print('Redirecting from %s to %s' % (path, path + '/')) return self.redirect(path + '/', permanent=True) + hidden_arg = self.request.arguments.get('hidden', [0])[-1] try: - show_hidden = int(self.request.arguments.get('hidden', [0])[-1]) + show_hidden = int(hidden_arg) except ValueError as e: show_hidden = False @@ -129,32 +161,33 @@ class BupRequestHandler(tornado.web.RequestHandler): 'list-directory.html', path=path, breadcrumbs=_compute_breadcrumbs(path, show_hidden), - files_hidden=_contains_hidden_files(n), + files_hidden=_contains_hidden_files(self.repo, resolution[-1][1]), hidden_shown=show_hidden, - dir_contents=_compute_dir_contents(n, path, show_hidden)) + dir_contents=_dir_contents(self.repo, resolution, + show_hidden=show_hidden)) @gen.coroutine - def _get_file(self, path, n): + def _get_file(self, repo, path, resolved): """Process a request on a file. Return value is either a file object, or None (indicating an error). In either case, the headers are sent. """ + file_item = resolved[-1][1] + file_item = vfs.augment_item_meta(repo, file_item, include_size=True) + meta = file_item.meta ctype = self._guess_type(path) - self.set_header("Last-Modified", self.date_time_string(n.mtime)) + self.set_header("Last-Modified", http_date_from_utc_ns(meta.mtime)) self.set_header("Content-Type", ctype) - size = n.size() - self.set_header("Content-Length", str(size)) - assert(len(n.hash) == 20) - self.set_header("Etag", n.hash.encode('hex')) + + self.set_header("Content-Length", str(meta.size)) + assert len(file_item.oid) == 20 + self.set_header("Etag", file_item.oid.encode('hex')) if self.request.method != 'HEAD': - f = n.open() - try: + with vfs.fopen(self.repo, file_item) as f: it = chunkyreader(f) for blob in chunkyreader(f): self.write(blob) - finally: - f.close() raise gen.Return() def _guess_type(self, path): @@ -189,9 +222,6 @@ class BupRequestHandler(tornado.web.RequestHandler): '.h': 'text/plain', }) - def date_time_string(self, t): - return time.strftime('%a, %d %b %Y %H:%M:%S', time.gmtime(t)) - io_loop = None @@ -242,7 +272,6 @@ else: address = InetAddress(host=host, port=port) git.check_repo_or_die() -top = vfs.RefList(None) settings = dict( debug = 1, @@ -254,7 +283,7 @@ settings = dict( sys.stdout = os.fdopen(sys.stdout.fileno(), 'w', 0) application = tornado.web.Application([ - (r"(?P/.*)", BupRequestHandler), + (r"(?P/.*)", BupRequestHandler, dict(repo=LocalRepo())), ], **settings) http_server = HTTPServer(application) @@ -266,14 +295,14 @@ if isinstance(address, InetAddress): sock = http_server._socket # tornado < 2.0 except AttributeError as e: sock = http_server._sockets.values()[0] - print "Serving HTTP on %s:%d..." % sock.getsockname() + print('Serving HTTP on %s:%d...' % sock.getsockname()) if opt.browser: browser_addr = 'http://' + address[0] + ':' + str(address[1]) io_loop_pending.add_callback(lambda : webbrowser.open(browser_addr)) elif isinstance(address, UnixAddress): unix_socket = bind_unix_socket(address.path) http_server.add_socket(unix_socket) - print "Serving HTTP on filesystem socket %r" % address.path + print('Serving HTTP on filesystem socket %r' % address.path) else: log('error: unexpected address %r', address) sys.exit(1) -- 2.39.2