)? # second date
\s*(?: )?
\s* )? # second date
\s*(?: )?
\s* ", "", ltriptext).strip()
entrytuple = (ldate, tripcave, triptitle, ltriptext,
trippeople, expedition, tu, "html", tripid1)
logentries.append(entrytuple)
# EnterLogIntoDbase(date = ldate, place = tripcave, title = triptitle, text = ltriptext,
# trippeople=trippeople, expedition=expedition, logtime_underground=0,
# entry_type="html", tid=tid)
# EnterLogIntoObjStore(year, ldate, tripcave, triptitle, ltriptext, trippeople, tu,
# "html", tripid1, logbook_entry_count, tid=tid)
# main parser for 1991 - 2001. simpler because the data has been hacked so much to fit it
# trying it out for years 1982 - 1990 too. Some logbook editing required by hand.. place
def Parseloghtml01(year, expedition, txt):
global logentries
global logdataissues
errorcount = 0
tripparas = re.findall(r" | ", "\n\n", ltriptext).strip()
ltriptext = re.sub(r"[^\s0-9a-zA-Z\-.,:;'!&()\[\]<>?=+*%]", "_NONASCII_", ltriptext)
entrytuple = (ldate, tripcave, triptitle, ltriptext,
trippeople, expedition, tu, "html03", tid)
logentries.append(entrytuple)
# EnterLogIntoDbase(date = ldate, place = tripcave, title = triptitle,
# text = ltriptext, trippeople=trippeople, expedition=expedition,
# logtime_underground=0, entry_type="html", tid=tid)
# EnterLogIntoObjStore(year, ldate, tripcave, triptitle, ltriptext, trippeople, tu,
# "html03", tid, logbook_entry_count, tid=tid)
def SetDatesFromLogbookEntries(expedition):
"""
Sets the date_from and date_to field for an expedition based on persontrips.
Then sets the expedition date_from and date_to based on the personexpeditions.
"""
# Probably a faster way to do this. This uses a lot of db queries, but we have all this
# in memory..
for personexpedition in expedition.personexpedition_set.all():
persontrips = personexpedition.persontrip_set.order_by('logbook_entry__date')
# sequencing is difficult to do
lprevpersontrip = None
for persontrip in persontrips:
persontrip.persontrip_prev = lprevpersontrip
if lprevpersontrip:
lprevpersontrip.persontrip_next = persontrip
lprevpersontrip.save()
persontrip.persontrip_next = None
lprevpersontrip = persontrip
#persontrip.save() # also saved in EnterLogIntoDbase. MERGE these to speed up import.
def LoadLogbookForExpedition(expedition):
""" Parses all logbook entries for one expedition
If a cache is found it uses it. If not found, or fails sanity checks, parses source file.
"""
# absolutely horrid. REFACTOR THIS (all my fault..)
global logentries
global logdataissues
global entries
logbook_parseable = False
logbook_cached = False
yearlinks = settings.LOGBOOK_PARSER_SETTINGS
expologbase = os.path.join(settings.EXPOWEB, "years")
logentries=[]
year = expedition.year
expect = entries[year]
# print(" - Logbook for: " + year)
def validcache(year,n, lbsize):
if year != expedition:
print(" ! cache loading: year != expedition ",year, expedition )
return False
currentsize = logbookpath.stat().st_size
if lbsize != currentsize:
print(f" ! cache loading: Logbook size {lbsize} != {currentsize} ")
return False
if len(logentries) != n:
print(" ! cache loading: len(logentries) != n ",len(logentries), n )
return False
if n != expect:
print(" ! cache loading: n != expect ",n, expect )
return False
return True
def cleanerrors(year):
global logdataissues
dataissues = DataIssue.objects.filter(parser='logbooks')
for di in dataissues:
ph = year
if re.search(ph, di.message) is not None:
#print(f' - CLEANING dataissue {di.message}')
di.delete()
#print(f' - CLEAN {year} {len(logdataissues)} {type(logdataissues)} data issues for this year')
dellist = []
for key, value in logdataissues.items():
#print(f' - CLEANING logdataissues [{key}]: {value}')
if key.startswith(year):
#print(f' - CLEANING logdataissues [{key:12}]: {value} ')
dellist.append(key)
for i in dellist:
del logdataissues[i]
cleanerrors(year)
if year in yearlinks:
logbookpath = Path(expologbase) / year / yearlinks[year][0]
expedition.logbookfile = yearlinks[year][0]
parsefunc = yearlinks[year][1]
else:
logbookpath = Path(expologbase) / year / settings.DEFAULT_LOGBOOK_FILE
expedition.logbookfile = settings.DEFAULT_LOGBOOK_FILE
parsefunc = settings.DEFAULT_LOGBOOK_PARSER
cache_filename = Path(str(logbookpath) + ".cache")
if not cache_filename.is_file():
print(" - Cache file does not exist \"" + str(cache_filename) +"\"")
expedition.save()
logbook_cached = False
if True: # enable cache system
now = time.time()
bad_cache = False # temporarily disable reading the cache - buggy
try:
cache_t = os.path.getmtime(cache_filename)
if os.path.getmtime(logbookpath) - cache_t > 2: # at least 2 secs later
print(" - ! Cache is older than the logbook file")
bad_cache= True
if now - cache_t > 30*24*60*60:
print(" - ! Cache is > 30 days old")
bad_cache= True
if bad_cache:
print(" - so cache is either stale or more than 30 days old. Deleting it.")
os.remove(cache_filename)
logentries=[]
print(" - Deleted stale or corrupt cache file")
raise
try:
# print(" - Reading cache: " + str(cache_filename), end='')
with open(cache_filename, "rb") as f:
year, lbsize, n, logentries = pickle.load(f)
if validcache(year, n, lbsize):
print(f" -- {year} : Loaded {len(logentries)} log entries")
logbook_cached = True
else:
print(" !- {year} : Cache failed validity checks")
raise
except:
print(" ! Failed to load corrupt cache (or I was told to ignore it). Deleting it.")
os.remove(cache_filename)
logentries=[]
raise
except :
print(" - Cache old or de-pickle failure \"" + str(cache_filename) +"\"")
try:
file_in = open(logbookpath,'rb')
txt = file_in.read().decode("utf-8")
file_in.close()
logbook_parseable = True
except (IOError):
logbook_parseable = False
print(" ! Couldn't open logbook as UTF-8 " + logbookpath)
if logbook_parseable:
parser = globals()[parsefunc]
print(f' - Using parser {parsefunc}')
parser(year, expedition, txt) # this launches the right parser for this year
print(" - Setting dates from logbook entries")
SetDatesFromLogbookEntries(expedition)
if len(logentries) >0:
print(" - Cacheing " , len(logentries), " log entries")
lbsize = logbookpath.stat().st_size
with open(cache_filename, "wb") as fc: # we much check that permission are g+w ! or expo can't delete the cache
logbk=(expedition,lbsize,len(logentries),logentries)
pickle.dump(logbk, fc, protocol=4)
else:
print(" ! NO TRIP entries found in logbook, check the syntax.")
i=0
for entrytuple in logentries:
try:
date, tripcave, triptitle, text, trippeople, expedition, logtime_underground, tripid1 = entrytuple
except ValueError: # cope with removal of entry_type but still in cache files. Remove in Sept. 2022.
date, tripcave, triptitle, text, trippeople, expedition, logtime_underground, entry_type, tripid1 = entrytuple
EnterLogIntoDbase(date, tripcave, triptitle, text, trippeople, expedition, 0,
tripid1)
EnterLogIntoObjStore(year, date, tripcave, triptitle, text, trippeople, logtime_underground,
tripid1, i)
i +=1
SetDatesFromLogbookEntries(expedition)
if len(logentries) == expect:
# print(f"OK {year} {len(logentries):5d} is {expect}\n")
pass
else:
print(f"BAD {year} {len(logentries):5d} is not {expect}\n")
return len(logentries)
def LoadLogbooks():
""" This is the master function for parsing all logbooks into the Troggle database.
Parser settings appropriate for each year are set in settings.py LOGBOOK_PARSER_SETTINGS.
This should be rewritten to use coroutines to load all logbooks from disc in parallel.
"""
global logdataissues
global entries
logdataissues = {}
DataIssue.objects.filter(parser='logbooks').delete()
expos = Expedition.objects.all()
if len(expos) <= 1:
message = f" ! - No expeditions found. Load 'people' first"
DataIssue.objects.create(parser='logbooks', message=message)
logdataissues[f"sqlfail 0000"]=message
print(message)
noexpo = ["1986", "2020", "2021",] #no expo
lostlogbook = ["1976", "1977", "1978", "1979", "1980", "1981"]
sqlfail = ["1987", "1988", "1989"] # breaks mysql with db constraint fail - debug locally first]
nologbook = noexpo + lostlogbook + sqlfail
nlbe={}
expd ={}
actuals = []
for expo in expos: # pointless as we explicitly know the years in this code.
year = expo.year
TROG['pagecache']['expedition'][year] = None # clear cache
if year in sqlfail:
print(" - Logbook for: " + year + " NO parsing attempted - known sql failures")
message = f" ! - Not even attempting to parse logbook for {year} until code fixed"
DataIssue.objects.create(parser='logbooks', message=message)
logdataissues[f"sqlfail {year}"]=message
print(message)
if year not in nologbook:
if year in entries:
actuals.append(expo)
else:
print(" - No Logbook yet for: " + year) # catch case when preparing for next expo
for ex in actuals:
nlbe[ex] = LoadLogbookForExpedition(ex) # this actually loads the logbook for one expo
# tried to use map with concurrent threads - but sqlite database is not concurrent, so failed with database lock
# yt = 0
# for r in map(LoadLogbookForExpedition, actuals):
# yt = r
yt = 0
for e in nlbe:
yt += nlbe[e]
print(f"total {yt:,} log entries parsed in all expeditions")
if yt != len(trips):
print(f"** total trips in ObjStore:{len(trips):,}")
try:
shelvfilenm = 'logbktrips.shelve' # ".db" automatically apended after python 3.8
with shelve.open(shelvfilenm, writeback=True) as odb:
for lbe in trips:
odb[lbe]=trips[lbe]
odb.sync()
odb.close()
except:
message = f" ! - Failed store cached logbooks in '{shelvfilenm}.db' - Delete old file and try again"
DataIssue.objects.create(parser='logbooks', message=message)
logdataissues["Shelve Fail"]=message
print(message)
# dateRegex = re.compile(r'
", "\n\n", ltriptext).strip()
ltriptext = re.sub(r"", "_", ltriptext)
ltriptext = re.sub(r"", "''", ltriptext)
ltriptext = re.sub(r"", "'''", ltriptext)
if ltriptext == "":
message = " ! - Zero content for logbook entry!: " + tid
DataIssue.objects.create(parser='logbooks', message=message)
logdataissues[tid]=message
print(message)
entrytuple = (ldate, tripcave, triptitle, ltriptext,
trippeople, expedition, tu, "html01", tid)
logentries.append(entrytuple)
# try:
# EnterLogIntoDbase(date=ldate, place=tripcave, title=triptitle, text=ltriptext,
# trippeople=trippeople, expedition=expedition, logtime_underground=0,
# entry_type="html", tid=tid)
# except:
# message = " ! - Enter log entry into database FAIL exception in: " + tid
# DataIssue.objects.create(parser='logbooks', message=message)
# logdataissues[tid]=message
# print(message)
# try:
# EnterLogIntoObjStore(year, ldate, tripcave, triptitle, ltriptext, trippeople, tu,
# "html01", tid, logbook_entry_count, tid=tid)
# except:
# message = " ! - Enter log entry into ObjectStore FAIL exception in: " + tid
# DataIssue.objects.create(parser='logbooks', message=message)
# logdataissues[tid]=message
# print(message)
except:
message = f" ! - Skipping logentry {year} due to exception in: {tid}"
DataIssue.objects.create(parser='logbooks', message=message)
logdataissues[tid]=message
print(message)
errorcount += 1
if errorcount >5 :
message = f" !!- TOO MANY ERRORS - aborting at '{tid}' logbook: {year}"
DataIssue.objects.create(parser='logbooks', message=message)
logdataissues[tid]=message
print(message)
return
# parser for 2003
def Parseloghtml03(year, expedition, txt):
global logentries
global logdataissues
tripparas = re.findall(r"(.*?)
', re.S)
# reportRegex = re.compile(r'