| 5f9a2c3c9f1deefa1da88dcf580bf28859b44aa2 |
|
22-Oct-2011 |
Daniel Lehmann <lehmannd@google.com> |
Match contacts if identities match.
Bug:5237405
Change-Id: I5def748ecd5ce7d9c847b06b5111217d5d577492
/packages/providers/ContactsProvider/src/com/android/providers/contacts/ContactMatcher.java
|
| d1f800d2712c9050f350bd940e3333c35f558e83 |
|
18-Oct-2011 |
Daniel Lehmann <lehmannd@google.com> |
Code cleanup. Fixes most of the IDE warnings
Change-Id: Ibeff9a520563f6e1fea3f409ada6f934e98e4814
/packages/providers/ContactsProvider/src/com/android/providers/contacts/ContactMatcher.java
|
| f266e8c568905337960b1fec5379841585af92a7 |
|
09-Dec-2010 |
Dmitri Plotnikov <dplotnikov@google.com> |
Excluding invisible contacts from aggregation
Bug: 3184559
Change-Id: If4b75548477cc296107d233fb6678c266fb0ca94
/packages/providers/ContactsProvider/src/com/android/providers/contacts/ContactMatcher.java
|
| 0dce6bf7a86a78d3073327419f17395c3a2d2688 |
|
12-Oct-2010 |
Dmitri Plotnikov <dplotnikov@google.com> |
Changing contact aggregation to not join from same account
Second attempt, now w/o an infinite loop.
The change addresses several requirements:
1. If "Michelle Lee" and "Michelle Lee" are two contacts
in the same account, we won't aggregate them.
2. If "Michelle Lee" and "Michelle Lee" are two contacts
from different accounts, they do get aggregated, but
if a third "Michelle Lee" shows up in one of those two
accounts - we bust the original aggregate.
3. If "Michelle Lee" and "Michelle Lee" are not aggregated
and a third "Michelle Lee shows up, it does not get
aggregated with either of the first two regardless of
the account.
4. Any manual joining overrides the above behavior.
5. The OTA with this change will bust all aggregates that
contain raw contacts from the same account.
Bug: 2650610
Change-Id: I413231af4cfa620f8d266a758c22bfc879aeae07
/packages/providers/ContactsProvider/src/com/android/providers/contacts/ContactMatcher.java
|
| 8638e1e999596ddfe12cbb529e58ca2345185af4 |
|
11-Oct-2010 |
Dmitri Plotnikov <dplotnikov@google.com> |
Revert "Changing contact aggregation to not join from same account"
This reverts commit 2a0d5f9c628e723261bc5198e0fd606076b76b74.
Bug: 3086301
Change-Id: I7f05eb14122687950ca7d22d591fc172c463045e
/packages/providers/ContactsProvider/src/com/android/providers/contacts/ContactMatcher.java
|
| 2a0d5f9c628e723261bc5198e0fd606076b76b74 |
|
11-Oct-2010 |
Dmitri Plotnikov <dplotnikov@google.com> |
Changing contact aggregation to not join from same account
The change addresses several requirements:
1. If "Michelle Lee" and "Michelle Lee" are two contacts
in the same account, we won't aggregate them.
2. If "Michelle Lee" and "Michelle Lee" are two contacts
from different accounts, they do get aggregated, but
if a third "Michelle Lee" shows up in one of those two
accounts - we bust the original aggregate.
3. If "Michelle Lee" and "Michelle Lee" are not aggregated
and a third "Michelle Lee shows up, it does not get
aggregated with either of the first two regardless of
the account.
4. Any manual joining overrides the above behavior.
5. The OTA with this change will bust all aggregates that
contain raw contacts from the same account.
Bug: 2650610
Change-Id: Idfba4dd330a4842310f3fed85e6ae0a3a634e95c
/packages/providers/ContactsProvider/src/com/android/providers/contacts/ContactMatcher.java
|
| 6581e1d13b9f310c2d7def2d2645f25158814107 |
|
09-Oct-2009 |
Dmitri Plotnikov <dplotnikov@google.com> |
[Issue 2176922] Raising aggregation suggestion threshold for email addresses from 0.82 to 0.95
This still does not require an exact match, but the difference should not exceed one or two characters.
Also, fixing the LegacyContactImporterTest (only the way test result is asserted is changing - not the test or the tested code)
Change-Id: I8d8e27f65a2e6ecaf73d50ded08b823090b31f02
/packages/providers/ContactsProvider/src/com/android/providers/contacts/ContactMatcher.java
|
| b38ed2c5ffeb20efc677b4a9229db4a00603aa8d |
|
29-Sep-2009 |
Dmitri Plotnikov <dplotnikov@google.com> |
Renaming OpenHelper to ContactsDatabaseHelper
Change-Id: Iccdf1ebcd8a3ab430941c957f712bb9e7dd2706c
/packages/providers/ContactsProvider/src/com/android/providers/contacts/ContactMatcher.java
|
| 2d89933b87a15ae5ed5d6b6ec4220ac085695ada |
|
22-Sep-2009 |
Dmitri Plotnikov <dplotnikov@google.com> |
[Issue 2097207] Adding support for filtering of contact aggregation suggestions
Change-Id: Iabfe6758d4d916bddd88cbc47a133e5bdfd1d722
/packages/providers/ContactsProvider/src/com/android/providers/contacts/ContactMatcher.java
|
| e1b656d725445663e443679af3fbedb5f4db3996 |
|
16-Sep-2009 |
Dmitri Plotnikov <dplotnikov@google.com> |
[Issue 2112473] Another round of optimizations in Contact Aggregator.
Pushing per-contact aggregation performance to about 20 ms.
Change-Id: I2f734feede5b5b1e5a762dcf44a9516822a4ae53
/packages/providers/ContactsProvider/src/com/android/providers/contacts/ContactMatcher.java
|
| c1778ef6fa53b6bf08fd715b3ad70c052c5f1ce9 |
|
10-Sep-2009 |
Dmitri Plotnikov <dplotnikov@google.com> |
[Issue 2102836] Fixing the false positive aggregating on multiple secondary matches
/packages/providers/ContactsProvider/src/com/android/providers/contacts/ContactMatcher.java
|
| 28245e1690abbab18b3224e4164a738f6e64d6c0 |
|
09-Sep-2009 |
Dmitri Plotnikov <dplotnikov@google.com> |
Making automatic contact aggregation even more conservative.
Also bringing back Jaro-Winkler distance calculator, which was emasculated by CL 22788
The Jaro-Winkler distance is actually very useful for generating
manual join suggestions. It will not be used for automatic aggregation.
Change-Id: I7844efbcc1e0f9265f9fc8c56af8772282879f5f
/packages/providers/ContactsProvider/src/com/android/providers/contacts/ContactMatcher.java
|
| cabac02a2416b495e030654accffcbb5ae526678 |
|
27-Aug-2009 |
Dmitri Plotnikov <dplotnikov@google.com> |
Solving aggregation of dissimilar names by parsing them explicitly.
/packages/providers/ContactsProvider/src/com/android/providers/contacts/ContactMatcher.java
|
| a276d23a70bdb8fdc7c8340a12a64e90696da115 |
|
26-Aug-2009 |
Dmitri Plotnikov <dplotnikov@google.com> |
Making name matching in aggregation much less lenient.
We are no longer using the Jaro-Winkler algorithm.
The new algorithm simply counts differences: character mismatches and transpositions
and aggressively caps the former and the total separately.
/packages/providers/ContactsProvider/src/com/android/providers/contacts/ContactMatcher.java
|
| 2a03d9c2bf627795963561a0f9406b23dc93fec5 |
|
15-Aug-2009 |
Dmitri Plotnikov <dplotnikov@google.com> |
Simplified and fixed aggregation on structured name.
Removed all of the over-engineered treatment of first name only,
last name only and similar use cases.
Now there are just three types of name matches:
1. Exact match
2. Variant match: first/last name flipped, Jeff/Jeffrey
3. Collation key match: this can be used for approximate matching
/packages/providers/ContactsProvider/src/com/android/providers/contacts/ContactMatcher.java
|
| d3fde755e73cd3912a488e7cb7d412d3c5f6ca94 |
|
28-Jul-2009 |
Dmitri Plotnikov <dplotnikov@google.com> |
ContactsProvider2: Renaming Aggregates to Contacts
/packages/providers/ContactsProvider/src/com/android/providers/contacts/ContactMatcher.java
|
| 28f8857b1b46bde18b85c6d3c2a63ac44c3c2e1c |
|
08-Jul-2009 |
Evan Millar <emillar@google.com> |
Merge commit 'goog/eclair-dev' into merge3
Merged the new contacts content provider into goog/master. The old and
new content providers now live side by side under separate authorities.
Conflicts:
Android.mk
AndroidManifest.xml
res/values/strings.xml
|