def doDailyClubs(infile, conn, cdate, firsttime=False):
""" infile is a file-like object """
global changecount
from datetime import datetime, timedelta
curs = conn.cursor()
reader = csv.reader(infile)
hline = next(reader)
headers = cleanheaders(hline)
try:
clubcol = headers.index("clubnumber")
except ValueError:
if not hline[0].startswith('{"Message"'):
print("'clubnumber' not in '%s'" % hline)
return
try:
prospectiveclubcol = headers.index("prospectiveclub")
except ValueError:
prospectiveclubcol = False
# Find out what fields we have in the database itself
dbfields = []
curs.execute("describe clubs")
for l in curs.fetchall():
dbfields.append(l[0])
inform("clubs for", cdate, suppress=1)
dbheaders = [p for p in headers]
# Convert between Toastmasters' names for address and location and ours; they've changed it a few times. *sigh*
if "address1" in dbheaders:
addrcol1 = dbheaders.index("address1")
else:
addrcol1 = dbheaders.index("location")
if "address2" in dbheaders:
addrcol2 = dbheaders.index("address2")
else:
addrcol2 = dbheaders.index("address")
dbheaders[addrcol1] = "place"
dbheaders[addrcol2] = "address"
expectedheaderscount = len(dbheaders)
dbheaders.append("firstdate")
dbheaders.append("lastdate") # For now...
areacol = dbheaders.index("area")
divisioncol = dbheaders.index("division")
statecol = dbheaders.index("state")
# Now, suppress anything in the file that's not in the database:
suppress = []
oldheaders = dbheaders
dbheaders = []
for i in range(len(oldheaders)):
if oldheaders[i] in dbfields:
dbheaders.append(oldheaders[i])
else:
suppress.append(i)
suppress.reverse() # We remove these columns from the input
Club.setfieldnames(dbheaders)
# We need to get clubs for the most recent update so we know whether to update an entry
# or start a new one.
yesterday = datetime.strftime(
datetime.strptime(cdate, "%Y-%m-%d") - timedelta(1), "%Y-%m-%d")
clubhist = Club.getClubsOn(curs, date=yesterday)
for row in reader:
if len(row) < expectedheaderscount:
break # we're finished
if prospectiveclubcol is not None and row[prospectiveclubcol]:
continue # Ignore prospective clubs
for i in suppress:
del row[i]
if len(row) > expectedheaderscount:
# Special case...Millbrae somehow snuck two club websites in!
row[16] = row[16] + "," + row[17]
del row[17]
# print row[addrcol1]
# print row[addrcol2]
# Now, clean up the address:
# Address line 1 is "place" information and can be multiple lines.
# Address line 2 is the real address and should be treated as one line, with spaces normalized.
place = "\n".join([x.strip() for x in row[addrcol1].split(" ")])
row[addrcol1] = place
address = normalize(row[addrcol2])
row[addrcol2] = address
# Toastmasters is currently reversing the "Area" and "Division" items. "Area" should be a
# number; if not, swap the two.
try:
thearea = row[areacol]
thedivision = row[divisioncol]
areanum = int(row[areacol])
except ValueError:
row[areacol] = thedivision
row[divisioncol] = thearea
# Collapse state names into their abbreviations
if row[statecol] in statelist:
row[statecol] = statelist[row[statecol]]
# Get the right number of items into the row by setting today as the
# tentative first and last date
row.append(cdate)
row.append(cdate)
# And create the object
club = Club(row)
# Now, clean up things coming from Toastmasters
if club.clubstatus.startswith("Open") or club.clubstatus.startswith(
"None"):
club.clubstatus = "Open"
else:
club.clubstatus = "Restricted"
# Clean up the club and district numbers and the area
club.clubnumber = cleanitem(club.clubnumber)
club.district = cleanitem(club.district)
club.area = cleanitem(club.area)
# If a club is partially unassigned, mark it as completely unassigned.
if ((club.area == "0A") or (club.area == "0D")
or (club.division == "0D") or (club.division == "0A")):
club.area = "0A"
club.division = "0D"
# Clean up the charter date
if not club.charterdate.strip():
continue # This is a prospective club that Toastmasters didn't mark properly; ignore it.
club.charterdate = cleandate(club.charterdate)
# Clean up advanced status
club.advanced = "1" if (club.advanced != "") else "0"
# Clean up online status
club.allowsonlineattendance = ("1" if
(club.allowsonlineattendance != "") else
"0")
# Add missing schemes to any URLs
club.fixURLSchemes()
# Now, take care of missing latitude/longitude
if ("latitude") in dbheaders:
try:
club.latitude = float(club.latitude)
except ValueError:
club.latitude = 0.0
else:
club.latitude = 0.0
if ("longitude") in dbheaders:
try:
club.longitude = float(club.longitude)
except ValueError:
club.longitude = 0.0
else:
club.longitude = 0.0
# Sometimes, Toastmasters gets the latitude and longitude backwards
# If that turns out to create an impossible location (which it will in California),
# let's swap them.
if abs(club.latitude) > 90.0:
(club.latitude, club.longitude) = (club.longitude, club.latitude)
# And put it into the database if need be
if club.clubnumber in clubhist:
changes = different(club, clubhist[club.clubnumber],
dbheaders[:-2])
else:
changes = []
if club.clubnumber not in clubhist and not firsttime:
# This is a new (or reinstated) club; note it in the changes database.
curs.execute(
'INSERT IGNORE INTO clubchanges (clubnumber, changedate, item, old, new) VALUES (%s, %s, "New Club", "", "")',
(club.clubnumber, cdate),
)
if club.clubnumber not in clubhist or changes:
club.firstdate = club.lastdate
# Encode newlines in the place as double-semicolons for the database
club.place = club.place.replace("\n", ";;")
values = [club.__dict__[x] for x in dbheaders]
# And then put the place back into normal form
club.place = club.place.replace(";;", "\n")
thestr = ("INSERT IGNORE INTO clubs (" + ",".join(dbheaders) +
") VALUES (" + ",".join(["%s"
for each in values]) + ");")
try:
changecount += curs.execute(thestr, values)
except Exception as e:
print(e)
# Capture changes
for (item, old, new) in changes:
if item == "place":
# Clean up the place (old and new) for the database
old = old.replace("\n", ";;")
new = new.replace("\n", ";;")
try:
curs.execute(
"INSERT IGNORE INTO clubchanges (clubnumber, changedate, item, old, new) VALUES (%s, %s, %s, %s, %s)",
(club.clubnumber, cdate, item, old, new),
)
except Exception as e:
print(e)
clubhist[club.clubnumber] = club
if different(club, clubhist[club.clubnumber], dbheaders[:-2]):
print("it's different after being set.")
sys.exit(3)
else:
# update the lastdate
changecount += curs.execute(
"UPDATE clubs SET lastdate = %s WHERE clubnumber = %s AND lastdate = %s;",
(cdate, club.clubnumber, clubhist[club.clubnumber].lastdate),
)
# If all the files were processed, today's work is done.
curs.execute(
'INSERT IGNORE INTO loaded (tablename, loadedfor) VALUES ("clubs", %s)',
(cdate, ),
)
def doDailyClubs(infile, conn, cdate, firsttime=False):
""" infile is a file-like object """
global changecount
from datetime import datetime, timedelta
curs = conn.cursor()
reader = csv.reader(infile)
hline = reader.next()
headers = cleanheaders(hline)
try:
clubcol = headers.index('clubnumber')
except ValueError:
if not hline[0].startswith('{"Message"'):
print "'clubnumber' not in '%s'" % hline
return
try:
prospectiveclubcol = headers.index('prospectiveclub')
except ValueError:
prospectiveclubcol = False
# Find out what fields we have in the database itself
dbfields = []
curs.execute("describe clubs")
for l in curs.fetchall():
dbfields.append(l[0])
inform("clubs for", cdate, suppress=1)
dbheaders = [p for p in headers]
# Convert between Toastmasters' names for address and location and ours; they've changed it a few times. *sigh*
if 'address1' in dbheaders:
addrcol1 = dbheaders.index('address1')
else:
addrcol1 = dbheaders.index('location')
if 'address2' in dbheaders:
addrcol2 = dbheaders.index('address2')
else:
addrcol2 = dbheaders.index('address')
dbheaders[addrcol1] = 'place'
dbheaders[addrcol2] = 'address'
expectedheaderscount = len(dbheaders)
dbheaders.append('firstdate')
dbheaders.append('lastdate') # For now...
areacol = dbheaders.index('area')
divisioncol = dbheaders.index('division')
statecol = dbheaders.index('state')
# Now, suppress anything in the file that's not in the database:
suppress = []
oldheaders = dbheaders
dbheaders = []
for i in xrange(len(oldheaders)):
if oldheaders[i] in dbfields:
dbheaders.append(oldheaders[i])
else:
suppress.append(i)
suppress.reverse() # We remove these columns from the input
Club.setfieldnames(dbheaders)
# We need to get clubs for the most recent update so we know whether to update an entry
# or start a new one.
yesterday = datetime.strftime(datetime.strptime(cdate, '%Y-%m-%d') - timedelta(1),'%Y-%m-%d')
clubhist = Club.getClubsOn(curs, date=yesterday)
for row in reader:
if len(row) < expectedheaderscount:
break # we're finished
if prospectiveclubcol is not None and row[prospectiveclubcol]:
continue # Ignore prospective clubs
for i in suppress:
del row[i]
# Convert to unicode. Toastmasters usually uses UTF-8 but occasionally uses Windows CP1252 on the wire.
try:
row = [unicode(t.strip(), "utf8") for t in row]
except UnicodeDecodeError:
row = [unicode(t.strip(), "CP1252") for t in row]
if len(row) > expectedheaderscount:
# Special case...Millbrae somehow snuck two club websites in!
row[16] = row[16] + ',' + row[17]
del row[17]
#print row[addrcol1]
#print row[addrcol2]
# Now, clean up the address:
# Address line 1 is "place" information and can be multiple lines.
# Address line 2 is the real address and should be treated as one line, with spaces normalized.
place = '\n'.join([x.strip() for x in row[addrcol1].split(' ')])
row[addrcol1] = place
address = normalize(row[addrcol2])
row[addrcol2] = address
# Toastmasters is currently reversing the "Area" and "Division" items. "Area" should be a
# number; if not, swap the two.
try:
thearea = row[areacol]
thedivision = row[divisioncol]
areanum = int(row[areacol])
except ValueError:
row[areacol] = thedivision
row[divisioncol] = thearea
# Collapse state names into their abbreviations
if row[statecol] in statelist:
row[statecol] = statelist[row[statecol]]
# Get the right number of items into the row by setting today as the
# tentative first and last date
row.append(cdate)
row.append(cdate)
# And create the object
club = Club(row)
# Now, clean up things coming from Toastmasters
if club.clubstatus.startswith('Open') or club.clubstatus.startswith('None'):
club.clubstatus = 'Open'
else:
club.clubstatus = 'Restricted'
# Clean up the club and district numbers and the area
club.clubnumber = cleanitem(club.clubnumber)
club.district = cleanitem(club.district)
club.area = cleanitem(club.area)
# If a club is partially unassigned, mark it as completely unassigned.
if (club.area == '0A') or (club.area == '0D') or (club.division == '0D') or (club.division == '0A'):
club.area = '0A'
club.division = '0D'
# Clean up the charter date
club.charterdate = cleandate(club.charterdate)
# Clean up advanced status
club.advanced = '1' if (club.advanced != '') else '0'
# Clean up online status
club.allowsonlineattendance = '1' if (club.allowsonlineattendance != '') else '0'
# Now, take care of missing latitude/longitude
if ('latitude') in dbheaders:
try:
club.latitude = float(club.latitude)
except ValueError:
club.latitude = 0.0
else:
club.latitude = 0.0
if ('longitude') in dbheaders:
try:
club.longitude = float(club.longitude)
except ValueError:
club.longitude = 0.0
else:
club.longitude = 0.0
# Sometimes, Toastmasters gets the latitude and longitude backwards
# If that turns out to create an impossible location (which it will in California),
# let's swap them.
if abs(club.latitude) > 90.0:
(club.latitude, club.longitude) = (club.longitude, club.latitude)
# And put it into the database if need be
if club.clubnumber in clubhist:
changes = different(club, clubhist[club.clubnumber], dbheaders[:-2])
else:
changes = []
if club.clubnumber not in clubhist and not firsttime:
# This is a new (or reinstated) club; note it in the changes database.
curs.execute('INSERT IGNORE INTO clubchanges (clubnumber, changedate, item, old, new) VALUES (%s, %s, "New Club", "", "")', (club.clubnumber, cdate))
if club.clubnumber not in clubhist or changes:
club.firstdate = club.lastdate
# Encode newlines in the place as double-semicolons for the database
club.place = club.place.replace('\n',';;')
values = [club.__dict__[x] for x in dbheaders]
# And then put the place back into normal form
club.place = club.place.replace(';;','\n')
thestr = 'INSERT IGNORE INTO clubs (' + ','.join(dbheaders) + ') VALUES (' + ','.join(['%s' for each in values]) + ');'
try:
changecount += curs.execute(thestr, values)
except Exception, e:
print e
# Capture changes
for (item, old, new) in changes:
if (item == 'place'):
# Clean up the place (old and new) for the database
old = old.replace('\n', ';;')
new = new.replace('\n', ';;')
try:
curs.execute('INSERT IGNORE INTO clubchanges (clubnumber, changedate, item, old, new) VALUES (%s, %s, %s, %s, %s)', (club.clubnumber, cdate, item, old, new))
except Exception, e:
print e
clubhist[club.clubnumber] = club
if different(club, clubhist[club.clubnumber], dbheaders[:-2]):
print 'it\'s different after being set.'
sys.exit(3)
def doDailyClubs(infile, conn, cdate, firsttime=False):
""" infile is a file-like object """
global changecount
from datetime import datetime, timedelta
curs = conn.cursor()
reader = csv.reader(infile)
hline = next(reader)
headers = cleanheaders(hline)
try:
clubcol = headers.index("clubnumber")
except ValueError:
if not hline[0].startswith('{"Message"'):
print("'clubnumber' not in '%s'" % hline)
return
try:
prospectiveclubcol = headers.index("prospectiveclub")
except ValueError:
prospectiveclubcol = False
# Find out what fields we have in the database itself
dbfields = []
curs.execute("describe clubs")
for l in curs.fetchall():
dbfields.append(l[0])
inform("clubs for", cdate, suppress=1)
dbheaders = [p for p in headers]
# Convert between Toastmasters' names for address and location and ours; they've changed it a few times. *sigh*
if "address1" in dbheaders:
addrcol1 = dbheaders.index("address1")
else:
addrcol1 = dbheaders.index("location")
if "address2" in dbheaders:
addrcol2 = dbheaders.index("address2")
else:
addrcol2 = dbheaders.index("address")
dbheaders[addrcol1] = "place"
dbheaders[addrcol2] = "address"
expectedheaderscount = len(dbheaders)
dbheaders.append("firstdate")
dbheaders.append("lastdate") # For now...
areacol = dbheaders.index("area")
divisioncol = dbheaders.index("division")
statecol = dbheaders.index("state")
# Now, suppress anything in the file that's not in the database:
suppress = []
oldheaders = dbheaders
dbheaders = []
for i in range(len(oldheaders)):
if oldheaders[i] in dbfields:
dbheaders.append(oldheaders[i])
else:
suppress.append(i)
suppress.reverse() # We remove these columns from the input
Club.setfieldnames(dbheaders)
# We need to get clubs for the most recent update so we know whether to update an entry
# or start a new one.
yesterday = datetime.strftime(
datetime.strptime(cdate, "%Y-%m-%d") - timedelta(1), "%Y-%m-%d"
)
clubhist = Club.getClubsOn(curs, date=yesterday)
for row in reader:
if len(row) < expectedheaderscount:
break # we're finished
if prospectiveclubcol is not None and row[prospectiveclubcol]:
continue # Ignore prospective clubs
for i in suppress:
del row[i]
if len(row) > expectedheaderscount:
# Special case...Millbrae somehow snuck two club websites in!
row[16] = row[16] + "," + row[17]
del row[17]
# print row[addrcol1]
# print row[addrcol2]
# Now, clean up the address:
# Address line 1 is "place" information and can be multiple lines.
# Address line 2 is the real address and should be treated as one line, with spaces normalized.
place = "\n".join([x.strip() for x in row[addrcol1].split(" ")])
row[addrcol1] = place
address = normalize(row[addrcol2])
row[addrcol2] = address
# Toastmasters is currently reversing the "Area" and "Division" items. "Area" should be a
# number; if not, swap the two.
try:
thearea = row[areacol]
thedivision = row[divisioncol]
areanum = int(row[areacol])
except ValueError:
row[areacol] = thedivision
row[divisioncol] = thearea
# Collapse state names into their abbreviations
if row[statecol] in statelist:
row[statecol] = statelist[row[statecol]]
# Get the right number of items into the row by setting today as the
# tentative first and last date
row.append(cdate)
row.append(cdate)
# And create the object
club = Club(row)
# Now, clean up things coming from Toastmasters
if club.clubstatus.startswith("Open") or club.clubstatus.startswith("None"):
club.clubstatus = "Open"
else:
club.clubstatus = "Restricted"
# Clean up the club and district numbers and the area
club.clubnumber = cleanitem(club.clubnumber)
club.district = cleanitem(club.district)
club.area = cleanitem(club.area)
# If a club is partially unassigned, mark it as completely unassigned.
if (
(club.area == "0A")
or (club.area == "0D")
or (club.division == "0D")
or (club.division == "0A")
):
club.area = "0A"
club.division = "0D"
# Clean up the charter date
club.charterdate = cleandate(club.charterdate)
# Clean up advanced status
club.advanced = "1" if (club.advanced != "") else "0"
# Clean up online status
club.allowsonlineattendance = (
"1" if (club.allowsonlineattendance != "") else "0"
)
# Add missing schemes to any URLs
club.fixURLSchemes()
# Now, take care of missing latitude/longitude
if ("latitude") in dbheaders:
try:
club.latitude = float(club.latitude)
except ValueError:
club.latitude = 0.0
else:
club.latitude = 0.0
if ("longitude") in dbheaders:
try:
club.longitude = float(club.longitude)
except ValueError:
club.longitude = 0.0
else:
club.longitude = 0.0
# Sometimes, Toastmasters gets the latitude and longitude backwards
# If that turns out to create an impossible location (which it will in California),
# let's swap them.
if abs(club.latitude) > 90.0:
(club.latitude, club.longitude) = (club.longitude, club.latitude)
# And put it into the database if need be
if club.clubnumber in clubhist:
changes = different(club, clubhist[club.clubnumber], dbheaders[:-2])
else:
changes = []
if club.clubnumber not in clubhist and not firsttime:
# This is a new (or reinstated) club; note it in the changes database.
curs.execute(
'INSERT IGNORE INTO clubchanges (clubnumber, changedate, item, old, new) VALUES (%s, %s, "New Club", "", "")',
(club.clubnumber, cdate),
)
if club.clubnumber not in clubhist or changes:
club.firstdate = club.lastdate
# Encode newlines in the place as double-semicolons for the database
club.place = club.place.replace("\n", ";;")
values = [club.__dict__[x] for x in dbheaders]
# And then put the place back into normal form
club.place = club.place.replace(";;", "\n")
thestr = (
"INSERT IGNORE INTO clubs ("
+ ",".join(dbheaders)
+ ") VALUES ("
+ ",".join(["%s" for each in values])
+ ");"
)
try:
changecount += curs.execute(thestr, values)
except Exception as e:
print(e)
# Capture changes
for (item, old, new) in changes:
if item == "place":
# Clean up the place (old and new) for the database
old = old.replace("\n", ";;")
new = new.replace("\n", ";;")
try:
curs.execute(
"INSERT IGNORE INTO clubchanges (clubnumber, changedate, item, old, new) VALUES (%s, %s, %s, %s, %s)",
(club.clubnumber, cdate, item, old, new),
)
except Exception as e:
print(e)
clubhist[club.clubnumber] = club
if different(club, clubhist[club.clubnumber], dbheaders[:-2]):
print("it's different after being set.")
sys.exit(3)
else:
# update the lastdate
changecount += curs.execute(
"UPDATE clubs SET lastdate = %s WHERE clubnumber = %s AND lastdate = %s;",
(cdate, club.clubnumber, clubhist[club.clubnumber].lastdate),
)
# If all the files were processed, today's work is done.
curs.execute(
'INSERT IGNORE INTO loaded (tablename, loadedfor) VALUES ("clubs", %s)',
(cdate,),
)
def overrideClubs(clubs, newAlignment, exclusive=True):
""" Updates 'clubs' to reflect the alignment in the newAlignment spreadsheet.
Typically used at a TM year boundary or for planning realignment.
newAlignment is a "test alignment" file - a CSV with the clubs in the new
alignment. Non-blank columns in the file override information from the database,
with the following exceptions:
"likelytoclose" is used during planning to identify clubs in danger.
"clubname" does NOT override what comes from Toastmasters.
"newarea" is of the form Xn (X is the Division; n is the area).
Specify exclusive=False to keep clubs not included in the spreadsheet;
otherwise, clubs not in the spreadsheet are discarded.
"""
pfile = open(newAlignment, 'r')
reader = csv.DictReader(pfile)
keepers = set()
# Sadly, the callers of this routine have different types of key.
# We have to be resilient against this.
if isinstance(list(clubs.keys())[0], int):
fixup = lambda x: int(x)
elif isinstance(list(clubs.keys())[0], (int, float)):
fixup = lambda x: float(x)
else:
fixup = lambda x: x
for row in reader:
clubnum = fixup(row['clubnumber'])
try:
club = clubs[clubnum]
except KeyError:
from simpleclub import Club
club = Club(list(row.values()),
fieldnames=list(row.keys()),
fillall=True)
if 'district' not in list(row.keys()):
club.district = globals.parms.district
clubs[clubnum] = club
keepers.add(clubnum)
if row['newarea'] and row['newarea'] != '0D0A':
club.division = row['newarea'][0]
club.area = row['newarea'][1:]
for item in reader.fieldnames:
if row[item].strip() and item not in [
'clubnumber', 'clubname', 'newarea'
]:
# Maintain the same type
if item in club.__dict__ and isinstance(
club.__dict__[item], numbers.Number):
if isinstance(club.__dict__[item], int):
club.__dict__[item] = int(row[item])
elif isinstance(club.__dict__[item], float):
club.__dict__[item] = float(row[item])
else:
club.__dict__[item] = int(row[item])
else:
club.__dict__[item] = row[item]
pfile.close()
# Remove clubs not in the file:
if exclusive:
clubnumbers = list(clubs.keys())
for c in clubnumbers:
if c not in keepers:
del clubs[c]
return clubs
def overrideClubs(clubs, newAlignment, exclusive=True):
""" Updates 'clubs' to reflect the alignment in the newAlignment spreadsheet.
Typically used at a TM year boundary or for planning realignment.
newAlignment is a "test alignment" file - a CSV with the clubs in the new
alignment. Non-blank columns in the file override information from the database,
with the following exceptions:
"likelytoclose" is used during planning to identify clubs in danger.
"clubname" does NOT override what comes from Toastmasters.
"newarea" is of the form Xn (X is the Division; n is the area).
Specify exclusive=False to keep clubs not included in the spreadsheet;
otherwise, clubs not in the spreadsheet are discarded.
"""
pfile = open(newAlignment, 'rbU')
reader = csv.DictReader(pfile)
keepers = set()
# Sadly, the callers of this routine have different types of key.
# We have to be resilient against this.
if isinstance(clubs.keys()[0], int):
fixup = lambda x:int(x)
elif isinstance(clubs.keys()[0], (long, float)):
fixup = lambda x:float(x)
else:
fixup = lambda x:x
for row in reader:
clubnum = fixup(row['clubnumber'])
try:
club = clubs[clubnum]
except KeyError:
from simpleclub import Club
club = Club(row.values(), fieldnames=row.keys(), fillall=True)
clubs[clubnum] = club
keepers.add(clubnum)
if row['newarea'] and row['newarea'] != '0D0A':
club.division = row['newarea'][0]
club.area = row['newarea'][1:]
for item in reader.fieldnames:
if row[item].strip() and item not in ['clubnumber', 'clubname', 'newarea']:
# Maintain the same type
if item in club.__dict__ and isinstance(club.__dict__[item], numbers.Number):
if isinstance(club.__dict__[item], int):
club.__dict__[item] = int(row[item])
elif isinstance(club.__dict__[item], float):
club.__dict__[item] = float(row[item])
else:
club.__dict__[item] = long(row[item])
else:
club.__dict__[item] = row[item]
pfile.close()
# Remove clubs not in the file:
if exclusive:
clubnumbers = clubs.keys()
for c in clubnumbers:
if c not in keepers:
del clubs[c]
return clubs
