-from bz2 import BZ2Decompressor
-from zlib import decompressobj, MAX_WBITS
-from gzip import FCOMMENT, FEXTRA, FHCRC, FNAME, FTEXT
from urlparse import urlparse
-from binascii import a2b_hex, b2a_hex
-import os, sha, md5
+import os
-from twisted.python import log, filepath
+from twisted.python import log
+from twisted.python.filepath import FilePath
from twisted.internet import defer
from twisted.trial import unittest
-from twisted.web2 import stream
from twisted.web2.http import splitHostPort
from AptPackages import AptPackages
-aptpkg_dir='.apt-dht'
-
-DECOMPRESS_EXTS = ['.gz', '.bz2']
-DECOMPRESS_FILES = ['release', 'sources', 'packages']
+aptpkg_dir='apt-packages'
class MirrorError(Exception):
"""Exception raised when there's a problem with the mirror."""
-class ProxyFileStream(stream.SimpleStream):
- """Saves a stream to a file while providing a new stream."""
-
- def __init__(self, stream, outFile, hashType = "sha1", decompress = None, decFile = None):
- """Initializes the proxy.
-
- @type stream: C{twisted.web2.stream.IByteStream}
- @param stream: the input stream to read from
- @type outFile: C{twisted.python.filepath.FilePath}
- @param outFile: the file to write to
- @type hashType: C{string}
- @param hashType: also hash the file using this hashing function
- (currently only 'sha1' and 'md5' are supported)
- @type decompress: C{string}
- @param decompress: also decompress the file as this type
- (currently only '.gz' and '.bz2' are supported)
- @type decFile: C{twisted.python.filepath.FilePath}
- @param decFile: the file to write the decompressed data to
- """
- self.stream = stream
- self.outFile = outFile.open('w')
- self.hasher = None
- if hashType == "sha1":
- self.hasher = sha.new()
- elif hashType == "md5":
- self.hasher = md5.new()
- self.gzfile = None
- self.bz2file = None
- if decompress == ".gz":
- self.gzheader = True
- self.gzfile = decFile.open('w')
- self.gzdec = decompressobj(-MAX_WBITS)
- elif decompress == ".bz2":
- self.bz2file = decFile.open('w')
- self.bz2dec = BZ2Decompressor()
- self.length = self.stream.length
- self.start = 0
- self.doneDefer = defer.Deferred()
-
- def _done(self):
- """Close the output file."""
- if not self.outFile.closed:
- self.outFile.close()
- fileHash = None
- if self.hasher:
- fileHash = self.hasher.digest()
- if self.gzfile:
- data_dec = self.gzdec.flush()
- self.gzfile.write(data_dec)
- self.gzfile.close()
- self.gzfile = None
- if self.bz2file:
- self.bz2file.close()
- self.bz2file = None
-
- self.doneDefer.callback(fileHash)
-
- def read(self):
- """Read some data from the stream."""
- if self.outFile.closed:
- return None
-
- data = self.stream.read()
- if isinstance(data, defer.Deferred):
- data.addCallbacks(self._write, self._done)
- return data
-
- self._write(data)
- return data
-
- def _write(self, data):
- """Write the stream data to the file and return it for others to use."""
- if data is None:
- self._done()
- return data
-
- self.outFile.write(data)
- if self.hasher:
- self.hasher.update(data)
- if self.gzfile:
- if self.gzheader:
- self.gzheader = False
- new_data = self._remove_gzip_header(data)
- dec_data = self.gzdec.decompress(new_data)
- else:
- dec_data = self.gzdec.decompress(data)
- self.gzfile.write(dec_data)
- if self.bz2file:
- dec_data = self.bz2dec.decompress(data)
- self.bz2file.write(dec_data)
- return data
-
- def _remove_gzip_header(self, data):
- if data[:2] != '\037\213':
- raise IOError, 'Not a gzipped file'
- if ord(data[2]) != 8:
- raise IOError, 'Unknown compression method'
- flag = ord(data[3])
- # modtime = self.fileobj.read(4)
- # extraflag = self.fileobj.read(1)
- # os = self.fileobj.read(1)
-
- skip = 10
- if flag & FEXTRA:
- # Read & discard the extra field, if present
- xlen = ord(data[10])
- xlen = xlen + 256*ord(data[11])
- skip = skip + 2 + xlen
- if flag & FNAME:
- # Read and discard a null-terminated string containing the filename
- while True:
- if not data[skip] or data[skip] == '\000':
- break
- skip += 1
- skip += 1
- if flag & FCOMMENT:
- # Read and discard a null-terminated string containing a comment
- while True:
- if not data[skip] or data[skip] == '\000':
- break
- skip += 1
- skip += 1
- if flag & FHCRC:
- skip += 2 # Read & discard the 16-bit header CRC
- return data[skip:]
-
- def close(self):
- """Clean everything up and return None to future reads."""
- self.length = 0
- self._done()
- self.stream.close()
-
class MirrorManager:
"""Manages all requests for mirror objects."""
- def __init__(self, cache_dir):
+ def __init__(self, cache_dir, unload_delay):
self.cache_dir = cache_dir
- self.cache = filepath.FilePath(self.cache_dir)
+ self.unload_delay = unload_delay
self.apt_caches = {}
def extractPath(self, url):
self.apt_caches[site] = {}
if baseDir not in self.apt_caches[site]:
- site_cache = os.path.join(self.cache_dir, aptpkg_dir, 'mirrors', site + baseDir.replace('/', '_'))
- self.apt_caches[site][baseDir] = AptPackages(site_cache)
+ site_cache = self.cache_dir.child(aptpkg_dir).child('mirrors').child(site + baseDir.replace('/', '_'))
+ site_cache.makedirs
+ self.apt_caches[site][baseDir] = AptPackages(site_cache, self.unload_delay)
def updatedFile(self, url, file_path):
site, baseDir, path = self.extractPath(url)
def findHash(self, url):
site, baseDir, path = self.extractPath(url)
if site in self.apt_caches and baseDir in self.apt_caches[site]:
- d = self.apt_caches[site][baseDir].findHash(path)
- d.addCallback(self.translateHash)
- return d
+ return self.apt_caches[site][baseDir].findHash(path)
d = defer.Deferred()
d.errback(MirrorError("Site Not Found"))
return d
- def translateHash(self, (hash, size)):
- """Translate a hash from apt's hex encoding to a string."""
- if hash:
- hash = a2b_hex(hash)
- return (hash, size)
-
- def save_file(self, response, hash, size, url):
- """Save a downloaded file to the cache and stream it."""
- log.msg('Returning file: %s' % url)
-
- parsed = urlparse(url)
- destFile = self.cache.preauthChild(parsed[1] + parsed[2])
- log.msg('Saving returned %r byte file to cache: %s' % (response.stream.length, destFile.path))
-
- if destFile.exists():
- log.msg('File already exists, removing: %s' % destFile.path)
- destFile.remove()
- else:
- destFile.parent().makedirs()
-
- root, ext = os.path.splitext(destFile.basename())
- if root.lower() in DECOMPRESS_FILES and ext.lower() in DECOMPRESS_EXTS:
- ext = ext.lower()
- decFile = destFile.sibling(root)
- log.msg('Decompressing to: %s' % decFile.path)
- if decFile.exists():
- log.msg('File already exists, removing: %s' % decFile.path)
- decFile.remove()
- else:
- ext = None
- decFile = None
-
- if hash and len(hash) == 16:
- hashType = "md5"
- else:
- hashType = "sha1"
-
- orig_stream = response.stream
- response.stream = ProxyFileStream(orig_stream, destFile, hashType, ext, decFile)
- response.stream.doneDefer.addCallback(self.save_complete, hash, size, url, destFile,
- response.headers.getHeader('Last-Modified'),
- ext, decFile)
- response.stream.doneDefer.addErrback(self.save_error, url)
- return response
-
- def save_complete(self, result, hash, size, url, destFile, modtime = None, ext = None, decFile = None):
- """Update the modification time and AptPackages."""
- if modtime:
- os.utime(destFile.path, (modtime, modtime))
- if ext:
- os.utime(decFile.path, (modtime, modtime))
-
- if not hash or result == hash:
- if hash:
- log.msg('Hashes match: %s' % url)
- else:
- log.msg('Hashed file to %s: %s' % (b2a_hex(result), url))
-
- self.updatedFile(url, destFile.path)
- if ext:
- self.updatedFile(url[:-len(ext)], decFile.path)
- else:
- log.msg("Hashes don't match %s != %s: %s" % (b2a_hex(hash), b2a_hex(result), url))
-
- def save_error(self, failure, url):
- """An error has occurred in downloadign or saving the file."""
- log.msg('Error occurred downloading %s' % url)
- log.err(failure)
- return failure
-
class TestMirrorManager(unittest.TestCase):
"""Unit tests for the mirror manager."""
client = None
def setUp(self):
- self.client = MirrorManager('/tmp')
+ self.client = MirrorManager(FilePath('/tmp/.apt-dht'), 300)
def test_extractPath(self):
site, baseDir, path = self.client.extractPath('http://ftp.us.debian.org/debian/dists/unstable/Release')
self.failUnless(path == "/dists/unstable/Release", "no match: %s" % path)
def verifyHash(self, found_hash, path, true_hash):
- self.failUnless(found_hash[0] == true_hash,
- "%s hashes don't match: %s != %s" % (path, found_hash[0], true_hash))
+ self.failUnless(found_hash.hexexpected() == true_hash,
+ "%s hashes don't match: %s != %s" % (path, found_hash.hexexpected(), true_hash))
def test_findHash(self):
self.packagesFile = os.popen('ls -Sr /var/lib/apt/lists/ | grep -E "_main_.*Packages$" | tail -n 1').read().rstrip('\n')
break
self.client.updatedFile('http://' + self.releaseFile.replace('_','/'),
- '/var/lib/apt/lists/' + self.releaseFile)
+ FilePath('/var/lib/apt/lists/' + self.releaseFile))
self.client.updatedFile('http://' + self.releaseFile[:self.releaseFile.find('_dists_')+1].replace('_','/') +
self.packagesFile[self.packagesFile.find('_dists_')+1:].replace('_','/'),
- '/var/lib/apt/lists/' + self.packagesFile)
+ FilePath('/var/lib/apt/lists/' + self.packagesFile))
self.client.updatedFile('http://' + self.releaseFile[:self.releaseFile.find('_dists_')+1].replace('_','/') +
self.sourcesFile[self.sourcesFile.find('_dists_')+1:].replace('_','/'),
- '/var/lib/apt/lists/' + self.sourcesFile)
+ FilePath('/var/lib/apt/lists/' + self.sourcesFile))
lastDefer = defer.Deferred()
idx_path = 'http://' + self.releaseFile.replace('_','/')[:-7] + 'main/binary-i386/Packages.bz2'
d = self.client.findHash(idx_path)
- d.addCallback(self.verifyHash, idx_path, a2b_hex(idx_hash))
+ d.addCallback(self.verifyHash, idx_path, idx_hash)
pkg_hash = os.popen('grep -A 30 -E "^Package: dpkg$" ' +
'/var/lib/apt/lists/' + self.packagesFile +
' | cut -d\ -f 2').read().rstrip('\n')
d = self.client.findHash(pkg_path)
- d.addCallback(self.verifyHash, pkg_path, a2b_hex(pkg_hash))
+ d.addCallback(self.verifyHash, pkg_path, pkg_hash)
src_dir = os.popen('grep -A 30 -E "^Package: dpkg$" ' +
'/var/lib/apt/lists/' + self.sourcesFile +
for i in range(len(src_hashes)):
src_path = 'http://' + self.releaseFile[:self.releaseFile.find('_dists_')+1].replace('_','/') + src_dir + '/' + src_paths[i]
d = self.client.findHash(src_path)
- d.addCallback(self.verifyHash, src_path, a2b_hex(src_hashes[i]))
+ d.addCallback(self.verifyHash, src_path, src_hashes[i])
idx_hash = os.popen('grep -A 3000 -E "^SHA1:" ' +
'/var/lib/apt/lists/' + self.releaseFile +
idx_path = 'http://' + self.releaseFile.replace('_','/')[:-7] + 'main/source/Sources.bz2'
d = self.client.findHash(idx_path)
- d.addCallback(self.verifyHash, idx_path, a2b_hex(idx_hash))
+ d.addCallback(self.verifyHash, idx_path, idx_hash)
d.addBoth(lastDefer.callback)
return lastDefer