#!/usr/bin/python -tt # This program is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Library General Public License for more details. # # You should have received a copy of the GNU General Public License # along with this program; if not, write to the Free Software # Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. # Copyright 2003 Duke University import libxml2 from mdErrors import PackageSackError import mdUtils class PackageSack: """represents sets (sacks) of Package Objects""" def __init__(self): self.nevra = {} #nevra[(Name, Epoch, Version, Release, Arch)] = [] self.obsoletes = {} #obs[obsoletename] = [pkg1, pkg2, pkg3] #the package lists are packages that obsolete the key name self.requires = {} #req[reqname] = [pkg1, pkg2, pkg3] #the package lists are packages that require the key name self.provides = {} #ditto of above but for provides self.conflicts = {} #ditto of above but for conflicts self.filenames = {} # duh self.pkgsByRepo = {} #pkgsByRepo['repoid']= [pkg1, pkg2, pkg3] self.pkgsByID = {} #pkgsById[pkgid] = [pkg1, pkg2] (should really only ever be one value but #you might have repos with the same package in them def __len__(self): return len(self.simplePkgList()) def searchNevra(self, name=None, epoch=None, ver=None, rel=None, arch=None): """return list of pkgobjects matching the nevra requested""" if self.nevra.has_key((name, epoch, ver, rel, arch)): return self.nevra[(name, epoch, ver, rel, arch)] else: return [] def searchID(self, pkgid): """return list of packages based on pkgid""" if self.pkgsByID.has_key(pkgid): return self.pkgsByID[pkgid] else: return [] def searchRequires(self, name): """return list of package requiring the name (any evr and flag)""" if self.requires.has_key(name): return self.requires[name] else: return [] def searchProvides(self, name): """return list of package providing the name (any evr and flag)""" # FIXME - should this do a pkgobj.checkPrco((name, flag, (e,v,r,))?? # has to do a searchFiles and a searchProvides for things starting with / returnList = [] if name[0] == '/': returnList.extend(self.searchFiles(name)) if self.provides.has_key(name): returnList.extend(self.provides[name]) return returnList def searchConflicts(self, name): """return list of package conflicting with the name (any evr and flag)""" if self.conflicts.has_key(name): return self.conflicts[name] else: return [] def searchObsoletes(self, name): """return list of package obsoleting the name (any evr and flag)""" if self.obsoletes.has_key(name): return self.obsoletes[name] else: return [] def searchFiles(self, file): """return list of packages by filename FIXME - need to add regex match against keys in file list """ if self.filenames.has_key(file): return self.filenames[file] else: return [] def _addToDictAsList(self, dict, key, data): if not dict.has_key(key): dict[key] = [] #if data not in dict[key]: - if I enable this the whole world grinds to a halt # need a faster way of looking for the object in any particular list dict[key].append(data) def _delFromListOfDict(self, dict, key, data): if not dict.has_key(key): dict[key] = [] try: dict[key].remove(data) except ValueError: pass if len(dict[key]) == 0: # if it's an empty list of the dict, then kill it del dict[key] def addPackage(self, obj): """add a pkgobject to the packageSack""" (name, epoch, ver, rel, arch) = obj.returnNevraTuple() self._addToDictAsList(self.nevra, (name, epoch, ver, rel, arch), obj) self._addToDictAsList(self.nevra, (name, None, None, None, None), obj) # store the things provided just on name, not the whole require+version # this lets us reduce the set of pkgs to search when we're trying to depSolve for (n, fl, (e,v,r)) in obj.returnPrco('obsoletes'): self._addToDictAsList(self.obsoletes, n, obj) for (n, fl, (e,v,r)) in obj.returnPrco('requires'): self._addToDictAsList(self.requires, n, obj) for (n, fl, (e,v,r)) in obj.returnPrco('provides'): self._addToDictAsList(self.provides, n, obj) for (n, fl, (e,v,r)) in obj.returnPrco('conflicts'): self._addToDictAsList(self.conflicts, n, obj) for ftype in obj.returnFileTypes(): for file in obj.returnFileEntries(ftype): self._addToDictAsList(self.filenames, file, obj) self._addToDictAsList(self.pkgsByID, obj.returnSimple('id'), obj) self._addToDictAsList(self.pkgsByRepo, obj.returnSimple('repoid'), obj) return def delPackage(self, obj): """delete a pkgobject""" # reverse addPackage # remove it from all the dicts (name, epoch, ver, rel, arch) = obj.returnNevraTuple() self._delFromListOfDict(self.nevra, (name, epoch, ver, rel, arch), obj) self._delFromListOfDict(self.nevra, (name, None, None, None, None), obj) for (n, fl, (e,v,r)) in obj.returnPrco('obsoletes'): self._delFromListOfDict(self.obsoletes, n, obj) for (n, fl, (e,v,r)) in obj.returnPrco('requires'): self._delFromListOfDict(self.requires, n, obj) for (n, fl, (e,v,r)) in obj.returnPrco('provides'): self._delFromListOfDict(self.provides, n, obj) for (n, fl, (e,v,r)) in obj.returnPrco('conflicts'): self._delFromListOfDict(self.conflicts, n, obj) for ftype in obj.returnFileTypes(): for file in obj.returnFileEntries(ftype): self._delFromListOfDict(self.filenames, file, obj) self._delFromListOfDict(self.pkgsByID, obj.returnSimple('id'), obj) self._delFromListOfDict(self.pkgsByRepo, obj.returnSimple('repoid'), obj) return def returnPackages(self, repoid=None): """return list of all packages, takes optional repoid""" returnList = [] if repoid is None: for repo in self.pkgsByRepo.keys(): returnList.extend(self.pkgsByRepo[repo]) else: returnlist = self.pkgsByRepo[repoid] return returnList def returnNewestByNameArch(self): """return list of newest packages based on name, arch matching this means(in name.arch form): foo.i386 and foo.noarch are not compared to each other for highest version only foo.i386 and foo.i386 will be compared""" highdict = {} for pkg in self.returnPackages(): (n, e, v ,r, a) = pkg.returnNevraTuple() if not highdict.has_key((n, a)): highdict[(n, a)] = pkg else: pkg2 = highdict[(n, a)] (e2, v2, r2) = pkg2.returnEVR() rc = mdUtils.compareEVR((e,v,r), (e2, v2, r2)) if rc > 0: highdict[(n, a)] = pkg return highdict.values() def returnNewestByName(self): """return list of newest packages based on name matching this means(in name.arch form): foo.i386 and foo.noarch will be compared to each other for highest version""" highdict = {} for pkg in self.returnPackages(): (n, e, v ,r, a) = pkg.returnNevraTuple() if not highdict.has_key(n): highdict[n] = pkg else: pkg2 = highdict[n] (e2, v2, r2) = pkg2.returnEVR() rc = mdUtils.compareEVR((e,v,r), (e2, v2, r2)) if rc > 0: highdict[n] = pkg #elif rc == 0: FIXME - this should do something to determine the best arch, I guess return highdict.values() def simplePkgList(self, repoid=None): """returns a list of pkg tuples (n, a, e, v, r) optionally from a single repoid""" simplelist = [] for pkg in self.returnPackages(repoid): (n, e, v, r, a) = pkg.returnNevraTuple() simplelist.append((n, a, e, v, r)) return simplelist def printPackages(self): for pkg in self.returnPackages(): print pkg.returnNevraPrintable() # packageSack should be a base class # two derived classes could be DBPackageSack and XMLPackageSack # one for importing this data from the localdb # another from XML metadata files class XMLPackageSack(PackageSack): """Derived class from PackageSack to build list from XML metadata file. Needs the Package Object Class passed to it for the Sack""" def __init__(self, pkgObjectClass): PackageSack.__init__(self) self.repoStatus = {} #[repoid]= [primary, filelist, other] (so you can tell #what things have been loaded or not - b/c w/o primary, #filelist and other really can't be loaded self.pkgObjectClass = pkgObjectClass def addFile(self, repoid, file, callback=None): """takes a repository id and an xml file. It populates whatever it can, if you try to populate with a filelist or other metadata file before the primary metadata you'll not like the results""" try: reader = libxml2.newTextReaderFilename(file) except libxml2.treeError: raise PackageSackError, "Invalid or non-existent file: %s" % (file) else: reader.Read() xmlfiletype=reader.Name() # - first node should be the type if xmlfiletype == 'metadata': self.loadPrimaryMD(reader, repoid, callback) elif xmlfiletype == 'filelists': if self._checkRepoStatus(repoid): self.loadFileMD(reader, repoid, callback) elif xmlfiletype == 'otherdata': if self._checkRepoStatus(repoid): self.loadOtherMD(reader, repoid, callback) else: print 'Error: other unknown root element %s' % xmlfiletype def _checkRepoStatus(self, repoid, itemcheck='primary'): """return 1 if itemcheck is in repo""" if self.repoStatus.has_key(repoid): if itemcheck in self.repoStatus[repoid]: return 1 return 0 def loadPrimaryMD(self, reader, repoid, callback=None): """load all the data from the primary metadata xml file""" """ nodes = [] total = 0 while node is not None: if node.type == 'element': total+=1 nodes.append(node) node = node.next continue """ ret = reader.Read() while ret: if reader.NodeType() == 14: ret = reader.Read() continue if reader.NodeType() == 1 and reader.Name() == 'package': if reader.HasAttributes(): if reader.GetAttribute('type') == 'rpm': #print 'a package' po = self.pkgObjectClass(reader, repoid) self.addPackage(po) ret = reader.Read() continue # update the repoStatus if not self.repoStatus.has_key(repoid): self.repoStatus[repoid] = [] if not 'primary' in self.repoStatus[repoid]: self.repoStatus[repoid].append('primary') def loadFileMD(self, reader, repoid, callback=None): """load all the filelist metadata from the file""" ret = reader.Read() while ret: if reader.NodeType() == 14: ret = reader.Read() continue if reader.NodeType() == 1 and reader.Name() == 'package': if reader.HasAttributes(): pkgid = reader.GetAttribute('pkgid') pkgs = self.searchID(pkgid) pkgmatch = 0 mydepth = reader.Depth() for pkg in pkgs: if pkg.returnSimple('repoid') == repoid: # check for matching repo pkgmatch+=1 reader.Read() while 1: if reader.NodeType() == 15 and reader.Depth() == mydepth: break elif reader.NodeType() == 14: ret = reader.Read() continue elif reader.NodeType() == 1: if reader.LocalName() == 'file': (ftype, file) = pkg.loadFileEntry(reader) #self._addToDictAsList(self.filenames, file, pkg) ret = reader.Read() continue if pkgmatch < 1: # FIXME - raise a warning? Emit error? bitch? moan? pass ret = reader.Read() continue # update the repostatus if not 'filelist' in self.repoStatus[repoid]: self.repoStatus[repoid].append('filelist') def loadOtherMD(self, reader, repoid, callback=None): """load the changelog, etc data from the other.xml file""" ret = reader.Read() while ret: if reader.NodeType() == 14: ret = reader.Read() continue if reader.NodeType() == 1 and reader.Name() == 'package': if reader.HasAttributes(): pkgid = reader.GetAttribute('pkgid') pkgs = self.searchID(pkgid) pkgmatch = 0 mydepth = reader.Depth() for pkg in pkgs: if pkg.returnSimple('repoid') == repoid: # check for matching repo pkgmatch+=1 reader.Read() while 1: if reader.NodeType() == 15 and reader.Depth() == mydepth: break elif reader.NodeType() == 14: ret = reader.Read() continue elif reader.NodeType() == 1: if reader.LocalName() == 'changelog': pkg.loadChangeLogEntry(reader) ret = reader.Read() continue if pkgmatch < 1: # FIXME - raise a warning? Emit error? bitch? moan? pass ret = reader.Read() continue if not 'other' in self.repoStatus[repoid]: self.repoStatus[repoid].append('other') class ListPackageSack(PackageSack): """Derived class from PackageSack to build new Sack from list of pkgObjects - like one returned from self.returnNewestByNameArch() or self.returnNewestByName()""" def __init__(self, Objlist=None): PackageSack.__init__(self) if Objlist is not None: self.addList(Objlist) def addList(self, ObjList): for pkgobj in ObjList: self.addPackage(pkgobj)