def dump_html(self): ''' dump HTML-formated revision content from RESTBase for the given wiki and date ''' dumpdir = MiscDumpDir(self.wiki.config, self.wiki.date) outputdir = dumpdir.get_dumpdir(self.wiki.db_name, self.wiki.date) htmlfile = HTMLFile(self.wiki.config, self.wiki.date, self.wiki.db_name) outputfile = htmlfile.get_filename(self.args['ns']) # /usr/bin/nodejs ./bin/dump_wiki --domain en.wikipedia.org --ns 0 \ # --apiURL http://en.wikipedia.org/w/api.php \ # --dataBase /srv/www/htmldumps/dumps/20160826/en.wikipedia.org.articles.ns0.sqlite3 domain = self.get_domain_from_wikidbname() # FIXME: the nodejs wrapper which will do the compress etc stuff for one wiki is # not yet written command = [self.wiki.config.nodejs] command.append(self.wiki.config.scriptpath) command.extend(["--domain", domain, "--ns", self.args['ns'], "--apiURL", "http://%s/w/api.php" % domain, "--dataBase", os.path.join(outputdir, outputfile), "--wiki=%s" % self.wiki.db_name, "--output=gzip:%s" % os.path.join(outputdir, outputfile)]) if self.dryrun: print("would run command for html dump:", command) else: success = RunSimpleCommand.run_with_no_output( command, shell=False, timeout=self.get_lock_timeout_interval(), timeout_callback=self.periodic_callback) if not success: self.log.warning("error producing html files for wiki %s", self.wiki.db_name) return False return True
def dump_stub(self, start_revid, end_revid): ''' dump stubs (metadata) for revs from start_revid up to but not including end_revid ''' if not self.steps['stubs']['run']: return True dumpdir = MiscDumpDir(self.wiki.config, self.wiki.date) outputdir = dumpdir.get_dumpdir(self.wiki.db_name, self.wiki.date) stubfile = StubFile(self.wiki.config, self.wiki.date, self.wiki.db_name) outputfile = stubfile.get_filename() script_command = MultiVersion.mw_script_as_array(self.wiki.config, "dumpBackup.php") command = [self.wiki.config.php] command.extend(script_command) command.extend(["--wiki=%s" % self.wiki.db_name, "--stub", "--quiet", "--output=gzip:%s" % os.path.join(outputdir, outputfile), "--revrange", "--revstart=%s" % start_revid, "--revend=%s" % end_revid]) if self.dryrun: print "would run command for stubs dump:", command else: log.info("running with no output: " + " ".join(command)) success = RunSimpleCommand.run_with_no_output( command, shell=False, timeout=self.get_lock_timeout_interval(), timeout_callback=self.periodic_callback) if not success: log.warning("error producing stub files for wiki %s", self.wiki.db_name) return False return True
def dump_revs(self): ''' dump revision content corresponding to previously-dumped stubs (revision metadata) ''' if not self.steps['revs']['run']: return True dumpdir = MiscDumpDir(self.wiki.config, self.wiki.date) outputdir = dumpdir.get_dumpdir(self.wiki.db_name, self.wiki.date) revsfile = RevsFile(self.wiki.config, self.wiki.date, self.wiki.db_name) outputfile = revsfile.get_filename() script_command = MultiVersion.mw_script_as_array(self.wiki.config, "dumpTextPass.php") command = [self.wiki.config.php] command.extend(script_command) stubfile = StubFile(self.wiki.config, self.wiki.date, self.wiki.db_name) stuboutputfile = stubfile.get_filename() command.extend(["--wiki=%s" % self.wiki.db_name, "--stub=gzip:%s" % os.path.join(outputdir, stuboutputfile), "--quiet", "--spawn=%s" % self.wiki.config.php, "--output=bzip2:%s" % os.path.join(outputdir, outputfile)]) if self.dryrun: print "would run command for revs dump:", command else: log.info("running with no output: " + " ".join(command)) success = RunSimpleCommand.run_with_no_output( command, shell=False, timeout=self.get_lock_timeout_interval(), timeout_callback=self.periodic_callback) if not success: log.warning("error producing revision text files" " for wiki %s", self.wiki.db_name) return False return True
def get_outputfile_indextxt(self, filenames_tocheck, expected, wikiname, dump_date): ''' generate and return a list of text strings that provide a link to the given files, along with filename, size and date. if the file does not exist, it will be silently excluded from the list. the expected list is a list of filenames that are expected to be produced by the dump; currently no errors are generated on this basis but this may change in the future. ''' dirinfo = MiscDumpDir(self.args['config'], dump_date) path = dirinfo.get_dumpdir(wikiname) output_fileinfo = {} for filename in filenames_tocheck: output_fileinfo[filename] = FileUtils.file_info(os.path.join(path, filename)) files_text = [] filenames = sorted(output_fileinfo.keys()) for filename in filenames: file_date, file_size = output_fileinfo[filename] self.log.info("output file %s for %s %s %s", filename, wikiname, safe(file_date), safe(file_size)) if filename in expected and file_date is None: # may do more with this sort of error in the future # for now, just get stats on the other files continue if file_date: files_text.append( "%s: %s (size %s)<br />" % (make_link( os.path.join( wikiname, dump_date, filename), os.path.basename(filename)), file_date, file_size)) return files_text
def do_one_wiki(self): """ run dump of specified type for one wiki, for given date unless it is among the wikis we skip, has already been run for the date, or some other process has the lock and is therefore presumably already dumping it """ if not skip_wiki(self.wiki.db_name, self.wiki.config): dumpdir = MiscDumpDir(self.args["config"], self.args["date"]) if not exists(dumpdir.get_dumpdir(self.wiki.db_name)): os.makedirs(dumpdir.get_dumpdir(self.wiki.db_name)) status_info = StatusInfo(self.args["config"], self.wiki.date, self.wiki.db_name) status = status_info.get_status() if status == "done:all" and not self.flags["forcerun"]: log.info("wiki %s skipped, adds/changes dump already" " complete", self.wiki.db_name) return STATUS_GOOD if not self.flags["dryrun"]: lock = MiscDumpLock(self.args["config"], self.wiki.date, self.wiki.db_name) # if lock is stale, remove it lock.remove_if_stale(self.wiki.config.lock_stale) # try to get the lock ourselves if not lock.get_lock(): log.info( "wiki %s skipped, wiki is locked," " another process should be doing the job", self.wiki.db_name ) return STATUS_TODO self.dumper.set_lockinfo(lock) dumps_dirs = MiscDumpDirs(self.wiki.config, self.wiki.db_name) dumps_dirs.cleanup_old_dumps(self.wiki.date) log.info("Doing run for wiki: %s", self.wiki.db_name) try: result = self.dumper.run() if not result: return STATUS_FAILED if not self.flags["dryrun"]: output_files, expected = self.dumper.get_output_files() if not md5sums(self.wiki, self.wiki.config.fileperms, output_files, expected): return STATUS_FAILED status_info.set_status("done:" + self.dumper.get_steps_done()) lock.unlock_if_owner() if self.flags["do_index"]: index = Index(self.args) index.do_all_wikis() except Exception as ex: log.warning("error from dump run" " for wiki %s", self.wiki.db_name, exc_info=ex) if not self.flags["dryrun"]: lock.unlock_if_owner() return STATUS_FAILED log.info("Success! Wiki %s %s dump complete.", self.wiki.db_name, self.args["dumptype"]) return STATUS_GOOD
def __init__(self, args): """ pass a dict of the standard args (config, date, dumptype, args) """ self.args = args self.indexfile = IndexFile(self.args["config"]) self.dumpdir = MiscDumpDir(self.args["config"])
def __init__(self, dryrun, args, log): ''' pass a dict of the standard args (config, date, dumptype, args) ''' self.args = args self.dryrun = dryrun self.log = log self.indexfile = IndexFile(self.args['config']) self.dumpdir = MiscDumpDir(self.args['config'])
def get_output_files(self): dumpdir = MiscDumpDir(self.wiki.config, self.wiki.date) outputdir = dumpdir.get_dumpdir(self.wiki.db_name, self.wiki.date) htmlfile = HTMLFile(self.wiki.config, self.wiki.date, self.wiki.db_name) filenames = [htmlfile.get_filename(self.args['ns'])] return [os.path.join(outputdir, filename) for filename in filenames]
class Index(): ''' generate index.html page containing information for the dump run of the specified date for all wikis ''' def __init__(self, dryrun, args, log): ''' pass a dict of the standard args (config, date, dumptype, args) ''' self.args = args self.dryrun = dryrun self.log = log self.indexfile = IndexFile(self.args['config']) self.dumpdir = MiscDumpDir(self.args['config']) def do_all_wikis(self): ''' generate index.html file for all wikis for the given date. This should only be done if the run is the most recent, otherwise your nice fresh index.html file willbe replaced with old data. ''' text = "" for wikiname in self.args['config'].all_wikis_list: result = self.do_one_wiki(wikiname) if result: self.log.info("result for wiki %s is %s", wikiname, result) text = text + "<li>" + result + "</li>\n" index_text = (self.args['config'].read_template(self.args['config'].indextmpl) % {"items": text}) if self.dryrun: print("would write {path} with index text".format(path=self.indexfile.get_path())) else: FileUtils.write_file_in_place(self.indexfile.get_path(), index_text, self.args['config'].fileperms) def get_outputfile_indextxt(self, filenames_tocheck, expected, wikiname, dump_date): ''' generate and return a list of text strings that provide a link to the given files, along with filename, size and date. if the file does not exist, it will be silently excluded from the list. the expected list is a list of filenames that are expected to be produced by the dump; currently no errors are generated on this basis but this may change in the future. ''' dirinfo = MiscDumpDir(self.args['config'], dump_date) path = dirinfo.get_dumpdir(wikiname) output_fileinfo = {} for filename in filenames_tocheck: output_fileinfo[filename] = FileUtils.file_info(os.path.join(path, filename)) files_text = [] filenames = sorted(output_fileinfo.keys()) for filename in filenames: file_date, file_size = output_fileinfo[filename] self.log.info("output file %s for %s %s %s", filename, wikiname, safe(file_date), safe(file_size)) if filename in expected and file_date is None: # may do more with this sort of error in the future # for now, just get stats on the other files continue if file_date: files_text.append( "%s: %s (size %s)<br />" % (make_link( os.path.join( wikiname, dump_date, filename), os.path.basename(filename)), file_date, file_size)) return files_text def get_stat_text(self, dump_date, wikiname): ''' generate and return the text string describing the status of the dump of the wiki for the given date ''' stat = StatusFile(self.args['config'], dump_date, wikiname) stat_contents = FileUtils.read_file(stat.get_path()) self.log.info("status for %s %s", wikiname, safe(stat_contents)) if stat_contents: stat_text = "(%s)" % (stat_contents) else: stat_text = None return stat_text def get_files_text(self, wiki): ''' given wiki object, return the list of links and descriptions for the output files for that wiki of the current dump type and date ''' dump_class = MiscDumpFactory.get_dumper(self.args['dumptype']) dumper = dump_class(wiki, self.log, False, self.args['args']) output_files, expected = dumper.get_output_files() files_text = self.get_outputfile_indextxt(output_files, expected, wiki.db_name, wiki.date) md5file = MD5File(wiki.config, wiki.date, wiki.db_name) md5file_text = self.get_outputfile_indextxt( [md5file.get_filename()], [], wiki.db_name, wiki.date) files_text.extend(md5file_text) return files_text def do_one_wiki(self, wikiname, date=None): ''' collect the text strings for one wiki to be inserted into the index.html file ''' if not skip_wiki(wikiname, self.args['config']): dumps_dirs = MiscDumpDirs(self.args['config'], wikiname, self.log) if not exists(self.dumpdir.get_dumpdir_no_date(wikiname)): self.log.info("No dump for wiki %s", wikiname) return None if date is not None: dump_date = date else: dump_date = dumps_dirs.get_latest_dump_date(True) if not dump_date: self.log.info("No dump for wiki %s", wikiname) return None other_runs_text = "other runs: %s<br />" % make_link(wikiname, wikiname) try: wiki = Wiki(self.args['config'], wikiname) wiki.set_date(dump_date) files_text = self.get_files_text(wiki) stat_text = self.get_stat_text(dump_date, wikiname) except Exception as ex: self.log.warning("Error encountered, no information available" " for wiki %s", wikiname, exc_info=ex) return ("<strong>%s</strong> Error encountered," " no information available | %s" % (wikiname, other_runs_text)) try: wikiname_text = "<strong>%s</strong>" % wikiname wiki_info = (" ".join([entry for entry in [wikiname_text, stat_text] if entry is not None]) + "<br />") wiki_info = (wiki_info + " " + "\n ".join(files_text)) wiki_info = wiki_info + "\n " + other_runs_text except Exception as ex: self.log.warning("Error encountered formatting information" " for wiki %s", wikiname, exc_info=ex) return ("Error encountered formatting information" " for wiki %s" % wikiname) return wiki_info return None