Initial commit

5 files changed, 343 insertions, 0 deletions

diff --git a/email_assistant/plugins/__init__.py b/email_assistant/plugins/__init__.py
new file mode 100644
index 0000000..10b28f0
--- /dev/null
+++ b/email_assistant/plugins/__init__.py
@@ -0,0 +1,29 @@
+#!/usr/bin/env python3
+
+# Copyright (C) 2019 James E. Blair <corvus@gnu.org>
+#
+# This file is part of Email-assistant.
+#
+# Email-assistant is free software: you can redistribute it and/or
+# modify it under the terms of the GNU Affero General Public License
+# as published by the Free Software Foundation, either version 3 of
+# the License, or (at your option) any later version.
+#
+# Email-assistant is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+# General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with Email-assistant.  If not, see
+# <https://www.gnu.org/licenses/>.
+
+from . import eventbrite
+from . import marriott
+from . import united
+
+plugins = [
+    eventbrite.Plugin,
+    marriott.Plugin,
+    united.Plugin,
+]
diff --git a/email_assistant/plugins/eventbrite.py b/email_assistant/plugins/eventbrite.py
new file mode 100644
index 0000000..9ef073a
--- /dev/null
+++ b/email_assistant/plugins/eventbrite.py
@@ -0,0 +1,90 @@
+#!/usr/bin/env python3
+
+# Copyright (C) 2019 James E. Blair <corvus@gnu.org>
+#
+# This file is part of Email-assistant.
+#
+# Email-assistant is free software: you can redistribute it and/or
+# modify it under the terms of the GNU Affero General Public License
+# as published by the Free Software Foundation, either version 3 of
+# the License, or (at your option) any later version.
+#
+# Email-assistant is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+# General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with Email-assistant.  If not, see
+# <https://www.gnu.org/licenses/>.
+
+import re
+import logging
+import hashlib
+import datetime
+import json
+
+from bs4 import BeautifulSoup
+import dateutil.parser
+import dateutil.tz
+import inscriptis
+import vobject
+
+from email_assistant import plugin
+
+class Plugin(plugin.Plugin):
+    name = 'eventbrite'
+
+    def match(self, msg):
+        if ('orders@eventbrite.com' in msg['From'] and
+            'Your Tickets for' in msg['Subject']):
+            return True
+
+    def get_events(self, msg):
+        events = []
+        for part in msg.walk():
+            if part.get_content_type() == 'text/html':
+                soup = BeautifulSoup(part.get_payload(decode=True).decode('utf8'), 'html.parser')
+                data = json.loads(soup.find('script', type="application/ld+json").string)
+
+                address = data['reservationFor']['location']['address']
+                location = ' '.join((address['streetAddress'],
+                                     address['addressLocality'],
+                                     address['addressRegion'],
+                                     address['postalCode'],
+                                     address['addressCountry']))
+                summary = data['reservationFor']['name']
+
+                start = dateutil.parser.parse(data['reservationFor']['startDate'])
+                end = dateutil.parser.parse(data['reservationFor']['endDate'])
+                if start.date() != end.date():
+                    # If it is a multi-day event, don't schedule a time.
+                    start = start.date()
+                    end = end.date()+datetime.timedelta(days=1)
+                else:
+                    # We do have a TZ offset, and could schedule this as a
+                    # UTC event.  It would generally display correctly,
+                    # but it would make it difficult to view the event
+                    # when the viewer is in a different timezone (as they
+                    # would have to mentally perform the UTC conversion,
+                    # instead of being able to see the event's own "local"
+                    # time.  To mitigate this, perform a location lookup.
+                    tzinfo = self.assistant.get_tzinfo(location)
+                    if tzinfo is not None:
+                        start = start.replace(tzinfo=tzinfo)
+                        end = end.replace(tzinfo=tzinfo)
+
+                cal = vobject.iCalendar()
+                event = cal.add('vevent')
+                event.add('dtstart').value = start
+                event.add('dtend').value = end
+                event.add('summary').value = summary
+                text = inscriptis.get_text(str(soup))
+                text = re.sub(r'([^ ]+)\s*\n', '\\1\n', text)
+                event.add('description').value = text
+                event.add('location').value = location
+                uid = hashlib.sha1((str(start) + summary).encode('utf8')).hexdigest()
+                event.add('uid').value = uid
+                events.append(cal)
+
+        return events
diff --git a/email_assistant/plugins/general.py b/email_assistant/plugins/general.py
new file mode 100644
index 0000000..c0517b8
--- /dev/null
+++ b/email_assistant/plugins/general.py
@@ -0,0 +1,58 @@
+
+# An unused sketch of a generalized plugin that uses commonregex to
+# find dates and addresses.
+
+import re
+import logging
+import hashlib
+import datetime
+
+from bs4 import BeautifulSoup
+import commonregex
+import dateutil.parser
+import dateutil.tz
+import inscriptis
+import vobject
+
+def match(msg):
+    if ('reservations@res-marriott.com' in msg['From'] and
+        'Reservation Confirmation' in msg['Subject']):
+        return True
+
+def get_events(msg):
+    events = []
+    log = logging.getLogger('assistant.marriott')
+    for part in msg.walk():
+        if part.get_content_type() == 'text/html':
+            soup = BeautifulSoup(part.get_payload(decode=True).decode('utf8'), 'html.parser')
+
+            start = end = location = None
+            for element in soup.descendants:
+                if not hasattr(element, 'children') and element.string:
+                    text = element.string.strip()
+                    if not text:
+                        continue
+                    found = commonregex.CommonRegex(text)
+                    for date in found.dates:
+                        if not start:
+                            start = dateutil.parser.parse(date).date()
+                            continue
+                        if not end:
+                            end = dateutil.parser.parse(date).date()+datetime.timedelta(days=1)
+                    if found.street_addresses and not location:
+                        location = text
+            if not (start and end and location):
+                continue
+            cal = vobject.iCalendar()
+            event = cal.add('vevent')
+            event.add('dtstart').value = start
+            event.add('dtend').value = end
+            summary = re.match('Reservation Confirmation .*? for (.*)', msg['Subject']).group(1).strip()
+            event.add('summary').value = summary
+            event.add('description').value = inscriptis.get_text(str(soup))
+            event.add('location').value = location
+            uid = hashlib.sha1((str(start) + summary).encode('utf8')).hexdigest()
+            event.add('uid').value = uid
+            events.append(cal)
+            print(repr(cal.serialize()))
+    return events
diff --git a/email_assistant/plugins/marriott.py b/email_assistant/plugins/marriott.py
new file mode 100644
index 0000000..52597cd
--- /dev/null
+++ b/email_assistant/plugins/marriott.py
@@ -0,0 +1,67 @@
+#!/usr/bin/env python3
+
+# Copyright (C) 2019 James E. Blair <corvus@gnu.org>
+#
+# This file is part of Email-assistant.
+#
+# Email-assistant is free software: you can redistribute it and/or
+# modify it under the terms of the GNU Affero General Public License
+# as published by the Free Software Foundation, either version 3 of
+# the License, or (at your option) any later version.
+#
+# Email-assistant is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+# General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with Email-assistant.  If not, see
+# <https://www.gnu.org/licenses/>.
+
+import re
+import logging
+import hashlib
+import datetime
+
+from bs4 import BeautifulSoup
+import dateutil.parser
+import dateutil.tz
+import inscriptis
+import vobject
+
+from email_assistant import plugin
+
+class Plugin(plugin.Plugin):
+    name = 'marriott'
+
+    def match(self, msg):
+        if ('reservations@res-marriott.com' in msg['From'] and
+            'Reservation Confirmation' in msg['Subject']):
+            return True
+
+    def get_events(self, msg):
+        events = []
+        for part in msg.walk():
+            if part.get_content_type() == 'text/html':
+                soup = BeautifulSoup(part.get_payload(decode=True).decode('utf8'), 'html.parser')
+                summary = soup.find_all('table')[7].a.string.strip()
+                location = soup.find_all('table')[9].a.string.strip()
+                start = (soup.find('th', string=re.compile('Check-In:')).
+                         find_next_sibling('th').string.strip())
+                end = (soup.find('td', string=re.compile('Check-Out:')).
+                         find_next_sibling('td').string.strip())
+                start = dateutil.parser.parse(start).date()
+                end = dateutil.parser.parse(end).date()+datetime.timedelta(days=1)
+                cal = vobject.iCalendar()
+                event = cal.add('vevent')
+                event.add('dtstart').value = start
+                event.add('dtend').value = end
+                event.add('summary').value = summary
+                text = inscriptis.get_text(str(soup))
+                text = re.sub(r'([^ ]+)\s*\n', '\\1\n', text)
+                event.add('description').value = text
+                event.add('location').value = location
+                uid = hashlib.sha1((str(start) + summary).encode('utf8')).hexdigest()
+                event.add('uid').value = uid
+                events.append(cal)
+        return events
diff --git a/email_assistant/plugins/united.py b/email_assistant/plugins/united.py
new file mode 100644
index 0000000..0b33cd9
--- /dev/null
+++ b/email_assistant/plugins/united.py
@@ -0,0 +1,99 @@
+#!/usr/bin/env python3
+
+# Copyright (C) 2019 James E. Blair <corvus@gnu.org>
+#
+# This file is part of Email-assistant.
+#
+# Email-assistant is free software: you can redistribute it and/or
+# modify it under the terms of the GNU Affero General Public License
+# as published by the Free Software Foundation, either version 3 of
+# the License, or (at your option) any later version.
+#
+# Email-assistant is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+# General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with Email-assistant.  If not, see
+# <https://www.gnu.org/licenses/>.
+
+import re
+import logging
+import hashlib
+
+from bs4 import BeautifulSoup
+import dateutil.parser
+import dateutil.tz
+import inscriptis
+import vobject
+
+from email_assistant import iata
+from email_assistant import plugin
+
+def parse_dep_arr(flight_date, dep_arr):
+    flight_year = dateutil.parser.parse(flight_date).year
+    city, br, code, flight_time = dep_arr.span.children
+    city = city.strip()
+    code = code.strip()[1:-1]
+    code = code.split()[0]
+    tz = iata.tzmap[code]
+    flight_time = flight_time.get_text().strip()
+    m = re.match(r'(.*) \((\d+[A-Z]+)\)', flight_time)
+    if m:
+        s = '%s%s %s' % (m.group(2), flight_year, m.group(1))
+        flight_time = dateutil.parser.parse(s)
+    else:
+        flight_time = dateutil.parser.parse(flight_date +' '+ flight_time)
+    flight_time = flight_time.replace(tzinfo=dateutil.tz.gettz(tz))
+    return (city, code, flight_time)
+
+class Plugin(plugin.Plugin):
+    name = 'united'
+
+    def match(self, msg):
+        if ('unitedairlines@united.com' in msg['From'] and
+            'Itinerary and Receipt' in msg['Subject']):
+            return True
+
+    def get_events(self, msg):
+        events = []
+        for part in msg.walk():
+            if part.get_content_type() == 'text/html':
+                soup = BeautifulSoup(part.get_payload(decode=True).decode('utf8'), 'html.parser')
+                # confirmation_number = soup.find(class_="eTicketConfirmation").string
+
+                index = 0
+                while True:
+                    info = soup.find(id="ShowSegments_ShowSegment_ctl%02i_Flight" % index)
+                    if not info:
+                        break
+                    for row in info.parents:
+                        if row.name == 'tr':
+                            break
+                    cols = row.find_all('td')
+                    flight_date, flight_num, flight_class, dep, arr, ac, meal = cols
+                    flight_date = flight_date.get_text().strip()
+                    flight_num = flight_num.get_text().strip()
+                    flight_class = flight_class.get_text().strip()
+
+                    dep_city, dep_code, dep_time = parse_dep_arr(flight_date, dep)
+                    arr_city, arr_code, arr_time = parse_dep_arr(flight_date, arr)
+                    self.log.debug("dep: %s %s %s", dep_city, dep_code, dep_time)
+                    self.log.debug("arr: %s %s %s", arr_city, arr_code, arr_time)
+
+                    cal = vobject.iCalendar()
+                    event = cal.add('vevent')
+                    event.add('dtstart').value = dep_time
+                    event.add('dtend').value = arr_time
+                    summary = "Flight from %s to %s" % (dep_code, arr_code)
+                    event.add('summary').value = summary
+                    text = inscriptis.get_text(str(soup))
+                    text = re.sub(r'([^ ]+)\s*\n', '\\1\n', text)
+                    event.add('description').value = text
+                    event.add('location').value = "%s airport" % (dep_code,)
+                    uid = hashlib.sha1((str(dep_time) + summary).encode('utf8')).hexdigest()
+                    event.add('uid').value = uid
+                    events.append(cal)
+                    index += 1
+        return events