"""
Assorted utility functions for yum.
"""

import types
import os
import os.path
from cStringIO import StringIO
import base64
import struct
import re
import errno
import pgpmsg
import tempfile
import glob
import pwd
import fnmatch
import bz2
from stat import *
try:
    import gpgme
    import gpgme.editutil
except ImportError:
    gpgme = None
try:
    import hashlib
    _available_checksums = ['md5', 'sha1', 'sha256', 'sha512']
    _default_checksums = ['sha256']
except ImportError:
    # Python-2.4.z ... gah!
    import sha
    import md5
    _available_checksums = ['md5', 'sha1']
    _default_checksums = ['sha1']
    class hashlib:

        @staticmethod
        def new(algo):
            if algo == 'md5':
                return md5.new()
            if algo == 'sha1':
                return sha.new()
            raise ValueError, "Bad checksum type"

from Errors import MiscError
# These are API things, so we can't remove them even if they aren't used here.
# pylint: disable-msg=W0611
from i18n import to_utf8, to_unicode, to_str
# pylint: enable-msg=W0611

_share_data_store   = {}
_share_data_store_u = {}
def share_data(value):
    """ Take a value and use the same value from the store,
        if the value isn't in the store this one becomes the shared version. """
    #  We don't want to change the types of strings, between str <=> unicode
    # and hash('a') == hash(u'a') ... so use different stores.
    #  In theory eventaully we'll have all of one type, but don't hold breath.
    store = _share_data_store
    if isinstance(value, unicode):
        store = _share_data_store_u
    # hahahah, of course the above means that:
    #   hash(('a', 'b')) == hash((u'a', u'b'))
    # ...which we have in deptuples, so just screw sharing those atm.
    if type(value) == types.TupleType:
        return value
    return store.setdefault(value, value)

def unshare_data():
    global _share_data_store
    global _share_data_store_u
    _share_data_store   = {}
    _share_data_store_u = {}

_re_compiled_glob_match = None
def re_glob(s):
    """ Tests if a string is a shell wildcard. """
    # re.match('.*[\*\?\[\]].*', name)
    global _re_compiled_glob_match
    if _re_compiled_glob_match is None:
        _re_compiled_glob_match = re.compile('.*[\*\?\[\]].*')
    return _re_compiled_glob_match.match(s)

_re_compiled_pri_fnames_match = None
def re_primary_filename(filename):
    global _re_compiled_pri_fnames_match
    if _re_compiled_pri_fnames_match is None:
        one   = re.compile('.*bin\/.*')
        two   = re.compile('^\/etc\/.*')
        three = re.compile('^\/usr\/lib\/sendmail$')
        _re_compiled_pri_fnames_match = (one, two, three)
    for rec in _re_compiled_pri_fnames_match:
        if rec.match(filename):
            return True
    return False

_re_compiled_pri_dnames_match = None
def re_primary_dirname(dirname):
    global _re_compiled_pri_dnames_match
    if _re_compiled_pri_dnames_match is None:
        one   = re.compile('.*bin\/.*')
        two   = re.compile('^\/etc\/.*')
        _re_compiled_pri_dnames_match = (one, two)
    for rec in _re_compiled_pri_dnames_match:
        if rec.match(dirname):
            return True
    return False

_re_compiled_full_match = None
def re_full_search_needed(s):
    """ Tests if a string needs a full nevra match, instead of just name. """
    global _re_compiled_full_match
    if _re_compiled_full_match is None:
        one   = re.compile('.*[-\.\*\?\[\]].*.$') # Any wildcard or - seperator
        two   = re.compile('^[0-9]')        # Any epoch, for envra
        _re_compiled_full_match = (one, two)
    for rec in _re_compiled_full_match:
        if rec.match(s):
            return True
    return False

###########
# Title: Remove duplicates from a sequence
# Submitter: Tim Peters 
# From: http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52560
def unique(s):
    """Return a list of the elements in s, but without duplicates.

    For example, unique([1,2,3,1,2,3]) is some permutation of [1,2,3],
    unique("abcabc") some permutation of ["a", "b", "c"], and
    unique(([1, 2], [2, 3], [1, 2])) some permutation of
    [[2, 3], [1, 2]].

    For best speed, all sequence elements should be hashable.  Then
    unique() will usually work in linear time.

    If not possible, the sequence elements should enjoy a total
    ordering, and if list(s).sort() doesn't raise TypeError it's
    assumed that they do enjoy a total ordering.  Then unique() will
    usually work in O(N*log2(N)) time.

    If that's not possible either, the sequence elements must support
    equality-testing.  Then unique() will usually work in quadratic
    time.
    """

    n = len(s)
    if n == 0:
        return []

    # Try using a set first, as that's the fastest and will usually
    # work.  If it doesn't work, it will usually fail quickly, so it
    # usually doesn't cost much to *try* it.  It requires that all the
    # sequence elements be hashable, and support equality comparison.
    try:
        u = set(s)
    except TypeError:
        pass
    else:
        return list(u)

    # We can't hash all the elements.  Second fastest is to sort,
    # which brings the equal elements together; then duplicates are
    # easy to weed out in a single pass.
    # NOTE:  Python's list.sort() was designed to be efficient in the
    # presence of many duplicate elements.  This isn't true of all
    # sort functions in all languages or libraries, so this approach
    # is more effective in Python than it may be elsewhere.
    try:
        t = list(s)
        t.sort()
    except TypeError:
        del t  # move on to the next method
    else:
        assert n > 0
        last = t[0]
        lasti = i = 1
        while i < n:
            if t[i] != last:
                t[lasti] = last = t[i]
                lasti += 1
            i += 1
        return t[:lasti]

    # Brute force is all that's left.
    u = []
    for x in s:
        if x not in u:
            u.append(x)
    return u

class Checksums:
    """ Generate checksum(s), on given pieces of data. Producing the
        Length and the result(s) when complete. """

    def __init__(self, checksums=None, ignore_missing=False):
        if checksums is None:
            checksums = _default_checksums
        self._sumalgos = []
        self._sumtypes = []
        self._len = 0

        done = set()
        for sumtype in checksums:
            if sumtype == 'sha':
                sumtype = 'sha1'
            if sumtype in done:
                continue

            if sumtype in _available_checksums:
                sumalgo = hashlib.new(sumtype)
            elif ignore_missing:
                continue
            else:
                raise MiscError, 'Error Checksumming, bad checksum type %s' % sumtype
            done.add(sumtype)
            self._sumtypes.append(sumtype)
            self._sumalgos.append(sumalgo)
        if not done:
            raise MiscError, 'Error Checksumming, no valid checksum type'

    def __len__(self):
        return self._len

    def update(self, data):
        self._len += len(data)
        for sumalgo in self._sumalgos:
            sumalgo.update(data)

    def read(self, fo, size=2**16):
        data = fo.read(size)
        self.update(data)
        return data

    def hexdigests(self):
        ret = {}
        for sumtype, sumdata in zip(self._sumtypes, self._sumalgos):
            ret[sumtype] = sumdata.hexdigest()
        return ret

    def hexdigest(self, checksum=None):
        if checksum is None:
            checksum = self._sumtypes[0]
        if checksum == 'sha':
            checksum = 'sha1'
        return self.hexdigests()[checksum]

    def digests(self):
        ret = {}
        for sumtype, sumdata in zip(self._sumtypes, self._sumalgos):
            ret[sumtype] = sumdata.digest()
        return ret

    def digest(self, checksum=None):
        if checksum is None:
            checksum = self._sumtypes[0]
        if checksum == 'sha':
            checksum = 'sha1'
        return self.digests()[checksum]


class AutoFileChecksums:
    """ Generate checksum(s), on given file/fileobject. Pretending to be a file
        object (overrrides read). """

    def __init__(self, fo, checksums, ignore_missing=False):
        self._fo       = fo
        self.checksums = Checksums(checksums, ignore_missing)

    def __getattr__(self, attr):
        return getattr(self._fo, attr)

    def read(self, size=-1):
        return self.checksums.read(self._fo, size)


def checksum(sumtype, file, CHUNK=2**16):
    """takes filename, hand back Checksum of it
       sumtype = md5 or sha/sha1/sha256/sha512 (note sha == sha1)
       filename = /path/to/file
       CHUNK=65536 by default"""
     
    # chunking brazenly lifted from Ryan Tomayko
    try:
        if type(file) not in types.StringTypes:
            fo = file # assume it's a file-like-object
        else:           
            fo = open(file, 'r', CHUNK)

        data = Checksums([sumtype])
        while data.read(fo, CHUNK):
            pass

        if type(file) is types.StringType:
            fo.close()
            del fo
            
        return data.hexdigest(sumtype)
    except (IOError, OSError), e:
        raise MiscError, 'Error opening file for checksum: %s' % file

def getFileList(path, ext, filelist):
    """Return all files in path matching ext, store them in filelist, 
       recurse dirs return list object"""
    
    extlen = len(ext)
    try:
        dir_list = os.listdir(path)
    except OSError, e:
        raise MiscError, ('Error accessing directory %s, %s') % (path, e)
        
    for d in dir_list:
        if os.path.isdir(path + '/' + d):
            filelist = getFileList(path + '/' + d, ext, filelist)
        else:
            if d[-extlen:].lower() == '%s' % (ext):
                newpath = os.path.normpath(path + '/' + d)
                filelist.append(newpath)
                    
    return filelist

class GenericHolder:
    """Generic Holder class used to hold other objects of known types
       It exists purely to be able to do object.somestuff, object.someotherstuff
       or object[key] and pass object to another function that will 
       understand it"""
       
    def __getitem__(self, item):
        if hasattr(self, item):
            return getattr(self, item)
        else:
            raise KeyError, item

def procgpgkey(rawkey):
    '''Convert ASCII armoured GPG key to binary
    '''
    # TODO: CRC checking? (will RPM do this anyway?)
    
    # Normalise newlines
    rawkey = re.compile('(\n|\r\n|\r)').sub('\n', rawkey)

    # Extract block
    block = StringIO()
    inblock = 0
    pastheaders = 0
    for line in rawkey.split('\n'):
        if line.startswith('-----BEGIN PGP PUBLIC KEY BLOCK-----'):
            inblock = 1
        elif inblock and line.strip() == '':
            pastheaders = 1
        elif inblock and line.startswith('-----END PGP PUBLIC KEY BLOCK-----'):
            # Hit the end of the block, get out
            break
        elif pastheaders and line.startswith('='):
            # Hit the CRC line, don't include this and stop
            break
        elif pastheaders:
            block.write(line+'\n')
  
    # Decode and return
    return base64.decodestring(block.getvalue())

def getgpgkeyinfo(rawkey, multiple=False):
    '''Return a dict of info for the given ASCII armoured key text

    Returned dict will have the following keys: 'userid', 'keyid', 'timestamp'

    Will raise ValueError if there was a problem decoding the key.
    '''
    # Catch all exceptions as there can be quite a variety raised by this call
    key_info_objs = []
    try:
        keys = pgpmsg.decode_multiple_keys(rawkey)
    except Exception, e:
        raise ValueError(str(e))
    if len(keys) == 0:
        raise ValueError('No key found in given key data')
    
    for key in keys:    
        keyid_blob = key.public_key.key_id()

        info = {
            'userid': key.user_id,
            'keyid': struct.unpack('>Q', keyid_blob)[0],
            'timestamp': key.public_key.timestamp,
            'fingerprint' : key.public_key.fingerprint,
            'raw_key' : key.raw_key,
        }

        # Retrieve the timestamp from the matching signature packet 
        # (this is what RPM appears to do) 
        for userid in key.user_ids[0]:
            if not isinstance(userid, pgpmsg.signature):
                continue

            if userid.key_id() == keyid_blob:
                # Get the creation time sub-packet if available
                if hasattr(userid, 'hashed_subpaks'):
                    tspkt = \
                        userid.get_hashed_subpak(pgpmsg.SIG_SUB_TYPE_CREATE_TIME)
                    if tspkt != None:
                        info['timestamp'] = int(tspkt[1])
                        break
        key_info_objs.append(info)
    if multiple:      
        return key_info_objs
    else:
        return key_info_objs[0]
        

def keyIdToRPMVer(keyid):
    '''Convert an integer representing a GPG key ID to the hex version string
    used by RPM
    '''
    return "%08x" % (keyid & 0xffffffffL)


def keyInstalled(ts, keyid, timestamp):
    '''
    Return if the GPG key described by the given keyid and timestamp are
    installed in the rpmdb.  

    The keyid and timestamp should both be passed as integers.
    The ts is an rpm transaction set object

    Return values:
        - -1      key is not installed
        - 0       key with matching ID and timestamp is installed
        - 1       key with matching ID is installed but has a older timestamp
        - 2       key with matching ID is installed but has a newer timestamp

    No effort is made to handle duplicates. The first matching keyid is used to 
    calculate the return result.
    '''
    # Convert key id to 'RPM' form
    keyid = keyIdToRPMVer(keyid)

    # Search
    for hdr in ts.dbMatch('name', 'gpg-pubkey'):
        if hdr['version'] == keyid:
            installedts = int(hdr['release'], 16)
            if installedts == timestamp:
                return 0
            elif installedts < timestamp:
                return 1    
            else:
                return 2

    return -1

def import_key_to_pubring(rawkey, keyid, cachedir=None, gpgdir=None):
    # FIXME - cachedir can be removed from this method when we break api
    if gpgme is None:
        return False
    
    if not gpgdir:
        gpgdir = '%s/gpgdir' % cachedir
    
    if not os.path.exists(gpgdir):
        os.makedirs(gpgdir)
    
    key_fo = StringIO(rawkey) 
    os.environ['GNUPGHOME'] = gpgdir
    # import the key
    ctx = gpgme.Context()
    fp = open(os.path.join(gpgdir, 'gpg.conf'), 'wb')
    fp.write('')
    fp.close()
    ctx.import_(key_fo)
    key_fo.close()
    # ultimately trust the key or pygpgme is definitionally stupid
    k = ctx.get_key(keyid)
    gpgme.editutil.edit_trust(ctx, k, gpgme.VALIDITY_ULTIMATE)
    return True
    
def return_keyids_from_pubring(gpgdir):
    if gpgme is None or not os.path.exists(gpgdir):
        return []

    os.environ['GNUPGHOME'] = gpgdir
    ctx = gpgme.Context()
    keyids = []
    for k in ctx.keylist():
        for subkey in k.subkeys:
            if subkey.can_sign:
                keyids.append(subkey.keyid)

    return keyids

def valid_detached_sig(sig_file, signed_file, gpghome=None):
    """takes signature , file that was signed and an optional gpghomedir"""

    if gpgme is None:
        return False

    if gpghome and os.path.exists(gpghome):
        os.environ['GNUPGHOME'] = gpghome

    sig = open(sig_file, 'r')
    signed_text = open(signed_file, 'r')
    plaintext = None
    ctx = gpgme.Context()

    try:
        sigs = ctx.verify(sig, signed_text, plaintext)
    except gpgme.GpgmeError, e:
        return False
    else:
        # is there ever a case where we care about a sig beyond the first one?
        thissig = sigs[0]
        if not thissig:
            return False

        if thissig.validity in (gpgme.VALIDITY_FULL, gpgme.VALIDITY_MARGINAL,
                                gpgme.VALIDITY_ULTIMATE):
            return True

    return False

def getCacheDir(tmpdir='/var/tmp', reuse=True):
    """return a path to a valid and safe cachedir - only used when not running
       as root or when --tempcache is set"""
    
    uid = os.geteuid()
    try:
        usertup = pwd.getpwuid(uid)
        username = usertup[0]
    except KeyError:
        return None # if it returns None then, well, it's bollocksed

    prefix = 'yum-'

    if reuse:
        # check for /var/tmp/yum-username-* - 
        prefix = 'yum-%s-' % username    
        dirpath = '%s/%s*' % (tmpdir, prefix)
        cachedirs = sorted(glob.glob(dirpath))
        for thisdir in cachedirs:
            stats = os.lstat(thisdir)
            if S_ISDIR(stats[0]) and S_IMODE(stats[0]) == 448 and stats[4] == uid:
                return thisdir

    # make the dir (tempfile.mkdtemp())
    cachedir = tempfile.mkdtemp(prefix=prefix, dir=tmpdir)
    return cachedir
        
def sortPkgObj(pkg1 ,pkg2):
    """sorts a list of yum package objects by name"""
    if pkg1.name > pkg2.name:
        return 1
    elif pkg1.name == pkg2.name:
        return 0
    else:
        return -1
        
def newestInList(pkgs):
    """ Return the newest in the list of packages. """
    ret = [ pkgs.pop() ]
    newest = ret[0]
    for pkg in pkgs:
        if pkg.verGT(newest):
            ret = [ pkg ]
            newest = pkg
        elif pkg.verEQ(newest):
            ret.append(pkg)
    return ret

def version_tuple_to_string(evrTuple):
    """
    Convert a tuple representing a package version to a string.

    @param evrTuple: A 3-tuple of epoch, version, and release.

    Return the string representation of evrTuple.
    """
    (e, v, r) = evrTuple
    s = ""
    
    if e not in [0, '0', None]:
        s += '%s:' % e
    if v is not None:
        s += '%s' % v
    if r is not None:
        s += '-%s' % r
    return s

def prco_tuple_to_string(prcoTuple):
    """returns a text string of the prco from the tuple format"""
    
    (name, flag, evr) = prcoTuple
    flags = {'GT':'>', 'GE':'>=', 'EQ':'=', 'LT':'<', 'LE':'<='}
    if flag is None:
        return name
    
    return '%s %s %s' % (name, flags[flag], version_tuple_to_string(evr))
    
def refineSearchPattern(arg):
    """Takes a search string from the cli for Search or Provides
       and cleans it up so it doesn't make us vomit"""
    
    if re.match('.*[\*,\[,\],\{,\},\?,\+].*', arg):
        restring = fnmatch.translate(arg)
    else:
        restring = re.escape(arg)
        
    return restring
    
def bunzipFile(source,dest):
    """ Extract the bzipped contents of source to dest. """
    s_fn = bz2.BZ2File(source, 'r')
    destination = open(dest, 'w')

    while True:
        try:
            data = s_fn.read(1024000)
        except IOError:
            break
        
        if not data: break
        destination.write(data)

    destination.close()
    s_fn.close()

def get_running_kernel_version_release(ts):
    """This takes the output of uname and figures out the (version, release)
    tuple for the running kernel."""
    ver = os.uname()[2]

    # we glob for the file that MIGHT have this kernel
    # and then look up the file in our rpmdb.
    fns = sorted(glob.glob('/boot/vmlinuz*%s*' % ver))
    for fn in fns:
        mi = ts.dbMatch('basenames', fn)
        for h in mi:
            return (h['version'], h['release'])
    
    return (None, None)
 

def find_unfinished_transactions(yumlibpath='/var/lib/yum'):
    """returns a list of the timestamps from the filenames of the unfinished 
       transactions remaining in the yumlibpath specified.
    """
    timestamps = []    
    tsallg = '%s/%s' % (yumlibpath, 'transaction-all*')
    tsdoneg = '%s/%s' % (yumlibpath, 'transaction-done*')
    tsalls = glob.glob(tsallg)
    tsdones = glob.glob(tsdoneg)

    for fn in tsalls:
        trans = os.path.basename(fn)
        timestamp = trans.replace('transaction-all.','')
        timestamps.append(timestamp)

    timestamps.sort()
    return timestamps
    
def find_ts_remaining(timestamp, yumlibpath='/var/lib/yum'):
    """this function takes the timestamp of the transaction to look at and 
       the path to the yum lib dir (defaults to /var/lib/yum)
       returns a list of tuples(action, pkgspec) for the unfinished transaction
       elements. Returns an empty list if none.

    """
    
    to_complete_items = []
    tsallpath = '%s/%s.%s' % (yumlibpath, 'transaction-all', timestamp)    
    tsdonepath = '%s/%s.%s' % (yumlibpath,'transaction-done', timestamp)
    tsdone_items = []

    if not os.path.exists(tsallpath):
        # something is wrong, here, probably need to raise _something_
        return to_complete_items    

            
    if os.path.exists(tsdonepath):
        tsdone_fo = open(tsdonepath, 'r')
        tsdone_items = tsdone_fo.readlines()
        tsdone_fo.close()     
    
    tsall_fo = open(tsallpath, 'r')
    tsall_items = tsall_fo.readlines()
    tsall_fo.close()
    
    for item in tsdone_items:
        # this probably shouldn't happen but it's worth catching anyway
        if item not in tsall_items:
            continue        
        tsall_items.remove(item)
        
    for item in tsall_items:
        item = item.replace('\n', '')
        if item == '':
            continue
        (action, pkgspec) = item.split()
        to_complete_items.append((action, pkgspec))
    
    return to_complete_items

def seq_max_split(seq, max_entries):
    """ Given a seq, split into a list of lists of length max_entries each. """
    ret = []
    num = len(seq)
    seq = list(seq) # Trying to use a set/etc. here is bad
    beg = 0
    while num > max_entries:
        end = beg + max_entries
        ret.append(seq[beg:end])
        beg += max_entries
        num -= max_entries
    ret.append(seq[beg:])
    return ret

def _ugly_utf8_string_hack(item):
    """hands back a unicoded string"""
    # this is backward compat for handling non-utf8 filenames 
    # and content inside packages. :(
    # content that xml can cope with but isn't really kosher

    # if we're anything obvious - do them first
    if item is None:
        return ''
    elif isinstance(item, unicode):    
        return item
    
    # this handles any bogon formats we see
    du = False
    try:
        x = unicode(item, 'ascii')
        du = True
    except UnicodeError:
        encodings = ['utf-8', 'iso-8859-1', 'iso-8859-15', 'iso-8859-2']
        for enc in encodings:
            try:
                x = unicode(item, enc)
            except UnicodeError:
                pass
                
            else:
                if x.encode(enc) == item:
                    if enc != 'utf-8':
                        print '\n%s encoding on %s\n' % (enc, item)
                    return x.encode('utf-8')
    
    
    # Kill bytes (or libxml will die) not in the small byte portion of:
    #  http://www.w3.org/TR/REC-xml/#NT-Char
    # we allow high bytes, if it passed the utf8 check above. Eg.
    # good chars = #x9 | #xA | #xD | [#x20-...]
    newitem = ''
    bad_small_bytes = range(0, 8) + [11, 12] + range(14, 32)
    for char in item:
        if ord(char) in bad_small_bytes:
            pass # Just ignore these bytes...
        elif not du and ord(char) > 127:
            newitem = newitem + '?' # byte by byte equiv of escape
        else:
            newitem = newitem + char
    return newitem

def to_xml(item, attrib=False):
    import xml.sax.saxutils
    item = _ugly_utf8_string_hack(item)
    item = to_utf8(item)
    item = item.rstrip()
    if attrib:
        item = xml.sax.saxutils.escape(item, entities={'"':"&quot;"})
    else:
        item = xml.sax.saxutils.escape(item)
    return item

def unlink_f(filename):
    """ Call os.unlink, but don't die if the file isn't there. This is the main
        difference between "rm -f" and plain "rm". """
    try:
        os.unlink(filename)
    except OSError, e:
        if e.errno != errno.ENOENT:
            raise

# ---------- i18n ----------
import locale
import sys
def setup_locale(override_codecs=True, override_time=False):
    # This test needs to be before locale.getpreferredencoding() as that
    # does setlocale(LC_CTYPE, "")
    try:
        locale.setlocale(locale.LC_ALL, '')
        # set time to C so that we output sane things in the logs (#433091)
        if override_time:
            locale.setlocale(locale.LC_TIME, 'C')
    except locale.Error, e:
        # default to C locale if we get a failure.
        print >> sys.stderr, 'Failed to set locale, defaulting to C'
        os.environ['LC_ALL'] = 'C'
        locale.setlocale(locale.LC_ALL, 'C')
        
    if override_codecs:
        import codecs
        sys.stdout = codecs.getwriter(locale.getpreferredencoding())(sys.stdout)
        sys.stdout.errors = 'replace'


def get_my_lang_code():
    mylang = locale.getlocale(locale.LC_MESSAGES)
    if mylang == (None, None): # odd :)
        mylang = 'C'
    else:
        mylang = '.'.join(mylang)
    
    return mylang