Commit graph

130 commits

Author SHA1 Message Date
Jean Chalard
46fe49fc05 Add a comment on a caveat for future reference.
Change-Id: I328a0cd4346275aac960e1369bf370688a004e11
2012-05-29 16:50:25 +09:00
Jean Chalard
18f650172d am bfba64bc: Merge "Compute the correct frequency for bigram prediction" into jb-dev
* commit 'bfba64bcfd70ce8e3932d52c4b9f3182d33e1f55':
  Compute the correct frequency for bigram prediction
2012-05-29 00:38:16 -07:00
Jean Chalard
e382bcc39a am 8f9f55dd: Merge "Fix a bug where the bigram freq would be underevaluated" into jb-dev
* commit '8f9f55dd9d27f6c759c12416c872138b9b4661dc':
  Fix a bug where the bigram freq would be underevaluated
2012-05-29 00:25:50 -07:00
Jean Chalard
c941bc9bd3 am 22cf6958: Merge "Split a method to reconstruct freq from uni/bi freq" into jb-dev
* commit '22cf695834009929a6c0fbcd75a9ff56efa04d83':
  Split a method to reconstruct freq from uni/bi freq
2012-05-29 00:25:49 -07:00
Jean Chalard
ca134ccc2a am bf61bba3: Merge "Fix two small possible bugs." into jb-dev
* commit 'bf61bba35feda09a2ad08475481ce042879ff5bb':
  Fix two small possible bugs.
2012-05-29 00:25:47 -07:00
Jean Chalard
e308459531 Compute the correct frequency for bigram prediction
Change-Id: I3196f48a0ca2ed5e94f430254d58e65d341398c8
2012-05-29 16:22:46 +09:00
Jean Chalard
cb99376307 Fix a bug where the bigram freq would be underevaluated
The difference in score is not large, but it's still a bug

Change-Id: Ie22c2b6e1206e829c1c8af096469df05af14d47b
2012-05-29 16:04:07 +09:00
Jean Chalard
19ebd93646 Split a method to reconstruct freq from uni/bi freq
This has no impact at all on the logic.

Change-Id: I3788c8335cc193433ad9a7512b211a49bb2ffb02
2012-05-29 16:00:25 +09:00
Jean Chalard
402b057050 Fix two small possible bugs.
None of these had any real impact, but they were potential
liabilities for the future

Change-Id: I2de581f8b638e423d47a6d99b1a3c96af4c6150d
2012-05-29 15:56:30 +09:00
Ken Wakasa
7d81f31871 am 7b1570e6: Merge "Cleanup Makefiles of LatinIME" into jb-dev
* commit '7b1570e60c2e04fe7d132df758476b34685eb709':
  Cleanup Makefiles of LatinIME
2012-05-23 20:22:04 -07:00
Ken Wakasa
dd58065733 Cleanup Makefiles of LatinIME
Change-Id: Id4c6700bc045825eb64fb2b7ae57f23a6211441d
2012-05-24 12:08:59 +09:00
satok
074e8c9206 am a0ac31fc: Fix the issue on multiple words suggestion
* commit 'a0ac31fcaa01c21592a6e7af243c14dada65cf3e':
  Fix the issue on multiple words suggestion
2012-05-23 05:06:08 -07:00
satok
a0ac31fcaa Fix the issue on multiple words suggestion
Bug: 6509844
Change-Id: I823074a2b29befc3e60c63699ab4dc7719105c63
2012-05-23 20:40:59 +09:00
Jean Chalard
7557d3c6f3 am bc77adef: Merge "Return the bigram frequency if available." into jb-dev
* commit 'bc77adefbb0305c5ec0e41ab01e3a085c47c21eb':
  Return the bigram frequency if available.
2012-05-17 03:31:15 -07:00
Jean Chalard
bc77adefbb Merge "Return the bigram frequency if available." into jb-dev 2012-05-17 03:15:40 -07:00
Jean-Baptiste Queru
cd7c41352f Fix build
Change-Id: I799811aa3afb59bba2e4086a063f5da03669bba3
2012-05-16 16:56:11 -07:00
Ken Wakasa
3b088a2f36 Add missing includes.
Change-Id: Ic7199045d0cffb208871f52cc167194013351d32
2012-05-16 23:05:32 +09:00
Jean Chalard
9416c81403 Return the bigram frequency if available.
This concludes the work on bug#6313806.
Don't submit it before the dictionaries are suitably amended.

Bug: 6313806
Change-Id: Icfea45bd52bb9d8cc68ba2266f80640e3942bb7f
2012-05-16 21:14:06 +09:00
satok
0028ed3627 Use "float" instead of "double"
Change-Id: I93ed4d88ede4058f081dd8d634b00dfff4e96d07
2012-05-16 20:45:05 +09:00
satok
f837b57bf5 Merge "Reorder suggestions result according to auto correction threshold" into jb-dev 2012-05-16 04:13:08 -07:00
satok
db1939dbaa Reorder suggestions result according to auto correction threshold
Bug: 5413904
Change-Id: I3aa3a8109ba45d2129b58d8242866fd3dd3473cb
2012-05-16 19:58:48 +09:00
satok
6804b8e0fd Fix a bug of handling single quote in the correction algorithm
Bug: 6096247
Change-Id: I5490bbdee4ce1e3e0729ec1510a2baab85eeaf05
2012-05-15 15:12:55 +09:00
Tom Ouyang
4d289d39ae Contacts dictionary rebuilds only when contact names have changed.
Bug: 6396600
Change-Id: Iad693ec4bab6351793d624e5c5b0a9f5c12a60e3
2012-05-11 18:43:53 -07:00
Jean Chalard
49ba135fde Perform the actual bigram frequency lookup.
This still returns the unigram frequency, because the values stored
for bigrams in the dictionary are not ready to be returned in-place
instead of unigram values. Aside from this, the code is complete.

Bug: 6313806
Change-Id: If7bb7b644730782277f0f6663334c170b7fe13fb
2012-05-10 20:01:44 +09:00
Jean Chalard
8950ce6c44 Replace the bigram list position with the map and filter
Passing the position will not allow us a reasonable lookup
time. Replace this with a map and bloom filter for very fast
lookup.

Bug: 6313806
Change-Id: I3a61c0001cbc987c1c3c7b8df635d4590a370144
2012-05-07 17:15:21 +09:00
Jean Chalard
f1634c872c Fill in the bloom filter for bigram lookup.
Bug: 6313806
Change-Id: Ib79e14f6f8b241f053da6069c15f19c71084317e
2012-05-07 15:38:43 +09:00
Jean Chalard
1ff8dc47be Fill up a map of bigram addresses for lookup.
We don't want to do a linear search on each terminal when there
may be 100+ bigrams for a given word because that would be
disastrous for performance. Also, we need to resolve each bigram
address anyway.
This change resolves the addresses at first and puts them in a
balanced tree so that lookup will be O(log(n)).

Bug: 6313806
Change-Id: Ibf088035870b9acb41e948f0ab7af4726f2cee24
2012-05-02 17:50:44 +09:00
Jean Chalard
351864b38a Fetch and pass the bigram position on suggestions.
This is a cherry-pick of change I2d81742f

Bug: 6313806
Change-Id: Ic1190b7980d032bc11b57841bca040d980889b6b
2012-04-26 15:24:42 +09:00
Jean Chalard
4d9b202c40 Pass the bigram list position from the top level
The position itself is still a const int = 0 until we have the previous
word passed to the function. This basically does the plumbing.

Bug: 6313806
Change-Id: Ib58995f334fe93e3ff5704d7c79f332017f101ac
2012-04-24 16:47:09 +09:00
Jean Chalard
171d1809ff Add methods to inverse compute the probability.
For now the probability is just returned with the same
value it had, but this is some ground work that needs to be
done anyway.

Bug: 6313806
Change-Id: I9bb8b96b294109771208ade558c9ad56932d2f8e
2012-04-24 09:40:44 +09:00
Jean Chalard
522a04ea5b Pass words as int[] to the native code.
We need to get the bigrams during the call to getSuggestions for
bug#6313806. We already give an int[] to getSuggestions and we
wanted to get rid of char[]'s anyway because it doesn't work with
surrogate pairs, so here we go.

Bug: 6313806
Change-Id: I56ce99f1db6b3302cdf42f0527343bded837091e
2012-04-23 16:05:36 +09:00
Tadashi G. Takaoka
a58ebc73ae Fix typo of some methods' name
Also changes some methods' argument type from Locale to String.

Change-Id: Ib68b528a450dc68a01546483403230f76500bee4
2012-04-18 16:40:50 +09:00
Jean Chalard
bde232dcaa Merge "Pass the previous word down to native code in getSuggestions" 2012-04-17 18:01:05 -07:00
Jean Chalard
80111f08e2 Pass the previous word down to native code in getSuggestions
Change-Id: I477b631d81ef58461e44954f3ae5fd895928bb97
2012-04-17 20:07:10 +09:00
Jean Chalard
fec6837ae1 Fix debug compilation + small cleanup
Change-Id: Ia89d84f62ba38dee05d25fbc94698e889cf27d2c
2012-04-17 17:33:25 +09:00
Jean Chalard
ee396df162 Fix a native crash
This was introduced by Ieb2e306a which failed to keep the return
behavior in case the word doesn't have a bigram.

Change-Id: I6d2f0b79d41c4335e94696690c8331e314961133
2012-04-17 16:57:42 +09:00
Jean Chalard
9c2a96aa6c Preparatory refactoring
Split out getting the pointer to the bigrams to a separate
function. This is a preparative change to bug#6313806

Change-Id: Ieb2e306a1151cd95dc1a16793c8dc2f7fed8b654
2012-04-17 11:46:20 +09:00
Ken Wakasa
db87fe4d5c Just cosmetic changes in jni code
Change-Id: I8628131b5a7ccdee4c158e891002c8b86623b0cd
2012-04-16 19:16:05 +09:00
Jean Chalard
3f675f7060 Fix a large native memory leak.
This leak was about 500k and would happen whenever a new binary
dictionary was opened/closed.

Bug: 6299535
Change-Id: I4fad5b4d9c556ca889f5ef62d9d083a2eff6346a
2012-04-16 16:48:53 +09:00
Jean Chalard
338d3ec725 Replace the flags in getSuggestions with a boolean.
Change-Id: I0ec44df1979cb1dc21017ea290d2151a2af0e7cd

Conflicts:

	java/src/com/android/inputmethod/latin/Suggest.java
2012-04-06 19:34:48 +09:00
Jean Chalard
aa8df59914 Enable using the flags read from the binary file.
Change-Id: Ib420c3e174ccc1a80c4b6fd066de3b7a2b6fb290
2012-04-06 18:54:20 +09:00
Jean Chalard
cd274b1469 Save the flags in a member in the unigram dictionary.
Change-Id: Ic8fad9110db6b97f98ace27af0f347b4e69de8c8
2012-04-06 18:34:59 +09:00
Jean Chalard
e81ac8baa0 Add a method to get the flags from a binary dictionary.
This method is not used yet

Change-Id: Ic15d3d423aff2c83c712bc0aa56571d30755e663
2012-04-06 18:34:22 +09:00
Jean Chalard
5b0761e6a9 Remove write-only stuff
Change-Id: I5ac8ab64c77a298502b3d063ea70db9b4da41716
2012-04-06 17:52:18 +09:00
Jean Chalard
9a933a742d Read shortcuts as strings in the dictionary.
This has no impact on performance.
Before:
(0)  9.61 (0.01%)
(1)  57514.58 (56.70%)
(2)  10.55 (0.01%)
(3)  10.79 (0.01%)
(4)  133.20 (0.13%)
(5)  43553.87 (42.94%)
(6)  10.03 (0.01%)
(20) 47.20 (0.05%)
Total 101431.47 (sum of others 101289.84)

After:
(0)  10.52 (0.01%)
(1)  56311.16 (56.66%)
(2)  13.40 (0.01%)
(3)  10.98 (0.01%)
(4)  136.72 (0.14%)
(5)  42707.92 (42.97%)
(6)  9.79 (0.01%)
(20) 51.35 (0.05%)
Total 99390.76 (sum of others 99251.84)

The difference is not significant with regard to measure imprecision

Change-Id: I2e4f1ef7a5e99082e67dd27f56cf4fc432bb48fa
2012-04-06 16:22:08 +09:00
Ken Wakasa
0c1a3ec629 Make LatinIME's native Makefile NDK-friendly
Change-Id: I55d430756b3a8251c9ff49dfabfcecb047d979a4
2012-03-31 05:07:32 +09:00
Ying Wang
32f0e24b33 Remove ".." in the native LOCAL_SRC_FILES.
The build system does not work well with ".." in the path of native
source code.
".." causes the object files to spill out of the module's intermediate
directory.

Change-Id: Ib4a473426be296a738e7facbaa091e56f0b7c5b8
2012-03-30 10:50:26 -07:00
Ken Wakasa
3ef3e24a12 Move the "src" directory as a preparation for Ib4a47342 and I66f6c5b9
Change-Id: I3ab65059f6e356530484bfd0bba26a634a4cba65
2012-03-30 09:53:51 +09:00
satok
6ba8de2a60 Good bye the proximity logic in Java code
Bug: 4343280
Change-Id: I82f7d08703647a3492ce6e2d3b741146df58927e
2012-03-28 18:42:30 +09:00
satok
0cb2097a45 Fix additional proximity in the native code
Bug: 4343280
Change-Id: I4164bb916b2dbdfb6bdc151b99d46a6171d9c355
2012-03-14 11:17:59 +09:00
satok
5eec574cf0 Use additional proximity chars in the native code
Bug: 4343280

Change-Id: Ida690fe246cea80a82fcdb3ad0c28e2907b882ac
2012-03-13 19:00:16 +09:00
satok
552c3c27f0 Implement additional proximity characters in the native code
Bug: 4343280

Change-Id: I9bbc5cab2fef1ee80c1fe32017df811ef8af10bc
2012-03-13 17:38:50 +09:00
satok
a70ee6e3b3 Add functions to calculate proximity characters in the native code
Bug: 4343280

Change-Id: I17f8f6295b01900948b98680d0267753f33a46cf
2012-03-08 12:55:15 +09:00
Jean-Baptiste Queru
26e315785d resolved conflicts for merge of 3ad1145a to master
Change-Id: I13159b95f90c5095373951bf9e91b7dbf8b14558
2012-02-14 13:09:49 -08:00
Jean-Baptiste Queru
11c41216f1 Merge 2577fca1
Change-Id: Ie2c9f6c2eafb59dff95db8954481ce49c87a6d44
2012-01-23 09:06:00 -08:00
satok
9fb6f47a6a New LOG lib
Change-Id: I977e7e10fa58c0a64ca0c3c7b5cb2272446e3efe
2012-01-13 18:04:15 +09:00
satok
be0cf72253 Move auto correction thresthold to the native code
bug: 5858137

Change-Id: Ic4b6270c6e51ef4ed25a6a1d8ddd7fdfa70fd78d
2012-01-12 19:56:15 +09:00
Ken Wakasa
d5e35acffb Reorganize LatinIME native build directory structure. Now it got NDK friendly.
Change-Id: I0f62ce8a6a4d2b6134db698f8b8d0576616e524d
2011-12-13 21:16:20 +09:00
Tadashi G. Takaoka
5fdcd7d5cd Fix JNI as 64bit ready
Change-Id: I7a3ed2cee3b20744869aa5cf89ff9fd95f5ab8fe
2011-10-31 20:44:01 +09:00
Tadashi G. Takaoka
94810dec21 Remove NULL from native/jni
Change-Id: I7c4e32d0d69876c7ea85d6997c9a40fa362152eb
2011-10-28 17:06:58 +09:00
Tadashi G. Takaoka
5049f5bbb4 Fix trivial compile issue
Change-Id: I9dbc51496dbdb2195bbe589b0c3b2cd7e3d2f1fc
2011-10-28 16:24:31 +09:00
Yusuke Nojima
ad35835bac Pass the touch position correction data to native.
Change-Id: I92958779377a530410d1682100f9d0a2ba267dea
2011-09-29 17:00:49 +09:00
Yusuke Nojima
1c923d8581 Extend JNI interface to pass the theme id of the keybaord to the native.
Change-Id: I705ec2fa1d32a673a8d84bc9afc2a46208a414a1
2011-09-28 11:56:28 +09:00
Yusuke Nojima
de2f8424ea Add explicit NULL check to latinime_Keyboard_setProximityInfo() and ProximityInfo::ProximityInfo().
Bug: 5375319
Change-Id: If3016ab56b6c62c8855b394f23e09e127a279be7
2011-09-27 14:51:41 +09:00
Yusuke Nojima
0e1f656c1b Extend JNI interface to pass the key coordinate information
Change-Id: Iec0d1415d6abba43b775a014516b961b167597b7
2011-09-22 18:39:34 +09:00
Jean Chalard
f0a9809662 Check the binary dictionary magic number
...and return NULL if it does not matched an expected value.

Bug: 5052486
Change-Id: I1dc7955d2785ee080bc5c22398be9befe332f096
2011-07-20 19:43:14 +09:00
Ken Wakasa
ce9e52a12a Clean up in LatinIME native code
Change-Id: I0062200a0181a491690115ac0fab8d11358e2f14
2011-06-18 23:52:09 +09:00
Jean Chalard
e93b1f2209 Allow reading a binary dictionary even without proximity info.
This prepares the way for spell checking, which is to be done
without context so without proximity info.

Bug: 4176026
Change-Id: I1b4bfaefe2611e1b484acdf3c33598cb80f81ff4
2011-06-02 12:10:13 +09:00
Jean Chalard
c2bbc6a449 Use translation of fallback umlauts digraphs for German.
For German : handle "ae", "oe" and "ue" to be alternate forms for
umlaut-bearing versions of "a", "o" and "u".

Issue: 3275926

Change-Id: I056c707cdacc464ceab63be56c016c7f8439196c
2011-03-03 11:52:23 +09:00
satok
8fbd552292 Add proximity info to native
Bug: 3311719

Change-Id: Ie596304070e321ad23fb67a13bf05e2b6af1b54b
2011-02-23 23:04:00 +09:00
Tadashi G. Takaoka
887f11ee43 Remove next letters frequency handling
Bug: 3428942
Change-Id: Id62f467ce4e50c60a56d59bf96770e799a4659e2
2011-02-17 13:59:41 +09:00
Ken Wakasa
da50e1e98d Fixes in close() in BinaryDictionary.
Avoid using 'synchronized' in finalizer as well.

bug: 3340837
Change-Id: I9b28f54e4490ecb844ba33a379f71b625e4246a2
2011-01-17 15:58:07 +09:00
Ken Wakasa
90d96615bc Clean up: Update variable names to comply with spec of ApplicationInfo.
ApplicationInfo.sourceDir may or may not be apk file name.  It can be a directory as well.
The spec just says it's "Full path to the location of this package".

Also, added error handling in loadDictionary().

Change-Id: I5e64d0aba4b1ec7634f4b3ac5537e7a774433ece
2011-01-09 16:53:33 +09:00
Ken Wakasa
e90b333017 Load main dic in native
Follow up to Id57dce51

bug: 3219819
Change-Id: I00e11ef21d0252ffa88c12dffb9c55b0f2e19a66
2011-01-07 19:51:45 +09:00
Ken Wakasa
a46683a47b Just code cleanup.
Change-Id: I796b47aa168b3b2cc29de65fc7dc5dab5a7f4b17
2010-12-03 17:13:04 +09:00
satok
3008825948 Fix parameters of native functions and refactor Dictionary
- created bigram/unigram dictionary classes

Change-Id: I233a28ed8d611870db3f4cf8f25fc45b5d41529b
2010-12-02 01:16:44 +09:00
Jae Yong Sung
80aa14fd43 - separate dict (uses xml)
- retrieve bigrams that only starts with character typed and neighbor keys
- contacts bigram
- performance measure

bug: 2873133

Change-Id: If97c005b18c82f3fafef50009dd2dfd972b0ab8f
2010-07-28 11:08:08 -07:00
Jae Yong Sung
937d5ad013 added bigram prediction
- after first character, only suggests bigram data (but doesn't autocomplete)
  - after second character, words from dictionary gets rearranged by using bigram
  - compatible with old dictionary
  - added preference option to disable bigram

Change-Id: Ia8f4e8fa55e797e86d858fd499887cd396388411
2010-07-13 11:33:39 -07:00
Ken Wakasa
826269c8ae Get rid of dependency on native AssetManager API. Confirmed the native code builds with the NDK r3.
Change-Id: I0d2d3a0e262847d6948a0336a35440e21e312ad2
2010-04-27 22:23:03 +09:00
Amith Yamasani
07b1603a3f Don't let the native code target be included twice when unbundling.
Move java code to a different directory so that the unbundled
version doesn't try to compile the native code again.

Change-Id: I05cf9e643824ddc448821f69805ccb0240c5b986
2010-03-09 15:01:09 -08:00