@kensio/pinyinjs 1.7.9 → 1.7.10
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/data/core.freq +0 -0
- package/data/full.freq +0 -0
- package/data/manifest.json +1 -1
- package/data/standard.freq +0 -0
- package/dist/dictionary/artifact.js +2 -2
- package/dist/dictionary/artifact.js.map +1 -1
- package/dist/dictionary/frequency-table.d.ts +57 -0
- package/dist/dictionary/frequency-table.d.ts.map +1 -1
- package/dist/dictionary/frequency-table.js +59 -0
- package/dist/dictionary/frequency-table.js.map +1 -1
- package/package.json +2 -2
package/data/core.freq
CHANGED
|
Binary file
|
package/data/full.freq
CHANGED
|
Binary file
|
package/data/manifest.json
CHANGED
package/data/standard.freq
CHANGED
|
Binary file
|
|
@@ -8,7 +8,7 @@
|
|
|
8
8
|
import { toCharacters } from "../script/characters.js";
|
|
9
9
|
export { claimKeys } from "./artifact-claims.js";
|
|
10
10
|
export { findRoundTripFailure, readArtifact, readingsByKey, } from "./artifact-read.js";
|
|
11
|
-
import { FrequencyTable } from "./frequency-table.js";
|
|
11
|
+
import { countForQuantising, FrequencyTable } from "./frequency-table.js";
|
|
12
12
|
import { KeyIndex } from "./key-index.js";
|
|
13
13
|
import { claimKeys, orderedClaims } from "./artifact-claims.js";
|
|
14
14
|
import { ALTERNATE, BOUNDARY_SEPARATOR, COLUMN, encodeReading, LINE, PROPER_NOUN_FLAG, } from "./artifact-format.js";
|
|
@@ -92,7 +92,7 @@ export function buildArtifact(entries) {
|
|
|
92
92
|
.map((alternate) => encodeReading(alternate))
|
|
93
93
|
.join(ALTERNATE),
|
|
94
94
|
]));
|
|
95
|
-
frequencies.push(entry.frequency);
|
|
95
|
+
frequencies.push(countForQuantising(entry.frequency, entry.isProperNoun));
|
|
96
96
|
}
|
|
97
97
|
return {
|
|
98
98
|
keys: index.serialise(),
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"artifact.js","sourceRoot":"","sources":["../../src/dictionary/artifact.ts"],"names":[],"mappings":"AAAA;;;;;;GAMG;AACH,OAAO,EAAE,YAAY,EAAE,MAAM,yBAAyB,CAAC;AACvD,OAAO,EAAE,SAAS,EAAE,MAAM,sBAAsB,CAAC;AACjD,OAAO,EACL,oBAAoB,EACpB,YAAY,EACZ,aAAa,GACd,MAAM,oBAAoB,CAAC;AAE5B,OAAO,EAAE,cAAc,EAAE,MAAM,sBAAsB,CAAC;
|
|
1
|
+
{"version":3,"file":"artifact.js","sourceRoot":"","sources":["../../src/dictionary/artifact.ts"],"names":[],"mappings":"AAAA;;;;;;GAMG;AACH,OAAO,EAAE,YAAY,EAAE,MAAM,yBAAyB,CAAC;AACvD,OAAO,EAAE,SAAS,EAAE,MAAM,sBAAsB,CAAC;AACjD,OAAO,EACL,oBAAoB,EACpB,YAAY,EACZ,aAAa,GACd,MAAM,oBAAoB,CAAC;AAE5B,OAAO,EAAE,kBAAkB,EAAE,cAAc,EAAE,MAAM,sBAAsB,CAAC;AAC1E,OAAO,EAAE,QAAQ,EAAE,MAAM,gBAAgB,CAAC;AAC1C,OAAO,EAAE,SAAS,EAAE,aAAa,EAAE,MAAM,sBAAsB,CAAC;AAChE,OAAO,EACL,SAAS,EACT,kBAAkB,EAClB,MAAM,EAEN,aAAa,EACb,IAAI,EACJ,gBAAgB,GACjB,MAAM,sBAAsB,CAAC;AAE9B,OAAO,EACL,aAAa,EAEb,aAAa,EACb,gBAAgB,GACjB,MAAM,sBAAsB,CAAC;AAE9B;;GAEG;AACH,SAAS,iBAAiB,CACxB,OAA6C;IAE7C,MAAM,QAAQ,GAAG,IAAI,GAAG,EAAkB,CAAC;IAC3C,KAAK,MAAM,CAAC,GAAG,EAAE,KAAK,CAAC,IAAI,OAAO,EAAE,CAAC;QACnC,IAAI,YAAY,CAAC,GAAG,CAAC,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;YACnC,QAAQ,CAAC,GAAG,CAAC,GAAG,EAAE,aAAa,CAAC,KAAK,CAAC,QAAQ,CAAC,EAAE,CAAC,CAAC,CAAC;QACtD,CAAC;IACH,CAAC;IACD,OAAO,QAAQ,CAAC;AAClB,CAAC;AAED;;;;;;;GAOG;AACH,SAAS,WAAW,CAClB,GAAW,EACX,OAAe,EACf,QAAqC;IAErC,MAAM,KAAK,GAAG,YAAY,CAAC,GAAG,CAAC,CAAC;IAChC,IAAI,KAAK,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QACrB,OAAO,KAAK,CAAC;IACf,CAAC;IACD,MAAM,OAAO,GAAa,EAAE,CAAC;IAC7B,KAAK,MAAM,SAAS,IAAI,KAAK,EAAE,CAAC;QAC9B,MAAM,KAAK,GAAG,QAAQ,CAAC,GAAG,CAAC,SAAS,CAAC,CAAC;QACtC,IAAI,KAAK,KAAK,SAAS,EAAE,CAAC;YACxB,OAAO,KAAK,CAAC;QACf,CAAC;QACD,OAAO,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;IACtB,CAAC;IACD,OAAO,OAAO,CAAC,IAAI,CAAC,GAAG,CAAC,KAAK,OAAO,CAAC;AACvC,CAAC;AAED;;GAEG;AACH,SAAS,WAAW,CAAC,OAA0B;IAC7C,IAAI,GAAG,GAAG,OAAO,CAAC,MAAM,CAAC;IACzB,OAAO,GAAG,GAAG,CAAC,IAAI,OAAO,CAAC,GAAG,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC;QAC1C,GAAG,EAAE,CAAC;IACR,CAAC;IACD,OAAO,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;AAC5C,CAAC;AAED;;;;;;;;;;;;GAYG;AACH,MAAM,UAAU,aAAa,CAC3B,OAAmC;IAEnC,MAAM,OAAO,GAAG,SAAS,CAAC,OAAO,CAAC,CAAC;IACnC,MAAM,QAAQ,GAAG,iBAAiB,CAAC,OAAO,CAAC,CAAC;IAC5C,MAAM,OAAO,GAAG,aAAa,CAAC,OAAO,CAAC,CAAC;IACvC,MAAM,KAAK,GAAG,QAAQ,CAAC,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,EAAE,EAAE,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC;IAEpE,MAAM,KAAK,GAAa,EAAE,CAAC;IAC3B,MAAM,WAAW,GAAa,EAAE,CAAC;IACjC,KAAK,MAAM,CAAC,GAAG,EAAE,KAAK,CAAC,IAAI,OAAO,EAAE,CAAC;QACnC,MAAM,OAAO,GAAG,aAAa,CAAC,KAAK,CAAC,QAAQ,CAAC,EAAE,CAAC,CAAC;QAEjD,KAAK,CAAC,IAAI,CACR,WAAW,CAAC;YACV,WAAW,CAAC,GAAG,EAAE,OAAO,EAAE,QAAQ,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,OAAO;YAClD,KAAK,CAAC,QAAQ,CAAC,EAAE,KAAK,SAAS,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,aAAa,CAAC,KAAK,CAAC,QAAQ,CAAC,EAAE,CAAC;YACvE,KAAK,CAAC,YAAY;YAClB,KAAK,CAAC,YAAY;gBAChB,CAAC,CAAC,gBAAgB;oBAChB,CAAC,KAAK,CAAC,cAAc,IAAI,EAAE,CAAC,CAAC,IAAI,CAAC,kBAAkB,CAAC;gBACvD,CAAC,CAAC,EAAE;YACN,CAAC,KAAK,CAAC,UAAU,IAAI,EAAE,CAAC;iBACrB,GAAG,CAAC,CAAC,SAAS,EAAE,EAAE,CAAC,aAAa,CAAC,SAAS,CAAC,CAAC;iBAC5C,IAAI,CAAC,SAAS,CAAC;SACnB,CAAC,CACH,CAAC;QACF,WAAW,CAAC,IAAI,CAAC,kBAAkB,CAAC,KAAK,CAAC,SAAS,EAAE,KAAK,CAAC,YAAY,CAAC,CAAC,CAAC;IAC5E,CAAC;IAED,OAAO;QACL,IAAI,EAAE,KAAK,CAAC,SAAS,EAAE;QACvB,OAAO,EAAE,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC;QACzB,WAAW,EAAE,cAAc,CAAC,KAAK,CAAC,WAAW,CAAC,CAAC,SAAS,EAAE;KAC3D,CAAC;AACJ,CAAC"}
|
|
@@ -46,6 +46,60 @@ export declare const FREQUENCY_BUCKETS = 16;
|
|
|
46
46
|
* bucket sum, re-broken by the order the floats happened to be added in.
|
|
47
47
|
*/
|
|
48
48
|
export declare const WORD_CHARGE = 4.62;
|
|
49
|
+
/**
|
|
50
|
+
* The count a name the corpus never counted is quantised at.
|
|
51
|
+
*
|
|
52
|
+
* A count of zero earns bucket 0, and {@link FrequencyTable.costOf} reads that
|
|
53
|
+
* as `−log P` with `P` of `1/N`. On jieba's corpus the flat cost it produces is
|
|
54
|
+
* 19.62, and two mid-frequency characters come to less. 脸书 cost 19.62 where
|
|
55
|
+
* 脸 and 书 together cost 18.24, and `decodeRun` read Facebook as a face and a
|
|
56
|
+
* book. 推特, 高雄, 乐高 and 网飞 all went the same way, and 671 of the 5,039
|
|
57
|
+
* uncounted names of two or three characters came apart when read on their
|
|
58
|
+
* own. 谷歌 survived on an exact tie.
|
|
59
|
+
*
|
|
60
|
+
* **Zero is the wrong count for a name jieba's list never held.** The list has
|
|
61
|
+
* no frequency floor to be under. It carries 一万一千五百二十颗 on two
|
|
62
|
+
* occurrences. A word missing from it was left out of the lexicon, and the
|
|
63
|
+
* corpus was never asked about it. Names are where that bites, since a name
|
|
64
|
+
* enters the language whenever something is given one. 微软 was counted 1757
|
|
65
|
+
* times and 脸书 not at all, and only one of the two existed when the list was
|
|
66
|
+
* drawn. 10,676 of the dictionary's 42,861 names are counted nowhere.
|
|
67
|
+
*
|
|
68
|
+
* **3 is jieba's own answer for a word it lists and cannot count.** Its counts
|
|
69
|
+
* bottom out at 2, and 3 carries 159,318 of 349,045 entries against 40,502 at 2
|
|
70
|
+
* and 12,679 at 4. Interpolating between the neighbours accounts for about
|
|
71
|
+
* 22,000 of those, and the other 137,000 carry a default. It is the default a
|
|
72
|
+
* hand-added entry gets (AT&T, B超 and C++ are all counted 3).
|
|
73
|
+
*
|
|
74
|
+
* The value is loose. Bucket 2 runs from 3 to 8, so any count jieba might have
|
|
75
|
+
* settled on in that range quantises the same way and lands the cost at 17.62.
|
|
76
|
+
* What the floor buys a name is two buckets against a split.
|
|
77
|
+
*
|
|
78
|
+
* **Names only.** The same floor over every uncounted key lifts two thirds of
|
|
79
|
+
* the dictionary by two buckets and re-tips
|
|
80
|
+
* {@link import("../decode/lattice-types.js").READING_CHARGE}, which was sized
|
|
81
|
+
* against bucket 0 at 19.62. Measured that way 從容地 joins, 都會 reads `dūhuì`
|
|
82
|
+
* and 過得 reads `guò dé`, and the gold harness goes from 3 misses to 6. The
|
|
83
|
+
* dictionary's tail is where 费耗, 块肉 and 拍张 live, and bucket 0 is what
|
|
84
|
+
* holds them off.
|
|
85
|
+
*
|
|
86
|
+
* 428 of the 671 names now stay whole. The ones left are compounds a split
|
|
87
|
+
* suits, 中国菜, 日本人, 上海话 and 美国人 among them, and each beats its own
|
|
88
|
+
* entry by more than two buckets. Over the 88,866 lines of Tatoeba and
|
|
89
|
+
* zh.wikipedia 86 of 139,682 Han runs move, the gold harness holds at 3 misses
|
|
90
|
+
* and CPP holds at 91.49%.
|
|
91
|
+
*/
|
|
92
|
+
export declare const UNCOUNTED_NAME = 3;
|
|
93
|
+
/**
|
|
94
|
+
* The count an entry is quantised at. Its own, unless it is a name the corpus
|
|
95
|
+
* never counted. See {@link UNCOUNTED_NAME}.
|
|
96
|
+
*
|
|
97
|
+
* Applied where the artifact is written and nowhere else, so `full.counts`
|
|
98
|
+
* keeps the corpus as it was met. The two files disagree by design over these
|
|
99
|
+
* 10,676 keys, and `data.test.ts` checks their alignment through this
|
|
100
|
+
* function.
|
|
101
|
+
*/
|
|
102
|
+
export declare function countForQuantising(frequency: number, isProperNoun: boolean): number;
|
|
49
103
|
/**
|
|
50
104
|
* The charge a corpus derives, for the build to hold {@link WORD_CHARGE} to.
|
|
51
105
|
*
|
|
@@ -71,6 +125,9 @@ export declare class FrequencyTable {
|
|
|
71
125
|
* the word was not attested, which still earns the lowest bucket rather than
|
|
72
126
|
* being excluded — an unattested word in the dictionary is rare, not
|
|
73
127
|
* impossible.
|
|
128
|
+
*
|
|
129
|
+
* Counts arrive through {@link countForQuantising}, which is where a name the
|
|
130
|
+
* corpus never counted is given jieba's own default for one.
|
|
74
131
|
*/
|
|
75
132
|
static build(frequencies: readonly number[]): FrequencyTable;
|
|
76
133
|
/**
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"frequency-table.d.ts","sourceRoot":"","sources":["../../src/dictionary/frequency-table.ts"],"names":[],"mappings":"AAAA;;;;;;;;GAQG;AACH,eAAO,MAAM,iBAAiB,KAAK,CAAC;AAIpC;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GAoCG;AACH,eAAO,MAAM,WAAW,OAAO,CAAC;AAEhC;;;;;;GAMG;AACH,wBAAgB,iBAAiB,CAAC,KAAK,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,GAAG,MAAM,CAExE;AAED;;;;;;GAMG;AACH,qBAAa,cAAc;;IACzB
|
|
1
|
+
{"version":3,"file":"frequency-table.d.ts","sourceRoot":"","sources":["../../src/dictionary/frequency-table.ts"],"names":[],"mappings":"AAAA;;;;;;;;GAQG;AACH,eAAO,MAAM,iBAAiB,KAAK,CAAC;AAIpC;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GAoCG;AACH,eAAO,MAAM,WAAW,OAAO,CAAC;AAEhC;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA0CG;AACH,eAAO,MAAM,cAAc,IAAI,CAAC;AAEhC;;;;;;;;GAQG;AACH,wBAAgB,kBAAkB,CAChC,SAAS,EAAE,MAAM,EACjB,YAAY,EAAE,OAAO,GACpB,MAAM,CAER;AAED;;;;;;GAMG;AACH,wBAAgB,iBAAiB,CAAC,KAAK,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,GAAG,MAAM,CAExE;AAED;;;;;;GAMG;AACH,qBAAa,cAAc;;IACzB;;;;;;;;;;;OAWG;IACH,MAAM,CAAC,KAAK,CAAC,WAAW,EAAE,SAAS,MAAM,EAAE,GAAG,cAAc,CA4B3D;IAED;;OAEG;IACH,MAAM,CAAC,IAAI,CAAC,MAAM,EAAE,UAAU,EAAE,IAAI,EAAE,MAAM,GAAG,cAAc,CAE5D;IAKD;;;OAGG;IACH,OAAO,eAGN;IAED;;OAEG;IACH,IAAI,IAAI,IAAI,MAAM,CAEjB;IAED;;;;;OAKG;IACH,QAAQ,CAAC,KAAK,EAAE,MAAM,GAAG,MAAM,CAM9B;IAED;;;;;;;;OAQG;IACH,MAAM,CAAC,KAAK,EAAE,MAAM,GAAG,MAAM,CAE5B;IAED;;OAEG;IACH,SAAS,IAAI,UAAU,CAEtB;CACF"}
|
|
@@ -47,6 +47,62 @@ const MAX_BUCKET = FREQUENCY_BUCKETS - 1;
|
|
|
47
47
|
* bucket sum, re-broken by the order the floats happened to be added in.
|
|
48
48
|
*/
|
|
49
49
|
export const WORD_CHARGE = 4.62;
|
|
50
|
+
/**
|
|
51
|
+
* The count a name the corpus never counted is quantised at.
|
|
52
|
+
*
|
|
53
|
+
* A count of zero earns bucket 0, and {@link FrequencyTable.costOf} reads that
|
|
54
|
+
* as `−log P` with `P` of `1/N`. On jieba's corpus the flat cost it produces is
|
|
55
|
+
* 19.62, and two mid-frequency characters come to less. 脸书 cost 19.62 where
|
|
56
|
+
* 脸 and 书 together cost 18.24, and `decodeRun` read Facebook as a face and a
|
|
57
|
+
* book. 推特, 高雄, 乐高 and 网飞 all went the same way, and 671 of the 5,039
|
|
58
|
+
* uncounted names of two or three characters came apart when read on their
|
|
59
|
+
* own. 谷歌 survived on an exact tie.
|
|
60
|
+
*
|
|
61
|
+
* **Zero is the wrong count for a name jieba's list never held.** The list has
|
|
62
|
+
* no frequency floor to be under. It carries 一万一千五百二十颗 on two
|
|
63
|
+
* occurrences. A word missing from it was left out of the lexicon, and the
|
|
64
|
+
* corpus was never asked about it. Names are where that bites, since a name
|
|
65
|
+
* enters the language whenever something is given one. 微软 was counted 1757
|
|
66
|
+
* times and 脸书 not at all, and only one of the two existed when the list was
|
|
67
|
+
* drawn. 10,676 of the dictionary's 42,861 names are counted nowhere.
|
|
68
|
+
*
|
|
69
|
+
* **3 is jieba's own answer for a word it lists and cannot count.** Its counts
|
|
70
|
+
* bottom out at 2, and 3 carries 159,318 of 349,045 entries against 40,502 at 2
|
|
71
|
+
* and 12,679 at 4. Interpolating between the neighbours accounts for about
|
|
72
|
+
* 22,000 of those, and the other 137,000 carry a default. It is the default a
|
|
73
|
+
* hand-added entry gets (AT&T, B超 and C++ are all counted 3).
|
|
74
|
+
*
|
|
75
|
+
* The value is loose. Bucket 2 runs from 3 to 8, so any count jieba might have
|
|
76
|
+
* settled on in that range quantises the same way and lands the cost at 17.62.
|
|
77
|
+
* What the floor buys a name is two buckets against a split.
|
|
78
|
+
*
|
|
79
|
+
* **Names only.** The same floor over every uncounted key lifts two thirds of
|
|
80
|
+
* the dictionary by two buckets and re-tips
|
|
81
|
+
* {@link import("../decode/lattice-types.js").READING_CHARGE}, which was sized
|
|
82
|
+
* against bucket 0 at 19.62. Measured that way 從容地 joins, 都會 reads `dūhuì`
|
|
83
|
+
* and 過得 reads `guò dé`, and the gold harness goes from 3 misses to 6. The
|
|
84
|
+
* dictionary's tail is where 费耗, 块肉 and 拍张 live, and bucket 0 is what
|
|
85
|
+
* holds them off.
|
|
86
|
+
*
|
|
87
|
+
* 428 of the 671 names now stay whole. The ones left are compounds a split
|
|
88
|
+
* suits, 中国菜, 日本人, 上海话 and 美国人 among them, and each beats its own
|
|
89
|
+
* entry by more than two buckets. Over the 88,866 lines of Tatoeba and
|
|
90
|
+
* zh.wikipedia 86 of 139,682 Han runs move, the gold harness holds at 3 misses
|
|
91
|
+
* and CPP holds at 91.49%.
|
|
92
|
+
*/
|
|
93
|
+
export const UNCOUNTED_NAME = 3;
|
|
94
|
+
/**
|
|
95
|
+
* The count an entry is quantised at. Its own, unless it is a name the corpus
|
|
96
|
+
* never counted. See {@link UNCOUNTED_NAME}.
|
|
97
|
+
*
|
|
98
|
+
* Applied where the artifact is written and nowhere else, so `full.counts`
|
|
99
|
+
* keeps the corpus as it was met. The two files disagree by design over these
|
|
100
|
+
* 10,676 keys, and `data.test.ts` checks their alignment through this
|
|
101
|
+
* function.
|
|
102
|
+
*/
|
|
103
|
+
export function countForQuantising(frequency, isProperNoun) {
|
|
104
|
+
return isProperNoun && frequency <= 0 ? UNCOUNTED_NAME : frequency;
|
|
105
|
+
}
|
|
50
106
|
/**
|
|
51
107
|
* The charge a corpus derives, for the build to hold {@link WORD_CHARGE} to.
|
|
52
108
|
*
|
|
@@ -73,6 +129,9 @@ export class FrequencyTable {
|
|
|
73
129
|
* the word was not attested, which still earns the lowest bucket rather than
|
|
74
130
|
* being excluded — an unattested word in the dictionary is rare, not
|
|
75
131
|
* impossible.
|
|
132
|
+
*
|
|
133
|
+
* Counts arrive through {@link countForQuantising}, which is where a name the
|
|
134
|
+
* corpus never counted is given jieba's own default for one.
|
|
76
135
|
*/
|
|
77
136
|
static build(frequencies) {
|
|
78
137
|
// Looped rather than `Math.max(0, ...frequencies)`, which passes one
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"frequency-table.js","sourceRoot":"","sources":["../../src/dictionary/frequency-table.ts"],"names":[],"mappings":"AAAA;;;;;;;;GAQG;AACH,MAAM,CAAC,MAAM,iBAAiB,GAAG,EAAE,CAAC;AAEpC,MAAM,UAAU,GAAG,iBAAiB,GAAG,CAAC,CAAC;AAEzC;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GAoCG;AACH,MAAM,CAAC,MAAM,WAAW,GAAG,IAAI,CAAC;AAEhC;;;;;;GAMG;AACH,MAAM,UAAU,iBAAiB,CAAC,KAAa,EAAE,OAAe;IAC9D,OAAO,UAAU,GAAG,CAAC,IAAI,CAAC,GAAG,CAAC,KAAK,CAAC,GAAG,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC;AAClE,CAAC;AAED;;;;;;GAMG;AACH,MAAM,OAAO,cAAc;IACzB
|
|
1
|
+
{"version":3,"file":"frequency-table.js","sourceRoot":"","sources":["../../src/dictionary/frequency-table.ts"],"names":[],"mappings":"AAAA;;;;;;;;GAQG;AACH,MAAM,CAAC,MAAM,iBAAiB,GAAG,EAAE,CAAC;AAEpC,MAAM,UAAU,GAAG,iBAAiB,GAAG,CAAC,CAAC;AAEzC;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GAoCG;AACH,MAAM,CAAC,MAAM,WAAW,GAAG,IAAI,CAAC;AAEhC;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA0CG;AACH,MAAM,CAAC,MAAM,cAAc,GAAG,CAAC,CAAC;AAEhC;;;;;;;;GAQG;AACH,MAAM,UAAU,kBAAkB,CAChC,SAAiB,EACjB,YAAqB;IAErB,OAAO,YAAY,IAAI,SAAS,IAAI,CAAC,CAAC,CAAC,CAAC,cAAc,CAAC,CAAC,CAAC,SAAS,CAAC;AACrE,CAAC;AAED;;;;;;GAMG;AACH,MAAM,UAAU,iBAAiB,CAAC,KAAa,EAAE,OAAe;IAC9D,OAAO,UAAU,GAAG,CAAC,IAAI,CAAC,GAAG,CAAC,KAAK,CAAC,GAAG,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC;AAClE,CAAC;AAED;;;;;;GAMG;AACH,MAAM,OAAO,cAAc;IACzB;;;;;;;;;;;OAWG;IACH,MAAM,CAAC,KAAK,CAAC,WAA8B;QACzC,qEAAqE;QACrE,4EAA4E;QAC5E,0EAA0E;QAC1E,IAAI,OAAO,GAAG,CAAC,CAAC;QAChB,KAAK,MAAM,SAAS,IAAI,WAAW,EAAE,CAAC;YACpC,IAAI,SAAS,GAAG,OAAO,EAAE,CAAC;gBACxB,OAAO,GAAG,SAAS,CAAC;YACtB,CAAC;QACH,CAAC;QACD,MAAM,KAAK,GAAG,OAAO,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,UAAU,GAAG,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC;QAEnE,MAAM,MAAM,GAAG,IAAI,UAAU,CAAC,IAAI,CAAC,IAAI,CAAC,WAAW,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC;QACjE,KAAK,MAAM,CAAC,KAAK,EAAE,SAAS,CAAC,IAAI,WAAW,CAAC,OAAO,EAAE,EAAE,CAAC;YACvD,MAAM,QAAQ,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,SAAS,CAAC,CAAC;YACxC,MAAM,MAAM,GAAG,IAAI,CAAC,GAAG,CACrB,UAAU,EACV,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,QAAQ,CAAC,GAAG,KAAK,CAAC,CACzC,CAAC;YACF,MAAM,SAAS,GAAG,KAAK,IAAI,CAAC,CAAC;YAC7B,MAAM,QAAQ,GAAG,MAAM,CAAC,SAAS,CAAC,IAAI,CAAC,CAAC;YACxC,MAAM,CAAC,SAAS,CAAC;gBACf,KAAK,GAAG,CAAC,KAAK,CAAC;oBACb,CAAC,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC,GAAG,CAAC,QAAQ,GAAG,GAAG,CAAC;oBAClC,CAAC,CAAC,CAAC,QAAQ,GAAG,IAAI,CAAC,GAAG,MAAM,CAAC;QACnC,CAAC;QAED,OAAO,IAAI,cAAc,CAAC,MAAM,EAAE,WAAW,CAAC,MAAM,CAAC,CAAC;IACxD,CAAC;IAED;;OAEG;IACH,MAAM,CAAC,IAAI,CAAC,MAAkB,EAAE,IAAY;QAC1C,OAAO,IAAI,cAAc,CAAC,MAAM,EAAE,IAAI,CAAC,CAAC;IAC1C,CAAC;IAEQ,OAAO,CAAa;IACpB,KAAK,CAAS;IAEvB;;;OAGG;IACH,YAAoB,MAAkB,EAAE,IAAY;QAClD,IAAI,CAAC,OAAO,GAAG,MAAM,CAAC;QACtB,IAAI,CAAC,KAAK,GAAG,IAAI,CAAC;IACpB,CAAC;IAED;;OAEG;IACH,IAAI,IAAI;QACN,OAAO,IAAI,CAAC,KAAK,CAAC;IACpB,CAAC;IAED;;;;;OAKG;IACH,QAAQ,CAAC,KAAa;QACpB,IAAI,KAAK,GAAG,CAAC,IAAI,KAAK,IAAI,IAAI,CAAC,KAAK,EAAE,CAAC;YACrC,OAAO,CAAC,CAAC;QACX,CAAC;QACD,MAAM,IAAI,GAAG,IAAI,CAAC,OAAO,CAAC,KAAK,IAAI,CAAC,CAAC,IAAI,CAAC,CAAC;QAC3C,OAAO,KAAK,GAAG,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,GAAG,GAAG,CAAC;IAClD,CAAC;IAED;;;;;;;;OAQG;IACH,MAAM,CAAC,KAAa;QAClB,OAAO,UAAU,GAAG,IAAI,CAAC,QAAQ,CAAC,KAAK,CAAC,GAAG,WAAW,CAAC;IACzD,CAAC;IAED;;OAEG;IACH,SAAS;QACP,OAAO,IAAI,CAAC,OAAO,CAAC;IACtB,CAAC;CACF"}
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "@kensio/pinyinjs",
|
|
3
|
-
"version": "1.7.
|
|
3
|
+
"version": "1.7.10",
|
|
4
4
|
"description": "Chinese hanzi to pinyin conversion",
|
|
5
5
|
"repository": "https://github.com/KensioSoftware/pinyinjs",
|
|
6
6
|
"homepage": "https://pinyinjs.dev",
|
|
@@ -69,7 +69,7 @@
|
|
|
69
69
|
"conventional-changelog-conventionalcommits": "9.3.1",
|
|
70
70
|
"fta-cli": "^3.0.0",
|
|
71
71
|
"jiti": "^2.7.0",
|
|
72
|
-
"oxfmt": "^0.
|
|
72
|
+
"oxfmt": "^0.64.0",
|
|
73
73
|
"oxlint": "^1.77.0",
|
|
74
74
|
"oxlint-tsgolint": "^7.0.2001",
|
|
75
75
|
"semantic-release": "25.0.9",
|