Mercurial > hg
view hgext/inotify/server.py @ 9115:b55d44719b47
inotify: server: new data structure to keep track of changes.
== Rationale for the new structure ==
Current structure was a dictionary tree. One directory was tracked
as a dictionary:
- keys: file/subdir name
- values:
- for a file, the status (a/r/m/...)
- for a subdir, the directory representing the subdir
It allowed efficient lookups, no matter of the type of the terminal leaf:
for part in path.split('/'): tree = tree[part]
However, there is no way to represent a directory and a file with the same name
because keys are conflicting in the dictionary. Concrete example:
Initial state:
root dir
|- foo (file)
|- bar (file)
# data state is: {'foo': 'n', 'bar': 'n'}
Remove foo:
root dir
|- bar (file)
# Data becomes {'foo': 'r'} until next commit.
Add foo, as a directory, and foo/barbar file:
root dir
|- bar (file)
|-> foo (dir)
|- barbar (file)
# New state should be represented as:
{'foo': {'barbar': 'a'}, 'bar': 'n'}
however, the key "foo" is already used and represents the old file.
The dirstate:
D foo
A foo/barbar
cannot be represented, hence the need for a new structure.
== The new structure ==
'directory' class. Represents one directory level.
* Notable attributes:
Two dictionaries:
- 'files' Maps filename -> status for the current dir.
- 'dirs' Maps subdir's name -> directory object representing the subdir
* methods
- walk(), formerly server.walk
- lookup(), old server.lookup
- dir(), old server.dir
This new class allows embedding all the tree walks/lookups in its own class,
instead of having everything mixed together in server.
Incidently, since files and directories are not stored in the same
dictionaries, we are solving the previous key conflict problem.
The small drawback is that lookup operation is a bit more complex:
for a path a/b/c/d/e we have to check twice the leaf, if e is a directory or a
file.
author | Nicolas Dumazet <nicdumz.commits@gmail.com> |
---|---|
date | Tue, 26 May 2009 23:00:35 +0900 |
parents | e1d119f450f0 |
children | f90bbf1ea09f |
line wrap: on
line source
# server.py - inotify status server # # Copyright 2006, 2007, 2008 Bryan O'Sullivan <bos@serpentine.com> # Copyright 2007, 2008 Brendan Cully <brendan@kublai.com> # # This software may be used and distributed according to the terms of the # GNU General Public License version 2, incorporated herein by reference. from mercurial.i18n import _ from mercurial import osutil, util import common import errno, os, select, socket, stat, struct, sys, tempfile, time try: import linux as inotify from linux import watcher except ImportError: raise class AlreadyStartedException(Exception): pass def join(a, b): if a: if a[-1] == '/': return a + b return a + '/' + b return b def split(path): c = path.rfind('/') if c == -1: return '', path return path[:c], path[c+1:] walk_ignored_errors = (errno.ENOENT, errno.ENAMETOOLONG) def walkrepodirs(repo): '''Iterate over all subdirectories of this repo. Exclude the .hg directory, any nested repos, and ignored dirs.''' rootslash = repo.root + os.sep def walkit(dirname, top): fullpath = rootslash + dirname try: for name, kind in osutil.listdir(fullpath): if kind == stat.S_IFDIR: if name == '.hg': if not top: return else: d = join(dirname, name) if repo.dirstate._ignore(d): continue for subdir in walkit(d, False): yield subdir except OSError, err: if err.errno not in walk_ignored_errors: raise yield fullpath return walkit('', True) def walk(repo, root): '''Like os.walk, but only yields regular files.''' # This function is critical to performance during startup. rootslash = repo.root + os.sep def walkit(root, reporoot): files, dirs = [], [] try: fullpath = rootslash + root for name, kind in osutil.listdir(fullpath): if kind == stat.S_IFDIR: if name == '.hg': if not reporoot: return else: dirs.append(name) path = join(root, name) if repo.dirstate._ignore(path): continue for result in walkit(path, False): yield result elif kind in (stat.S_IFREG, stat.S_IFLNK): files.append(name) yield fullpath, dirs, files except OSError, err: if err.errno not in walk_ignored_errors: raise return walkit(root, root == '') def _explain_watch_limit(ui, repo): path = '/proc/sys/fs/inotify/max_user_watches' try: limit = int(file(path).read()) except IOError, err: if err.errno != errno.ENOENT: raise raise util.Abort(_('this system does not seem to ' 'support inotify')) ui.warn(_('*** the current per-user limit on the number ' 'of inotify watches is %s\n') % limit) ui.warn(_('*** this limit is too low to watch every ' 'directory in this repository\n')) ui.warn(_('*** counting directories: ')) ndirs = len(list(walkrepodirs(repo))) ui.warn(_('found %d\n') % ndirs) newlimit = min(limit, 1024) while newlimit < ((limit + ndirs) * 1.1): newlimit *= 2 ui.warn(_('*** to raise the limit from %d to %d (run as root):\n') % (limit, newlimit)) ui.warn(_('*** echo %d > %s\n') % (newlimit, path)) raise util.Abort(_('cannot watch %s until inotify watch limit is raised') % repo.root) class pollable(object): """ Interface to support polling. The file descriptor returned by fileno() is registered to a polling object. Usage: Every tick, check if an event has happened since the last tick: * If yes, call handle_events * If no, call handle_timeout """ poll_events = select.POLLIN instances = {} poll = select.poll() def fileno(self): raise NotImplementedError def handle_events(self, events): raise NotImplementedError def handle_timeout(self): raise NotImplementedError def shutdown(self): raise NotImplementedError def register(self, timeout): fd = self.fileno() pollable.poll.register(fd, pollable.poll_events) pollable.instances[fd] = self self.registered = True self.timeout = timeout def unregister(self): pollable.poll.unregister(self) self.registered = False @classmethod def run(cls): while True: timeout = None timeobj = None for obj in cls.instances.itervalues(): if obj.timeout is not None and (timeout is None or obj.timeout < timeout): timeout, timeobj = obj.timeout, obj try: events = cls.poll.poll(timeout) except select.error, err: if err[0] == errno.EINTR: continue raise if events: by_fd = {} for fd, event in events: by_fd.setdefault(fd, []).append(event) for fd, events in by_fd.iteritems(): cls.instances[fd].handle_pollevents(events) elif timeobj: timeobj.handle_timeout() def eventaction(code): """ Decorator to help handle events in repowatcher """ def decorator(f): def wrapper(self, wpath): if code == 'm' and wpath in self.lastevent and \ self.lastevent[wpath] in 'cm': return self.lastevent[wpath] = code self.timeout = 250 f(self, wpath) wrapper.func_name = f.func_name return wrapper return decorator class directory(object): """ Representing a directory * path is the relative path from repo root to this directory * files is a dict listing the files in this directory - keys are file names - values are file status * dirs is a dict listing the subdirectories - key are subdirectories names - values are directory objects """ def __init__(self, relpath=''): self.path = relpath self.files = {} self.dirs = {} def dir(self, relpath): """ Returns the directory contained at the relative path relpath. Creates the intermediate directories if necessary. """ if not relpath: return self l = relpath.split('/') ret = self while l: next = l.pop(0) try: ret = ret.dirs[next] except KeyError: d = directory(join(ret.path, next)) ret.dirs[next] = d ret = d return ret def walk(self, states): """ yield (filename, status) pairs for items in the trees that have status in states. filenames are relative to the repo root """ for file, st in self.files.iteritems(): if st in states: yield join(self.path, file), st for dir in self.dirs.itervalues(): for e in dir.walk(states): yield e def lookup(self, states, path): """ yield root-relative filenames that match path, and whose status are in states: * if path is a file, yield path * if path is a directory, yield directory files * if path is not tracked, yield nothing """ if path[-1] == '/': path = path[:-1] paths = path.split('/') # we need to check separately for last node last = paths.pop() tree = self try: for dir in paths: tree = tree.dirs[dir] except KeyError: # path is not tracked return try: # if path is a directory, walk it for file, st in tree.dirs[last].walk(states): yield file except KeyError: try: if tree.files[last] in states: # path is a file yield path except KeyError: # path is not tracked pass class repowatcher(pollable): """ Watches inotify events """ statuskeys = 'almr!?' mask = ( inotify.IN_ATTRIB | inotify.IN_CREATE | inotify.IN_DELETE | inotify.IN_DELETE_SELF | inotify.IN_MODIFY | inotify.IN_MOVED_FROM | inotify.IN_MOVED_TO | inotify.IN_MOVE_SELF | inotify.IN_ONLYDIR | inotify.IN_UNMOUNT | 0) def __init__(self, ui, repo): self.ui = ui self.repo = repo self.wprefix = self.repo.wjoin('') try: self.watcher = watcher.watcher() except OSError, err: raise util.Abort(_('inotify service not available: %s') % err.strerror) self.threshold = watcher.threshold(self.watcher) self.fileno = self.watcher.fileno self.tree = directory() self.statcache = {} self.statustrees = dict([(s, directory()) for s in self.statuskeys]) self.last_event = None self.lastevent = {} self.register(timeout=None) self.ds_info = self.dirstate_info() self.handle_timeout() self.scan() def event_time(self): last = self.last_event now = time.time() self.last_event = now if last is None: return 'start' delta = now - last if delta < 5: return '+%.3f' % delta if delta < 50: return '+%.2f' % delta return '+%.1f' % delta def dirstate_info(self): try: st = os.lstat(self.repo.join('dirstate')) return st.st_mtime, st.st_ino except OSError, err: if err.errno != errno.ENOENT: raise return 0, 0 def add_watch(self, path, mask): if not path: return if self.watcher.path(path) is None: if self.ui.debugflag: self.ui.note(_('watching %r\n') % path[len(self.wprefix):]) try: self.watcher.add(path, mask) except OSError, err: if err.errno in (errno.ENOENT, errno.ENOTDIR): return if err.errno != errno.ENOSPC: raise _explain_watch_limit(self.ui, self.repo) def setup(self): self.ui.note(_('watching directories under %r\n') % self.repo.root) self.add_watch(self.repo.path, inotify.IN_DELETE) self.check_dirstate() def filestatus(self, fn, st): try: type_, mode, size, time = self.repo.dirstate._map[fn][:4] except KeyError: type_ = '?' if type_ == 'n': st_mode, st_size, st_mtime = st if size == -1: return 'l' if size and (size != st_size or (mode ^ st_mode) & 0100): return 'm' if time != int(st_mtime): return 'l' return 'n' if type_ == '?' and self.repo.dirstate._ignore(fn): return 'i' return type_ def updatefile(self, wfn, osstat): ''' update the file entry of an existing file. osstat: (mode, size, time) tuple, as returned by os.lstat(wfn) ''' self._updatestatus(wfn, self.filestatus(wfn, osstat)) def deletefile(self, wfn, oldstatus): ''' update the entry of a file which has been deleted. oldstatus: char in statuskeys, status of the file before deletion ''' if oldstatus == 'r': newstatus = 'r' elif oldstatus in 'almn': newstatus = '!' else: newstatus = None self.statcache.pop(wfn, None) self._updatestatus(wfn, newstatus) def _updatestatus(self, wfn, newstatus): ''' Update the stored status of a file. newstatus: - char in (statuskeys + 'ni'), new status to apply. - or None, to stop tracking wfn ''' root, fn = split(wfn) d = self.tree.dir(root) oldstatus = d.files.get(fn) # oldstatus can be either: # - None : fn is new # - a char in statuskeys: fn is a (tracked) file if self.ui.debugflag and oldstatus != newstatus: self.ui.note(_('status: %r %s -> %s\n') % (wfn, oldstatus, newstatus)) if oldstatus and oldstatus in self.statuskeys \ and oldstatus != newstatus: del self.statustrees[oldstatus].dir(root).files[fn] if newstatus and newstatus != 'i': d.files[fn] = newstatus if newstatus in self.statuskeys: dd = self.statustrees[newstatus].dir(root) if oldstatus != newstatus or fn not in dd.files: dd.files[fn] = newstatus else: d.files.pop(fn, None) def check_deleted(self, key): # Files that had been deleted but were present in the dirstate # may have vanished from the dirstate; we must clean them up. nuke = [] for wfn, ignore in self.statustrees[key].walk(key): if wfn not in self.repo.dirstate: nuke.append(wfn) for wfn in nuke: root, fn = split(wfn) del self.statustrees[key].dir(root).files[fn] del self.tree.dir(root).files[fn] def scan(self, topdir=''): ds = self.repo.dirstate._map.copy() self.add_watch(join(self.repo.root, topdir), self.mask) for root, dirs, files in walk(self.repo, topdir): for d in dirs: self.add_watch(join(root, d), self.mask) wroot = root[len(self.wprefix):] for fn in files: wfn = join(wroot, fn) self.updatefile(wfn, self.getstat(wfn)) ds.pop(wfn, None) wtopdir = topdir if wtopdir and wtopdir[-1] != '/': wtopdir += '/' for wfn, state in ds.iteritems(): if not wfn.startswith(wtopdir): continue try: st = self.stat(wfn) except OSError: status = state[0] self.deletefile(wfn, status) else: self.updatefile(wfn, st) self.check_deleted('!') self.check_deleted('r') def check_dirstate(self): ds_info = self.dirstate_info() if ds_info == self.ds_info: return self.ds_info = ds_info if not self.ui.debugflag: self.last_event = None self.ui.note(_('%s dirstate reload\n') % self.event_time()) self.repo.dirstate.invalidate() self.handle_timeout() self.scan() self.ui.note(_('%s end dirstate reload\n') % self.event_time()) def update_hgignore(self): # An update of the ignore file can potentially change the # states of all unknown and ignored files. # XXX If the user has other ignore files outside the repo, or # changes their list of ignore files at run time, we'll # potentially never see changes to them. We could get the # client to report to us what ignore data they're using. # But it's easier to do nothing than to open that can of # worms. if '_ignore' in self.repo.dirstate.__dict__: delattr(self.repo.dirstate, '_ignore') self.ui.note(_('rescanning due to .hgignore change\n')) self.handle_timeout() self.scan() def getstat(self, wpath): try: return self.statcache[wpath] except KeyError: try: return self.stat(wpath) except OSError, err: if err.errno != errno.ENOENT: raise def stat(self, wpath): try: st = os.lstat(join(self.wprefix, wpath)) ret = st.st_mode, st.st_size, st.st_mtime self.statcache[wpath] = ret return ret except OSError: self.statcache.pop(wpath, None) raise @eventaction('c') def created(self, wpath): if wpath == '.hgignore': self.update_hgignore() try: st = self.stat(wpath) if stat.S_ISREG(st[0]): self.updatefile(wpath, st) except OSError: pass @eventaction('m') def modified(self, wpath): if wpath == '.hgignore': self.update_hgignore() try: st = self.stat(wpath) if stat.S_ISREG(st[0]): if self.repo.dirstate[wpath] in 'lmn': self.updatefile(wpath, st) except OSError: pass @eventaction('d') def deleted(self, wpath): if wpath == '.hgignore': self.update_hgignore() elif wpath.startswith('.hg/'): if wpath == '.hg/wlock': self.check_dirstate() return self.deletefile(wpath, self.repo.dirstate[wpath]) def process_create(self, wpath, evt): if self.ui.debugflag: self.ui.note(_('%s event: created %s\n') % (self.event_time(), wpath)) if evt.mask & inotify.IN_ISDIR: self.scan(wpath) else: self.created(wpath) def process_delete(self, wpath, evt): if self.ui.debugflag: self.ui.note(_('%s event: deleted %s\n') % (self.event_time(), wpath)) if evt.mask & inotify.IN_ISDIR: tree = self.tree.dir(wpath) todelete = [wfn for wfn, ignore in tree.walk('?')] for fn in todelete: self.deletefile(fn, '?') self.scan(wpath) else: self.deleted(wpath) def process_modify(self, wpath, evt): if self.ui.debugflag: self.ui.note(_('%s event: modified %s\n') % (self.event_time(), wpath)) if not (evt.mask & inotify.IN_ISDIR): self.modified(wpath) def process_unmount(self, evt): self.ui.warn(_('filesystem containing %s was unmounted\n') % evt.fullpath) sys.exit(0) def handle_pollevents(self, events): if self.ui.debugflag: self.ui.note(_('%s readable: %d bytes\n') % (self.event_time(), self.threshold.readable())) if not self.threshold(): if self.registered: if self.ui.debugflag: self.ui.note(_('%s below threshold - unhooking\n') % (self.event_time())) self.unregister() self.timeout = 250 else: self.read_events() def read_events(self, bufsize=None): events = self.watcher.read(bufsize) if self.ui.debugflag: self.ui.note(_('%s reading %d events\n') % (self.event_time(), len(events))) for evt in events: assert evt.fullpath.startswith(self.wprefix) wpath = evt.fullpath[len(self.wprefix):] if evt.mask & inotify.IN_UNMOUNT: self.process_unmount(wpath, evt) elif evt.mask & (inotify.IN_MODIFY | inotify.IN_ATTRIB): self.process_modify(wpath, evt) elif evt.mask & (inotify.IN_DELETE | inotify.IN_DELETE_SELF | inotify.IN_MOVED_FROM): self.process_delete(wpath, evt) elif evt.mask & (inotify.IN_CREATE | inotify.IN_MOVED_TO): self.process_create(wpath, evt) self.lastevent.clear() def handle_timeout(self): if not self.registered: if self.ui.debugflag: self.ui.note(_('%s hooking back up with %d bytes readable\n') % (self.event_time(), self.threshold.readable())) self.read_events(0) self.register(timeout=None) self.timeout = None def shutdown(self): self.watcher.close() def debug(self): """ Returns a sorted list of relatives paths currently watched, for debugging purposes. """ return sorted(tuple[0][len(self.wprefix):] for tuple in self.watcher) class server(pollable): """ Listens for client queries on unix socket inotify.sock """ def __init__(self, ui, repo, repowatcher, timeout): self.ui = ui self.repo = repo self.repowatcher = repowatcher self.sock = socket.socket(socket.AF_UNIX) self.sockpath = self.repo.join('inotify.sock') self.realsockpath = None try: self.sock.bind(self.sockpath) except socket.error, err: if err[0] == errno.EADDRINUSE: raise AlreadyStartedException(_('could not start server: %s') % err[1]) if err[0] == "AF_UNIX path too long": tempdir = tempfile.mkdtemp(prefix="hg-inotify-") self.realsockpath = os.path.join(tempdir, "inotify.sock") try: self.sock.bind(self.realsockpath) os.symlink(self.realsockpath, self.sockpath) except (OSError, socket.error), inst: try: os.unlink(self.realsockpath) except: pass os.rmdir(tempdir) if inst.errno == errno.EEXIST: raise AlreadyStartedException(_('could not start server: %s') % inst.strerror) raise else: raise self.sock.listen(5) self.fileno = self.sock.fileno self.register(timeout=timeout) def handle_timeout(self): pass def answer_stat_query(self, cs): names = cs.read().split('\0') states = names.pop() self.ui.note(_('answering query for %r\n') % states) if self.repowatcher.timeout: # We got a query while a rescan is pending. Make sure we # rescan before responding, or we could give back a wrong # answer. self.repowatcher.handle_timeout() if not names: def genresult(states, tree): for fn, state in tree.walk(states): yield fn else: def genresult(states, tree): for fn in names: for f in tree.lookup(states, fn): yield f return ['\0'.join(r) for r in [ genresult('l', self.repowatcher.statustrees['l']), genresult('m', self.repowatcher.statustrees['m']), genresult('a', self.repowatcher.statustrees['a']), genresult('r', self.repowatcher.statustrees['r']), genresult('!', self.repowatcher.statustrees['!']), '?' in states and genresult('?', self.repowatcher.statustrees['?']) or [], [], 'c' in states and genresult('n', self.repowatcher.tree) or [], ]] def answer_dbug_query(self): return ['\0'.join(self.repowatcher.debug())] def handle_pollevents(self, events): for e in events: self.handle_pollevent() def handle_pollevent(self): sock, addr = self.sock.accept() cs = common.recvcs(sock) version = ord(cs.read(1)) if version != common.version: self.ui.warn(_('received query from incompatible client ' 'version %d\n') % version) try: # try to send back our version to the client # this way, the client too is informed of the mismatch sock.sendall(chr(common.version)) except: pass return type = cs.read(4) if type == 'STAT': results = self.answer_stat_query(cs) elif type == 'DBUG': results = self.answer_dbug_query() else: self.ui.warn(_('unrecognized query type: %s\n') % type) return try: try: v = chr(common.version) sock.sendall(v + type + struct.pack(common.resphdrfmts[type], *map(len, results))) sock.sendall(''.join(results)) finally: sock.shutdown(socket.SHUT_WR) except socket.error, err: if err[0] != errno.EPIPE: raise def shutdown(self): self.sock.close() try: os.unlink(self.sockpath) if self.realsockpath: os.unlink(self.realsockpath) os.rmdir(os.path.dirname(self.realsockpath)) except OSError, err: if err.errno != errno.ENOENT: raise class master(object): def __init__(self, ui, repo, timeout=None): self.ui = ui self.repo = repo self.repowatcher = repowatcher(ui, repo) self.server = server(ui, repo, self.repowatcher, timeout) def shutdown(self): for obj in pollable.instances.itervalues(): obj.shutdown() def run(self): self.repowatcher.setup() self.ui.note(_('finished setup\n')) if os.getenv('TIME_STARTUP'): sys.exit(0) pollable.run() def start(ui, repo): def closefds(ignore): # (from python bug #1177468) # close all inherited file descriptors # Python 2.4.1 and later use /dev/urandom to seed the random module's RNG # a file descriptor is kept internally as os._urandomfd (created on demand # the first time os.urandom() is called), and should not be closed try: os.urandom(4) urandom_fd = getattr(os, '_urandomfd', None) except AttributeError: urandom_fd = None ignore.append(urandom_fd) for fd in range(3, 256): if fd in ignore: continue try: os.close(fd) except OSError: pass m = master(ui, repo) sys.stdout.flush() sys.stderr.flush() pid = os.fork() if pid: return pid closefds(pollable.instances.keys()) os.setsid() fd = os.open('/dev/null', os.O_RDONLY) os.dup2(fd, 0) if fd > 0: os.close(fd) fd = os.open(ui.config('inotify', 'log', '/dev/null'), os.O_RDWR | os.O_CREAT | os.O_TRUNC) os.dup2(fd, 1) os.dup2(fd, 2) if fd > 2: os.close(fd) try: m.run() finally: m.shutdown() os._exit(0)