5c6366432eadc33cfdc7c48c1308bce156479905 lrnassar Wed Aug 5 17:24:23 2026 -0700 Harden error handling in trackCountsParse per CR feedback. refs #37975 The report pipeline had no pipefail, so it reported awk's exit status rather than makeUsageReport's. awk is happy with truncated input, so a report that died partway through produced a non-empty partial count file with a success status, and the v2 cache then reused that partial file on every later run. The empty-file check only caught a total failure. Adds pipefail, and removes the partial file when the command fails so a failed run cannot leave something behind for the cache to pick up. The mergeStderr=False path added in the last commit dropped stderr from the error message, since CalledProcessError reports stdout in e.output. That is the one call site using the flag, the bulk tdbQuery, so a failure there lost its reason entirely. The RuntimeError now carries stderr too. Also aborts when tdbQuery returns no trackDb entries at all. tdbQuery exits 0 and prints nothing for a database it does not know, so a mistyped db name used to run to completion treating every track as notInTrackDb before failing with a message that blamed the usage logs. Pins the subshell to bash, which pipefail needs and which the existing sort -t $'\t' already quietly relied on. diff --git src/utils/qa/trackCountsParse src/utils/qa/trackCountsParse index 1e401b2e48d..4a2291329d5 100755 --- src/utils/qa/trackCountsParse +++ src/utils/qa/trackCountsParse @@ -1,468 +1,489 @@ #!/usr/bin/env python3 import subprocess from collections import OrderedDict from time import localtime, strftime from datetime import datetime from dateutil.relativedelta import relativedelta import calendar import math import subprocess,sys,argparse,os def parseArgs(): """ Parse the command line arguments. """ parser = argparse.ArgumentParser(description = __doc__, formatter_class=argparse.RawDescriptionHelpFormatter) optional = parser._action_groups.pop() required = parser.add_argument_group('required arguments') required.add_argument ("dbs", help = "Database to query for track counts, e.g. hg19, hg38, mm10.") required.add_argument ("workDir", help = "Work directory to use for processing and final output. Use full path with '/' at the end.") #The %% is doubled because argparse percent-formats help strings, a bare % aborts -h optional.add_argument ("-c", "--cutOffThreshhold", dest = "cutOffThreshhold", type = float, default = .3, help = "Optional: The %% value, as compared to the trackCounts median, to be used " + \ "as a threshhold to choose what tracks should be filtered. Default is .3.") optional.add_argument ("-n", "--numOfMonthsToCompare", dest = "numOfMonthsToCompare", type = int, default = 6, help = "Optional: The number of months to compare for filtering. " + \ "Default is 6.") optional.add_argument ("-r", "--singleReport", default = False, action = "store_true", help = "Optional: Run as a singleReport. This generates track counts for the specified " + \ "dates. This is useful for seeing track counts over a period of time. Requires the vars below") optional.add_argument ("-s", "--startDate", dest = "startDate", help = "Optional: The start date when running in singleReport mode. " + \ "Date should be formatted as YYYY-MM-DD.") optional.add_argument ("-e", "--endDate", dest = "endDate", help = "Optional: The end date when running in singleReport mode. " + \ "Date should be formatted as YYYY-MM-DD.") optional.add_argument ("-a", "--asOfDate", dest = "asOfDate", help = "Optional: Run as if today were this month, formatted as YYYY-MM. Defaults to " + \ "the current month. This moves the window of months searched, it does not move " + \ "trackDb, which is always read as it is today. Tracks retired since the window " + \ "are reported with a group of 'notInTrackDb'.") if (len(sys.argv) == 1): parser.print_usage() print("\nGenerates track counts based on the log parsing paring script '/hive/users/chmalee/logs/byDate/makeUsageReport'.\n" + \ "The default behavior looks at track counts over the last 6 months and generates a list of tracks that\n" + \ "were below 30% of the median track usage every month, allowing for a floor(n/4) exception where n is\n" + \ "the number of months searched. Using the default 6 months, that allows for 1 exemption.\n" + \ "This output can be used in order to identify seldomly used tracks for archiving, retiring or restructuring.\n\n" + \ "Alternatively, the script can be run in 'singleReport(-r)' mode where it will generate a sorted list\n" + \ "of track counts over a period of time. This could be useful for reporting purposes.\n\n" + \ "Example runs:\n" + \ " trackCountsParse hg38 /hive/users/lrnassar/trackCounts/\n" + \ " trackCountsParse hg38 /hive/users/lrnassar/trackCounts/ -c .5 -n 12\n" + \ " trackCountsParse hg38 /hive/users/lrnassar/trackCounts/ -r -s 2023-01-01 -e 2023-12-31\n") exit(0) parser._action_groups.append(optional) options = parser.parse_args() if options.singleReport: if options.asOfDate: parser.error("-a/--asOfDate does not apply in -r/--singleReport mode, use -s and -e") if not options.startDate or not options.endDate: parser.error("-r/--singleReport needs both -s/--startDate and -e/--endDate") if options.asOfDate: try: datetime.strptime(options.asOfDate, '%Y-%m') except ValueError: parser.error("-a/--asOfDate must be formatted as YYYY-MM, got '"+options.asOfDate+"'") if options.numOfMonthsToCompare < 1: parser.error("-n/--numOfMonthsToCompare must be at least 1") return options def bash(cmd,mergeStderr=True): - """Run the cmd in bash subprocess. Callers that parse the output should pass - mergeStderr=False so that a warning cannot splice itself into a data line.""" + """Run the cmd in a bash subprocess. Callers that parse the output should pass + mergeStderr=False so that a warning cannot splice itself into a data line, a + failing command still reports its stderr either way. executable is pinned to bash + because the commands here use bash features, 'set -o pipefail' and ANSI-C quoting + in the sort, that plain sh does not have.""" try: - rawBashOutput = subprocess.run(cmd, check=True, shell=True,\ + rawBashOutput = subprocess.run(cmd, check=True, shell=True, executable="/bin/bash",\ stdout=subprocess.PIPE, universal_newlines=True,\ stderr=subprocess.STDOUT if mergeStderr else subprocess.PIPE) bashStdoutt = rawBashOutput.stdout except subprocess.CalledProcessError as e: - raise RuntimeError("command '{}' return with error (code {}): {}".format(e.cmd, e.returncode, e.output)) + errorText = e.output if e.output else "" + if e.stderr: + errorText += e.stderr + raise RuntimeError("command '{}' return with error (code {}): {}".format(e.cmd, e.returncode, errorText)) return(bashStdoutt) def file_exists(filepath): return os.path.isfile(filepath) #Tracks whose parent chain loops, so the warning is printed once and not once per month per track loopsAlreadyWarnedAbout = set() def buildTrackDbCache(dbs): """ Dump the whole trackDb for a db in a single tdbQuery call and return a dictionary of trackName -> {'shortLabel','group','parentName'}. Querying tdbQuery once per track takes hours on a large assembly, one dump takes under a second. A track with no 'parentName' key is top level. """ trackDbCache = {} selectedKeys = ("track","shortLabel","group","parent","superTrack") query = "tdbQuery -oneLine \"select track,shortLabel,group,parent,superTrack from "+dbs+"\"" for line in bash(query,mergeStderr=False).split("\n"): if not line.startswith("track "): continue settings = {} lastKey = None for field in line.split("|"): if field == "": continue key = field.split(" ")[0] #tdbQuery does not escape a '|' inside a value, so a shortLabel containing one #splits into extra fields. Only a field starting with one of the keys we asked #for begins a new setting, anything else is the rest of the previous value. if key in selectedKeys and key not in settings: settings[key] = field[len(key)+1:] lastKey = key elif lastKey: settings[lastKey] += "|" + field trackName = settings["track"] trackDbCache[trackName] = {'shortLabel':settings.get("shortLabel",trackName), 'group':settings.get("group","No group")} #A subtrack points at its container with 'parent', a member of a superTrack points #at it with 'superTrack <name>'. A superTrack itself carries 'superTrack on'. if "parent" in settings: trackDbCache[trackName]['parentName'] = settings["parent"].split(" ")[0] elif "superTrack" in settings and settings["superTrack"].split(" ")[0] != "on": trackDbCache[trackName]['parentName'] = settings["superTrack"].split(" ")[0] + #tdbQuery exits 0 and prints nothing for a database it does not know, so an empty cache + #means a bad db name. Without this the run carries on, treats every track as notInTrackDb, + #and only fails later with a message blaming the usage logs. + if not trackDbCache: + sys.exit("Error: tdbQuery returned no trackDb entries for '"+dbs+"'. Check the database name.") print("trackDb entries cached for "+dbs+": "+str(len(trackDbCache))) return(trackDbCache) def findTopLevelTrack(trackDbCache,trackName): """ Walk up the parent chain and return the track at the top. Tracks with no parent, and tracks that are no longer in trackDb at all, are their own top level. This is a loop rather than a fixed number of hops so that adding another layer of nesting to trackDb does not silently strand counts on a middle container. """ currentTrack = trackName visited = set() while currentTrack in trackDbCache and 'parentName' in trackDbCache[currentTrack]: if currentTrack in visited: if currentTrack not in loopsAlreadyWarnedAbout: loopsAlreadyWarnedAbout.add(currentTrack) print("WARNING: parent loop in trackDb at track "+currentTrack) break visited.add(currentTrack) currentTrack = trackDbCache[currentTrack]['parentName'] return(currentTrack) def generateTrackCounts(dbs,workDir,startDate,endDate): #Generate track usage report binned by month for a specific time frame and dbs #Format date format is XXXX-XX-XX, e.g. 2023-11-01 #The v2 in the name marks counts made with the field anchored awk filter below. Older #.trackCounts.txt files were filtered differently, and since this function reuses any #file it finds, sharing the name would silently serve the old filtering forever. outputFileName = workDir+dbs+"."+startDate+"to"+endDate+".trackCounts.v2.txt" #Run the script and remove the ct lines (ct_), hub lines (hub_), dup lines (dup_) #and header line (^#), as well as remove the first column of repeating database. #The track name is the second column, so the prefixes are matched against the start #of that field only. Matching them anywhere on the line, as this used to, also drops #real tracks whose names merely contain the string, such as dbVar_conflict_pathogenic #containing 'ct_'. This is awk rather than grep because grep -P silently failed to #drop a few dozen hub_ lines out of the 650k that makeUsageReport emits for a month, #while awk field matching is exact. The awk also does what the old 'cut -f2-' did. awkFilter = "awk -F'\\t' 'BEGIN{OFS=\"\\t\"} $1!~/^#/ && $2!~/^(hub_|ct_|dup_)/ " + \ "{out=$2; for(i=3;i<=NF;i++) out=out OFS $i; print out}'" - reportCmd = "/hive/users/chmalee/logs/byDate/makeUsageReport -t -db " + dbs + \ + #pipefail so that makeUsageReport dying partway through counts as a failure. Without it + #the pipeline reports awk's status, awk is perfectly happy with truncated input, and the + #half written file would be treated as a valid cache by every later run. + reportCmd = "set -o pipefail; /hive/users/chmalee/logs/byDate/makeUsageReport -t -db " + dbs + \ " --bin-months -s " + startDate + " -e " + endDate + \ " | " + awkFilter + " > " + outputFileName if not file_exists(outputFileName): print("Generating new file") print(reportCmd) + try: bash(reportCmd) + except RuntimeError: + #A run that died partway still leaves what it managed to write, so clear it out + #rather than leave a partial file for the next run to pick up as cached. + if file_exists(outputFileName): + os.remove(outputFileName) + raise else: print("File already generated: " + outputFileName) #An empty report means the log data does not cover these dates. Leaving the file behind #would cache that emptiness, and an empty month aborts later on an out of range index. if os.path.getsize(outputFileName) == 0: os.remove(outputFileName) sys.exit("Error: no track counts for "+startDate+" to "+endDate+". The usage logs may " + \ "not reach back that far, or the weekly files for those dates may not be built yet. " + \ "Check /hive/users/chmalee/logs/byDate/result/*/ for the dates you asked for.") return(outputFileName) def createDicFromTrackCountsFile(trackCountsFilePath): """ This function reads a file containing track counts and creates a dictionary with track names as keys and their corresponding counts as values. Args: - trackCountsFilePath (str): Path to the file containing track counts. Returns: - dict: A dictionary containing track names as keys and their counts as values. """ trackList = open(trackCountsFilePath, 'r') trackCountsDic = {} for line in trackList: parsedLine = line.rstrip().split("\t") if parsedLine[0].endswith(":"): currentTrackName = parsedLine[0][0:len(parsedLine[0])-1] else: currentTrackName = parsedLine[0] if currentTrackName not in trackCountsDic.keys(): trackCountsDic[currentTrackName] = {} trackCountsDic[currentTrackName]["Count"] = parsedLine[1] else: trackCountsDic[currentTrackName]["Count"] = str(int(trackCountsDic[currentTrackName]["Count"]) + int(parsedLine[1])) trackList.close() totalCount = len(trackCountsDic.keys()) print("Total tracks to parse: "+str(totalCount)) return(trackCountsDic,totalCount) def buildFinalDicWithOnlyTopLevelTrackCounts(trackCountsDic,trackDbCache): """ Roll every track in the log counts up to its top level container and return a dictionary of topLevelTrack -> {'max','sum'}. 'max' is the busiest single member and is the metric the cutoff has always used. 'sum' adds every member together. A wide composite nearly always has one busy child, so 'max' flatters it relative to a standalone track, while 'sum' counts a single page view once per displayed subtrack. Neither is right on its own, so both are carried through to the report. """ finalDicOfTopLevelTracksAndCounts = {} for trackName in trackCountsDic.keys(): topLevelTrack = findTopLevelTrack(trackDbCache,trackName) count = int(trackCountsDic[trackName]['Count']) if topLevelTrack not in finalDicOfTopLevelTracksAndCounts.keys(): finalDicOfTopLevelTracksAndCounts[topLevelTrack] = {'max':count,'sum':count} else: if count > finalDicOfTopLevelTracksAndCounts[topLevelTrack]['max']: finalDicOfTopLevelTracksAndCounts[topLevelTrack]['max'] = count finalDicOfTopLevelTracksAndCounts[topLevelTrack]['sum'] += count return(finalDicOfTopLevelTracksAndCounts) def makeFinalFileOnTopLevelTrackCounts(finalDicOfTopLevelTracksAndCounts,pathUrl,trackDbCache): """ This function creates a final output file containing details of top-level tracks, including their short labels (if available) and counts. The file is saved at the specified path URL. Args: - finalDicOfTopLevelTracksAndCounts (dict): A dictionary containing top-level track names and their respective counts. - pathUrl (str): The path where the final output file will be saved. """ outputFile = open(pathUrl, 'w') n=0 for key in finalDicOfTopLevelTracksAndCounts.keys(): if key != "": n+=1 shortLabel = trackDbCache.get(key,{}).get('shortLabel',key) outputFile.write(key+"\t"+shortLabel+"\t"+str(finalDicOfTopLevelTracksAndCounts[key]['max'])+"\n") print("Final file completed. Total number of tracks: "+str(n)) outputFile.close() #Order and sort final file bash("sort -t $'\t' -k3 -rn "+pathUrl+" > "+pathUrl+".sorted") print("Final sorted file: "+pathUrl+".sorted") def get_count(dicField): return dicField['trackCounts'] def makeOrderedDicForSpecificTime(finalDicOfTopLevelTracksAndCounts,trackDbCache): listOfCountsToSort = [] n=0 #Attach the shortLabels and make a list where each entry is a dic with trackName, shortLabel, and trackCount for key in finalDicOfTopLevelTracksAndCounts.keys(): if key != "": n+=1 shortLabel = trackDbCache.get(key,{}).get('shortLabel',key) listOfCountsToSort.append({'trackName':key,'shortLabel':shortLabel, 'trackCounts':finalDicOfTopLevelTracksAndCounts[key]['max'], 'summedCounts':finalDicOfTopLevelTracksAndCounts[key]['sum']}) print("Total number of tracks: "+str(n)) #Sort the trackCounts list to return in order to find data that meets cutoff threshhold listOfCountsToSort.sort(key=get_count, reverse=True) return(listOfCountsToSort) def refineTrackCountsBasedOnCutOff(listOfTracks,cutOffThreshhold,period): """ Take an ordered list containing dics of track counts and filter it based on a set threshhold. Then return a new ordered dictionary that contains the tracks below the threshhold with trackNames as keys and shortLabel and counts as values. The cutoff itself is returned as well so it can be recorded, the median moves a lot from year to year as the track list grows. """ trackCountCutoff = listOfTracks[int(len(listOfTracks)/2)]['trackCounts']*cutOffThreshhold finalTrackCountsDic = OrderedDict() for track in listOfTracks: if track['trackCounts'] < trackCountCutoff: finalTrackCountsDic[track['trackName']]={'shortLabel':track['shortLabel'],'trackCounts':track['trackCounts'],'countComparedToMaxForPeriod':track['trackCounts']/listOfTracks[0]["trackCounts"],'countComparedToCutoff':track['trackCounts']/trackCountCutoff} print("The trackCount cutoff for "+period+" is: "+str(trackCountCutoff)) return(finalTrackCountsDic,trackCountCutoff) def getDateRangesForComparison(numOfMonthsToCompare,asOfDate): """ Based on a number of months to compare given, find the year + month combination followed by the last day of each month to be used in the log query script. Return an ordered dictionary with the date ranges as keys, which will be used as the titles of the respective final ouputs, and the start/end dates as the content. **Note** This subtracts an additional month from the latest month in order to ensure that the logs chosen are complete. """ dateRanges = OrderedDict() if asOfDate: date = datetime.strptime(asOfDate, '%Y-%m').strftime('%Y-%m') else: date = datetime.today().strftime('%Y-%m') for number in range(numOfMonthsToCompare): monthToParse = datetime.strftime(datetime.strptime(date, '%Y-%m') - relativedelta(months=number+1), '%Y-%m') year = monthToParse.split('-')[0] month = monthToParse.split('-')[1] lastDateOfMonth = calendar.monthrange(int(year), int(month))[1] startDate = year+"-"+month+"-01" endDate = year+"-"+month+"-"+str(lastDateOfMonth) dateRanges[startDate+"-"+endDate] = {'startDate':startDate,'endDate':endDate} return(dateRanges) def createFinalListOfTracksThatMeetCutoffEveryMonth(finalDicWithCutOffDics,numOfMonthsToCompare): """ Iterates through all of the monthly dictionaries and creates a final list where only tracks present in every period are present. This is to filter out monthly outliers. It then reports which ones were filtered, if any. Due to data being weird, if we are checking at least 4 months allow for outliers of floor(n/4) """ initialTrackList = {} listOfTracksInEligiblePeriods = [] listOfTracksFilteredOut = [] #Create initial list of all tracks present in these periods for period in finalDicWithCutOffDics.keys(): for track in finalDicWithCutOffDics[period].keys(): if track not in initialTrackList: initialTrackList[track] = 0 #Due to data being weird, if we are checking at least 4 months allow #for outliers of floor(n/4) outlierMonthsExcemption = math.floor(numOfMonthsToCompare/4) #Go over the list and add a penalty of 1 for every period in which #the track is missing, then create a final list of tracks that #pass through the filter for period in finalDicWithCutOffDics.keys(): for track in initialTrackList.keys(): if track not in finalDicWithCutOffDics[period].keys(): initialTrackList[track] = initialTrackList[track] + 1 for track in initialTrackList.keys(): if initialTrackList[track] <= outlierMonthsExcemption: listOfTracksInEligiblePeriods.append(track) else: listOfTracksFilteredOut.append(track) if listOfTracksFilteredOut != []: print("The following tracks were filtered out because they did not meet") print("the cutoff in all of the months specified:\n") for track in listOfTracksFilteredOut: print(track) return(listOfTracksInEligiblePeriods) def get_avCount(dicField): """Helper function to help sort dict""" return dicField['averageTrackCount'] def constructSortedFinalTrackDicWithAllData(finalDicWithCutOffDics,allPeriodTopLevelCounts,listOfTracksInEligiblePeriods,numOfMonthsToCompare,trackDbCache): """ Takes in a final list of track names which have met all the conditions for potential archiving and constructs a final dictionary with all of the data sorted. This includes averages over the time period for the track counts as well as the comparison to median/max. The track group is also reported for use in the ultimate decision. The count averages cover every month searched, using the full monthly counts rather than only the months in which the track fell below the cutoff. Averaging only the qualifying months hides the one good month a track was granted an exemption for, which is the month a reader most needs to see. The two ratio averages still cover only the qualifying months, since a track above the cutoff has no ratio recorded that month, so monthsBelowCutoff is reported next to them. """ listOfTracksToReport = [] for track in listOfTracksInEligiblePeriods: #A name in the logs with no trackDb entry is either a track retired since, or noise #from a malformed request. Either way it is not something anyone can act on, so say so. if track in trackDbCache: shortLabel = trackDbCache[track]['shortLabel'] group = trackDbCache[track]['group'] else: shortLabel = track group = "notInTrackDb" addedTrackCounts = 0 addedSummedCounts = 0 addedCountComparedToMaxForPeriod = 0 addedCountComparedToCutoff = 0 monthsBelowCutoff = 0 for period in finalDicWithCutOffDics.keys(): #Full counts come from every month, whether or not the track was under the cutoff if track in allPeriodTopLevelCounts[period].keys(): addedTrackCounts+=allPeriodTopLevelCounts[period][track]['max'] addedSummedCounts+=allPeriodTopLevelCounts[period][track]['sum'] if track in finalDicWithCutOffDics[period].keys(): monthsBelowCutoff+=1 addedCountComparedToMaxForPeriod+=finalDicWithCutOffDics[period][track]['countComparedToMaxForPeriod'] addedCountComparedToCutoff+=finalDicWithCutOffDics[period][track]['countComparedToCutoff'] averageTrackCount = round(addedTrackCounts/numOfMonthsToCompare,2) averageSummedCount = round(addedSummedCounts/numOfMonthsToCompare,2) averageCountComparedToMaxForPeriod = round(addedCountComparedToMaxForPeriod/monthsBelowCutoff,4) averageCountComparedToCutoff = round(addedCountComparedToCutoff/monthsBelowCutoff,2) listOfTracksToReport.append({'trackName':track,'shortLabel':shortLabel,'group':group,'averageTrackCount':averageTrackCount,'averageSummedCount':averageSummedCount,'averageCountComparedToMaxForPeriod':averageCountComparedToMaxForPeriod,'averageCountComparedToCutoff':averageCountComparedToCutoff,'monthsBelowCutoff':monthsBelowCutoff}) listOfTracksToReport.sort(key=get_avCount, reverse=True) return(listOfTracksToReport) def writeFinalTrackListToFile(finalOutputTrackDicToReport,workDir,dbs,cutOffThreshhold,numOfMonthsToCompare,dateRanges,cutoffsPerPeriod): """ Take the final processed dictionary and write it out to a tsv file including the vars used in the data generation. The file is named for the months actually searched. """ #dateRanges runs newest month first, so the oldest month searched is the last entry newestMonth = dateRanges[next(iter(dateRanges))]['startDate'][0:7] oldestMonth = dateRanges[next(reversed(dateRanges))]['startDate'][0:7] fileNamePathStartToEndDate = workDir+oldestMonth+"-"+newestMonth+"."+dbs+".tracksToArchive.tsv" finalOutputFile = open(fileNamePathStartToEndDate,'w') finalOutputFile.write("#Variables used in this file generation: dbs="+dbs+" numOfMonthsToCompare="+str(numOfMonthsToCompare)+" cutOffThreshhold="+str(cutOffThreshhold)+"\n") finalOutputFile.write("#averageTrackCount averages every month searched. Reports made before " + \ "2026-07 averaged only the months a track was under the cutoff, so the two are not directly comparable.\n") finalOutputFile.write("#Monthly cutoffs, that month's median top level track count times the threshhold: "+ \ ", ".join([period+"="+str(round(cutoffsPerPeriod[period],1)) for period in reversed(cutoffsPerPeriod)])+"\n") finalOutputFile.write("#trackName\tshortLabel\tgroup\taverageTrackCount\taverageSummedCount\taverageCountComparedToMaxForPeriod\taverageCountComparedToCutoff\tmonthsBelowCutoff\n") for track in finalOutputTrackDicToReport: finalOutputFile.write(track['trackName']+"\t"+track['shortLabel']+"\t"+track['group']+"\t"+str(track['averageTrackCount'])+"\t"+str(track['averageSummedCount'])+"\t"+str(track['averageCountComparedToMaxForPeriod'])+"\t"+str(track['averageCountComparedToCutoff'])+"\t"+str(track['monthsBelowCutoff'])+"\n") finalOutputFile.close() print("\nCutoff tracks file complete: "+fileNamePathStartToEndDate) print("\nYou nicely format the output as such: tail -n +4 outputFilePath.tsv | tabFmt stdin") def main(): """Initialize options and call other functions""" options = parseArgs() dbs,workDir,cutOffThreshhold,numOfMonthsToCompare,singleReport = options.dbs,options.workDir,\ options.cutOffThreshhold,options.numOfMonthsToCompare,options.singleReport trackDbCache = buildTrackDbCache(dbs) if singleReport == True: startDate,endDate = options.startDate,options.endDate logFile = generateTrackCounts(dbs,workDir,startDate,endDate) trackCountsDic,totalCount = createDicFromTrackCountsFile(logFile) finalDicOfTopLevelTracksAndCounts = buildFinalDicWithOnlyTopLevelTrackCounts(trackCountsDic,trackDbCache) makeFinalFileOnTopLevelTrackCounts(finalDicOfTopLevelTracksAndCounts,workDir+"trackCounts.tsv",trackDbCache) else: print("Script started: "+strftime("%Y-%m-%d %H:%M:%S", localtime())) dateRanges = getDateRangesForComparison(numOfMonthsToCompare,options.asOfDate) finalDicWithCutOffDics,allPeriodTopLevelCounts,cutoffsPerPeriod = OrderedDict(),{},OrderedDict() for period in dateRanges: logFile = generateTrackCounts(dbs,workDir,dateRanges[period]['startDate'],dateRanges[period]['endDate']) trackCountsDic,totalCount = createDicFromTrackCountsFile(logFile) finalDicOfTopLevelTracksAndCounts = buildFinalDicWithOnlyTopLevelTrackCounts(trackCountsDic,trackDbCache) allPeriodTopLevelCounts[period] = finalDicOfTopLevelTracksAndCounts listOfTracks = makeOrderedDicForSpecificTime(finalDicOfTopLevelTracksAndCounts,trackDbCache) finalDicWithCutOffDics[period],cutoffsPerPeriod[period] = refineTrackCountsBasedOnCutOff(listOfTracks,cutOffThreshhold,period) print("The number of tracks that met the criteria for "+period+" is: "+str(len(finalDicWithCutOffDics[period].keys()))) listOfTracksInEligiblePeriods = createFinalListOfTracksThatMeetCutoffEveryMonth(finalDicWithCutOffDics,numOfMonthsToCompare) finalOutputTrackDicToReport = constructSortedFinalTrackDicWithAllData(finalDicWithCutOffDics,allPeriodTopLevelCounts,listOfTracksInEligiblePeriods,numOfMonthsToCompare,trackDbCache) writeFinalTrackListToFile(finalOutputTrackDicToReport,workDir,dbs,cutOffThreshhold,numOfMonthsToCompare,dateRanges,cutoffsPerPeriod) print("Script finished: "+strftime("%Y-%m-%d %H:%M:%S", localtime())) main()