def test_mentioned_data(self) -> None: slack_user_map = {"U08RGD1RD": 540, "U0CBK5KAT": 554, "U09TYF5SK": 571} # For this test, only relevant keys are 'id', 'name', 'deleted' # and 'real_name' users = [ { "id": "U0CBK5KAT", "name": "aaron.anzalone", "deleted": False, "is_mirror_dummy": False, "real_name": "", }, { "id": "U08RGD1RD", "name": "john", "deleted": False, "is_mirror_dummy": False, "real_name": "John Doe", }, { "id": "U09TYF5Sk", "name": "Jane", "is_mirror_dummy": False, "deleted": True, # Deleted users don't have 'real_name' key in Slack }, ] channel_map = {"general": ("C5Z73A7RA", 137)} message = "Hi <@U08RGD1RD|john>: How are you? <#C5Z73A7RA|general>" text, mentioned_users, has_link = convert_to_zulip_markdown( message, users, channel_map, slack_user_map) full_name = get_user_full_name(users[1]) self.assertEqual(full_name, "John Doe") self.assertEqual(get_user_full_name(users[2]), "Jane") self.assertEqual(text, f"Hi @**{full_name}**: How are you? #**general**") self.assertEqual(mentioned_users, [540]) # multiple mentioning message = "Hi <@U08RGD1RD|john>: How are you?<@U0CBK5KAT> asked." text, mentioned_users, has_link = convert_to_zulip_markdown( message, users, channel_map, slack_user_map) self.assertEqual( text, "Hi @**John Doe**: How are you?@**aaron.anzalone** asked.") self.assertEqual(mentioned_users, [540, 554]) # Check wrong mentioning message = "Hi <@U08RGD1RD|jon>: How are you?" text, mentioned_users, has_link = convert_to_zulip_markdown( message, users, channel_map, slack_user_map) self.assertEqual(text, message) self.assertEqual(mentioned_users, [])
def test_mentioned_data(self) -> None: slack_user_map = {'U08RGD1RD': 540, 'U0CBK5KAT': 554, 'U09TYF5SK': 571} # For this test, only relevant keys are 'id', 'name', 'deleted' # and 'real_name' users = [{ "id": "U0CBK5KAT", "name": "aaron.anzalone", "deleted": False, "is_mirror_dummy": False, "real_name": "" }, { "id": "U08RGD1RD", "name": "john", "deleted": False, "is_mirror_dummy": False, "real_name": "John Doe" }, { "id": "U09TYF5Sk", "name": "Jane", "is_mirror_dummy": False, "deleted": True }] # Deleted users don't have 'real_name' key in Slack channel_map = {'general': ('C5Z73A7RA', 137)} message = 'Hi <@U08RGD1RD|john>: How are you? <#C5Z73A7RA|general>' text, mentioned_users, has_link = convert_to_zulip_markdown( message, users, channel_map, slack_user_map) full_name = get_user_full_name(users[1]) self.assertEqual(full_name, 'John Doe') self.assertEqual(get_user_full_name(users[2]), 'Jane') self.assertEqual( text, 'Hi @**%s**: How are you? #**general**' % (full_name, )) self.assertEqual(mentioned_users, [540]) # multiple mentioning message = 'Hi <@U08RGD1RD|john>: How are you?<@U0CBK5KAT> asked.' text, mentioned_users, has_link = convert_to_zulip_markdown( message, users, channel_map, slack_user_map) self.assertEqual( text, 'Hi @**%s**: How are you?@**%s** asked.' % ('John Doe', 'aaron.anzalone')) self.assertEqual(mentioned_users, [540, 554]) # Check wrong mentioning message = 'Hi <@U08RGD1RD|jon>: How are you?' text, mentioned_users, has_link = convert_to_zulip_markdown( message, users, channel_map, slack_user_map) self.assertEqual(text, message) self.assertEqual(mentioned_users, [])
def test_has_link(self) -> None: slack_user_map = {} # type: Dict[str, int] message = '<http://journals.plos.org/plosone/article>' text, mentioned_users, has_link = convert_to_zulip_markdown(message, [], {}, slack_user_map) self.assertEqual(text, 'http://journals.plos.org/plosone/article') self.assertEqual(has_link, True) message = '<mailto:[email protected]>' text, mentioned_users, has_link = convert_to_zulip_markdown(message, [], {}, slack_user_map) self.assertEqual(text, 'mailto:[email protected]') self.assertEqual(has_link, True) message = 'random message' text, mentioned_users, has_link = convert_to_zulip_markdown(message, [], {}, slack_user_map) self.assertEqual(has_link, False)
def test_has_link(self) -> None: slack_user_map = {} # type: Dict[str, int] message = '<http://journals.plos.org/plosone/article>' text, mentioned_users, has_link = convert_to_zulip_markdown(message, [], {}, slack_user_map) self.assertEqual(text, 'http://journals.plos.org/plosone/article') self.assertEqual(has_link, True) message = '<mailto:[email protected]>' text, mentioned_users, has_link = convert_to_zulip_markdown(message, [], {}, slack_user_map) self.assertEqual(text, 'mailto:[email protected]') self.assertEqual(has_link, True) message = 'random message' text, mentioned_users, has_link = convert_to_zulip_markdown(message, [], {}, slack_user_map) self.assertEqual(has_link, False)
def test_has_link(self) -> None: slack_user_map: Dict[str, int] = {} message = "<http://journals.plos.org/plosone/article>" text, mentioned_users, has_link = convert_to_zulip_markdown( message, [], {}, slack_user_map) self.assertEqual(text, "http://journals.plos.org/plosone/article") self.assertEqual(has_link, True) message = "<mailto:[email protected]>" text, mentioned_users, has_link = convert_to_zulip_markdown( message, [], {}, slack_user_map) self.assertEqual(text, "mailto:[email protected]") self.assertEqual(has_link, True) message = "random message" text, mentioned_users, has_link = convert_to_zulip_markdown( message, [], {}, slack_user_map) self.assertEqual(has_link, False)
def test_mentioned_data(self) -> None: slack_user_map = {'U08RGD1RD': 540, 'U0CBK5KAT': 554, 'U09TYF5SK': 571} # For this test, only relevant keys are 'id', 'name', 'deleted' # and 'real_name' users = [{"id": "U0CBK5KAT", "name": "aaron.anzalone", "deleted": False, "real_name": ""}, {"id": "U08RGD1RD", "name": "john", "deleted": False, "real_name": "John Doe"}, {"id": "U09TYF5Sk", "name": "Jane", "deleted": True}] # Deleted users don't have 'real_name' key in Slack channel_map = {'general': ('C5Z73A7RA', 137)} message = 'Hi <@U08RGD1RD|john>: How are you? <#C5Z73A7RA|general>' text, mentioned_users, has_link = convert_to_zulip_markdown(message, users, channel_map, slack_user_map) full_name = get_user_full_name(users[1]) self.assertEqual(full_name, 'John Doe') self.assertEqual(get_user_full_name(users[2]), 'Jane') self.assertEqual(text, 'Hi @**%s**: How are you? #**general**' % (full_name)) self.assertEqual(mentioned_users, [540]) # multiple mentioning message = 'Hi <@U08RGD1RD|john>: How are you?<@U0CBK5KAT> asked.' text, mentioned_users, has_link = convert_to_zulip_markdown(message, users, channel_map, slack_user_map) self.assertEqual(text, 'Hi @**%s**: How are you?@**%s** asked.' % ('John Doe', 'aaron.anzalone')) self.assertEqual(mentioned_users, [540, 554]) # Check wrong mentioning message = 'Hi <@U08RGD1RD|jon>: How are you?' text, mentioned_users, has_link = convert_to_zulip_markdown(message, users, channel_map, slack_user_map) self.assertEqual(text, message) self.assertEqual(mentioned_users, [])
def test_message_conversion_fixtures(self) -> None: format_tests = self.load_slack_message_conversion_tests() valid_keys = {'name', "input", "conversion_output"} for name, test in format_tests.items(): # Check that there aren't any unexpected keys as those are often typos self.assertEqual(len(set(test.keys()) - valid_keys), 0) slack_user_map = {} # type: Dict[str, int] users = [{}] # type: List[Dict[str, Any]] channel_map = {} # type: Dict[str, Tuple[str, int]] converted = convert_to_zulip_markdown(test['input'], users, channel_map, slack_user_map) converted_text = converted[0] with self.subTest(slack_message_conversion=name): self.assertEqual(converted_text, test['conversion_output'])
def test_message_conversion_fixtures(self) -> None: format_tests = self.load_slack_message_conversion_tests() valid_keys = set(['name', "input", "conversion_output"]) for name, test in format_tests.items(): # Check that there aren't any unexpected keys as those are often typos self.assertEqual(len(set(test.keys()) - valid_keys), 0) slack_user_map = {} # type: Dict[str, int] users = [{}] # type: List[Dict[str, Any]] channel_map = {} # type: Dict[str, Tuple[str, int]] converted = convert_to_zulip_markdown(test['input'], users, channel_map, slack_user_map) converted_text = converted[0] print("Running Slack Message Conversion test: %s" % (name,)) self.assertEqual(converted_text, test['conversion_output'])
def channel_message_to_zerver_message(realm_id: int, users: List[ZerverFieldsT], added_users: AddedUsersT, added_recipient: AddedRecipientsT, all_messages: List[ZerverFieldsT], zerver_realmemoji: List[ZerverFieldsT], zerver_subscription: List[ZerverFieldsT], added_channels: AddedChannelsT, id_list: Tuple[int, int, int, int], domain_name: str) -> Tuple[List[ZerverFieldsT], List[ZerverFieldsT], List[ZerverFieldsT], List[ZerverFieldsT], List[ZerverFieldsT], Tuple[int, int, int, int]]: """ Returns: 1. zerver_message, which is a list of the messages 2. zerver_usermessage, which is a list of the usermessages 3. zerver_attachment, which is a list of the attachments 4. uploads_list, which is a list of uploads to be mapped in uploads records.json 5. reaction_list, which is a list of all user reactions 6. id_list, which is a tuple of max ids of messages, usermessages, reactions and attachments """ message_id_count, usermessage_id_count, reaction_id_count, attachment_id_count = id_list zerver_message = [] zerver_usermessage = [] # type: List[ZerverFieldsT] uploads_list = [] # type: List[ZerverFieldsT] zerver_attachment = [] # type: List[ZerverFieldsT] reaction_list = [] # type: List[ZerverFieldsT] # For unicode emoji with open(NAME_TO_CODEPOINT_PATH) as fp: name_to_codepoint = ujson.load(fp) for message in all_messages: user = get_message_sending_user(message) if not user: # Ignore messages without user names # These are Sometimes produced by slack continue subtype = message.get('subtype', False) if subtype in [ # Zulip doesn't have a pinned_item concept "pinned_item", "unpinned_item", # Slack's channel join/leave notices are spammy "channel_join", "channel_leave", "channel_name" ]: continue has_attachment = has_image = False try: content, mentioned_users_id, has_link = convert_to_zulip_markdown( message['text'], users, added_channels, added_users) except Exception: print("Slack message unexpectedly missing text representation:") print(json.dumps(message, indent=4)) continue rendered_content = None recipient_id = added_recipient[message['channel_name']] message_id = message_id_count # Process message reactions if 'reactions' in message.keys(): reaction_id_count = build_reactions(reaction_list, message['reactions'], added_users, message_id, reaction_id_count, name_to_codepoint, zerver_realmemoji) # Process different subtypes of slack messages # Subtypes which have only the action in the message should # be rendered with '/me' in the content initially # For example "sh_room_created" has the message 'started a call' # which should be displayed as '/me started a call' if subtype in ["bot_add", "sh_room_created", "me_message"]: content = ('/me %s' % (content)) files = message.get('files', []) if subtype == 'file_share': # In Slack messages, uploads can either have the subtype as 'file_share' or # have the upload information in 'files' keyword files = [message['file']] for fileinfo in files: url = fileinfo['url_private'] # For attachments with slack download link if 'files.slack.com' in url: has_attachment = has_link = True has_image = True if 'image' in fileinfo['mimetype'] else False file_user = [iterate_user for iterate_user in users if message['user'] == iterate_user['id']] file_user_email = get_user_email(file_user[0], domain_name) s3_path, content = get_attachment_path_and_content(fileinfo, realm_id) # construct attachments build_uploads(added_users[user], realm_id, file_user_email, fileinfo, s3_path, uploads_list) attachment_id = attachment_id_count build_attachment(realm_id, message_id, attachment_id, added_users[user], fileinfo, s3_path, zerver_attachment) attachment_id_count += 1 # For attachments with link not from slack # Example: Google drive integration else: has_link = True if 'title' in fileinfo: file_name = fileinfo['title'] else: file_name = fileinfo['name'] content = '[%s](%s)' % (file_name, fileinfo['url_private']) # construct message subject = 'imported from slack' zulip_message = build_message(subject, float(message['ts']), message_id, content, rendered_content, added_users[user], recipient_id, has_image, has_link, has_attachment) zerver_message.append(zulip_message) # construct usermessages usermessage_id_count = build_usermessages( zerver_usermessage, usermessage_id_count, zerver_subscription, recipient_id, mentioned_users_id, message_id) message_id_count += 1 id_list = (message_id_count, usermessage_id_count, reaction_id_count, attachment_id_count) return zerver_message, zerver_usermessage, zerver_attachment, uploads_list, \ reaction_list, id_list
def channel_message_to_zerver_message(realm_id: int, users: List[ZerverFieldsT], added_users: AddedUsersT, added_recipient: AddedRecipientsT, all_messages: List[ZerverFieldsT], zerver_realmemoji: List[ZerverFieldsT], subscriber_map: Dict[int, Set[int]], added_channels: AddedChannelsT, domain_name: str) -> Tuple[List[ZerverFieldsT], List[ZerverFieldsT], List[ZerverFieldsT], List[ZerverFieldsT], List[ZerverFieldsT]]: """ Returns: 1. zerver_message, which is a list of the messages 2. zerver_usermessage, which is a list of the usermessages 3. zerver_attachment, which is a list of the attachments 4. uploads_list, which is a list of uploads to be mapped in uploads records.json 5. reaction_list, which is a list of all user reactions """ zerver_message = [] zerver_usermessage = [] # type: List[ZerverFieldsT] uploads_list = [] # type: List[ZerverFieldsT] zerver_attachment = [] # type: List[ZerverFieldsT] reaction_list = [] # type: List[ZerverFieldsT] # For unicode emoji with open(NAME_TO_CODEPOINT_PATH) as fp: name_to_codepoint = ujson.load(fp) for message in all_messages: user = get_message_sending_user(message) if not user: # Ignore messages without user names # These are Sometimes produced by slack continue subtype = message.get('subtype', False) if subtype in [ # Zulip doesn't have a pinned_item concept "pinned_item", "unpinned_item", # Slack's channel join/leave notices are spammy "channel_join", "channel_leave", "channel_name" ]: continue try: content, mentioned_user_ids, has_link = convert_to_zulip_markdown( message['text'], users, added_channels, added_users) except Exception: print("Slack message unexpectedly missing text representation:") print(ujson.dumps(message, indent=4)) continue rendered_content = None recipient_id = added_recipient[message['channel_name']] message_id = NEXT_ID('message') # Process message reactions if 'reactions' in message.keys(): build_reactions(reaction_list, message['reactions'], added_users, message_id, name_to_codepoint, zerver_realmemoji) # Process different subtypes of slack messages # Subtypes which have only the action in the message should # be rendered with '/me' in the content initially # For example "sh_room_created" has the message 'started a call' # which should be displayed as '/me started a call' if subtype in ["bot_add", "sh_room_created", "me_message"]: content = ('/me %s' % (content)) if subtype == 'file_comment': # The file_comment message type only indicates the # responsible user in a subfield. message['user'] = message['comment']['user'] file_info = process_message_files( message=message, domain_name=domain_name, realm_id=realm_id, message_id=message_id, user=user, users=users, added_users=added_users, zerver_attachment=zerver_attachment, uploads_list=uploads_list, ) content += file_info['content'] has_link = has_link or file_info['has_link'] has_attachment = file_info['has_attachment'] has_image = file_info['has_image'] # construct message subject = 'imported from slack' zulip_message = build_message(subject, float(message['ts']), message_id, content, rendered_content, added_users[user], recipient_id, has_image, has_link, has_attachment) zerver_message.append(zulip_message) # construct usermessages build_usermessages( zerver_usermessage=zerver_usermessage, subscriber_map=subscriber_map, recipient_id=recipient_id, mentioned_user_ids=mentioned_user_ids, message_id=message_id, ) return zerver_message, zerver_usermessage, zerver_attachment, uploads_list, \ reaction_list
def channel_message_to_zerver_message( realm_id: int, users: List[ZerverFieldsT], slack_user_id_to_zulip_user_id: SlackToZulipUserIDT, slack_recipient_name_to_zulip_recipient_id: SlackToZulipRecipientT, all_messages: List[ZerverFieldsT], zerver_realmemoji: List[ZerverFieldsT], subscriber_map: Dict[int, Set[int]], added_channels: AddedChannelsT, dm_members: DMMembersT, domain_name: str, long_term_idle: Set[int] ) -> Tuple[List[ZerverFieldsT], List[ZerverFieldsT], List[ZerverFieldsT], List[ZerverFieldsT], List[ZerverFieldsT]]: """ Returns: 1. zerver_message, which is a list of the messages 2. zerver_usermessage, which is a list of the usermessages 3. zerver_attachment, which is a list of the attachments 4. uploads_list, which is a list of uploads to be mapped in uploads records.json 5. reaction_list, which is a list of all user reactions """ zerver_message = [] zerver_usermessage = [] # type: List[ZerverFieldsT] uploads_list = [] # type: List[ZerverFieldsT] zerver_attachment = [] # type: List[ZerverFieldsT] reaction_list = [] # type: List[ZerverFieldsT] # For unicode emoji with open(NAME_TO_CODEPOINT_PATH) as fp: name_to_codepoint = ujson.load(fp) total_user_messages = 0 total_skipped_user_messages = 0 for message in all_messages: slack_user_id = get_message_sending_user(message) if not slack_user_id: # Ignore messages without slack_user_id # These are Sometimes produced by slack continue subtype = message.get('subtype', False) if subtype in [ # Zulip doesn't have a pinned_item concept "pinned_item", "unpinned_item", # Slack's channel join/leave notices are spammy "channel_join", "channel_leave", "channel_name" ]: continue try: content, mentioned_user_ids, has_link = convert_to_zulip_markdown( message['text'], users, added_channels, slack_user_id_to_zulip_user_id) except Exception: print("Slack message unexpectedly missing text representation:") print(ujson.dumps(message, indent=4)) continue rendered_content = None if "channel_name" in message: is_private = False recipient_id = slack_recipient_name_to_zulip_recipient_id[ message['channel_name']] elif "mpim_name" in message: is_private = True recipient_id = slack_recipient_name_to_zulip_recipient_id[ message['mpim_name']] elif "pm_name" in message: is_private = True sender = get_message_sending_user(message) members = dm_members[message['pm_name']] if sender == members[0]: recipient_id = slack_recipient_name_to_zulip_recipient_id[ members[1]] sender_recipient_id = slack_recipient_name_to_zulip_recipient_id[ members[0]] else: recipient_id = slack_recipient_name_to_zulip_recipient_id[ members[0]] sender_recipient_id = slack_recipient_name_to_zulip_recipient_id[ members[1]] message_id = NEXT_ID('message') if 'reactions' in message.keys(): build_reactions(reaction_list, message['reactions'], slack_user_id_to_zulip_user_id, message_id, name_to_codepoint, zerver_realmemoji) # Process different subtypes of slack messages # Subtypes which have only the action in the message should # be rendered with '/me' in the content initially # For example "sh_room_created" has the message 'started a call' # which should be displayed as '/me started a call' if subtype in ["bot_add", "sh_room_created", "me_message"]: content = '/me %s' % (content, ) if subtype == 'file_comment': # The file_comment message type only indicates the # responsible user in a subfield. message['user'] = message['comment']['user'] file_info = process_message_files( message=message, domain_name=domain_name, realm_id=realm_id, message_id=message_id, slack_user_id=slack_user_id, users=users, slack_user_id_to_zulip_user_id=slack_user_id_to_zulip_user_id, zerver_attachment=zerver_attachment, uploads_list=uploads_list, ) content += file_info['content'] has_link = has_link or file_info['has_link'] has_attachment = file_info['has_attachment'] has_image = file_info['has_image'] topic_name = 'imported from slack' zulip_message = build_message( topic_name, float(message['ts']), message_id, content, rendered_content, slack_user_id_to_zulip_user_id[slack_user_id], recipient_id, has_image, has_link, has_attachment) zerver_message.append(zulip_message) (num_created, num_skipped) = build_usermessages( zerver_usermessage=zerver_usermessage, subscriber_map=subscriber_map, recipient_id=recipient_id, mentioned_user_ids=mentioned_user_ids, message_id=message_id, is_private=is_private, long_term_idle=long_term_idle, ) total_user_messages += num_created total_skipped_user_messages += num_skipped if "pm_name" in message and recipient_id != sender_recipient_id: (num_created, num_skipped) = build_usermessages( zerver_usermessage=zerver_usermessage, subscriber_map=subscriber_map, recipient_id=sender_recipient_id, mentioned_user_ids=mentioned_user_ids, message_id=message_id, is_private=is_private, long_term_idle=long_term_idle, ) total_user_messages += num_created total_skipped_user_messages += num_skipped logging.debug( "Created %s UserMessages; deferred %s due to long-term idle" % (total_user_messages, total_skipped_user_messages)) return zerver_message, zerver_usermessage, zerver_attachment, uploads_list, \ reaction_list
def channel_message_to_zerver_message( realm_id: int, users: List[ZerverFieldsT], added_users: AddedUsersT, added_recipient: AddedRecipientsT, all_messages: List[ZerverFieldsT], zerver_realmemoji: List[ZerverFieldsT], zerver_subscription: List[ZerverFieldsT], added_channels: AddedChannelsT, id_list: Tuple[int, int, int, int], domain_name: str ) -> Tuple[List[ZerverFieldsT], List[ZerverFieldsT], List[ZerverFieldsT], List[ZerverFieldsT], List[ZerverFieldsT], Tuple[int, int, int, int]]: """ Returns: 1. zerver_message, which is a list of the messages 2. zerver_usermessage, which is a list of the usermessages 3. zerver_attachment, which is a list of the attachments 4. uploads_list, which is a list of uploads to be mapped in uploads records.json 5. reaction_list, which is a list of all user reactions 6. id_list, which is a tuple of max ids of messages, usermessages, reactions and attachments """ message_id_count, usermessage_id_count, reaction_id_count, attachment_id_count = id_list zerver_message = [] zerver_usermessage = [] # type: List[ZerverFieldsT] uploads_list = [] # type: List[ZerverFieldsT] zerver_attachment = [] # type: List[ZerverFieldsT] reaction_list = [] # type: List[ZerverFieldsT] # For unicode emoji with open(NAME_TO_CODEPOINT_PATH) as fp: name_to_codepoint = ujson.load(fp) for message in all_messages: user = get_message_sending_user(message) if not user: # Ignore messages without user names # These are Sometimes produced by slack continue subtype = message.get('subtype', False) if subtype in [ # Zulip doesn't have a pinned_item concept "pinned_item", "unpinned_item", # Slack's channel join/leave notices are spammy "channel_join", "channel_leave", "channel_name" ]: continue has_attachment = has_image = False try: content, mentioned_users_id, has_link = convert_to_zulip_markdown( message['text'], users, added_channels, added_users) except Exception: print("Slack message unexpectedly missing text representation:") print(json.dumps(message, indent=4)) continue rendered_content = None recipient_id = added_recipient[message['channel_name']] message_id = message_id_count # Process message reactions if 'reactions' in message.keys(): reaction_id_count = build_reactions( reaction_list, message['reactions'], added_users, message_id, reaction_id_count, name_to_codepoint, zerver_realmemoji) # Process different subtypes of slack messages # Subtypes which have only the action in the message should # be rendered with '/me' in the content initially # For example "sh_room_created" has the message 'started a call' # which should be displayed as '/me started a call' if subtype in ["bot_add", "sh_room_created", "me_message"]: content = ('/me %s' % (content)) files = message.get('files', []) if subtype == 'file_share': # In Slack messages, uploads can either have the subtype as 'file_share' or # have the upload information in 'files' keyword files = [message['file']] for fileinfo in files: url = fileinfo['url_private'] # For attachments with slack download link if 'files.slack.com' in url: has_attachment = has_link = True has_image = True if 'image' in fileinfo['mimetype'] else False file_user = [ iterate_user for iterate_user in users if message['user'] == iterate_user['id'] ] file_user_email = get_user_email(file_user[0], domain_name) s3_path, content = get_attachment_path_and_content( fileinfo, realm_id) # construct attachments build_uploads(added_users[user], realm_id, file_user_email, fileinfo, s3_path, uploads_list) attachment_id = attachment_id_count build_attachment(realm_id, message_id, attachment_id, added_users[user], fileinfo, s3_path, zerver_attachment) attachment_id_count += 1 # For attachments with link not from slack # Example: Google drive integration else: has_link = True if 'title' in fileinfo: file_name = fileinfo['title'] else: file_name = fileinfo['name'] content = '[%s](%s)' % (file_name, fileinfo['url_private']) # construct message zulip_message = dict( sending_client=1, rendered_content_version=1, # This is Zulip-specific has_image=has_image, subject='imported from slack', # This is Zulip-specific pub_date=float(message['ts']), id=message_id, has_attachment= has_attachment, # attachment will be posted in the subsequent message; # this is how Slack does it, i.e. less like email edit_history=None, sender=added_users[user], # map slack id to zulip id content=content, rendered_content=rendered_content, # slack doesn't cache this recipient=recipient_id, last_edit_time=None, has_link=has_link) zerver_message.append(zulip_message) # construct usermessages usermessage_id_count = build_usermessages( zerver_usermessage, usermessage_id_count, zerver_subscription, recipient_id, mentioned_users_id, message_id) message_id_count += 1 id_list = (message_id_count, usermessage_id_count, reaction_id_count, attachment_id_count) return zerver_message, zerver_usermessage, zerver_attachment, uploads_list, \ reaction_list, id_list
def channel_message_to_zerver_message(realm_id: int, users: List[ZerverFieldsT], added_users: AddedUsersT, added_recipient: AddedRecipientsT, all_messages: List[ZerverFieldsT], zerver_realmemoji: List[ZerverFieldsT], subscriber_map: Dict[int, Set[int]], added_channels: AddedChannelsT, domain_name: str, long_term_idle: Set[int]) -> Tuple[List[ZerverFieldsT], List[ZerverFieldsT], List[ZerverFieldsT], List[ZerverFieldsT], List[ZerverFieldsT]]: """ Returns: 1. zerver_message, which is a list of the messages 2. zerver_usermessage, which is a list of the usermessages 3. zerver_attachment, which is a list of the attachments 4. uploads_list, which is a list of uploads to be mapped in uploads records.json 5. reaction_list, which is a list of all user reactions """ zerver_message = [] zerver_usermessage = [] # type: List[ZerverFieldsT] uploads_list = [] # type: List[ZerverFieldsT] zerver_attachment = [] # type: List[ZerverFieldsT] reaction_list = [] # type: List[ZerverFieldsT] # For unicode emoji with open(NAME_TO_CODEPOINT_PATH) as fp: name_to_codepoint = ujson.load(fp) total_user_messages = 0 total_skipped_user_messages = 0 for message in all_messages: user = get_message_sending_user(message) if not user: # Ignore messages without user names # These are Sometimes produced by slack continue subtype = message.get('subtype', False) if subtype in [ # Zulip doesn't have a pinned_item concept "pinned_item", "unpinned_item", # Slack's channel join/leave notices are spammy "channel_join", "channel_leave", "channel_name" ]: continue try: content, mentioned_user_ids, has_link = convert_to_zulip_markdown( message['text'], users, added_channels, added_users) except Exception: print("Slack message unexpectedly missing text representation:") print(ujson.dumps(message, indent=4)) continue rendered_content = None recipient_id = added_recipient[message['channel_name']] message_id = NEXT_ID('message') # Process message reactions if 'reactions' in message.keys(): build_reactions(reaction_list, message['reactions'], added_users, message_id, name_to_codepoint, zerver_realmemoji) # Process different subtypes of slack messages # Subtypes which have only the action in the message should # be rendered with '/me' in the content initially # For example "sh_room_created" has the message 'started a call' # which should be displayed as '/me started a call' if subtype in ["bot_add", "sh_room_created", "me_message"]: content = ('/me %s' % (content)) if subtype == 'file_comment': # The file_comment message type only indicates the # responsible user in a subfield. message['user'] = message['comment']['user'] file_info = process_message_files( message=message, domain_name=domain_name, realm_id=realm_id, message_id=message_id, user=user, users=users, added_users=added_users, zerver_attachment=zerver_attachment, uploads_list=uploads_list, ) content += file_info['content'] has_link = has_link or file_info['has_link'] has_attachment = file_info['has_attachment'] has_image = file_info['has_image'] # construct message topic_name = 'imported from slack' zulip_message = build_message(topic_name, float(message['ts']), message_id, content, rendered_content, added_users[user], recipient_id, has_image, has_link, has_attachment) zerver_message.append(zulip_message) # construct usermessages (num_created, num_skipped) = build_usermessages( zerver_usermessage=zerver_usermessage, subscriber_map=subscriber_map, recipient_id=recipient_id, mentioned_user_ids=mentioned_user_ids, message_id=message_id, long_term_idle=long_term_idle, ) total_user_messages += num_created total_skipped_user_messages += num_skipped logging.debug("Created %s UserMessages; deferred %s due to long-term idle" % ( total_user_messages, total_skipped_user_messages)) return zerver_message, zerver_usermessage, zerver_attachment, uploads_list, \ reaction_list