vmario/ical2csv-fork
0
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Updated ical2csv.py to match ical2txt.py, including recursive events and HTML removal

Browse Source 
Handled events that returned date objects in stead of datetime objects
Added information on how to get the output in the timezone of your choice to the help
This commit is contained in:
Martin Moeller 2021-05-02 17:43:32 +02:00
parent eb6fdf7a19
commit 026d6a6a6d
2 changed files with 94 additions and 22 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

100
ical2csv.py
View File

@ -3,8 +3,24 @@
import sys import sys
import os.path import os.path
from icalendar import Calendar from icalendar import Calendar
import recurring_ical_events
from bs4 import BeautifulSoup
import warnings
from dateutil.parser import parse
import datetime
import csv import csv
warnings.filterwarnings("ignore", category=UserWarning, module='bs4') # We don't want warnings about URL's. We just what the URL printed, if there.
if len(sys.argv) <= 1:
print("Please call this script with an ics-file as parameter.\n")
print("Even better, call it with start and end dates:\n")
print(sys.argv[0] + " myexport.ics 20210101 20210201")
print(sys.argv[0] + " myexport.ics 2021-01-01T00:00:00 2021-01-31T23:59:59\n")
print("NOTE: If you need data in another timezone than the system is set to, override like this before running the script:")
print("export TZ=\"Europe/Copenhagen\"\n")
exit(1)
filename = sys.argv[1] filename = sys.argv[1]
# TODO: use regex to get file extension (chars after last period), in case it's not exactly 3 chars. # TODO: use regex to get file extension (chars after last period), in case it's not exactly 3 chars.
file_extension = str(sys.argv[1])[-3:] file_extension = str(sys.argv[1])[-3:]
@ -25,6 +41,24 @@ class CalendarEvent:
events = [] events = []
def removehtml(html):
# Almost word for word copy from here: https://stackoverflow.com/questions/328356/extracting-text-from-html-file-using-python
soup = BeautifulSoup(html, features="html.parser")
# kill all script and style elements
for script in soup(["script", "style"]):
script.extract() # remove it
text = soup.get_text() # Get plain text
# break into lines and remove leading and trailing space on each
lines = (line.strip() for line in text.splitlines())
# break multi-headlines into a line each
chunks = (phrase.strip() for line in lines for phrase in line.split(" "))
# drop blank lines
text = '\n'.join(chunk for chunk in chunks if chunk)
return text
def open_cal(): def open_cal():
if os.path.isfile(filename): if os.path.isfile(filename):
@ -32,24 +66,33 @@ def open_cal():
print("Extracting events from file:", filename, "\n") print("Extracting events from file:", filename, "\n")
f = open(sys.argv[1], 'rb') f = open(sys.argv[1], 'rb')
gcal = Calendar.from_ical(f.read()) gcal = Calendar.from_ical(f.read())
revents = recurring_ical_events.of(gcal).between(istart,istop)
for component in gcal.walk(): # for component in gcal.walk():
for component in revents:
event = CalendarEvent("event") event = CalendarEvent("event")
if component.get('TRANSP') == 'TRANSPARENT': continue #skip event that have not been accepted v=(dir(component).count('get')) # Only proces data if object is a valid event
if component.get('SUMMARY') == None: continue #skip blank items if (v != 0):
event.summary = component.get('SUMMARY') if component.get('TRANSP') == 'TRANSPARENT': continue #skip all day events and the like
event.uid = component.get('UID') if component.get('SUMMARY') == None: continue #skip blank items
if component.get('DESCRIPTION') == None: continue #skip blank items event.summary = component.get('SUMMARY')
event.description = component.get('DESCRIPTION') event.uid = component.get('UID')
event.location = component.get('LOCATION') if component.get('DESCRIPTION') == None: continue #skip blank items
if hasattr(component.get('dtstart'), 'dt'): event.description = component.get('DESCRIPTION')
event.start = component.get('dtstart').dt event.location = component.get('LOCATION')
if hasattr(component.get('dtend'), 'dt'): if hasattr(component.get('dtstart'), 'dt'):
event.end = component.get('dtend').dt event.start = component.get('dtstart').dt
if hasattr(component.get('dtend'), 'dt'):
event.end = component.get('dtend').dt
if type(now) != type(event.start): # If we get a datetime.date object, convert to datetime.datetime
event.url = component.get('URL') event.start=datetime.datetime.combine(event.start, datetime.time.min)
events.append(event) if type(now) != type(event.end): # If we get a datetime.date object, convert to datetime.datetime
event.end=datetime.datetime.combine(event.end, datetime.time.max)
event.start = event.start.astimezone()
event.end = event.end.astimezone()
event.url = component.get('URL')
events.append(event)
f.close() f.close()
else: else:
print("You entered ", filename, ". ") print("You entered ", filename, ". ")
@ -63,14 +106,21 @@ def open_cal():
def csv_write(icsfile): def csv_write(icsfile):
csvfile = icsfile[:-3] + "csv" csvfile = icsfile[:-3] + "csv"
spent=0
evcount=0
evskip=0
sys.stdout.write("Processing events : ")
try: try:
with open(csvfile, 'w') as myfile: with open(csvfile, 'w') as myfile:
wr = csv.writer(myfile, quoting=csv.QUOTE_ALL) wr = csv.writer(myfile, quoting=csv.QUOTE_ALL)
wr.writerow(headers) wr.writerow(headers)
for event in sortedevents: for event in sortedevents:
values = (event.summary.encode('utf8').decode(), event.uid, event.description.encode('uft8').decode(), event.location, event.start, event.end, event.url) values = (event.summary.encode('utf-8').decode(), event.uid, removehtml(event.description.encode('utf-8').decode()), event.location.encode('utf-8').decode(), event.start, event.end, event.url)
wr.writerow(values) wr.writerow(values)
print("Wrote to ", csvfile, "\n") sys.stdout.write(".")
sys.stdout.flush()
evcount+=1
print("\n\nWrote " + str(evcount) + " events to ", csvfile, "\n")
except IOError: except IOError:
print("Could not open file! Please close Excel!") print("Could not open file! Please close Excel!")
exit(0) exit(0)
@ -86,7 +136,19 @@ def debug_event(class_name):
print(class_name.end) print(class_name.end)
print(class_name.url, "\n") print(class_name.url, "\n")
open_cal() now=datetime.datetime.now()
sortedevents=sorted(events, key=lambda obj: obj.start) # Needed to sort events. They are not fully chronological in a Google Calendard export ... istart=datetime.datetime.fromtimestamp(0) # Start of UNIX epoch (1970-01-01T00:00:00)
istop=now+datetime.timedelta(seconds=157680000) # Stop 5 years in the future, if no enddate is given, to make sure reucurring events don't go on forever ...
if len(sys.argv) > 3:
if sys.argv[2] != '':
istart=parse(sys.argv[2])
if sys.argv[3] != '':
istop=parse(sys.argv[3])
print("Opening ics file\n")
open_cal() # Open ics file and do initial parsing of events
print("Sorting events\n")
sortedevents=sorted(events, key=lambda obj: obj.start) # Make sure events are in chronological order
csv_write(filename) csv_write(filename)
#debug_event(event) #debug_event(event)

16
ical2txt.py
View File

@ -16,6 +16,8 @@ if len(sys.argv) <= 1:
print("Even better, call it with start and end dates:\n") print("Even better, call it with start and end dates:\n")
print(sys.argv[0] + " myexport.ics 20210101 20210201") print(sys.argv[0] + " myexport.ics 20210101 20210201")
print(sys.argv[0] + " myexport.ics 2021-01-01T00:00:00 2021-01-31T23:59:59\n") print(sys.argv[0] + " myexport.ics 2021-01-01T00:00:00 2021-01-31T23:59:59\n")
print("NOTE: If you need data in another timezone than the system is set to, override like this before running the script:")
print("export TZ=\"Europe/Copenhagen\"\n")
exit(1) exit(1)
filename = sys.argv[1] filename = sys.argv[1]
@ -78,10 +80,16 @@ def open_cal():
event.description = component.get('DESCRIPTION') event.description = component.get('DESCRIPTION')
event.location = component.get('LOCATION') event.location = component.get('LOCATION')
if hasattr(component.get('dtstart'), 'dt'): if hasattr(component.get('dtstart'), 'dt'):
event.start = component.get('dtstart').dt event.start = component.get('dtstart').dt.astimezone()
if hasattr(component.get('dtend'), 'dt'): if hasattr(component.get('dtend'), 'dt'):
event.end = component.get('dtend').dt event.end = component.get('dtend').dt.astimezone()
if type(now) != type(event.start): # If we get a datetime.date object, convert to datetime.datetime
event.start=datetime.datetime.combine(event.start, datetime.time.min)
if type(now) != type(event.end): # If we get a datetime.date object, convert to datetime.datetime
event.end=datetime.datetime.combine(event.end, datetime.time.max)
event.start = event.start.astimezone()
event.end = event.end.astimezone()
event.url = component.get('URL') event.url = component.get('URL')
events.append(event) events.append(event)
f.close() f.close()
@ -124,7 +132,7 @@ def txt_write(icsfile):
minutes = divmod(ds,3600)[1]/60 minutes = divmod(ds,3600)[1]/60
description=removehtml(event.description.encode('utf-8').decode()) description=removehtml(event.description.encode('utf-8').decode())
values = event.start.strftime("%H:%M") + " - " + event.end.strftime("%H:%M") + " (" + '{:02.0f}'.format(hours) + ":" + '{:02.0f}'.format(minutes) + ") " + event.summary.encode('utf-8').decode() values = event.start.strftime("%H:%M") + " - " + event.end.strftime("%H:%M") + " (" + '{:02.0f}'.format(hours) + ":" + '{:02.0f}'.format(minutes) + ") " + event.summary.encode('utf-8').decode()
if event.location != '': values = values + " [" + event.location + "]" # Only include location if there is one if event.location != '': values = values + " [" + event.location.encode('utf-8').decode() + "]" # Only include location if there is one
# Remove Google Meet and Skype Meeting part of description # Remove Google Meet and Skype Meeting part of description
trimmed=description.split('-::~')[0].split('......')[0] trimmed=description.split('-::~')[0].split('......')[0]
@ -171,7 +179,9 @@ if len(sys.argv) > 3:
if sys.argv[3] != '': if sys.argv[3] != '':
istop=parse(sys.argv[3]) istop=parse(sys.argv[3])
print("Opening ics file\n")
open_cal() # Open ics file and do initial parsing of events open_cal() # Open ics file and do initial parsing of events
print("Sorting events\n")
sortedevents=sorted(events, key=lambda obj: obj.start) # Make sure events are in chronological order sortedevents=sorted(events, key=lambda obj: obj.start) # Make sure events are in chronological order
txt_write(filename) # Write the matching events to the textfile. With recurring_ical_events, scoping is already done. txt_write(filename) # Write the matching events to the textfile. With recurring_ical_events, scoping is already done.
#debug_event(event) #debug_event(event)