TheArchive/session-android
1
0
Fork 0
mirror of https://github.com/oxen-io/session-android.git synced 2024-11-28 12:35:17 +00:00
Code Issues Packages Projects Releases Wiki Activity
f3fdde6040
session-android/src/org/thoughtcrime/securesms/database/MmsDatabase.java

883 lines
37 KiB
Java
Raw Normal View History

Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
/**
Initial Project Import
2011-12-20 18:20:44 +00:00
* Copyright (C) 2011 Whisper Systems
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
*
Initial Project Import
2011-12-20 18:20:44 +00:00
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* This program 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.
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
*
Initial Project Import
2011-12-20 18:20:44 +00:00
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
package org.thoughtcrime.securesms.database;
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
import android.content.ContentValues;
import android.content.Context;
import android.database.Cursor;
import android.database.sqlite.SQLiteDatabase;
import android.database.sqlite.SQLiteOpenHelper;
import android.net.Uri;
import android.util.Log;
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
import android.util.Pair;
Initial Project Import
2011-12-20 18:20:44 +00:00
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
import org.thoughtcrime.securesms.R;
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
import org.thoughtcrime.securesms.contacts.ContactPhotoFactory;
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
import org.thoughtcrime.securesms.crypto.MasterCipher;
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
import org.thoughtcrime.securesms.crypto.MasterSecret;
Locally encrypted messages no longer show as ciphertext notifications.
2013-04-30 18:14:01 +00:00
import org.thoughtcrime.securesms.database.model.DisplayRecord;
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
import org.thoughtcrime.securesms.database.model.MediaMmsMessageRecord;
import org.thoughtcrime.securesms.database.model.MessageRecord;
import org.thoughtcrime.securesms.database.model.NotificationMmsMessageRecord;
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
import org.thoughtcrime.securesms.mms.PartParser;
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
import org.thoughtcrime.securesms.mms.SlideDeck;
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
import org.thoughtcrime.securesms.mms.TextSlide;
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
import org.thoughtcrime.securesms.recipients.Recipient;
Initial Project Import
2011-12-20 18:20:44 +00:00
import org.thoughtcrime.securesms.recipients.RecipientFactory;
import org.thoughtcrime.securesms.recipients.RecipientFormattingException;
import org.thoughtcrime.securesms.recipients.Recipients;
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
import org.thoughtcrime.securesms.util.InvalidMessageException;
Replace LinkedHashMap with a SoftReference LRUCache. Add Slide cache.
2013-05-06 19:22:03 +00:00
import org.thoughtcrime.securesms.util.LRUCache;
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
import org.thoughtcrime.securesms.util.ListenableFutureTask;
Support for auto-deleting old messages beyond a certain conversation thread length.
2013-01-10 05:06:56 +00:00
import org.thoughtcrime.securesms.util.Trimmer;
Sometimes querying the APN DB throws this exception as well.
2013-02-27 06:55:23 +00:00
import org.thoughtcrime.securesms.util.Util;
Initial Project Import
2011-12-20 18:20:44 +00:00
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
import java.io.UnsupportedEncodingException;
Replace LinkedHashMap with a SoftReference LRUCache. Add Slide cache.
2013-05-06 19:22:03 +00:00
import java.lang.ref.SoftReference;
import java.util.Collections;
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
import java.util.HashSet;
Replace LinkedHashMap with a SoftReference LRUCache. Add Slide cache.
2013-05-06 19:22:03 +00:00
import java.util.Map;
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
import java.util.Set;
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
import java.util.concurrent.Callable;
import java.util.concurrent.ExecutorService;
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
import ws.com.google.android.mms.InvalidHeaderValueException;
import ws.com.google.android.mms.MmsException;
import ws.com.google.android.mms.pdu.CharacterSets;
import ws.com.google.android.mms.pdu.EncodedStringValue;
import ws.com.google.android.mms.pdu.MultimediaMessagePdu;
import ws.com.google.android.mms.pdu.NotificationInd;
import ws.com.google.android.mms.pdu.PduBody;
import ws.com.google.android.mms.pdu.PduHeaders;
import ws.com.google.android.mms.pdu.RetrieveConf;
import ws.com.google.android.mms.pdu.SendReq;
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
// XXXX Clean up MMS efficiency:
// 1) We need to be careful about how much memory we're using for parts. SoftRefereences.
// 2) How many queries do we make? calling getMediaMessageForId() from within an existing query
// seems wasteful.
Initial Project Import
2011-12-20 18:20:44 +00:00
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
public class MmsDatabase extends Database implements MmsSmsColumns {
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
public static final String TABLE_NAME = "mms";
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
static final String DATE_SENT = "date";
static final String DATE_RECEIVED = "date_received";
Initial Project Import
2011-12-20 18:20:44 +00:00
public static final String MESSAGE_BOX = "msg_box";
private static final String MESSAGE_ID = "m_id";
private static final String SUBJECT = "sub";
private static final String SUBJECT_CHARSET = "sub_cs";
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
static final String CONTENT_TYPE = "ct_t";
static final String CONTENT_LOCATION = "ct_l";
static final String EXPIRY = "exp";
Initial Project Import
2011-12-20 18:20:44 +00:00
private static final String MESSAGE_CLASS = "m_cls";
public static final String MESSAGE_TYPE = "m_type";
private static final String MMS_VERSION = "v";
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
static final String MESSAGE_SIZE = "m_size";
Initial Project Import
2011-12-20 18:20:44 +00:00
private static final String PRIORITY = "pri";
private static final String READ_REPORT = "rr";
private static final String REPORT_ALLOWED = "rpt_a";
private static final String RESPONSE_STATUS = "resp_st";
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
static final String STATUS = "st";
static final String TRANSACTION_ID = "tr_id";
Initial Project Import
2011-12-20 18:20:44 +00:00
private static final String RETRIEVE_STATUS = "retr_st";
private static final String RETRIEVE_TEXT = "retr_txt";
private static final String RETRIEVE_TEXT_CS = "retr_txt_cs";
private static final String READ_STATUS = "read_status";
private static final String CONTENT_CLASS = "ct_cls";
private static final String RESPONSE_TEXT = "resp_txt";
private static final String DELIVERY_TIME = "d_tm";
private static final String DELIVERY_REPORT = "d_rpt";
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
static final String PART_COUNT = "part_count";
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
public static final String CREATE_TABLE = "CREATE TABLE " + TABLE_NAME + " (" + ID + " INTEGER PRIMARY KEY, " +
Fix for DB creation and migration with new date schema changes.
2013-01-07 02:47:20 +00:00
THREAD_ID + " INTEGER, " + DATE_SENT + " INTEGER, " + DATE_RECEIVED + " INTEGER, " + MESSAGE_BOX + " INTEGER, " +
Initial Project Import
2011-12-20 18:20:44 +00:00
READ + " INTEGER DEFAULT 0, " + MESSAGE_ID + " TEXT, " + SUBJECT + " TEXT, " +
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
SUBJECT_CHARSET + " INTEGER, " + BODY + " TEXT, " + PART_COUNT + " INTEGER, " +
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
CONTENT_TYPE + " TEXT, " + CONTENT_LOCATION + " TEXT, " + ADDRESS + " TEXT, " +
Initial Project Import
2011-12-20 18:20:44 +00:00
EXPIRY + " INTEGER, " + MESSAGE_CLASS + " TEXT, " + MESSAGE_TYPE + " INTEGER, " +
MMS_VERSION + " INTEGER, " + MESSAGE_SIZE + " INTEGER, " + PRIORITY + " INTEGER, " +
READ_REPORT + " INTEGER, " + REPORT_ALLOWED + " INTEGER, " + RESPONSE_STATUS + " INTEGER, " +
STATUS + " INTEGER, " + TRANSACTION_ID + " TEXT, " + RETRIEVE_STATUS + " INTEGER, " +
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
RETRIEVE_TEXT + " TEXT, " + RETRIEVE_TEXT_CS + " INTEGER, " + READ_STATUS + " INTEGER, " +
Initial Project Import
2011-12-20 18:20:44 +00:00
CONTENT_CLASS + " INTEGER, " + RESPONSE_TEXT + " TEXT, " + DELIVERY_TIME + " INTEGER, " +
DELIVERY_REPORT + " INTEGER);";
Create DB indexes for optimized queries.
2012-10-30 00:41:06 +00:00
public static final String[] CREATE_INDEXS = {
"CREATE INDEX IF NOT EXISTS mms_thread_id_index ON " + TABLE_NAME + " (" + THREAD_ID + ");",
"CREATE INDEX IF NOT EXISTS mms_read_index ON " + TABLE_NAME + " (" + READ + ");",
"CREATE INDEX IF NOT EXISTS mms_read_and_thread_id_index ON " + TABLE_NAME + "(" + READ + "," + THREAD_ID + ");",
"CREATE INDEX IF NOT EXISTS mms_message_box_index ON " + TABLE_NAME + " (" + MESSAGE_BOX + ");"
};
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
private static final String[] MMS_PROJECTION = new String[] {
ID, THREAD_ID, DATE_SENT + " * 1000 AS " + NORMALIZED_DATE_SENT,
DATE_RECEIVED + " * 1000 AS " + NORMALIZED_DATE_RECEIVED,
MESSAGE_BOX, READ, MESSAGE_ID, SUBJECT, SUBJECT_CHARSET, CONTENT_TYPE,
CONTENT_LOCATION, EXPIRY, MESSAGE_CLASS, MESSAGE_TYPE, MMS_VERSION,
MESSAGE_SIZE, PRIORITY, REPORT_ALLOWED, STATUS, TRANSACTION_ID, RETRIEVE_STATUS,
RETRIEVE_TEXT, RETRIEVE_TEXT_CS, READ_STATUS, CONTENT_CLASS, RESPONSE_TEXT,
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
DELIVERY_TIME, DELIVERY_REPORT, BODY, PART_COUNT, ADDRESS
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
};
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
public static final ExecutorService slideResolver = Util.newSingleThreadedLifoExecutor();
Replace LinkedHashMap with a SoftReference LRUCache. Add Slide cache.
2013-05-06 19:22:03 +00:00
private static final Map<Long, SoftReference<SlideDeck>> slideCache =
Collections.synchronizedMap(new LRUCache<Long, SoftReference<SlideDeck>>(20));
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
public MmsDatabase(Context context, SQLiteOpenHelper databaseHelper) {
super(context, databaseHelper);
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
public int getMessageCountForThread(long threadId) {
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
SQLiteDatabase db = databaseHelper.getReadableDatabase();
Initial Project Import
2011-12-20 18:20:44 +00:00
Cursor cursor = null;
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
try {
cursor = db.query(TABLE_NAME, new String[] {"COUNT(*)"}, THREAD_ID + " = ?", new String[] {threadId+""}, null, null, null);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
if (cursor != null && cursor.moveToFirst())
return cursor.getInt(0);
} finally {
if (cursor != null)
cursor.close();
}
return 0;
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
public long getThreadIdForMessage(long id) {
String sql = "SELECT " + THREAD_ID + " FROM " + TABLE_NAME + " WHERE " + ID + " = ?";
String[] sqlArgs = new String[] {id+""};
SQLiteDatabase db = databaseHelper.getReadableDatabase();
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
Cursor cursor = null;
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
try {
cursor = db.rawQuery(sql, sqlArgs);
if (cursor != null && cursor.moveToFirst())
return cursor.getLong(0);
else
return -1;
} finally {
if (cursor != null)
cursor.close();
}
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
First cut at group messaging refactor.
2013-04-26 01:59:49 +00:00
private long getThreadIdFor(RetrieveConf retrieved) throws RecipientFormattingException {
Initial Project Import
2011-12-20 18:20:44 +00:00
try {
First cut at group messaging refactor.
2013-04-26 01:59:49 +00:00
Set<String> group = new HashSet<String>();
EncodedStringValue encodedFrom = retrieved.getFrom();
group.add(new String(encodedFrom.getTextString(), CharacterSets.MIMENAME_ISO_8859_1));
EncodedStringValue[] encodedCcList = retrieved.getCc();
if (encodedCcList != null) {
for (EncodedStringValue encodedCc : encodedCcList) {
group.add(new String(encodedCc.getTextString(), CharacterSets.MIMENAME_ISO_8859_1));
}
}
StringBuilder sb = new StringBuilder();
for (String recipient : group) {
sb.append(recipient);
sb.append(",");
}
Recipients recipients = RecipientFactory.getRecipientsFromString(context, sb.toString(), false);
return DatabaseFactory.getThreadDatabase(context).getThreadIdFor(recipients);
} catch (UnsupportedEncodingException e) {
throw new AssertionError(e);
}
}
private long getThreadIdFor(NotificationInd notification) throws RecipientFormattingException {
try {
EncodedStringValue encodedString = notification.getFrom();
Initial Project Import
2011-12-20 18:20:44 +00:00
String fromString = new String(encodedString.getTextString(), CharacterSets.MIMENAME_ISO_8859_1);
Refactor recipient access. 1) Refactor recipient class to support asynchronous loading operations. 2) Refactor recipient factory to simplify recipient access. 3) Consoliate everything into one recipient provider that is capable of doing async lookups and intelligent caching.
2012-12-24 16:40:37 +00:00
Recipients recipients = RecipientFactory.getRecipientsFromString(context, fromString, false);
Initial Project Import
2011-12-20 18:20:44 +00:00
return DatabaseFactory.getThreadDatabase(context).getThreadIdFor(recipients);
} catch (UnsupportedEncodingException e) {
throw new AssertionError(e);
}
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
public void updateResponseStatus(long messageId, int status) {
SQLiteDatabase database = databaseHelper.getWritableDatabase();
ContentValues contentValues = new ContentValues();
contentValues.put(RESPONSE_STATUS, status);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Locally encrypted messages no longer show as ciphertext notifications.
2013-04-30 18:14:01 +00:00
database.update(TABLE_NAME, contentValues, ID_WHERE, new String[] {messageId + ""});
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
private void updateMailboxBitmask(long id, long maskOff, long maskOn) {
SQLiteDatabase db = databaseHelper.getWritableDatabase();
db.execSQL("UPDATE " + TABLE_NAME +
" SET " + MESSAGE_BOX + " = (" + MESSAGE_BOX + " & " + (Types.TOTAL_MASK - maskOff) + " | " + maskOn + " )" +
Locally encrypted messages no longer show as ciphertext notifications.
2013-04-30 18:14:01 +00:00
" WHERE " + ID + " = ?", new String[] {id + ""});
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Allow to resend a failed message Add an option to the context menu of the message to resend it if it has failed.
2013-10-13 10:53:41 +00:00
public void markAsSending(long messageId) {
updateMailboxBitmask(messageId, Types.BASE_TYPE_MASK, Types.BASE_SENDING_TYPE);
notifyConversationListeners(getThreadIdForMessage(messageId));
}
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
public void markAsSentFailed(long messageId) {
updateMailboxBitmask(messageId, Types.BASE_TYPE_MASK, Types.BASE_SENT_FAILED_TYPE);
Initial Project Import
2011-12-20 18:20:44 +00:00
notifyConversationListeners(getThreadIdForMessage(messageId));
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
public void markAsSent(long messageId, byte[] mmsId, long status) {
Initial Project Import
2011-12-20 18:20:44 +00:00
SQLiteDatabase database = databaseHelper.getWritableDatabase();
ContentValues contentValues = new ContentValues();
contentValues.put(RESPONSE_STATUS, status);
contentValues.put(MESSAGE_ID, new String(mmsId));
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
database.update(TABLE_NAME, contentValues, ID_WHERE, new String[] {messageId+""});
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
updateMailboxBitmask(messageId, Types.BASE_TYPE_MASK, Types.BASE_SENT_TYPE);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
notifyConversationListeners(getThreadIdForMessage(messageId));
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
public void markDownloadState(long messageId, long state) {
SQLiteDatabase database = databaseHelper.getWritableDatabase();
ContentValues contentValues = new ContentValues();
contentValues.put(STATUS, state);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Locally encrypted messages no longer show as ciphertext notifications.
2013-04-30 18:14:01 +00:00
database.update(TABLE_NAME, contentValues, ID_WHERE, new String[] {messageId + ""});
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
notifyConversationListeners(getThreadIdForMessage(messageId));
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
public void markAsNoSession(long messageId, long threadId) {
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
updateMailboxBitmask(messageId, Types.ENCRYPTION_MASK, Types.ENCRYPTION_REMOTE_NO_SESSION_BIT);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
notifyConversationListeners(threadId);
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
public void markAsDecryptFailed(long messageId, long threadId) {
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
updateMailboxBitmask(messageId, Types.ENCRYPTION_MASK, Types.ENCRYPTION_REMOTE_FAILED_BIT);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
notifyConversationListeners(threadId);
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
public void setMessagesRead(long threadId) {
SQLiteDatabase database = databaseHelper.getWritableDatabase();
ContentValues contentValues = new ContentValues();
contentValues.put(READ, 1);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
database.update(TABLE_NAME, contentValues, THREAD_ID + " = ?", new String[] {threadId+""});
}
Add 'mark all as read' option
2013-05-06 20:59:40 +00:00
public void setAllMessagesRead() {
SQLiteDatabase database = databaseHelper.getWritableDatabase();
ContentValues contentValues = new ContentValues();
contentValues.put(READ, 1);
database.update(TABLE_NAME, contentValues, null, null);
}
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
public SendReq[] getOutgoingMessages(MasterSecret masterSecret, long messageId)
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
throws MmsException
{
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
MmsAddressDatabase addr = DatabaseFactory.getMmsAddressDatabase(context);
PartDatabase parts = getPartDatabase(masterSecret);
SQLiteDatabase database = databaseHelper.getReadableDatabase();
MasterCipher masterCipher = masterSecret == null ? null : new MasterCipher(masterSecret);
Cursor cursor = null;
String selection;
String[] selectionArgs;
if (messageId > 0) {
selection = ID_WHERE;
selectionArgs = new String[]{messageId + ""};
} else {
selection = MESSAGE_BOX + " & " + Types.BASE_TYPE_MASK + " = ?";
selectionArgs = new String[]{Types.BASE_OUTBOX_TYPE + ""};
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
try {
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
cursor = database.query(TABLE_NAME, MMS_PROJECTION, selection, selectionArgs, null, null, null);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
if (cursor == null || cursor.getCount() == 0)
return new SendReq[0];
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
SendReq[] requests = new SendReq[cursor.getCount()];
int i = 0;
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
while (cursor.moveToNext()) {
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
messageId = cursor.getLong(cursor.getColumnIndexOrThrow(ID));
Removed tabs
2012-10-01 02:56:29 +00:00
long outboxType = cursor.getLong(cursor.getColumnIndexOrThrow(MESSAGE_BOX));
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
String messageText = cursor.getString(cursor.getColumnIndexOrThrow(BODY));
PduHeaders headers = getHeadersFromCursor(cursor);
Removed tabs
2012-10-01 02:56:29 +00:00
addr.getAddressesForId(messageId, headers);
PduBody body = parts.getParts(messageId, true);
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
try {
if (!Util.isEmpty(messageText) && Types.isSymmetricEncryption(outboxType)) {
body.addPart(new TextSlide(context, masterCipher.decryptBody(messageText)).getPart());
} else if (!Util.isEmpty(messageText)) {
body.addPart(new TextSlide(context, messageText).getPart());
}
} catch (InvalidMessageException e) {
Log.w("MmsDatabase", e);
}
requests[i++] = new SendReq(headers, body, messageId, outboxType);
Initial Project Import
2011-12-20 18:20:44 +00:00
}
return requests;
} finally {
if (cursor != null)
cursor.close();
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
}
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
private Pair<Long, Long> insertMessageInbox(MasterSecret masterSecret, RetrieveConf retrieved,
String contentLocation, long threadId, long mailbox)
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
throws MmsException
{
Initial Project Import
2011-12-20 18:20:44 +00:00
PduHeaders headers = retrieved.getPduHeaders();
ContentValues contentValues = getContentValuesFromHeader(headers);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
First cut at group messaging refactor.
2013-04-26 01:59:49 +00:00
if (!Util.isEmpty(retrieved.getCc())) {
try {
threadId = getThreadIdFor(retrieved);
} catch (RecipientFormattingException e) {
Log.w("MmsDatabase", e);
}
}
Initial Project Import
2011-12-20 18:20:44 +00:00
contentValues.put(MESSAGE_BOX, mailbox);
contentValues.put(THREAD_ID, threadId);
contentValues.put(CONTENT_LOCATION, contentLocation);
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
contentValues.put(STATUS, Status.DOWNLOAD_INITIALIZED);
Display both sent and received time in message details. 1) We record time sent in SMS database (date_sent). 2) We record time received in MMS database (date_received). 3) We union this information correctly in MmsSmsDatabase.
2013-01-06 21:13:14 +00:00
contentValues.put(DATE_RECEIVED, System.currentTimeMillis() / 1000);
if (!contentValues.containsKey(DATE_SENT))
contentValues.put(DATE_SENT, contentValues.getAsLong(DATE_RECEIVED));
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
long messageId = insertMediaMessage(masterSecret, retrieved, contentValues);
DatabaseFactory.getThreadDatabase(context).setUnread(threadId);
Fixed group in-thread notifications
2013-05-05 20:14:23 +00:00
DatabaseFactory.getThreadDatabase(context).update(threadId);
notifyConversationListeners(threadId);
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
Trimmer.trimThread(context, threadId);
Fixed group in-thread notifications
2013-05-05 20:14:23 +00:00
return new Pair<Long, Long>(messageId, threadId);
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
public Pair<Long, Long> insertMessageInbox(MasterSecret masterSecret, RetrieveConf retrieved,
String contentLocation, long threadId)
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
throws MmsException
{
return insertMessageInbox(masterSecret, retrieved, contentLocation, threadId,
Types.BASE_INBOX_TYPE | Types.ENCRYPTION_SYMMETRIC_BIT);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
}
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
public Pair<Long, Long> insertSecureMessageInbox(MasterSecret masterSecret, RetrieveConf retrieved,
String contentLocation, long threadId)
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
throws MmsException
{
return insertMessageInbox(masterSecret, retrieved, contentLocation, threadId,
Types.BASE_INBOX_TYPE | Types.SECURE_MESSAGE_BIT | Types.ENCRYPTION_REMOTE_BIT);
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
public Pair<Long, Long> insertSecureDecryptedMessageInbox(MasterSecret masterSecret,
RetrieveConf retrieved,
long threadId)
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
throws MmsException
{
return insertMessageInbox(masterSecret, retrieved, "", threadId,
Types.BASE_INBOX_TYPE | Types.SECURE_MESSAGE_BIT | Types.ENCRYPTION_SYMMETRIC_BIT);
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
public Pair<Long, Long> insertMessageInbox(NotificationInd notification) {
Initial Project Import
2011-12-20 18:20:44 +00:00
try {
SQLiteDatabase db = databaseHelper.getWritableDatabase();
PduHeaders headers = notification.getPduHeaders();
ContentValues contentValues = getContentValuesFromHeader(headers);
First cut at group messaging refactor.
2013-04-26 01:59:49 +00:00
long threadId = getThreadIdFor(notification);
Initial Project Import
2011-12-20 18:20:44 +00:00
MmsAddressDatabase addressDatabase = DatabaseFactory.getMmsAddressDatabase(context);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
Log.w("MmsDatabse", "Message received type: " + headers.getOctet(PduHeaders.MESSAGE_TYPE));
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
contentValues.put(MESSAGE_BOX, Types.BASE_INBOX_TYPE);
Initial Project Import
2011-12-20 18:20:44 +00:00
contentValues.put(THREAD_ID, threadId);
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
contentValues.put(STATUS, Status.DOWNLOAD_INITIALIZED);
Display both sent and received time in message details. 1) We record time sent in SMS database (date_sent). 2) We record time received in MMS database (date_received). 3) We union this information correctly in MmsSmsDatabase.
2013-01-06 21:13:14 +00:00
contentValues.put(DATE_RECEIVED, System.currentTimeMillis() / 1000);
if (!contentValues.containsKey(DATE_SENT))
contentValues.put(DATE_SENT, contentValues.getAsLong(DATE_RECEIVED));
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
long messageId = db.insert(TABLE_NAME, null, contentValues);
Initial Project Import
2011-12-20 18:20:44 +00:00
addressDatabase.insertAddressesForId(messageId, headers);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
// notifyConversationListeners(threadId);
// DatabaseFactory.getThreadDatabase(context).update(threadId);
// DatabaseFactory.getThreadDatabase(context).setUnread(threadId);
// Trimmer.trimThread(context, threadId);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
return new Pair<Long, Long>(messageId, threadId);
Initial Project Import
2011-12-20 18:20:44 +00:00
} catch (RecipientFormattingException rfe) {
Log.w("MmsDatabase", rfe);
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
return new Pair<Long, Long>(-1L, -1L);
Initial Project Import
2011-12-20 18:20:44 +00:00
}
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
public void markIncomingNotificationReceived(long threadId) {
notifyConversationListeners(threadId);
DatabaseFactory.getThreadDatabase(context).update(threadId);
DatabaseFactory.getThreadDatabase(context).setUnread(threadId);
Trimmer.trimThread(context, threadId);
}
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
public long insertMessageOutbox(MasterSecret masterSecret, SendReq sendRequest,
long threadId, boolean isSecure)
throws MmsException
{
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
long type = Types.BASE_OUTBOX_TYPE | Types.ENCRYPTION_SYMMETRIC_BIT;
Initial Project Import
2011-12-20 18:20:44 +00:00
PduHeaders headers = sendRequest.getPduHeaders();
ContentValues contentValues = getContentValuesFromHeader(headers);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
if (isSecure) {
type |= Types.SECURE_MESSAGE_BIT;
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
contentValues.put(MESSAGE_BOX, type);
Initial Project Import
2011-12-20 18:20:44 +00:00
contentValues.put(THREAD_ID, threadId);
contentValues.put(READ, 1);
Display both sent and received time in message details. 1) We record time sent in SMS database (date_sent). 2) We record time received in MMS database (date_received). 3) We union this information correctly in MmsSmsDatabase.
2013-01-06 21:13:14 +00:00
contentValues.put(DATE_RECEIVED, contentValues.getAsLong(DATE_SENT));
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
contentValues.remove(ADDRESS);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
long messageId = insertMediaMessage(masterSecret, sendRequest, contentValues);
Support for auto-deleting old messages beyond a certain conversation thread length.
2013-01-10 05:06:56 +00:00
Trimmer.trimThread(context, threadId);
Initial Project Import
2011-12-20 18:20:44 +00:00
return messageId;
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
private long insertMediaMessage(MasterSecret masterSecret,
MultimediaMessagePdu message,
ContentValues contentValues)
throws MmsException
{
Initial Project Import
2011-12-20 18:20:44 +00:00
SQLiteDatabase db = databaseHelper.getWritableDatabase();
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
PartDatabase partsDatabase = getPartDatabase(masterSecret);
Initial Project Import
2011-12-20 18:20:44 +00:00
MmsAddressDatabase addressDatabase = DatabaseFactory.getMmsAddressDatabase(context);
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
PduBody body = message.getBody();
if (Types.isSymmetricEncryption(contentValues.getAsLong(MESSAGE_BOX))) {
String messageText = PartParser.getMessageText(body);
body = PartParser.getNonTextParts(body);
if (!Util.isEmpty(messageText)) {
contentValues.put(BODY, new MasterCipher(masterSecret).encryptBody(messageText));
}
}
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
contentValues.put(PART_COUNT, PartParser.getDisplayablePartCount(body));
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
long messageId = db.insert(TABLE_NAME, null, contentValues);
Initial Project Import
2011-12-20 18:20:44 +00:00
addressDatabase.insertAddressesForId(messageId, message.getPduHeaders());
partsDatabase.insertParts(messageId, body);
notifyConversationListeners(contentValues.getAsLong(THREAD_ID));
DatabaseFactory.getThreadDatabase(context).update(contentValues.getAsLong(THREAD_ID));
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
return messageId;
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
public void delete(long messageId) {
long threadId = getThreadIdForMessage(messageId);
MmsAddressDatabase addrDatabase = DatabaseFactory.getMmsAddressDatabase(context);
PartDatabase partDatabase = DatabaseFactory.getPartDatabase(context);
partDatabase.deleteParts(messageId);
addrDatabase.deleteAddressesForId(messageId);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
SQLiteDatabase database = databaseHelper.getWritableDatabase();
database.delete(TABLE_NAME, ID_WHERE, new String[] {messageId+""});
DatabaseFactory.getThreadDatabase(context).update(threadId);
notifyConversationListeners(threadId);
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
public void deleteThread(long threadId) {
Set<Long> singleThreadSet = new HashSet<Long>();
singleThreadSet.add(threadId);
deleteThreads(singleThreadSet);
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
/*package*/ void deleteThreads(Set<Long> threadIds) {
SQLiteDatabase db = databaseHelper.getWritableDatabase();
String where = "";
Cursor cursor = null;
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
for (long threadId : threadIds) {
where += THREAD_ID + " = '" + threadId + "' OR ";
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
where = where.substring(0, where.length() - 4);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
try {
cursor = db.query(TABLE_NAME, new String[] {ID}, where, null, null, null, null);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
while (cursor != null && cursor.moveToNext()) {
delete(cursor.getLong(0));
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
} finally {
if (cursor != null)
cursor.close();
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
}
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Support for auto-deleting old messages beyond a certain conversation thread length.
2013-01-10 05:06:56 +00:00
/*package*/void deleteMessagesInThreadBeforeDate(long threadId, long date) {
date = date / 1000;
Cursor cursor = null;
try {
SQLiteDatabase db = databaseHelper.getReadableDatabase();
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
String where = THREAD_ID + " = ? AND (CASE (" + MESSAGE_BOX + " & " + Types.BASE_TYPE_MASK + ") ";
Support for auto-deleting old messages beyond a certain conversation thread length.
2013-01-10 05:06:56 +00:00
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
for (long outgoingType : Types.OUTGOING_MESSAGE_TYPES) {
Support for auto-deleting old messages beyond a certain conversation thread length.
2013-01-10 05:06:56 +00:00
where += " WHEN " + outgoingType + " THEN " + DATE_SENT + " < " + date;
}
where += (" ELSE " + DATE_RECEIVED + " < " + date + " END)");
Log.w("MmsDatabase", "Executing trim query: " + where);
cursor = db.query(TABLE_NAME, new String[] {ID}, where, new String[] {threadId+""}, null, null, null);
while (cursor != null && cursor.moveToNext()) {
Log.w("MmsDatabase", "Trimming: " + cursor.getLong(0));
delete(cursor.getLong(0));
}
} finally {
if (cursor != null)
cursor.close();
}
}
Initial Project Import
2011-12-20 18:20:44 +00:00
public void deleteAllThreads() {
DatabaseFactory.getPartDatabase(context).deleteAllParts();
DatabaseFactory.getMmsAddressDatabase(context).deleteAllAddresses();
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
SQLiteDatabase database = databaseHelper.getWritableDatabase();
database.delete(TABLE_NAME, null, null);
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial fixes for MMS retrieval. 1) Parse the APN information based on what the ConnectionManager tells us. 2) Accept email addresses as a valid Recipient format.
2012-09-30 18:46:45 +00:00
public Cursor getCarrierMmsInformation(String apn) {
We need READ_CALL_LOG if targetSDK is > 15. What a drag.
2012-11-21 03:09:46 +00:00
Uri uri = Uri.withAppendedPath(Uri.parse("content://telephony/carriers"), "current");
Sometimes querying the APN DB throws this exception as well.
2013-02-27 06:55:23 +00:00
String selection = Util.isEmpty(apn) ? null : "apn = ?";
String[] selectionArgs = Util.isEmpty(apn) ? null : new String[] {apn.trim()};
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Catch NPE thrown by APN provider.
2013-03-17 18:19:36 +00:00
try {
return context.getContentResolver().query(uri, null, selection, selectionArgs, null);
} catch (NullPointerException npe) {
// NOTE - This is dumb, but on some devices there's an NPE in the Android framework
// for the provider of this call, which gets rethrown back to here through a binder
// call.
throw new IllegalArgumentException(npe);
}
Initial Project Import
2011-12-20 18:20:44 +00:00
}
private PduHeaders getHeadersFromCursor(Cursor cursor) throws InvalidHeaderValueException {
PduHeaders headers = new PduHeaders();
PduHeadersBuilder phb = new PduHeadersBuilder(headers, cursor);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
phb.add(RETRIEVE_TEXT, RETRIEVE_TEXT_CS, PduHeaders.RETRIEVE_TEXT);
phb.add(SUBJECT, SUBJECT_CHARSET, PduHeaders.SUBJECT);
phb.addText(CONTENT_LOCATION, PduHeaders.CONTENT_LOCATION);
phb.addText(CONTENT_TYPE, PduHeaders.CONTENT_TYPE);
phb.addText(MESSAGE_CLASS, PduHeaders.MESSAGE_CLASS);
phb.addText(MESSAGE_ID, PduHeaders.MESSAGE_ID);
phb.addText(RESPONSE_TEXT, PduHeaders.RESPONSE_TEXT);
phb.addText(TRANSACTION_ID, PduHeaders.TRANSACTION_ID);
phb.addOctet(CONTENT_CLASS, PduHeaders.CONTENT_CLASS);
phb.addOctet(DELIVERY_REPORT, PduHeaders.DELIVERY_REPORT);
phb.addOctet(MESSAGE_TYPE, PduHeaders.MESSAGE_TYPE);
phb.addOctet(MMS_VERSION, PduHeaders.MMS_VERSION);
phb.addOctet(PRIORITY, PduHeaders.PRIORITY);
phb.addOctet(READ_STATUS, PduHeaders.READ_STATUS);
phb.addOctet(REPORT_ALLOWED, PduHeaders.REPORT_ALLOWED);
phb.addOctet(RETRIEVE_STATUS, PduHeaders.RETRIEVE_STATUS);
phb.addOctet(STATUS, PduHeaders.STATUS);
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
phb.addLong(NORMALIZED_DATE_SENT, PduHeaders.DATE);
Initial Project Import
2011-12-20 18:20:44 +00:00
phb.addLong(DELIVERY_TIME, PduHeaders.DELIVERY_TIME);
phb.addLong(EXPIRY, PduHeaders.EXPIRY);
phb.addLong(MESSAGE_SIZE, PduHeaders.MESSAGE_SIZE);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
headers.setLongInteger(headers.getLongInteger(PduHeaders.DATE) / 1000L, PduHeaders.DATE);
return headers;
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
private ContentValues getContentValuesFromHeader(PduHeaders headers) {
ContentValues contentValues = new ContentValues();
ContentValuesBuilder cvb = new ContentValuesBuilder(contentValues);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
cvb.add(RETRIEVE_TEXT, RETRIEVE_TEXT_CS, headers.getEncodedStringValue(PduHeaders.RETRIEVE_TEXT));
cvb.add(SUBJECT, SUBJECT_CHARSET, headers.getEncodedStringValue(PduHeaders.SUBJECT));
cvb.add(CONTENT_LOCATION, headers.getTextString(PduHeaders.CONTENT_LOCATION));
cvb.add(CONTENT_TYPE, headers.getTextString(PduHeaders.CONTENT_TYPE));
cvb.add(MESSAGE_CLASS, headers.getTextString(PduHeaders.MESSAGE_CLASS));
cvb.add(MESSAGE_ID, headers.getTextString(PduHeaders.MESSAGE_ID));
cvb.add(RESPONSE_TEXT, headers.getTextString(PduHeaders.RESPONSE_TEXT));
cvb.add(TRANSACTION_ID, headers.getTextString(PduHeaders.TRANSACTION_ID));
cvb.add(CONTENT_CLASS, headers.getOctet(PduHeaders.CONTENT_CLASS));
cvb.add(DELIVERY_REPORT, headers.getOctet(PduHeaders.DELIVERY_REPORT));
cvb.add(MESSAGE_TYPE, headers.getOctet(PduHeaders.MESSAGE_TYPE));
cvb.add(MMS_VERSION, headers.getOctet(PduHeaders.MMS_VERSION));
cvb.add(PRIORITY, headers.getOctet(PduHeaders.PRIORITY));
cvb.add(READ_REPORT, headers.getOctet(PduHeaders.READ_REPORT));
cvb.add(READ_STATUS, headers.getOctet(PduHeaders.READ_STATUS));
cvb.add(REPORT_ALLOWED, headers.getOctet(PduHeaders.REPORT_ALLOWED));
cvb.add(RETRIEVE_STATUS, headers.getOctet(PduHeaders.RETRIEVE_STATUS));
cvb.add(STATUS, headers.getOctet(PduHeaders.STATUS));
Display both sent and received time in message details. 1) We record time sent in SMS database (date_sent). 2) We record time received in MMS database (date_received). 3) We union this information correctly in MmsSmsDatabase.
2013-01-06 21:13:14 +00:00
cvb.add(DATE_SENT, headers.getLongInteger(PduHeaders.DATE));
Initial Project Import
2011-12-20 18:20:44 +00:00
cvb.add(DELIVERY_TIME, headers.getLongInteger(PduHeaders.DELIVERY_TIME));
cvb.add(EXPIRY, headers.getLongInteger(PduHeaders.EXPIRY));
cvb.add(MESSAGE_SIZE, headers.getLongInteger(PduHeaders.MESSAGE_SIZE));
Fix for NPE on null FROM in MMS header.
2013-05-31 03:29:13 +00:00
if (headers.getEncodedStringValue(PduHeaders.FROM) != null)
cvb.add(ADDRESS, headers.getEncodedStringValue(PduHeaders.FROM).getTextString());
else
cvb.add(ADDRESS, null);
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
return cvb.getContentValues();
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
protected PartDatabase getPartDatabase(MasterSecret masterSecret) {
if (masterSecret == null)
return DatabaseFactory.getPartDatabase(context);
else
return DatabaseFactory.getEncryptingPartDatabase(context, masterSecret);
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
public Reader readerFor(MasterSecret masterSecret, Cursor cursor) {
return new Reader(masterSecret, cursor);
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
public static class Status {
Initial Project Import
2011-12-20 18:20:44 +00:00
public static final int DOWNLOAD_INITIALIZED = 1;
public static final int DOWNLOAD_NO_CONNECTIVITY = 2;
public static final int DOWNLOAD_CONNECTING = 3;
Removed tabs
2012-10-01 02:56:29 +00:00
public static final int DOWNLOAD_SOFT_FAILURE = 4;
public static final int DOWNLOAD_HARD_FAILURE = 5;
Initial Project Import
2011-12-20 18:20:44 +00:00
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
public static boolean isDisplayDownloadButton(int status) {
return
status == DOWNLOAD_INITIALIZED ||
status == DOWNLOAD_NO_CONNECTIVITY ||
status == DOWNLOAD_SOFT_FAILURE;
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
public static String getLabelForStatus(Context context, int status) {
switch (status) {
case DOWNLOAD_CONNECTING: return context.getString(R.string.MmsDatabase_connecting_to_mms_server);
case DOWNLOAD_INITIALIZED: return context.getString(R.string.MmsDatabase_downloading_mms);
case DOWNLOAD_HARD_FAILURE: return context.getString(R.string.MmsDatabase_mms_download_failed);
Support for auto-deleting old messages beyond a certain conversation thread length.
2013-01-10 05:06:56 +00:00
}
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
return context.getString(R.string.MmsDatabase_downloading);
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
public static boolean isHardError(int status) {
return status == DOWNLOAD_HARD_FAILURE;
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
public class Reader {
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
private final Cursor cursor;
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
private final MasterSecret masterSecret;
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
private final MasterCipher masterCipher;
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
public Reader(MasterSecret masterSecret, Cursor cursor) {
this.cursor = cursor;
this.masterSecret = masterSecret;
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
if (masterSecret != null) masterCipher = new MasterCipher(masterSecret);
else masterCipher = null;
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
public MessageRecord getNext() {
if (cursor == null || !cursor.moveToNext())
return null;
return getCurrent();
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
public MessageRecord getCurrent() {
long mmsType = cursor.getLong(cursor.getColumnIndexOrThrow(MmsDatabase.MESSAGE_TYPE));
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
if (mmsType == PduHeaders.MESSAGE_TYPE_NOTIFICATION_IND) {
return getNotificationMmsMessageRecord(cursor);
} else {
return getMediaMmsMessageRecord(cursor);
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
}
Extract TextSecure strings for i18n. 1) Change all instances which use concatenation to build strings with variables in them to use string formatting instead. 2) Extract all string literals from layouts and menus into strings.xml 3) Extract all string literals from code into strings.xml
2012-09-08 03:03:23 +00:00
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
private NotificationMmsMessageRecord getNotificationMmsMessageRecord(Cursor cursor) {
long id = cursor.getLong(cursor.getColumnIndexOrThrow(MmsDatabase.ID));
long dateSent = cursor.getLong(cursor.getColumnIndexOrThrow(MmsDatabase.NORMALIZED_DATE_SENT));
long dateReceived = cursor.getLong(cursor.getColumnIndexOrThrow(MmsDatabase.NORMALIZED_DATE_RECEIVED));
long threadId = cursor.getLong(cursor.getColumnIndexOrThrow(MmsDatabase.THREAD_ID));
long mailbox = cursor.getLong(cursor.getColumnIndexOrThrow(MmsDatabase.MESSAGE_BOX));
String address = cursor.getString(cursor.getColumnIndexOrThrow(MmsDatabase.ADDRESS));
Recipients recipients = getRecipientsFor(address);
String contentLocation = cursor.getString(cursor.getColumnIndexOrThrow(MmsDatabase.CONTENT_LOCATION));
String transactionId = cursor.getString(cursor.getColumnIndexOrThrow(MmsDatabase.TRANSACTION_ID));
long messageSize = cursor.getLong(cursor.getColumnIndexOrThrow(MmsDatabase.MESSAGE_SIZE));
long expiry = cursor.getLong(cursor.getColumnIndexOrThrow(MmsDatabase.EXPIRY));
int status = cursor.getInt(cursor.getColumnIndexOrThrow(MmsDatabase.STATUS));
byte[]contentLocationBytes = null;
byte[]transactionIdBytes = null;
if (!Util.isEmpty(contentLocation))
contentLocationBytes = Util.toIsoBytes(contentLocation);
if (!Util.isEmpty(transactionId))
transactionIdBytes = Util.toIsoBytes(transactionId);
return new NotificationMmsMessageRecord(context, id, recipients, recipients.getPrimaryRecipient(),
dateSent, dateReceived, threadId, contentLocationBytes,
messageSize, expiry, status, transactionIdBytes, mailbox);
}
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
private MediaMmsMessageRecord getMediaMmsMessageRecord(Cursor cursor) {
Locally encrypted messages no longer show as ciphertext notifications.
2013-04-30 18:14:01 +00:00
long id = cursor.getLong(cursor.getColumnIndexOrThrow(MmsDatabase.ID));
long dateSent = cursor.getLong(cursor.getColumnIndexOrThrow(MmsDatabase.NORMALIZED_DATE_SENT));
long dateReceived = cursor.getLong(cursor.getColumnIndexOrThrow(MmsDatabase.NORMALIZED_DATE_RECEIVED));
long box = cursor.getLong(cursor.getColumnIndexOrThrow(MmsDatabase.MESSAGE_BOX));
long threadId = cursor.getLong(cursor.getColumnIndexOrThrow(MmsDatabase.THREAD_ID));
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
String address = cursor.getString(cursor.getColumnIndexOrThrow(MmsDatabase.ADDRESS));
Locally encrypted messages no longer show as ciphertext notifications.
2013-04-30 18:14:01 +00:00
DisplayRecord.Body body = getBody(cursor);
int partCount = cursor.getInt(cursor.getColumnIndexOrThrow(MmsDatabase.PART_COUNT));
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
Recipients recipients = getRecipientsFor(address);
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
ListenableFutureTask<SlideDeck> slideDeck = getSlideDeck(masterSecret, id);
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
return new MediaMmsMessageRecord(context, id, recipients, recipients.getPrimaryRecipient(),
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
dateSent, dateReceived, threadId, body,
slideDeck, partCount, box);
}
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
private Recipients getRecipientsFor(String address) {
try {
Filter MMS weirdness
2013-05-06 19:33:42 +00:00
if (Util.isEmpty(address) || address.equals("insert-address-token")) {
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
return new Recipients(new Recipient("Unknown", "Unknown", null,
ContactPhotoFactory.getDefaultContactPhoto(context)));
}
Recipients recipients = RecipientFactory.getRecipientsFromString(context, address, false);
if (recipients == null || recipients.isEmpty()) {
return new Recipients(new Recipient("Unknown", "Unknown", null,
ContactPhotoFactory.getDefaultContactPhoto(context)));
}
return recipients;
} catch (RecipientFormattingException e) {
Log.w("MmsDatabase", e);
return new Recipients(new Recipient("Unknown", "Unknown", null,
ContactPhotoFactory.getDefaultContactPhoto(context)));
}
}
Locally encrypted messages no longer show as ciphertext notifications.
2013-04-30 18:14:01 +00:00
private DisplayRecord.Body getBody(Cursor cursor) {
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
try {
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
String body = cursor.getString(cursor.getColumnIndexOrThrow(MmsDatabase.BODY));
long box = cursor.getLong(cursor.getColumnIndexOrThrow(MmsDatabase.MESSAGE_BOX));
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
Locally encrypted messages no longer show as ciphertext notifications.
2013-04-30 18:14:01 +00:00
if (!Util.isEmpty(body) && masterCipher != null && Types.isSymmetricEncryption(box)) {
return new DisplayRecord.Body(masterCipher.decryptBody(body), true);
} else if (!Util.isEmpty(body) && masterCipher == null && Types.isSymmetricEncryption(box)) {
return new DisplayRecord.Body(body, false);
} else {
return new DisplayRecord.Body(body == null ? "" : body, true);
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
}
} catch (InvalidMessageException e) {
Log.w("MmsDatabase", e);
Locally encrypted messages no longer show as ciphertext notifications.
2013-04-30 18:14:01 +00:00
return new DisplayRecord.Body("Error decrypting message.", true);
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
}
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
private ListenableFutureTask<SlideDeck> getSlideDeck(final MasterSecret masterSecret,
final long id)
{
Replace LinkedHashMap with a SoftReference LRUCache. Add Slide cache.
2013-05-06 19:22:03 +00:00
ListenableFutureTask<SlideDeck> future = getCachedSlideDeck(id);
if (future != null) {
return future;
}
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
Callable<SlideDeck> task = new Callable<SlideDeck>() {
@Override
public SlideDeck call() throws Exception {
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
if (masterSecret == null)
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
return null;
MMS and Group Messaging Improvements 1) Display the individual sender name in a group conversation. 2) Add an "address" column to MmsDatabase and keep FROM there. 3) Remove all blocking operations from MmsDatabase.Reader path. 4) Strip SMIL and other undisplayable parts from part count. 5) Fix places where messages weren't being correctly decrypted.
2013-05-05 19:51:36 +00:00
PduBody body = getPartDatabase(masterSecret).getParts(id, false);
Replace LinkedHashMap with a SoftReference LRUCache. Add Slide cache.
2013-05-06 19:22:03 +00:00
SlideDeck slideDeck = new SlideDeck(context, masterSecret, body);
slideCache.put(id, new SoftReference<SlideDeck>(slideDeck));
return slideDeck;
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
}
};
Replace LinkedHashMap with a SoftReference LRUCache. Add Slide cache.
2013-05-06 19:22:03 +00:00
future = new ListenableFutureTask<SlideDeck>(task, null);
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
slideResolver.execute(future);
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
Make MMS more asynchronous and consistent with new SMS types. 1) We now delay MMS notifications until a payload is received, or there's an error downloading the payload. This makes group messages more consistent. 2) All "text" parts of an MMS are combined into a second text record, which is stored in the MMS row directly rather than as a distinct part. This allows for immediate text loading, which means there's no chance a ConversationItem will resize. To do this, we need to include MMS in the big DB migration that's already staged for this application update. It's also an "application-level" migration, because we need the MasterSecret to do it. 3) On conversation display, all image-based parts now have their thumbnails loaded asynchronously. This allows for smooth-scrolling. The thumbnails are also scaled more accurately.
2013-04-26 18:23:43 +00:00
return future;
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
}
Replace LinkedHashMap with a SoftReference LRUCache. Add Slide cache.
2013-05-06 19:22:03 +00:00
private ListenableFutureTask<SlideDeck> getCachedSlideDeck(final long id) {
SoftReference<SlideDeck> reference = slideCache.get(id);
if (reference != null) {
final SlideDeck slideDeck = reference.get();
if (slideDeck != null) {
Callable<SlideDeck> task = new Callable<SlideDeck>() {
@Override
public SlideDeck call() throws Exception {
return slideDeck;
}
};
ListenableFutureTask<SlideDeck> future = new ListenableFutureTask<SlideDeck>(task, null);
future.run();
return future;
}
}
return null;
}
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
public void close() {
cursor.close();
Initial Project Import
2011-12-20 18:20:44 +00:00
}
}
Major storage layer refactoring to set the stage for clean GCM. 1) We now try to hand out cursors at a minimum. There has always been a fairly clean insertion layer that handles encrypting message bodies, but the process of decrypting message bodies has always been less than ideal. Here we introduce a "Reader" interface that will decrypt message bodies when appropriate and return objects that encapsulate record state. No more MessageDisplayHelper. The MmsSmsDatabase interface is also more sane. 2) We finally rid ourselves of the technical debt associated with TextSecure's initial usage of the default SMS DB. In that world, we weren't able to use anything other than the default "Inbox, Outbox, Sent" types to describe a message, and had to overload the message content itself with a set of local "prefixes" to describe what it was (encrypted, asymetric encrypted, remote encrypted, a key exchange, procssed key exchange), and so on. This includes a major schema update that transforms the "type" field into a bitmask that describes everything that used to be encoded in a prefix, and prefixes have been completely eliminated from the system. No more Prefix.java 3) Refactoring of the MultipartMessageHandler code. It's less of a mess, and hopefully more clear as to what's going on. The next step is to remove what we can from SmsTransportDetails and genericize that interface for a GCM equivalent.
2013-04-20 19:22:04 +00:00
Initial Project Import
2011-12-20 18:20:44 +00:00
}
Reference in New Issue Copy Permalink