@postalsys/email-content 0.1.1 → 0.1.3
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/README.md +27 -19
- package/dist/.tsbuildinfo +1 -1
- package/dist/analyze.d.ts.map +1 -1
- package/dist/analyze.js +8 -6
- package/dist/analyze.js.map +1 -1
- package/dist/annotate/prompt.d.ts +2 -2
- package/dist/annotate/prompt.d.ts.map +1 -1
- package/dist/annotate/prompt.js +32 -13
- package/dist/annotate/prompt.js.map +1 -1
- package/dist/annotate/store.d.ts +8 -2
- package/dist/annotate/store.d.ts.map +1 -1
- package/dist/annotate/store.js +19 -5
- package/dist/annotate/store.js.map +1 -1
- package/dist/classify/rules.d.ts.map +1 -1
- package/dist/classify/rules.js +6 -6
- package/dist/classify/rules.js.map +1 -1
- package/dist/cli/annotate.d.ts.map +1 -1
- package/dist/cli/annotate.js +33 -11
- package/dist/cli/annotate.js.map +1 -1
- package/dist/cli/disagreements.d.ts.map +1 -1
- package/dist/cli/disagreements.js +18 -2
- package/dist/cli/disagreements.js.map +1 -1
- package/dist/cli/evaluate.d.ts +5 -3
- package/dist/cli/evaluate.d.ts.map +1 -1
- package/dist/cli/evaluate.js +28 -10
- package/dist/cli/evaluate.js.map +1 -1
- package/dist/detect/disclaimer.d.ts.map +1 -1
- package/dist/detect/disclaimer.js +9 -2
- package/dist/detect/disclaimer.js.map +1 -1
- package/dist/detect/patterns/disclaimers.json +47 -7
- package/dist/detect/patterns/html-clients.json +18 -0
- package/dist/detect/patterns/reply-headers.json +6 -10
- package/dist/detect/patterns/signatures.json +13 -5
- package/dist/detect/patterns.d.ts.map +1 -1
- package/dist/detect/patterns.js +11 -2
- package/dist/detect/patterns.js.map +1 -1
- package/dist/detect/reply-header.d.ts.map +1 -1
- package/dist/detect/reply-header.js +8 -2
- package/dist/detect/reply-header.js.map +1 -1
- package/dist/detect/signature.d.ts.map +1 -1
- package/dist/detect/signature.js +7 -2
- package/dist/detect/signature.js.map +1 -1
- package/dist/eval/metrics.d.ts +1 -0
- package/dist/eval/metrics.d.ts.map +1 -1
- package/dist/eval/metrics.js +16 -9
- package/dist/eval/metrics.js.map +1 -1
- package/dist/segment/html-blocks.d.ts.map +1 -1
- package/dist/segment/html-blocks.js +122 -3
- package/dist/segment/html-blocks.js.map +1 -1
- package/dist/types.d.ts +2 -2
- package/dist/types.js +2 -2
- package/package.json +9 -8
package/dist/cli/evaluate.js
CHANGED
|
@@ -7,9 +7,11 @@
|
|
|
7
7
|
* a change to the rules is actually measured instead of being compared against its own past
|
|
8
8
|
* output.
|
|
9
9
|
*
|
|
10
|
-
* The reference here is LLM produced. It is treated as a reference, not as truth
|
|
11
|
-
* models
|
|
12
|
-
*
|
|
10
|
+
* The reference here is LLM produced. It is treated as a reference, not as truth. Two strong
|
|
11
|
+
* models given identical input agree on 97.2 percent of blocks under the current conventions,
|
|
12
|
+
* 98.5 excluding three messages where one message-level judgment flips every block
|
|
13
|
+
* (docs/annotation.md), so that is the effective ceiling; the earlier 80 percent figure came from
|
|
14
|
+
* pairing a strong model with a weak one and measured the weak model, not annotation noise.
|
|
13
15
|
*/
|
|
14
16
|
var __importDefault = (this && this.__importDefault) || function (mod) {
|
|
15
17
|
return (mod && mod.__esModule) ? mod : { "default": mod };
|
|
@@ -71,9 +73,9 @@ function renderReport(report, title) {
|
|
|
71
73
|
out.push('');
|
|
72
74
|
if (report.byGroup.length > 1) {
|
|
73
75
|
out.push('BY GROUP');
|
|
74
|
-
out.push(' group blocks agreement new content retained');
|
|
76
|
+
out.push(' group blocks agreement new content retained history hidden');
|
|
75
77
|
for (const group of report.byGroup.slice(0, 15)) {
|
|
76
|
-
out.push(` ${group.group.padEnd(24)} ${String(group.blocks).padStart(6)} ${pct(group.agreement).padStart(10)} ${pct(group.newContentCharRecall).padStart(21)}`);
|
|
78
|
+
out.push(` ${group.group.padEnd(24)} ${String(group.blocks).padStart(6)} ${pct(group.agreement).padStart(10)} ${pct(group.newContentCharRecall).padStart(21)} ${pct(group.historyHiddenShare).padStart(15)}`);
|
|
77
79
|
}
|
|
78
80
|
out.push('');
|
|
79
81
|
}
|
|
@@ -86,6 +88,7 @@ async function evaluateCommand(args) {
|
|
|
86
88
|
return 2;
|
|
87
89
|
}
|
|
88
90
|
const model = (0, args_1.flagString)(args, 'model', 'gpt-5.6-sol');
|
|
91
|
+
const promptVersion = (0, args_1.flagString)(args, 'prompt-version', '');
|
|
89
92
|
const useHistory = (0, args_1.flagBool)(args, 'history', false);
|
|
90
93
|
const groupBy = (0, args_1.flagString)(args, 'group-by', 'client');
|
|
91
94
|
const limit = (0, args_1.flagNonNegativeInt)(args, 'limit', 0);
|
|
@@ -100,16 +103,23 @@ async function evaluateCommand(args) {
|
|
|
100
103
|
return 2;
|
|
101
104
|
}
|
|
102
105
|
// Latest annotation per (path, block) for the chosen model, so re-annotated messages do not
|
|
103
|
-
// count twice.
|
|
106
|
+
// count twice. An explicit --prompt-version pins the reference to labels made under one
|
|
107
|
+
// convention; without it, the newest label per block wins, whatever prompt produced it, and
|
|
108
|
+
// messages labelled only under an older convention are still included. The annotation guide's
|
|
109
|
+
// rule that runs made under different conventions must never be silently compared is the
|
|
110
|
+
// reason the flag exists; use it whenever more than one prompt version is present.
|
|
111
|
+
const versionClause = promptVersion ? ' AND r.prompt_version = ?' : '';
|
|
112
|
+
const versionClauseInner = promptVersion ? ' AND r2.prompt_version = ?' : '';
|
|
113
|
+
const params = promptVersion ? [model, promptVersion, promptVersion] : [model];
|
|
104
114
|
const reference = store.query(`SELECT a.path, a.block_index, a.mime_part, a.label, a.char_count, m.client, m.lang, m.reply_style
|
|
105
115
|
FROM annotations a
|
|
106
116
|
JOIN annotation_runs r ON r.id = a.run_id
|
|
107
117
|
JOIN messages m ON m.path = a.path
|
|
108
|
-
WHERE r.model = ? AND a.source = 'llm'
|
|
118
|
+
WHERE r.model = ? AND a.source = 'llm'${versionClause}
|
|
109
119
|
AND a.id = (SELECT MAX(a2.id) FROM annotations a2
|
|
110
120
|
JOIN annotation_runs r2 ON r2.id = a2.run_id
|
|
111
|
-
WHERE a2.path = a.path AND a2.block_index = a.block_index AND r2.model = r.model)
|
|
112
|
-
ORDER BY a.path, a.block_index`,
|
|
121
|
+
WHERE a2.path = a.path AND a2.block_index = a.block_index AND r2.model = r.model${versionClauseInner})
|
|
122
|
+
ORDER BY a.path, a.block_index`, ...params);
|
|
113
123
|
if (!reference.length) {
|
|
114
124
|
process.stderr.write(`no annotations found for model ${model}. Run "annotate" first.\n`);
|
|
115
125
|
store.close();
|
|
@@ -204,7 +214,15 @@ async function evaluateCommand(args) {
|
|
|
204
214
|
return best === null ? null : (predictions.get(key)?.get(best) ?? null);
|
|
205
215
|
};
|
|
206
216
|
const first = references[0];
|
|
207
|
-
const group = groupBy === '
|
|
217
|
+
const group = groupBy === 'history'
|
|
218
|
+
? historyDocuments?.length
|
|
219
|
+
? 'history available'
|
|
220
|
+
: 'history unavailable'
|
|
221
|
+
: groupBy === 'lang'
|
|
222
|
+
? (first?.lang ?? 'unknown')
|
|
223
|
+
: groupBy === 'reply-style'
|
|
224
|
+
? (first?.reply_style ?? 'unknown')
|
|
225
|
+
: (first?.client ?? 'unknown');
|
|
208
226
|
for (const item of references) {
|
|
209
227
|
const prediction = lookup(item.mime_part, item.block_index);
|
|
210
228
|
rows.push({
|
package/dist/cli/evaluate.js.map
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"evaluate.js","sourceRoot":"","sources":["../../src/cli/evaluate.ts"],"names":[],"mappings":";AAAA;;;;;;;;;;;GAWG;;;;;AA0FH,0CA+KC;AAvQD,sDAAyB;AACzB,0DAA6B;AAC7B,oCAAmC;AACnC,2CAA8C;AAC9C,yDAA8D;AAC9D,6CAAsF;AACtF,yCAA6C;AAC7C,oCAAwE;AACxE,iCAAmF;AAcnF,SAAS,GAAG,CAAC,KAAa;IACtB,OAAO,GAAG,CAAC,GAAG,GAAG,KAAK,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,GAAG,CAAC;AAC1C,CAAC;AAED,SAAS,YAAY,CAAC,MAAwB,EAAE,KAAa;IACzD,MAAM,GAAG,GAAa,EAAE,CAAC;IACzB,GAAG,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;IAChB,GAAG,CAAC,IAAI,CAAC,cAAc,MAAM,CAAC,QAAQ,YAAY,MAAM,CAAC,MAAM,2BAA2B,GAAG,CAAC,MAAM,CAAC,cAAc,CAAC,EAAE,CAAC,CAAC;IACxH,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IAEb,GAAG,CAAC,IAAI,CAAC,2CAA2C,CAAC,CAAC;IACtD,MAAM,CAAC,GAAG,MAAM,CAAC,SAAS,CAAC;IAC3B,GAAG,CAAC,IAAI,CAAC,uCAAuC,GAAG,CAAC,CAAC,CAAC,oBAAoB,CAAC,OAAO,CAAC,CAAC,uBAAuB,OAAO,CAAC,CAAC,eAAe,GAAG,CAAC,CAAC;IACxI,GAAG,CAAC,IAAI,CAAC,uCAAuC,GAAG,CAAC,CAAC,CAAC,qBAAqB,CAAC,OAAO,CAAC,CAAC,wBAAwB,OAAO,CAAC,CAAC,gBAAgB,GAAG,CAAC,CAAC;IAC3I,GAAG,CAAC,IAAI,CAAC,uCAAuC,GAAG,CAAC,CAAC,CAAC,eAAe,CAAC,OAAO,CAAC,CAAC,2BAA2B,OAAO,CAAC,CAAC,WAAW,GAAG,CAAC,CAAC;IACnI,GAAG,CAAC,IAAI,CAAC,uCAAuC,GAAG,CAAC,CAAC,CAAC,wBAAwB,CAAC,OAAO,CAAC,CAAC,wBAAwB,OAAO,CAAC,CAAC,YAAY,GAAG,CAAC,CAAC;IAC1I,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IAEb,GAAG,CAAC,IAAI,CAAC,WAAW,CAAC,CAAC;IACtB,GAAG,CAAC,IAAI,CAAC,oEAAoE,CAAC,CAAC;IAC/E,KAAK,MAAM,MAAM,IAAI,MAAM,CAAC,QAAQ,EAAE,CAAC;QACnC,IAAI,CAAC,MAAM,CAAC,OAAO,IAAI,CAAC,MAAM,CAAC,SAAS,EAAE,CAAC;YACvC,SAAS;QACb,CAAC;QACD,GAAG,CAAC,IAAI,CACJ,KAAK,MAAM,CAAC,KAAK,CAAC,MAAM,CAAC,EAAE,CAAC,IAAI,MAAM,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,IAAI,MAAM,CAAC,MAAM,CAAC,SAAS,CAAC,CAAC,QAAQ,CAAC,EAAE,CAAC,GAAG;YAC1G,GAAG,GAAG,CAAC,MAAM,CAAC,SAAS,CAAC,CAAC,QAAQ,CAAC,EAAE,CAAC,IAAI,GAAG,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,MAAM,CAAC,EAAE,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,EAAE,CAC9G,CAAC;IACN,CAAC;IACD,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IAEb,GAAG,CAAC,IAAI,CAAC,kDAAkD,CAAC,CAAC;IAC7D,MAAM,OAAO,GAAG,cAAM,CAAC,MAAM,CAAC,KAAK,CAAC,EAAE,CAAC,MAAM,CAAC,QAAQ,CAAC,IAAI,CAAC,MAAM,CAAC,EAAE,CAAC,MAAM,CAAC,KAAK,KAAK,KAAK,IAAI,CAAC,MAAM,CAAC,OAAO,IAAI,MAAM,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC;IACvI,GAAG,CAAC,IAAI,CAAC,KAAK,GAAG,CAAC,MAAM,CAAC,EAAE,CAAC,GAAG,OAAO,CAAC,GAAG,CAAC,KAAK,CAAC,EAAE,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,EAAE,CAAC,CAAC;IAC/F,KAAK,MAAM,SAAS,IAAI,OAAO,EAAE,CAAC;QAC9B,MAAM,KAAK,GAAG,OAAO,CAAC,GAAG,CAAC,SAAS,CAAC,EAAE,CAAC,MAAM,CAAC,MAAM,CAAC,SAAS,CAAC,GAAG,CAAC,GAAG,SAAS,IAAI,SAAS,EAAE,CAAC,IAAI,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QAC5H,GAAG,CAAC,IAAI,CAAC,KAAK,SAAS,CAAC,MAAM,CAAC,EAAE,CAAC,GAAG,KAAK,EAAE,CAAC,CAAC;IAClD,CAAC;IACD,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IAEb,GAAG,CAAC,IAAI,CAAC,4DAA4D,CAAC,CAAC;IACvE,MAAM,CAAC,GAAG,MAAM,CAAC,YAAY,CAAC;IAC9B,MAAM,IAAI,GAAG,CAAC,KAAa,EAAE,KAAa,EAAE,EAAE,CAAC,KAAK,KAAK,CAAC,MAAM,CAAC,EAAE,CAAC,IAAI,MAAM,CAAC,KAAK,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,QAAQ,CAAC,CAAC,EAAE,CAAC;IAC7I,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,yBAAyB,EAAE,CAAC,CAAC,oBAAoB,CAAC,CAAC,CAAC;IAClE,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,uBAAuB,EAAE,CAAC,CAAC,mBAAmB,CAAC,CAAC,CAAC;IAC/D,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,wBAAwB,EAAE,CAAC,CAAC,oBAAoB,CAAC,CAAC,CAAC;IACjE,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,6BAA6B,EAAE,CAAC,CAAC,kBAAkB,CAAC,CAAC,CAAC;IACpE,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,wBAAwB,EAAE,CAAC,CAAC,kBAAkB,CAAC,CAAC,CAAC;IAC/D,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,yCAAyC,EAAE,CAAC,CAAC,eAAe,CAAC,CAAC,CAAC;IAC7E,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,+BAA+B,EAAE,CAAC,CAAC,aAAa,CAAC,CAAC,CAAC;IACjE,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IAEb,IAAI,MAAM,CAAC,OAAO,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QAC5B,GAAG,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC;QACrB,GAAG,CAAC,IAAI,CAAC,qEAAqE,CAAC,CAAC;QAChF,KAAK,MAAM,KAAK,IAAI,MAAM,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,EAAE,CAAC;YAC9C,GAAG,CAAC,IAAI,CACJ,KAAK,KAAK,CAAC,KAAK,CAAC,MAAM,CAAC,EAAE,CAAC,IAAI,MAAM,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,KAAK,CAAC,SAAS,CAAC,CAAC,QAAQ,CAAC,EAAE,CAAC,IAAI,GAAG,CAAC,KAAK,CAAC,oBAAoB,CAAC,CAAC,QAAQ,CAAC,EAAE,CAAC,EAAE,CACzJ,CAAC;QACN,CAAC;QACD,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACjB,CAAC;IAED,OAAO,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;AAC1B,CAAC;AAEM,KAAK,UAAU,eAAe,CAAC,IAAgB;IAClD,MAAM,MAAM,GAAG,IAAA,iBAAU,EAAC,IAAI,EAAE,IAAI,EAAE,oBAAoB,CAAC,CAAC;IAC5D,IAAI,CAAC,iBAAE,CAAC,UAAU,CAAC,MAAM,CAAC,EAAE,CAAC;QACzB,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,yBAAyB,MAAM,IAAI,CAAC,CAAC;QAC1D,OAAO,CAAC,CAAC;IACb,CAAC;IAED,MAAM,KAAK,GAAG,IAAA,iBAAU,EAAC,IAAI,EAAE,OAAO,EAAE,aAAa,CAAC,CAAC;IACvD,MAAM,UAAU,GAAG,IAAA,eAAQ,EAAC,IAAI,EAAE,SAAS,EAAE,KAAK,CAAC,CAAC;IACpD,MAAM,OAAO,GAAG,IAAA,iBAAU,EAAC,IAAI,EAAE,UAAU,EAAE,QAAQ,CAAC,CAAC;IACvD,MAAM,KAAK,GAAG,IAAA,yBAAkB,EAAC,IAAI,EAAE,OAAO,EAAE,CAAC,CAAC,CAAC;IAEnD,MAAM,KAAK,GAAG,IAAI,mBAAW,CAAC,MAAM,CAAC,CAAC;IACtC,2FAA2F;IAC3F,6FAA6F;IAC7F,wDAAwD;IACxD,MAAM,IAAI,GAAG,KAAK,CAAC,kBAAkB,EAAE,EAAE,IAAI,CAAC;IAC9C,IAAI,CAAC,IAAI,EAAE,CAAC;QACR,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,+BAA+B,CAAC,CAAC;QACtD,KAAK,CAAC,KAAK,EAAE,CAAC;QACd,OAAO,CAAC,CAAC;IACb,CAAC;IAED,4FAA4F;IAC5F,eAAe;IACf,MAAM,SAAS,GAAG,KAAK,CAAC,KAAK,CACzB;;;;;;;;wCAQgC,EAChC,KAAK,CACR,CAAC;IAEF,IAAI,CAAC,SAAS,CAAC,MAAM,EAAE,CAAC;QACpB,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,kCAAkC,KAAK,2BAA2B,CAAC,CAAC;QACzF,KAAK,CAAC,KAAK,EAAE,CAAC;QACd,OAAO,CAAC,CAAC;IACb,CAAC;IAED,MAAM,MAAM,GAAG,IAAI,GAAG,EAA0B,CAAC;IACjD,KAAK,MAAM,GAAG,IAAI,SAAS,EAAE,CAAC;QAC1B,MAAM,IAAI,GAAG,MAAM,CAAC,GAAG,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC;QAClC,IAAI,IAAI,EAAE,CAAC;YACP,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QACnB,CAAC;aAAM,CAAC;YACJ,MAAM,CAAC,GAAG,CAAC,GAAG,CAAC,IAAI,EAAE,CAAC,GAAG,CAAC,CAAC,CAAC;QAChC,CAAC;IACL,CAAC;IAED,MAAM,KAAK,GAAG,CAAC,GAAG,MAAM,CAAC,IAAI,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,KAAK,IAAI,SAAS,CAAC,CAAC;IAC9D,MAAM,QAAQ,GAAG,UAAU,CAAC,CAAC,CAAC,IAAA,mCAAoB,EAAC,KAAK,EAAE,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC;IAEvE,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,cAAc,KAAK,CAAC,MAAM,qBAAqB,KAAK,GAAG,UAAU,CAAC,CAAC,CAAC,uBAAuB,CAAC,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC;IAE3H,MAAM,IAAI,GAAoB,EAAE,CAAC;IACjC,IAAI,QAAQ,GAAG,CAAC,CAAC;IACjB,IAAI,WAAW,GAAG,CAAC,CAAC;IAEpB,KAAK,MAAM,OAAO,IAAI,KAAK,EAAE,CAAC;QAC1B,MAAM,UAAU,GAAG,MAAM,CAAC,GAAG,CAAC,OAAO,CAAmB,CAAC;QACzD,IAAI,GAAW,CAAC;QAChB,IAAI,CAAC;YACD,GAAG,GAAG,iBAAE,CAAC,YAAY,CAAC,mBAAI,CAAC,IAAI,CAAC,IAAI,EAAE,OAAO,CAAC,CAAC,CAAC;QACpD,CAAC;QAAC,MAAM,CAAC;YACL,QAAQ,EAAE,CAAC;YACX,SAAS;QACb,CAAC;QAED,IAAI,gBAAgB,CAAC;QACrB,IAAI,QAAQ,EAAE,CAAC;YACX,IAAI,CAAC;gBACD,MAAM,OAAO,GAAG,MAAM,IAAA,oBAAY,EAAC,GAAG,CAAC,CAAC;gBACxC,gBAAgB,GAAG,MAAM,QAAQ,CAAC,OAAO,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC;gBAC3D,IAAI,gBAAgB,CAAC,MAAM,EAAE,CAAC;oBAC1B,WAAW,EAAE,CAAC;gBAClB,CAAC;YACL,CAAC;YAAC,MAAM,CAAC;gBACL,gBAAgB,GAAG,SAAS,CAAC;YACjC,CAAC;QACL,CAAC;QAED,IAAI,MAAM,CAAC;QACX,IAAI,CAAC;YACD,MAAM,GAAG,MAAM,IAAA,eAAO,EAAC,GAAG,EAAE,gBAAgB,CAAC,CAAC,CAAC,EAAE,gBAAgB,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC;QAC9E,CAAC;QAAC,MAAM,CAAC;YACL,QAAQ,EAAE,CAAC;YACX,SAAS;QACb,CAAC;QAED,2FAA2F;QAC3F,6EAA6E;QAC7E,EAAE;QACF,0FAA0F;QAC1F,2FAA2F;QAC3F,yFAAyF;QACzF,4FAA4F;QAC5F,4DAA4D;QAC5D,MAAM,WAAW,GAAG,IAAI,GAAG,EAA0D,CAAC;QACtF,MAAM,SAAS,GAAG,CAAC,IAAY,EAAE,EAAE;YAC/B,IAAI,MAAM,GAAG,WAAW,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC;YACnC,IAAI,CAAC,MAAM,EAAE,CAAC;gBACV,MAAM,GAAG,IAAI,GAAG,EAAE,CAAC;gBACnB,WAAW,CAAC,GAAG,CAAC,IAAI,EAAE,MAAM,CAAC,CAAC;YAClC,CAAC;YACD,OAAO,MAAM,CAAC;QAClB,CAAC,CAAC;QACF,KAAK,MAAM,OAAO,IAAI,CAAC,GAAG,MAAM,CAAC,QAAQ,EAAE,GAAG,MAAM,CAAC,cAAc,CAAC,EAAE,CAAC;YACnE,MAAM,UAAU,GAAG,IAAA,oBAAY,EAAC,OAAO,CAAC,MAAM,CAAC,CAAC;YAChD,IAAI,UAAU,KAAK,IAAI,EAAE,CAAC;gBACtB,SAAS,CAAC,IAAA,kBAAU,EAAC,OAAO,CAAC,MAAM,CAAC,IAAI,EAAE,CAAC,CAAC,GAAG,CAAC,UAAU,EAAE,EAAE,KAAK,EAAE,OAAO,CAAC,IAAI,EAAE,MAAM,EAAE,OAAO,CAAC,UAAU,KAAK,MAAM,EAAE,CAAC,CAAC;YAChI,CAAC;QACL,CAAC;QACD,MAAM,OAAO,GAAG,IAAI,GAAG,CAAC,CAAC,GAAG,WAAW,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,IAAI,EAAE,MAAM,CAAC,EAAE,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC,GAAG,MAAM,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;QAEpH,4FAA4F;QAC5F,MAAM,MAAM,GAAG,CAAC,IAAmB,EAAE,KAAa,EAA4C,EAAE;YAC5F,uFAAuF;YACvF,+DAA+D;YAC/D,MAAM,GAAG,GAAG,IAAI,KAAK,IAAI,IAAI,WAAW,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,WAAW,CAAC,IAAI,EAAE,CAAC,IAAI,EAAE,CAAC,KAAK,IAAI,EAAE,CAAC,CAAC;YACpG,MAAM,MAAM,GAAG,OAAO,CAAC,GAAG,CAAC,GAAG,CAAC,IAAI,EAAE,CAAC;YACtC,IAAI,IAAI,GAAkB,IAAI,CAAC;YAC/B,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;gBACzB,IAAI,KAAK,IAAI,KAAK,EAAE,CAAC;oBACjB,IAAI,GAAG,KAAK,CAAC;gBACjB,CAAC;qBAAM,CAAC;oBACJ,MAAM;gBACV,CAAC;YACL,CAAC;YACD,OAAO,IAAI,KAAK,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,WAAW,CAAC,GAAG,CAAC,GAAG,CAAC,EAAE,GAAG,CAAC,IAAI,CAAC,IAAI,IAAI,CAAC,CAAC;QAC5E,CAAC,CAAC;QAEF,MAAM,KAAK,GAAG,UAAU,CAAC,CAAC,CAAC,CAAC;QAC5B,MAAM,KAAK,GACP,OAAO,KAAK,MAAM,CAAC,CAAC,CAAC,CAAC,KAAK,EAAE,IAAI,IAAI,SAAS,CAAC,CAAC,CAAC,CAAC,OAAO,KAAK,aAAa,CAAC,CAAC,CAAC,CAAC,KAAK,EAAE,WAAW,IAAI,SAAS,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,EAAE,MAAM,IAAI,SAAS,CAAC,CAAC;QAEnJ,KAAK,MAAM,IAAI,IAAI,UAAU,EAAE,CAAC;YAC5B,MAAM,UAAU,GAAG,MAAM,CAAC,IAAI,CAAC,SAAS,EAAE,IAAI,CAAC,WAAW,CAAC,CAAC;YAC5D,IAAI,CAAC,IAAI,CAAC;gBACN,IAAI,EAAE,OAAO;gBACb,UAAU,EAAE,IAAI,CAAC,WAAW;gBAC5B,SAAS,EAAE,IAAI,CAAC,UAAU;gBAC1B,SAAS,EAAE,IAAI,CAAC,KAAK;gBACrB,gFAAgF;gBAChF,+EAA+E;gBAC/E,+BAA+B;gBAC/B,SAAS,EAAE,UAAU,EAAE,KAAK,IAAI,SAAS;gBACzC,MAAM,EAAE,UAAU,EAAE,MAAM,IAAI,KAAK;gBACnC,KAAK;aACR,CAAC,CAAC;QACP,CAAC;IACL,CAAC;IAED,KAAK,CAAC,KAAK,EAAE,CAAC;IAEd,IAAI,QAAQ,EAAE,CAAC;QACX,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,GAAG,QAAQ,sCAAsC,CAAC,CAAC;IAC5E,CAAC;IACD,IAAI,QAAQ,EAAE,CAAC;QACX,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,gCAAgC,WAAW,OAAO,KAAK,CAAC,MAAM,aAAa,CAAC,CAAC;IACtG,CAAC;IAED,MAAM,MAAM,GAAG,IAAA,kBAAQ,EAAC,IAAI,CAAC,CAAC;IAE9B,IAAI,IAAA,eAAQ,EAAC,IAAI,EAAE,MAAM,EAAE,KAAK,CAAC,EAAE,CAAC;QAChC,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,GAAG,IAAI,CAAC,SAAS,CAAC,EAAE,GAAG,MAAM,EAAE,SAAS,EAAE,MAAM,CAAC,WAAW,CAAC,MAAM,CAAC,SAAS,CAAC,EAAE,EAAE,IAAI,EAAE,CAAC,CAAC,IAAI,CAAC,CAAC;QACrH,OAAO,CAAC,CAAC;IACb,CAAC;IAED,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,GAAG,YAAY,CAAC,MAAM,EAAE,iBAAiB,KAAK,GAAG,UAAU,CAAC,CAAC,CAAC,iBAAiB,CAAC,CAAC,CAAC,eAAe,EAAE,CAAC,IAAI,CAAC,CAAC;IAC/H,OAAO,CAAC,CAAC;AACb,CAAC"}
|
|
1
|
+
{"version":3,"file":"evaluate.js","sourceRoot":"","sources":["../../src/cli/evaluate.ts"],"names":[],"mappings":";AAAA;;;;;;;;;;;;;GAaG;;;;;AA4FH,0CA+LC;AAzRD,sDAAyB;AACzB,0DAA6B;AAC7B,oCAAmC;AACnC,2CAA8C;AAC9C,yDAA8D;AAC9D,6CAAsF;AACtF,yCAA6C;AAC7C,oCAAwE;AACxE,iCAAmF;AAcnF,SAAS,GAAG,CAAC,KAAa;IACtB,OAAO,GAAG,CAAC,GAAG,GAAG,KAAK,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,GAAG,CAAC;AAC1C,CAAC;AAED,SAAS,YAAY,CAAC,MAAwB,EAAE,KAAa;IACzD,MAAM,GAAG,GAAa,EAAE,CAAC;IACzB,GAAG,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;IAChB,GAAG,CAAC,IAAI,CAAC,cAAc,MAAM,CAAC,QAAQ,YAAY,MAAM,CAAC,MAAM,2BAA2B,GAAG,CAAC,MAAM,CAAC,cAAc,CAAC,EAAE,CAAC,CAAC;IACxH,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IAEb,GAAG,CAAC,IAAI,CAAC,2CAA2C,CAAC,CAAC;IACtD,MAAM,CAAC,GAAG,MAAM,CAAC,SAAS,CAAC;IAC3B,GAAG,CAAC,IAAI,CAAC,uCAAuC,GAAG,CAAC,CAAC,CAAC,oBAAoB,CAAC,OAAO,CAAC,CAAC,uBAAuB,OAAO,CAAC,CAAC,eAAe,GAAG,CAAC,CAAC;IACxI,GAAG,CAAC,IAAI,CAAC,uCAAuC,GAAG,CAAC,CAAC,CAAC,qBAAqB,CAAC,OAAO,CAAC,CAAC,wBAAwB,OAAO,CAAC,CAAC,gBAAgB,GAAG,CAAC,CAAC;IAC3I,GAAG,CAAC,IAAI,CAAC,uCAAuC,GAAG,CAAC,CAAC,CAAC,eAAe,CAAC,OAAO,CAAC,CAAC,2BAA2B,OAAO,CAAC,CAAC,WAAW,GAAG,CAAC,CAAC;IACnI,GAAG,CAAC,IAAI,CAAC,uCAAuC,GAAG,CAAC,CAAC,CAAC,wBAAwB,CAAC,OAAO,CAAC,CAAC,wBAAwB,OAAO,CAAC,CAAC,YAAY,GAAG,CAAC,CAAC;IAC1I,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IAEb,GAAG,CAAC,IAAI,CAAC,WAAW,CAAC,CAAC;IACtB,GAAG,CAAC,IAAI,CAAC,oEAAoE,CAAC,CAAC;IAC/E,KAAK,MAAM,MAAM,IAAI,MAAM,CAAC,QAAQ,EAAE,CAAC;QACnC,IAAI,CAAC,MAAM,CAAC,OAAO,IAAI,CAAC,MAAM,CAAC,SAAS,EAAE,CAAC;YACvC,SAAS;QACb,CAAC;QACD,GAAG,CAAC,IAAI,CACJ,KAAK,MAAM,CAAC,KAAK,CAAC,MAAM,CAAC,EAAE,CAAC,IAAI,MAAM,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,IAAI,MAAM,CAAC,MAAM,CAAC,SAAS,CAAC,CAAC,QAAQ,CAAC,EAAE,CAAC,GAAG;YAC1G,GAAG,GAAG,CAAC,MAAM,CAAC,SAAS,CAAC,CAAC,QAAQ,CAAC,EAAE,CAAC,IAAI,GAAG,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,MAAM,CAAC,EAAE,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,EAAE,CAC9G,CAAC;IACN,CAAC;IACD,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IAEb,GAAG,CAAC,IAAI,CAAC,kDAAkD,CAAC,CAAC;IAC7D,MAAM,OAAO,GAAG,cAAM,CAAC,MAAM,CAAC,KAAK,CAAC,EAAE,CAAC,MAAM,CAAC,QAAQ,CAAC,IAAI,CAAC,MAAM,CAAC,EAAE,CAAC,MAAM,CAAC,KAAK,KAAK,KAAK,IAAI,CAAC,MAAM,CAAC,OAAO,IAAI,MAAM,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC;IACvI,GAAG,CAAC,IAAI,CAAC,KAAK,GAAG,CAAC,MAAM,CAAC,EAAE,CAAC,GAAG,OAAO,CAAC,GAAG,CAAC,KAAK,CAAC,EAAE,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,EAAE,CAAC,CAAC;IAC/F,KAAK,MAAM,SAAS,IAAI,OAAO,EAAE,CAAC;QAC9B,MAAM,KAAK,GAAG,OAAO,CAAC,GAAG,CAAC,SAAS,CAAC,EAAE,CAAC,MAAM,CAAC,MAAM,CAAC,SAAS,CAAC,GAAG,CAAC,GAAG,SAAS,IAAI,SAAS,EAAE,CAAC,IAAI,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QAC5H,GAAG,CAAC,IAAI,CAAC,KAAK,SAAS,CAAC,MAAM,CAAC,EAAE,CAAC,GAAG,KAAK,EAAE,CAAC,CAAC;IAClD,CAAC;IACD,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IAEb,GAAG,CAAC,IAAI,CAAC,4DAA4D,CAAC,CAAC;IACvE,MAAM,CAAC,GAAG,MAAM,CAAC,YAAY,CAAC;IAC9B,MAAM,IAAI,GAAG,CAAC,KAAa,EAAE,KAAa,EAAE,EAAE,CAAC,KAAK,KAAK,CAAC,MAAM,CAAC,EAAE,CAAC,IAAI,MAAM,CAAC,KAAK,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,QAAQ,CAAC,CAAC,EAAE,CAAC;IAC7I,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,yBAAyB,EAAE,CAAC,CAAC,oBAAoB,CAAC,CAAC,CAAC;IAClE,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,uBAAuB,EAAE,CAAC,CAAC,mBAAmB,CAAC,CAAC,CAAC;IAC/D,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,wBAAwB,EAAE,CAAC,CAAC,oBAAoB,CAAC,CAAC,CAAC;IACjE,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,6BAA6B,EAAE,CAAC,CAAC,kBAAkB,CAAC,CAAC,CAAC;IACpE,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,wBAAwB,EAAE,CAAC,CAAC,kBAAkB,CAAC,CAAC,CAAC;IAC/D,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,yCAAyC,EAAE,CAAC,CAAC,eAAe,CAAC,CAAC,CAAC;IAC7E,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,+BAA+B,EAAE,CAAC,CAAC,aAAa,CAAC,CAAC,CAAC;IACjE,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IAEb,IAAI,MAAM,CAAC,OAAO,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QAC5B,GAAG,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC;QACrB,GAAG,CAAC,IAAI,CAAC,qFAAqF,CAAC,CAAC;QAChG,KAAK,MAAM,KAAK,IAAI,MAAM,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,EAAE,CAAC;YAC9C,GAAG,CAAC,IAAI,CACJ,KAAK,KAAK,CAAC,KAAK,CAAC,MAAM,CAAC,EAAE,CAAC,IAAI,MAAM,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,KAAK,CAAC,SAAS,CAAC,CAAC,QAAQ,CAAC,EAAE,CAAC,IAAI,GAAG,CAAC,KAAK,CAAC,oBAAoB,CAAC,CAAC,QAAQ,CAAC,EAAE,CAAC,IAAI,GAAG,CACvJ,KAAK,CAAC,kBAAkB,CAC3B,CAAC,QAAQ,CAAC,EAAE,CAAC,EAAE,CACnB,CAAC;QACN,CAAC;QACD,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACjB,CAAC;IAED,OAAO,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;AAC1B,CAAC;AAEM,KAAK,UAAU,eAAe,CAAC,IAAgB;IAClD,MAAM,MAAM,GAAG,IAAA,iBAAU,EAAC,IAAI,EAAE,IAAI,EAAE,oBAAoB,CAAC,CAAC;IAC5D,IAAI,CAAC,iBAAE,CAAC,UAAU,CAAC,MAAM,CAAC,EAAE,CAAC;QACzB,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,yBAAyB,MAAM,IAAI,CAAC,CAAC;QAC1D,OAAO,CAAC,CAAC;IACb,CAAC;IAED,MAAM,KAAK,GAAG,IAAA,iBAAU,EAAC,IAAI,EAAE,OAAO,EAAE,aAAa,CAAC,CAAC;IACvD,MAAM,aAAa,GAAG,IAAA,iBAAU,EAAC,IAAI,EAAE,gBAAgB,EAAE,EAAE,CAAC,CAAC;IAC7D,MAAM,UAAU,GAAG,IAAA,eAAQ,EAAC,IAAI,EAAE,SAAS,EAAE,KAAK,CAAC,CAAC;IACpD,MAAM,OAAO,GAAG,IAAA,iBAAU,EAAC,IAAI,EAAE,UAAU,EAAE,QAAQ,CAAC,CAAC;IACvD,MAAM,KAAK,GAAG,IAAA,yBAAkB,EAAC,IAAI,EAAE,OAAO,EAAE,CAAC,CAAC,CAAC;IAEnD,MAAM,KAAK,GAAG,IAAI,mBAAW,CAAC,MAAM,CAAC,CAAC;IACtC,2FAA2F;IAC3F,6FAA6F;IAC7F,wDAAwD;IACxD,MAAM,IAAI,GAAG,KAAK,CAAC,kBAAkB,EAAE,EAAE,IAAI,CAAC;IAC9C,IAAI,CAAC,IAAI,EAAE,CAAC;QACR,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,+BAA+B,CAAC,CAAC;QACtD,KAAK,CAAC,KAAK,EAAE,CAAC;QACd,OAAO,CAAC,CAAC;IACb,CAAC;IAED,4FAA4F;IAC5F,wFAAwF;IACxF,4FAA4F;IAC5F,8FAA8F;IAC9F,yFAAyF;IACzF,mFAAmF;IACnF,MAAM,aAAa,GAAG,aAAa,CAAC,CAAC,CAAC,2BAA2B,CAAC,CAAC,CAAC,EAAE,CAAC;IACvE,MAAM,kBAAkB,GAAG,aAAa,CAAC,CAAC,CAAC,4BAA4B,CAAC,CAAC,CAAC,EAAE,CAAC;IAC7E,MAAM,MAAM,GAAa,aAAa,CAAC,CAAC,CAAC,CAAC,KAAK,EAAE,aAAa,EAAE,aAAa,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC;IACzF,MAAM,SAAS,GAAG,KAAK,CAAC,KAAK,CACzB;;;;iDAIyC,aAAa;;;yGAG2C,kBAAkB;wCACnF,EAChC,GAAG,MAAM,CACZ,CAAC;IAEF,IAAI,CAAC,SAAS,CAAC,MAAM,EAAE,CAAC;QACpB,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,kCAAkC,KAAK,2BAA2B,CAAC,CAAC;QACzF,KAAK,CAAC,KAAK,EAAE,CAAC;QACd,OAAO,CAAC,CAAC;IACb,CAAC;IAED,MAAM,MAAM,GAAG,IAAI,GAAG,EAA0B,CAAC;IACjD,KAAK,MAAM,GAAG,IAAI,SAAS,EAAE,CAAC;QAC1B,MAAM,IAAI,GAAG,MAAM,CAAC,GAAG,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC;QAClC,IAAI,IAAI,EAAE,CAAC;YACP,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QACnB,CAAC;aAAM,CAAC;YACJ,MAAM,CAAC,GAAG,CAAC,GAAG,CAAC,IAAI,EAAE,CAAC,GAAG,CAAC,CAAC,CAAC;QAChC,CAAC;IACL,CAAC;IAED,MAAM,KAAK,GAAG,CAAC,GAAG,MAAM,CAAC,IAAI,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,KAAK,IAAI,SAAS,CAAC,CAAC;IAC9D,MAAM,QAAQ,GAAG,UAAU,CAAC,CAAC,CAAC,IAAA,mCAAoB,EAAC,KAAK,EAAE,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC;IAEvE,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,cAAc,KAAK,CAAC,MAAM,qBAAqB,KAAK,GAAG,UAAU,CAAC,CAAC,CAAC,uBAAuB,CAAC,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC;IAE3H,MAAM,IAAI,GAAoB,EAAE,CAAC;IACjC,IAAI,QAAQ,GAAG,CAAC,CAAC;IACjB,IAAI,WAAW,GAAG,CAAC,CAAC;IAEpB,KAAK,MAAM,OAAO,IAAI,KAAK,EAAE,CAAC;QAC1B,MAAM,UAAU,GAAG,MAAM,CAAC,GAAG,CAAC,OAAO,CAAmB,CAAC;QACzD,IAAI,GAAW,CAAC;QAChB,IAAI,CAAC;YACD,GAAG,GAAG,iBAAE,CAAC,YAAY,CAAC,mBAAI,CAAC,IAAI,CAAC,IAAI,EAAE,OAAO,CAAC,CAAC,CAAC;QACpD,CAAC;QAAC,MAAM,CAAC;YACL,QAAQ,EAAE,CAAC;YACX,SAAS;QACb,CAAC;QAED,IAAI,gBAA0F,CAAC;QAC/F,IAAI,QAAQ,EAAE,CAAC;YACX,IAAI,CAAC;gBACD,MAAM,OAAO,GAAG,MAAM,IAAA,oBAAY,EAAC,GAAG,CAAC,CAAC;gBACxC,gBAAgB,GAAG,MAAM,QAAQ,CAAC,OAAO,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC;gBAC3D,IAAI,gBAAgB,CAAC,MAAM,EAAE,CAAC;oBAC1B,WAAW,EAAE,CAAC;gBAClB,CAAC;YACL,CAAC;YAAC,MAAM,CAAC;gBACL,gBAAgB,GAAG,SAAS,CAAC;YACjC,CAAC;QACL,CAAC;QAED,IAAI,MAAM,CAAC;QACX,IAAI,CAAC;YACD,MAAM,GAAG,MAAM,IAAA,eAAO,EAAC,GAAG,EAAE,gBAAgB,CAAC,CAAC,CAAC,EAAE,gBAAgB,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC;QAC9E,CAAC;QAAC,MAAM,CAAC;YACL,QAAQ,EAAE,CAAC;YACX,SAAS;QACb,CAAC;QAED,2FAA2F;QAC3F,6EAA6E;QAC7E,EAAE;QACF,0FAA0F;QAC1F,2FAA2F;QAC3F,yFAAyF;QACzF,4FAA4F;QAC5F,4DAA4D;QAC5D,MAAM,WAAW,GAAG,IAAI,GAAG,EAA0D,CAAC;QACtF,MAAM,SAAS,GAAG,CAAC,IAAY,EAAE,EAAE;YAC/B,IAAI,MAAM,GAAG,WAAW,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC;YACnC,IAAI,CAAC,MAAM,EAAE,CAAC;gBACV,MAAM,GAAG,IAAI,GAAG,EAAE,CAAC;gBACnB,WAAW,CAAC,GAAG,CAAC,IAAI,EAAE,MAAM,CAAC,CAAC;YAClC,CAAC;YACD,OAAO,MAAM,CAAC;QAClB,CAAC,CAAC;QACF,KAAK,MAAM,OAAO,IAAI,CAAC,GAAG,MAAM,CAAC,QAAQ,EAAE,GAAG,MAAM,CAAC,cAAc,CAAC,EAAE,CAAC;YACnE,MAAM,UAAU,GAAG,IAAA,oBAAY,EAAC,OAAO,CAAC,MAAM,CAAC,CAAC;YAChD,IAAI,UAAU,KAAK,IAAI,EAAE,CAAC;gBACtB,SAAS,CAAC,IAAA,kBAAU,EAAC,OAAO,CAAC,MAAM,CAAC,IAAI,EAAE,CAAC,CAAC,GAAG,CAAC,UAAU,EAAE,EAAE,KAAK,EAAE,OAAO,CAAC,IAAI,EAAE,MAAM,EAAE,OAAO,CAAC,UAAU,KAAK,MAAM,EAAE,CAAC,CAAC;YAChI,CAAC;QACL,CAAC;QACD,MAAM,OAAO,GAAG,IAAI,GAAG,CAAC,CAAC,GAAG,WAAW,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,IAAI,EAAE,MAAM,CAAC,EAAE,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC,GAAG,MAAM,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;QAEpH,4FAA4F;QAC5F,MAAM,MAAM,GAAG,CAAC,IAAmB,EAAE,KAAa,EAA4C,EAAE;YAC5F,uFAAuF;YACvF,+DAA+D;YAC/D,MAAM,GAAG,GAAG,IAAI,KAAK,IAAI,IAAI,WAAW,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,WAAW,CAAC,IAAI,EAAE,CAAC,IAAI,EAAE,CAAC,KAAK,IAAI,EAAE,CAAC,CAAC;YACpG,MAAM,MAAM,GAAG,OAAO,CAAC,GAAG,CAAC,GAAG,CAAC,IAAI,EAAE,CAAC;YACtC,IAAI,IAAI,GAAkB,IAAI,CAAC;YAC/B,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;gBACzB,IAAI,KAAK,IAAI,KAAK,EAAE,CAAC;oBACjB,IAAI,GAAG,KAAK,CAAC;gBACjB,CAAC;qBAAM,CAAC;oBACJ,MAAM;gBACV,CAAC;YACL,CAAC;YACD,OAAO,IAAI,KAAK,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,WAAW,CAAC,GAAG,CAAC,GAAG,CAAC,EAAE,GAAG,CAAC,IAAI,CAAC,IAAI,IAAI,CAAC,CAAC;QAC5E,CAAC,CAAC;QAEF,MAAM,KAAK,GAAG,UAAU,CAAC,CAAC,CAAC,CAAC;QAC5B,MAAM,KAAK,GACP,OAAO,KAAK,SAAS;YACjB,CAAC,CAAC,gBAAgB,EAAE,MAAM;gBACtB,CAAC,CAAC,mBAAmB;gBACrB,CAAC,CAAC,qBAAqB;YAC3B,CAAC,CAAC,OAAO,KAAK,MAAM;gBAClB,CAAC,CAAC,CAAC,KAAK,EAAE,IAAI,IAAI,SAAS,CAAC;gBAC5B,CAAC,CAAC,OAAO,KAAK,aAAa;oBACzB,CAAC,CAAC,CAAC,KAAK,EAAE,WAAW,IAAI,SAAS,CAAC;oBACnC,CAAC,CAAC,CAAC,KAAK,EAAE,MAAM,IAAI,SAAS,CAAC,CAAC;QAE3C,KAAK,MAAM,IAAI,IAAI,UAAU,EAAE,CAAC;YAC5B,MAAM,UAAU,GAAG,MAAM,CAAC,IAAI,CAAC,SAAS,EAAE,IAAI,CAAC,WAAW,CAAC,CAAC;YAC5D,IAAI,CAAC,IAAI,CAAC;gBACN,IAAI,EAAE,OAAO;gBACb,UAAU,EAAE,IAAI,CAAC,WAAW;gBAC5B,SAAS,EAAE,IAAI,CAAC,UAAU;gBAC1B,SAAS,EAAE,IAAI,CAAC,KAAK;gBACrB,gFAAgF;gBAChF,+EAA+E;gBAC/E,+BAA+B;gBAC/B,SAAS,EAAE,UAAU,EAAE,KAAK,IAAI,SAAS;gBACzC,MAAM,EAAE,UAAU,EAAE,MAAM,IAAI,KAAK;gBACnC,KAAK;aACR,CAAC,CAAC;QACP,CAAC;IACL,CAAC;IAED,KAAK,CAAC,KAAK,EAAE,CAAC;IAEd,IAAI,QAAQ,EAAE,CAAC;QACX,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,GAAG,QAAQ,sCAAsC,CAAC,CAAC;IAC5E,CAAC;IACD,IAAI,QAAQ,EAAE,CAAC;QACX,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,gCAAgC,WAAW,OAAO,KAAK,CAAC,MAAM,aAAa,CAAC,CAAC;IACtG,CAAC;IAED,MAAM,MAAM,GAAG,IAAA,kBAAQ,EAAC,IAAI,CAAC,CAAC;IAE9B,IAAI,IAAA,eAAQ,EAAC,IAAI,EAAE,MAAM,EAAE,KAAK,CAAC,EAAE,CAAC;QAChC,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,GAAG,IAAI,CAAC,SAAS,CAAC,EAAE,GAAG,MAAM,EAAE,SAAS,EAAE,MAAM,CAAC,WAAW,CAAC,MAAM,CAAC,SAAS,CAAC,EAAE,EAAE,IAAI,EAAE,CAAC,CAAC,IAAI,CAAC,CAAC;QACrH,OAAO,CAAC,CAAC;IACb,CAAC;IAED,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,GAAG,YAAY,CAAC,MAAM,EAAE,iBAAiB,KAAK,GAAG,UAAU,CAAC,CAAC,CAAC,iBAAiB,CAAC,CAAC,CAAC,eAAe,EAAE,CAAC,IAAI,CAAC,CAAC;IAC/H,OAAO,CAAC,CAAC;AACb,CAAC"}
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"disclaimer.d.ts","sourceRoot":"","sources":["../../src/detect/disclaimer.ts"],"names":[],"mappings":"AAAA;;;;;;;GAOG;AAGH,OAAO,EAAE,KAAK,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AACjD,OAAO,EAAE,KAAK,aAAa,EAAE,MAAM,kBAAkB,CAAC;
|
|
1
|
+
{"version":3,"file":"disclaimer.d.ts","sourceRoot":"","sources":["../../src/detect/disclaimer.ts"],"names":[],"mappings":"AAAA;;;;;;;GAOG;AAGH,OAAO,EAAE,KAAK,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AACjD,OAAO,EAAE,KAAK,aAAa,EAAE,MAAM,kBAAkB,CAAC;AAYtD,wBAAgB,gBAAgB,CAAC,IAAI,EAAE,MAAM,EAAE,QAAQ,EAAE,aAAa,GAAG,QAAQ,EAAE,CAYlF"}
|
|
@@ -10,8 +10,15 @@
|
|
|
10
10
|
Object.defineProperty(exports, "__esModule", { value: true });
|
|
11
11
|
exports.detectDisclaimer = detectDisclaimer;
|
|
12
12
|
const patterns_1 = require("./patterns");
|
|
13
|
-
/**
|
|
14
|
-
|
|
13
|
+
/**
|
|
14
|
+
* Disclaimers live after the sender's own words, but not necessarily late in the document: in a
|
|
15
|
+
* notification reply the footer sits between a short comment and a long quoted thread, which puts
|
|
16
|
+
* it early by character position. The old 0.4 gate looked like it worked only because HTML
|
|
17
|
+
* positions were measured against raw markup, where a large style block pushed every visible block
|
|
18
|
+
* past the gate. Once positions became honest the gate cost half of measured recall while the
|
|
19
|
+
* pattern tightening already carried the precision, so only the opening slice is excluded now.
|
|
20
|
+
*/
|
|
21
|
+
const MIN_POSITION = 0.1;
|
|
15
22
|
function detectDisclaimer(text, features) {
|
|
16
23
|
if (features.position < MIN_POSITION || features.isAllBlank) {
|
|
17
24
|
return [];
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"disclaimer.js","sourceRoot":"","sources":["../../src/detect/disclaimer.ts"],"names":[],"mappings":";AAAA;;;;;;;GAOG;;
|
|
1
|
+
{"version":3,"file":"disclaimer.js","sourceRoot":"","sources":["../../src/detect/disclaimer.ts"],"names":[],"mappings":";AAAA;;;;;;;GAOG;;AAgBH,4CAYC;AA1BD,yCAA0C;AAI1C;;;;;;;GAOG;AACH,MAAM,YAAY,GAAG,GAAG,CAAC;AAEzB,SAAgB,gBAAgB,CAAC,IAAY,EAAE,QAAuB;IAClE,IAAI,QAAQ,CAAC,QAAQ,GAAG,YAAY,IAAI,QAAQ,CAAC,UAAU,EAAE,CAAC;QAC1D,OAAO,EAAE,CAAC;IACd,CAAC;IACD,MAAM,QAAQ,GAAG,IAAA,uBAAY,GAAE,CAAC;IAChC,MAAM,KAAK,GAAe,EAAE,CAAC;IAC7B,KAAK,MAAM,OAAO,IAAI,QAAQ,CAAC,iBAAiB,EAAE,CAAC;QAC/C,IAAI,OAAO,CAAC,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;YAC3B,KAAK,CAAC,IAAI,CAAC,EAAE,EAAE,EAAE,cAAc,OAAO,CAAC,EAAE,EAAE,EAAE,IAAI,EAAE,YAAY,EAAE,QAAQ,EAAE,OAAO,CAAC,QAAQ,EAAE,CAAC,CAAC;QACnG,CAAC;IACL,CAAC;IACD,OAAO,KAAK,CAAC;AACjB,CAAC"}
|
|
@@ -39,15 +39,17 @@
|
|
|
39
39
|
},
|
|
40
40
|
{
|
|
41
41
|
"id": "confidentiality_es",
|
|
42
|
+
"comment": "The bare word destinatario used to be an alternative on its own, so any Spanish sentence mentioning a recipient counted as a 0.9 disclaimer. The recipient reference now needs its boilerplate context.",
|
|
42
43
|
"strength": 0.9,
|
|
43
44
|
"flags": "i",
|
|
44
|
-
"pattern": "(Este mensaje .{0,60}confidencial|informaci.n confidencial|destinatario)"
|
|
45
|
+
"pattern": "(Este (mensaje|correo) .{0,60}confidencial|informaci.n confidencial|uso exclusivo de(l)? destinatario|si (usted )?no es el destinatario)"
|
|
45
46
|
},
|
|
46
47
|
{
|
|
47
48
|
"id": "gdpr_notice",
|
|
49
|
+
"comment": "The bare token GDPR and the bare phrase privacy policy both appear in ordinary work prose, so each alternative now needs its footer context.",
|
|
48
50
|
"strength": 0.7,
|
|
49
51
|
"flags": "i",
|
|
50
|
-
"pattern": "(privacy (notice|policy)|
|
|
52
|
+
"pattern": "((see|view|read|review|check) (our|the) privacy (notice|policy)|privacy (notice|policy)\\s*[:|]|in accordance with (the )?(GDPR|General Data Protection Regulation)|process(es|ing)? (of )?your personal data|isikuandmete töötlemise)"
|
|
51
53
|
},
|
|
52
54
|
{
|
|
53
55
|
"id": "company_registration",
|
|
@@ -57,9 +59,10 @@
|
|
|
57
59
|
},
|
|
58
60
|
{
|
|
59
61
|
"id": "unsubscribe_footer",
|
|
62
|
+
"comment": "The bare word unsubscribe also appears in prose about mailing lists, so it only counts standing alone on its own line, which is the bare-link-text shape, or with its imperative footer context.",
|
|
60
63
|
"strength": 0.8,
|
|
61
|
-
"flags": "
|
|
62
|
-
"pattern": "(unsubscribe|opt out of these emails|manage your (email )?preferences|loobu (uudiskirjast|teavitustest))"
|
|
64
|
+
"flags": "im",
|
|
65
|
+
"pattern": "(to unsubscribe|unsubscribe (from|at|here|now|at any time)|click here to unsubscribe|^\\s*unsubscribe\\s*\\.?\\s*$|if you (no longer |don'?t )?(wish|want) to (receive|get)|opt out of these emails|manage your (email )?preferences|loobu (uudiskirjast|teavitustest))"
|
|
63
66
|
},
|
|
64
67
|
{
|
|
65
68
|
"id": "confidentiality_et",
|
|
@@ -67,6 +70,43 @@
|
|
|
67
70
|
"flags": "i",
|
|
68
71
|
"pattern": "(Käesolev (e-?kiri|sõnum) .{0,60}(konfidentsiaal|asutusesisene)|kui te ei ole .{0,40}saaja)"
|
|
69
72
|
},
|
|
73
|
+
{
|
|
74
|
+
"id": "confidentiality_it",
|
|
75
|
+
"comment": "Italian, Portuguese, Dutch, Polish, Russian, Finnish and Scandinavian mail all had reply-header coverage but no disclaimer coverage at all. These are the standard legalese openings, same shape as the English and German entries.",
|
|
76
|
+
"strength": 0.9,
|
|
77
|
+
"flags": "i",
|
|
78
|
+
"pattern": "(Le informazioni contenute .{0,60}(riservat|confidenzial)|Questo messaggio .{0,60}(riservat|confidenzial)|Se non siete il destinatario)"
|
|
79
|
+
},
|
|
80
|
+
{
|
|
81
|
+
"id": "confidentiality_pt",
|
|
82
|
+
"strength": 0.9,
|
|
83
|
+
"flags": "i",
|
|
84
|
+
"pattern": "(Esta mensagem .{0,60}confidencial|informa..o confidencial|n.o (é|e|for) o destinat.rio)"
|
|
85
|
+
},
|
|
86
|
+
{
|
|
87
|
+
"id": "confidentiality_nl",
|
|
88
|
+
"strength": 0.9,
|
|
89
|
+
"flags": "i",
|
|
90
|
+
"pattern": "(Dit bericht .{0,60}vertrouwelijk|De informatie .{0,60}vertrouwelijk|niet de beoogde ontvanger)"
|
|
91
|
+
},
|
|
92
|
+
{
|
|
93
|
+
"id": "confidentiality_pl",
|
|
94
|
+
"strength": 0.9,
|
|
95
|
+
"flags": "i",
|
|
96
|
+
"pattern": "(Ta wiadomo.. .{0,60}poufn|informacje poufne|nie jeste. .{0,40}adresatem)"
|
|
97
|
+
},
|
|
98
|
+
{
|
|
99
|
+
"id": "confidentiality_ru",
|
|
100
|
+
"strength": 0.9,
|
|
101
|
+
"flags": "i",
|
|
102
|
+
"pattern": "(сообщение .{0,60}конфиденциальн|конфиденциальную информацию|не являетесь .{0,40}(получателем|адресатом))"
|
|
103
|
+
},
|
|
104
|
+
{
|
|
105
|
+
"id": "confidentiality_fi_sv_da",
|
|
106
|
+
"strength": 0.9,
|
|
107
|
+
"flags": "i",
|
|
108
|
+
"pattern": "(T.m. viesti .{0,60}luottamuksellinen|luottamuksellista tietoa|Detta meddelande .{0,60}konfidentiellt|Denne (e-?mail|meddelelse|melding) .{0,60}fortrolig)"
|
|
109
|
+
},
|
|
70
110
|
{
|
|
71
111
|
"id": "automated_do_not_reply",
|
|
72
112
|
"strength": 0.85,
|
|
@@ -98,9 +138,9 @@
|
|
|
98
138
|
{
|
|
99
139
|
"id": "powered_by_footer",
|
|
100
140
|
"strength": 0.7,
|
|
101
|
-
"flags": "
|
|
102
|
-
"comment": "Provider branding appended by the sending platform.",
|
|
103
|
-
"pattern": "(
|
|
141
|
+
"flags": "m",
|
|
142
|
+
"comment": "Provider branding appended by the sending platform. Case sensitive on purpose: under the i flag the capital brand class matched any word, so prose like 'I sent via courier yesterday' counted as branding. The line anchor keeps it to the standalone footer shape.",
|
|
143
|
+
"pattern": "^\\s*([Pp]owered by|[Ss]ent (via|with|using)|[Dd]elivered by) [A-Z][\\w.'-]+.{0,40}$"
|
|
104
144
|
},
|
|
105
145
|
{
|
|
106
146
|
"id": "support_contact_footer",
|
|
@@ -65,6 +65,24 @@
|
|
|
65
65
|
"selector": "p.MsoNormal",
|
|
66
66
|
"prevalence": 0.16
|
|
67
67
|
},
|
|
68
|
+
{
|
|
69
|
+
"id": "outlook_signature",
|
|
70
|
+
"client": "outlook-desktop",
|
|
71
|
+
"role": "signature",
|
|
72
|
+
"strength": 0.9,
|
|
73
|
+
"selector": "div#Signature",
|
|
74
|
+
"prevalence": 0.02,
|
|
75
|
+
"comment": "Outlook desktop and Outlook Web both wrap the composed signature in an element with id Signature. Only Gmail, Thunderbird and Proton carried a signature selector before this."
|
|
76
|
+
},
|
|
77
|
+
{
|
|
78
|
+
"id": "outlook_mobile_signature",
|
|
79
|
+
"client": "outlook-mobile",
|
|
80
|
+
"role": "signature",
|
|
81
|
+
"strength": 0.9,
|
|
82
|
+
"selector": "div.ms-outlook-mobile-signature",
|
|
83
|
+
"prevalence": 0.005,
|
|
84
|
+
"comment": "Outlook mobile marks its Get Outlook for iOS/Android footer with this class."
|
|
85
|
+
},
|
|
68
86
|
{
|
|
69
87
|
"id": "outlook_word_section",
|
|
70
88
|
"client": "outlook-desktop",
|
|
@@ -19,9 +19,10 @@
|
|
|
19
19
|
{
|
|
20
20
|
"id": "x_wrote_trailing",
|
|
21
21
|
"lang": "en",
|
|
22
|
+
"comment": "Attribution without an On prefix, such as 'Bob Smith wrote:'. The lookaheads reject first person prose like 'As I wrote:' or 'Like I wrote:', which is the sender talking about themselves, not a client inserted attribution.",
|
|
22
23
|
"strength": 0.8,
|
|
23
24
|
"flags": "i",
|
|
24
|
-
"pattern": "^\\s
|
|
25
|
+
"pattern": "^\\s*(?!(as|like|just as|so|and|but|because|remember)\\b)(?!I\\b).{3,200}?\\bwrote\\s*:\\s*$"
|
|
25
26
|
},
|
|
26
27
|
{
|
|
27
28
|
"id": "at_x_wrote",
|
|
@@ -96,16 +97,18 @@
|
|
|
96
97
|
{
|
|
97
98
|
"id": "den_x_skrev_sv",
|
|
98
99
|
"lang": "sv",
|
|
100
|
+
"comment": "Two real shapes: 'Den <date> skrev <name>:' with the name after the verb, and the bare 'X skrev:' form matching the Estonian and Finnish family. The old '(Den|.{0,60})' alternation made Den dead and accepted any line with skrev anywhere before a colon.",
|
|
99
101
|
"strength": 0.95,
|
|
100
102
|
"flags": "i",
|
|
101
|
-
"pattern": "^\\s*(Den
|
|
103
|
+
"pattern": "^\\s*(Den\\s+.{3,200}?\\bskrev\\s+.{0,80}:|.{3,200}?\\bskrev\\s*:)\\s*$"
|
|
102
104
|
},
|
|
103
105
|
{
|
|
104
106
|
"id": "x_pisal_ru",
|
|
105
107
|
"lang": "ru",
|
|
108
|
+
"comment": "The lookahead rejects first person prose such as 'как я писал:', which is the sender referring to themselves, not an attribution.",
|
|
106
109
|
"strength": 0.95,
|
|
107
110
|
"flags": "i",
|
|
108
|
-
"pattern": "^\\s
|
|
111
|
+
"pattern": "^\\s*(?!(как|я)\\s).{3,200}?(писал|написал)\\s*.{0,40}:\\s*$"
|
|
109
112
|
},
|
|
110
113
|
{
|
|
111
114
|
"id": "date_addr_colon",
|
|
@@ -127,13 +130,6 @@
|
|
|
127
130
|
"strength": 0.95,
|
|
128
131
|
"flags": "i",
|
|
129
132
|
"pattern": "^\\s*-{2,}\\s*(Reply Message|Antwort|Svar)\\s*-{2,}\\s*$"
|
|
130
|
-
},
|
|
131
|
-
{
|
|
132
|
-
"id": "outlook_underscore_rule",
|
|
133
|
-
"lang": "*",
|
|
134
|
-
"strength": 0.6,
|
|
135
|
-
"flags": "",
|
|
136
|
-
"pattern": "^_{10,}\\s*$"
|
|
137
133
|
}
|
|
138
134
|
]
|
|
139
135
|
}
|
|
@@ -31,14 +31,15 @@
|
|
|
31
31
|
"id": "best_regards",
|
|
32
32
|
"strength": 0.35,
|
|
33
33
|
"flags": "i",
|
|
34
|
-
"pattern": "^\\s*(Best regards|Kind regards|Warm regards|Best wishes|Regards|Cheers|Thanks|Thank you|Sincerely|Best|Yours( sincerely| truly)?|Mit freundlichen Grüßen|Viele Grüße|Beste Grüße|Cordialement|Bien à vous|Saludos|Atentamente|Cordiali saluti|Groeten|Met vriendelijke groet|Med vänliga hälsningar|Vänliga hälsningar|Mvh|Ystävällisin terveisin|Terveisin|Parimat|Parimate soovidega|Heade soovidega|Lugupidamisega|Tervitades|Tervitused|Terv|Kõike head|Ilusat päeva|Head aega|Austusega|С уважением|Z poważaniem|Pozdrawiam)\\s*[
|
|
35
|
-
"comment": "Closing greeting. Mined from blocks the reference labelled SIGNATURE that this baseline missed, which is why the Estonian forms are here: the working corpus is roughly ten percent Estonian and those accounted for the largest group of misses."
|
|
34
|
+
"pattern": "^\\s*(Best regards|Kind regards|Warm regards|Warmest regards|With (kind|best|warm) regards|Best wishes|All the best|Take care|Regards|Cheers|Thanks|Many thanks|Thanks (again|in advance)|Thank you|Sincerely|Best|Yours( sincerely| truly)?|Br|Rgds|Mit freundlichen Grüßen|Viele Grüße|Beste Grüße|Cordialement|Bien à vous|Saludos|Atentamente|Cordiali saluti|Groeten|Met vriendelijke groet|Med vänliga hälsningar|Vänliga hälsningar|Med venlig hilsen|Venlig hilsen|Med vennlig hilsen|Vennlig hilsen|Mvh|S pozdravem|Ystävällisin terveisin|Terveisin|Parimat|Parimate soovidega|Heade soovidega|Lugupidamisega|Tervitades|Tervitused|Terv|Kõike head|Ilusat päeva|Head aega|Austusega|Aitäh|Tänan|Tänud|Ette tänades|С уважением|Z poważaniem|Pozdrawiam)\\s*[,.!;]?\\s*[-\\u2013\\u2014]?\\s*$",
|
|
35
|
+
"comment": "Closing greeting. Mined from blocks the reference labelled SIGNATURE that this baseline missed, which is why the Estonian forms are here: the working corpus is roughly ten percent Estonian and those accounted for the largest group of misses. The greeting must still be the whole line; only a trailing comma, period or dash is tolerated, because 'Thanks, that works.' is prose, not a sign-off."
|
|
36
36
|
},
|
|
37
37
|
{
|
|
38
38
|
"id": "contact_line",
|
|
39
|
+
"comment": "Labelled phone, email or web line. The email and web alternatives need their own value shape: an address or URL never starts with a digit, so folding them into the phone branch made them unmatchable.",
|
|
39
40
|
"strength": 0.3,
|
|
40
41
|
"flags": "i",
|
|
41
|
-
"pattern": "^\\s*(Tel|Phone|Mobile|Mob|Fax|Cell|M|T
|
|
42
|
+
"pattern": "^\\s*((Tel|Phone|Mobile|Mob|Fax|Cell|GSM|M|T)\\s*[:.]?\\s*[+0-9(][-0-9()\\s.]{6,}(\\s*(x|ext\\.?)\\s*\\d{1,5})?|(E|E-?mail|Mail|W|Web|WWW)\\s*[:.]\\s*(\\S+@\\S+\\.\\S+|https?:\\/\\/\\S+|www\\.\\S+))\\s*$"
|
|
42
43
|
},
|
|
43
44
|
{
|
|
44
45
|
"id": "company_registration_line",
|
|
@@ -49,10 +50,17 @@
|
|
|
49
50
|
},
|
|
50
51
|
{
|
|
51
52
|
"id": "postal_address_line",
|
|
52
|
-
"comment": "A street
|
|
53
|
+
"comment": "A street line in a trailing block. Weak on its own, but it pairs with a name or a closing greeting to identify a corporate signature.",
|
|
53
54
|
"strength": 0.4,
|
|
54
55
|
"flags": "i",
|
|
55
|
-
"pattern": "^\\s*([A-ZÄÖÜÕŠŽ][\\w.-]*\\s+)?(Str\\.|Straße|Strasse|Street|Avenue|Ave\\.|Road|Rd\\.|tee|tänav|pst)\\s*\\.?\\s*\\d
|
|
56
|
+
"pattern": "^\\s*([A-ZÄÖÜÕŠŽ][\\w.-]*\\s+)?(Str\\.|Straße|Strasse|Street|Avenue|Ave\\.|Road|Rd\\.|tee|tänav|pst)\\s*\\.?\\s*\\d"
|
|
57
|
+
},
|
|
58
|
+
{
|
|
59
|
+
"id": "postal_code_line",
|
|
60
|
+
"comment": "A postal code followed by a capitalised city name. Deliberately case sensitive: under the i flag the capital letter class matched any word, so a line like '12345 units shipped' counted as an address.",
|
|
61
|
+
"strength": 0.4,
|
|
62
|
+
"flags": "",
|
|
63
|
+
"pattern": "^\\s*[A-Z]{0,2}-?\\d{4,6}\\s+[A-ZÄÖÜÕ]"
|
|
56
64
|
}
|
|
57
65
|
]
|
|
58
66
|
}
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"patterns.d.ts","sourceRoot":"","sources":["../../src/detect/patterns.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;GAUG;AAKH,MAAM,WAAW,WAAW;IACxB,EAAE,EAAE,MAAM,CAAC;IACX,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,QAAQ,EAAE,MAAM,CAAC;IACjB,KAAK,EAAE,MAAM,CAAC;CACjB;AAED,MAAM,WAAW,eAAe;IAC5B,EAAE,EAAE,MAAM,CAAC;IACX,MAAM,EAAE,MAAM,CAAC;IACf,IAAI,EAAE,OAAO,GAAG,aAAa,GAAG,WAAW,GAAG,WAAW,GAAG,YAAY,GAAG,KAAK,GAAG,MAAM,CAAC;IAC1F,QAAQ,EAAE,MAAM,CAAC;IACjB,QAAQ,EAAE,MAAM,CAAC;IACjB,UAAU,EAAE,MAAM,CAAC;CACtB;AAED,MAAM,WAAW,WAAW;IACxB,YAAY,EAAE,WAAW,EAAE,CAAC;IAC5B,cAAc,EAAE,WAAW,EAAE,CAAC;IAC9B,kBAAkB,EAAE,WAAW,EAAE,CAAC;IAClC,mBAAmB,EAAE,WAAW,EAAE,CAAC;IACnC,gBAAgB,EAAE,WAAW,EAAE,CAAC;IAChC,iBAAiB,EAAE,WAAW,EAAE,CAAC;IACjC,aAAa,EAAE,eAAe,EAAE,CAAC;IACjC,+DAA+D;IAC/D,OAAO,EAAE,MAAM,EAAE,CAAC;CACrB;AAcD;;;;;;;;GAQG;AACH,qBAAa,iBAAkB,SAAQ,KAAK;IACxC,QAAQ,CAAC,IAAI,6BAA6B;IAC1C,QAAQ,CAAC,IAAI,EAAE,MAAM,GAAG,IAAI,CAAC;gBAEjB,OAAO,EAAE,MAAM,EAAE,IAAI,GAAE,MAAM,GAAG,IAAW;CAK1D;
|
|
1
|
+
{"version":3,"file":"patterns.d.ts","sourceRoot":"","sources":["../../src/detect/patterns.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;GAUG;AAKH,MAAM,WAAW,WAAW;IACxB,EAAE,EAAE,MAAM,CAAC;IACX,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,QAAQ,EAAE,MAAM,CAAC;IACjB,KAAK,EAAE,MAAM,CAAC;CACjB;AAED,MAAM,WAAW,eAAe;IAC5B,EAAE,EAAE,MAAM,CAAC;IACX,MAAM,EAAE,MAAM,CAAC;IACf,IAAI,EAAE,OAAO,GAAG,aAAa,GAAG,WAAW,GAAG,WAAW,GAAG,YAAY,GAAG,KAAK,GAAG,MAAM,CAAC;IAC1F,QAAQ,EAAE,MAAM,CAAC;IACjB,QAAQ,EAAE,MAAM,CAAC;IACjB,UAAU,EAAE,MAAM,CAAC;CACtB;AAED,MAAM,WAAW,WAAW;IACxB,YAAY,EAAE,WAAW,EAAE,CAAC;IAC5B,cAAc,EAAE,WAAW,EAAE,CAAC;IAC9B,kBAAkB,EAAE,WAAW,EAAE,CAAC;IAClC,mBAAmB,EAAE,WAAW,EAAE,CAAC;IACnC,gBAAgB,EAAE,WAAW,EAAE,CAAC;IAChC,iBAAiB,EAAE,WAAW,EAAE,CAAC;IACjC,aAAa,EAAE,eAAe,EAAE,CAAC;IACjC,+DAA+D;IAC/D,OAAO,EAAE,MAAM,EAAE,CAAC;CACrB;AAcD;;;;;;;;GAQG;AACH,qBAAa,iBAAkB,SAAQ,KAAK;IACxC,QAAQ,CAAC,IAAI,6BAA6B;IAC1C,QAAQ,CAAC,IAAI,EAAE,MAAM,GAAG,IAAI,CAAC;gBAEjB,OAAO,EAAE,MAAM,EAAE,IAAI,GAAE,MAAM,GAAG,IAAW;CAK1D;AA0GD,gDAAgD;AAChD,wBAAgB,aAAa,CAAC,IAAI,EAAE,WAAW,GAAG,MAAM,CAUvD;AAED;;;;;;;;;;;GAWG;AACH,wBAAgB,oBAAoB,CAAC,IAAI,EAAE,WAAW,EAAE,MAAM,EAAE,MAAM,GAAG,IAAI,CAQ5E;AAID,sEAAsE;AACtE,wBAAgB,YAAY,CAAC,WAAW,CAAC,EAAE,MAAM,GAAG,IAAI,GAAG,WAAW,CA+BrE;AAED,oEAAoE;AACpE,wBAAgB,iBAAiB,IAAI,IAAI,CAExC"}
|
package/dist/detect/patterns.js
CHANGED
|
@@ -94,14 +94,23 @@ function compileLinePatterns(input, invalid) {
|
|
|
94
94
|
invalid.push(String(item.id ?? 'unnamed'));
|
|
95
95
|
continue;
|
|
96
96
|
}
|
|
97
|
+
// A strength outside [0, 1] would outrank every structural anchor, and the g and y flags
|
|
98
|
+
// make the shared regex object stateful, so test() alternates results across lines. Both
|
|
99
|
+
// are treated the same as a regex that does not compile.
|
|
100
|
+
const strength = typeof item.strength === 'number' ? item.strength : 0.5;
|
|
101
|
+
const flags = typeof item.flags === 'string' ? item.flags : '';
|
|
102
|
+
if (strength < 0 || strength > 1 || !Number.isFinite(strength) || /[gy]/.test(flags)) {
|
|
103
|
+
invalid.push(id);
|
|
104
|
+
continue;
|
|
105
|
+
}
|
|
97
106
|
try {
|
|
98
107
|
out.push({
|
|
99
108
|
id,
|
|
100
109
|
lang: typeof item.lang === 'string' ? item.lang : undefined,
|
|
101
110
|
role: typeof item.role === 'string' ? item.role : undefined,
|
|
102
111
|
client: typeof item.client === 'string' ? item.client : undefined,
|
|
103
|
-
strength
|
|
104
|
-
regex: new RegExp(source,
|
|
112
|
+
strength,
|
|
113
|
+
regex: new RegExp(source, flags)
|
|
105
114
|
});
|
|
106
115
|
}
|
|
107
116
|
catch {
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"patterns.js","sourceRoot":"","sources":["../../src/detect/patterns.ts"],"names":[],"mappings":";AAAA;;;;;;;;;;GAUG;;;;;;
|
|
1
|
+
{"version":3,"file":"patterns.js","sourceRoot":"","sources":["../../src/detect/patterns.ts"],"names":[],"mappings":";AAAA;;;;;;;;;;GAUG;;;;;;AA4KH,sCAUC;AAcD,oDAQC;AAKD,oCA+BC;AAGD,8CAEC;AAnPD,sDAAyB;AACzB,0DAA6B;AA0C7B,MAAM,WAAW,GAAG,mBAAI,CAAC,IAAI,CAAC,SAAS,EAAE,UAAU,CAAC,CAAC;AAErD;;;;;;;;GAQG;AACH,MAAa,iBAAkB,SAAQ,KAAK;IAC/B,IAAI,GAAG,yBAAyB,CAAC;IACjC,IAAI,CAAgB;IAE7B,YAAY,OAAe,EAAE,OAAsB,IAAI;QACnD,KAAK,CAAC,OAAO,CAAC,CAAC;QACf,IAAI,CAAC,IAAI,GAAG,mBAAmB,CAAC;QAChC,IAAI,CAAC,IAAI,GAAG,IAAI,CAAC;IACrB,CAAC;CACJ;AATD,8CASC;AAID,SAAS,QAAQ,CAAC,GAAW,EAAE,IAAY;IACvC,IAAI,GAAW,CAAC;IAChB,IAAI,CAAC;QACD,GAAG,GAAG,iBAAE,CAAC,YAAY,CAAC,mBAAI,CAAC,IAAI,CAAC,GAAG,EAAE,IAAI,CAAC,EAAE,MAAM,CAAC,CAAC;IACxD,CAAC;IAAC,MAAM,CAAC;QACL,OAAO,EAAE,EAAE,EAAE,KAAK,EAAE,MAAM,EAAE,SAAS,EAAE,CAAC;IAC5C,CAAC;IACD,+FAA+F;IAC/F,uFAAuF;IACvF,IAAI,CAAC;QACD,MAAM,MAAM,GAAY,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC;QACxC,IAAI,CAAC,MAAM,IAAI,OAAO,MAAM,KAAK,QAAQ,EAAE,CAAC;YACxC,OAAO,EAAE,EAAE,EAAE,KAAK,EAAE,MAAM,EAAE,SAAS,EAAE,CAAC;QAC5C,CAAC;QACD,OAAO,EAAE,EAAE,EAAE,IAAI,EAAE,KAAK,EAAE,MAAiC,EAAE,CAAC;IAClE,CAAC;IAAC,MAAM,CAAC;QACL,OAAO,EAAE,EAAE,EAAE,KAAK,EAAE,MAAM,EAAE,SAAS,EAAE,CAAC;IAC5C,CAAC;AACL,CAAC;AAED,SAAS,QAAQ,CAAC,IAAY,EAAE,WAA0B;IACtD,IAAI,WAAW,EAAE,CAAC;QACd,MAAM,QAAQ,GAAG,QAAQ,CAAC,WAAW,EAAE,IAAI,CAAC,CAAC;QAC7C,IAAI,QAAQ,CAAC,EAAE,EAAE,CAAC;YACd,OAAO,QAAQ,CAAC,KAAK,CAAC;QAC1B,CAAC;QACD,IAAI,QAAQ,CAAC,MAAM,KAAK,SAAS,EAAE,CAAC;YAChC,MAAM,IAAI,iBAAiB,CAAC,gBAAgB,IAAI,OAAO,WAAW,qBAAqB,EAAE,IAAI,CAAC,CAAC;QACnG,CAAC;QACD,wFAAwF;IAC5F,CAAC;IACD,MAAM,OAAO,GAAG,QAAQ,CAAC,WAAW,EAAE,IAAI,CAAC,CAAC;IAC5C,IAAI,OAAO,CAAC,EAAE,EAAE,CAAC;QACb,OAAO,OAAO,CAAC,KAAK,CAAC;IACzB,CAAC;IACD,IAAI,OAAO,CAAC,MAAM,KAAK,SAAS,EAAE,CAAC;QAC/B,MAAM,IAAI,iBAAiB,CAAC,wBAAwB,IAAI,qCAAqC,EAAE,IAAI,CAAC,CAAC;IACzG,CAAC;IACD,OAAO,EAAE,CAAC;AACd,CAAC;AAED,SAAS,mBAAmB,CAAC,KAAc,EAAE,OAAiB;IAC1D,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC,EAAE,CAAC;QACxB,OAAO,EAAE,CAAC;IACd,CAAC;IACD,MAAM,GAAG,GAAkB,EAAE,CAAC;IAC9B,KAAK,MAAM,IAAI,IAAI,KAAqB,EAAE,CAAC;QACvC,MAAM,EAAE,GAAG,OAAO,IAAI,CAAC,EAAE,KAAK,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC;QACxD,MAAM,MAAM,GAAG,OAAO,IAAI,CAAC,OAAO,KAAK,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,CAAC,IAAI,CAAC;QACtE,IAAI,CAAC,EAAE,IAAI,CAAC,MAAM,EAAE,CAAC;YACjB,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,SAAS,CAAC,CAAC,CAAC;YAC3C,SAAS;QACb,CAAC;QACD,yFAAyF;QACzF,yFAAyF;QACzF,yDAAyD;QACzD,MAAM,QAAQ,GAAG,OAAO,IAAI,CAAC,QAAQ,KAAK,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,GAAG,CAAC;QACzE,MAAM,KAAK,GAAG,OAAO,IAAI,CAAC,KAAK,KAAK,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,CAAC;QAC/D,IAAI,QAAQ,GAAG,CAAC,IAAI,QAAQ,GAAG,CAAC,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,QAAQ,CAAC,IAAI,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC;YACnF,OAAO,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;YACjB,SAAS;QACb,CAAC;QACD,IAAI,CAAC;YACD,GAAG,CAAC,IAAI,CAAC;gBACL,EAAE;gBACF,IAAI,EAAE,OAAO,IAAI,CAAC,IAAI,KAAK,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,SAAS;gBAC3D,IAAI,EAAE,OAAO,IAAI,CAAC,IAAI,KAAK,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,SAAS;gBAC3D,MAAM,EAAE,OAAO,IAAI,CAAC,MAAM,KAAK,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,SAAS;gBACjE,QAAQ;gBACR,KAAK,EAAE,IAAI,MAAM,CAAC,MAAM,EAAE,KAAK,CAAC;aACnC,CAAC,CAAC;QACP,CAAC;QAAC,MAAM,CAAC;YACL,OAAO,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QACrB,CAAC;IACL,CAAC;IACD,OAAO,GAAG,CAAC;AACf,CAAC;AAED,SAAS,gBAAgB,CAAC,KAAc,EAAE,OAAiB;IACvD,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC,EAAE,CAAC;QACxB,OAAO,EAAE,CAAC;IACd,CAAC;IACD,MAAM,GAAG,GAAsB,EAAE,CAAC;IAClC,KAAK,MAAM,IAAI,IAAI,KAAkC,EAAE,CAAC;QACpD,MAAM,EAAE,GAAG,OAAO,IAAI,CAAC,EAAE,KAAK,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC;QACxD,MAAM,QAAQ,GAAG,OAAO,IAAI,CAAC,QAAQ,KAAK,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC;QAC1E,IAAI,CAAC,EAAE,IAAI,CAAC,QAAQ,EAAE,CAAC;YACnB,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,SAAS,CAAC,CAAC,CAAC;YAC3C,SAAS;QACb,CAAC;QACD,GAAG,CAAC,IAAI,CAAC;YACL,EAAE;YACF,MAAM,EAAE,OAAO,IAAI,CAAC,MAAM,KAAK,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,GAAG;YAC3D,IAAI,EAAE,CAAC,OAAO,IAAI,CAAC,IAAI,KAAK,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,MAAM,CAA4B;YACrF,QAAQ,EAAE,OAAO,IAAI,CAAC,QAAQ,KAAK,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,GAAG;YACjE,QAAQ;YACR,UAAU,EAAE,OAAO,IAAI,CAAC,UAAU,KAAK,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC;SACxE,CAAC,CAAC;IACP,CAAC;IACD,OAAO,GAAG,CAAC;AACf,CAAC;AAED,gDAAgD;AAChD,SAAgB,aAAa,CAAC,IAAiB;IAC3C,OAAO,CACH,IAAI,CAAC,YAAY,CAAC,MAAM;QACxB,IAAI,CAAC,cAAc,CAAC,MAAM;QAC1B,IAAI,CAAC,kBAAkB,CAAC,MAAM;QAC9B,IAAI,CAAC,mBAAmB,CAAC,MAAM;QAC/B,IAAI,CAAC,gBAAgB,CAAC,MAAM;QAC5B,IAAI,CAAC,iBAAiB,CAAC,MAAM;QAC7B,IAAI,CAAC,aAAa,CAAC,MAAM,CAC5B,CAAC;AACN,CAAC;AAED;;;;;;;;;;;GAWG;AACH,SAAgB,oBAAoB,CAAC,IAAiB,EAAE,MAAc;IAClE,IAAI,aAAa,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC;QAC5B,MAAM,IAAI,iBAAiB,CACvB,6CAA6C,MAAM,IAAI;YACnD,0EAA0E;YAC1E,gDAAgD,CACvD,CAAC;IACN,CAAC;AACL,CAAC;AAED,IAAI,MAAM,GAAuB,IAAI,CAAC;AAEtC,sEAAsE;AACtE,SAAgB,YAAY,CAAC,WAA2B;IACpD,IAAI,MAAM,IAAI,WAAW,KAAK,SAAS,EAAE,CAAC;QACtC,OAAO,MAAM,CAAC;IAClB,CAAC;IACD,MAAM,GAAG,GAAG,WAAW,KAAK,SAAS,CAAC,CAAC,CAAC,OAAO,CAAC,GAAG,CAAC,sBAAsB,IAAI,IAAI,CAAC,CAAC,CAAC,WAAW,CAAC;IACjG,MAAM,OAAO,GAAa,EAAE,CAAC;IAE7B,MAAM,KAAK,GAAG,QAAQ,CAAC,oBAAoB,EAAE,GAAG,CAAC,CAAC;IAClD,MAAM,OAAO,GAAG,QAAQ,CAAC,sBAAsB,EAAE,GAAG,CAAC,CAAC;IACtD,MAAM,MAAM,GAAG,QAAQ,CAAC,qBAAqB,EAAE,GAAG,CAAC,CAAC;IACpD,MAAM,UAAU,GAAG,QAAQ,CAAC,iBAAiB,EAAE,GAAG,CAAC,CAAC;IACpD,MAAM,WAAW,GAAG,QAAQ,CAAC,kBAAkB,EAAE,GAAG,CAAC,CAAC;IACtD,MAAM,IAAI,GAAG,QAAQ,CAAC,mBAAmB,EAAE,GAAG,CAAC,CAAC;IAEhD,MAAM,IAAI,GAAgB;QACtB,YAAY,EAAE,mBAAmB,CAAC,KAAK,CAAC,QAAQ,EAAE,OAAO,CAAC;QAC1D,cAAc,EAAE,mBAAmB,CAAC,OAAO,CAAC,QAAQ,EAAE,OAAO,CAAC;QAC9D,kBAAkB,EAAE,mBAAmB,CAAC,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC;QAC/D,mBAAmB,EAAE,mBAAmB,CAAC,UAAU,CAAC,UAAU,EAAE,OAAO,CAAC;QACxE,gBAAgB,EAAE,mBAAmB,CAAC,UAAU,CAAC,OAAO,EAAE,OAAO,CAAC;QAClE,iBAAiB,EAAE,mBAAmB,CAAC,WAAW,CAAC,OAAO,EAAE,OAAO,CAAC;QACpE,aAAa,EAAE,gBAAgB,CAAC,IAAI,CAAC,SAAS,EAAE,OAAO,CAAC;QACxD,OAAO;KACV,CAAC;IAEF,oBAAoB,CAAC,IAAI,EAAE,GAAG,IAAI,WAAW,CAAC,CAAC;IAE/C,IAAI,WAAW,KAAK,SAAS,EAAE,CAAC;QAC5B,MAAM,GAAG,IAAI,CAAC;IAClB,CAAC;IACD,OAAO,IAAI,CAAC;AAChB,CAAC;AAED,oEAAoE;AACpE,SAAgB,iBAAiB;IAC7B,MAAM,GAAG,IAAI,CAAC;AAClB,CAAC"}
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"reply-header.d.ts","sourceRoot":"","sources":["../../src/detect/reply-header.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;GAaG;AAGH,OAAO,EAAE,KAAK,QAAQ,EAAE,MAAM,kBAAkB,CAAC;
|
|
1
|
+
{"version":3,"file":"reply-header.d.ts","sourceRoot":"","sources":["../../src/detect/reply-header.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;GAaG;AAGH,OAAO,EAAE,KAAK,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AAkCjD,MAAM,WAAW,gBAAgB;IAC7B,iDAAiD;IACjD,SAAS,EAAE,MAAM,CAAC;IAClB,gDAAgD;IAChD,OAAO,EAAE,MAAM,CAAC;IAChB,QAAQ,EAAE,QAAQ,CAAC;CACtB;AAED;;;;;GAKG;AACH,wBAAgB,qBAAqB,CAAC,IAAI,EAAE,MAAM,GAAG,QAAQ,GAAG,IAAI,CAgBnE;AASD;;;;;;;;;;;;;;;;;;;GAmBG;AACH,wBAAgB,uBAAuB,CAAC,KAAK,EAAE,SAAS,MAAM,EAAE,EAAE,IAAI,SAAI,GAAG,gBAAgB,GAAG,IAAI,CA8EnG;AAED;;;;;GAKG;AACH,wBAAgB,mBAAmB,CAAC,IAAI,EAAE,MAAM,GAAG,QAAQ,GAAG,IAAI,CASjE"}
|
|
@@ -32,8 +32,14 @@ const MIN_FIELDS = 3;
|
|
|
32
32
|
* headers is classified as reply history and disappears.
|
|
33
33
|
*/
|
|
34
34
|
const ORIGINATOR_VALUE = /(@[\w-]+\.[\w.-]+|<[^>]*@[^>]*>|mailto:|\[[^\]]*@[^\]]*\])/i;
|
|
35
|
-
/**
|
|
36
|
-
|
|
35
|
+
/**
|
|
36
|
+
* A date field carries something date shaped: a numeric date, a month name, or a clock time.
|
|
37
|
+
*
|
|
38
|
+
* The month stems cover the languages the quoted header field names already cover, because a Sent
|
|
39
|
+
* field whose value the date check cannot read contributes no date role, and the block then only
|
|
40
|
+
* qualifies through a subject field.
|
|
41
|
+
*/
|
|
42
|
+
const DATE_VALUE = /(\d{1,4}[./-]\d{1,2}[./-]\d{1,4}|\b\d{1,2}:\d{2}\b|\b(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec|mon|tue|wed|thu|fri|sat|sun|maj|okt|sty|lut|kwi|cze|lip|sie|wrz|paz|lis|gru|tammi|helmi|maalis|huhti|touko|kesä|heinä|elo|syys|loka|marras|joulu)[a-zäö]*\.?\b|(^|\s)(янв|фев|мар|апр|ма[йя]|июн|июл|авг|сен|окт|ноя|дек)|\b(esmaspäev|teisipäev|kolmapäev|neljapäev|reede|laupäev|pühapäev|montag|dienstag|mittwoch|donnerstag|freitag|samstag|sonntag|lundi|mardi|mercredi|jeudi|vendredi|samedi|dimanche)\b)/i;
|
|
37
43
|
function valueOf(line) {
|
|
38
44
|
const colon = line.indexOf(':');
|
|
39
45
|
return colon === -1 ? '' : line.slice(colon + 1).trim();
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"reply-header.js","sourceRoot":"","sources":["../../src/detect/reply-header.ts"],"names":[],"mappings":";AAAA;;;;;;;;;;;;;GAaG;;
|
|
1
|
+
{"version":3,"file":"reply-header.js","sourceRoot":"","sources":["../../src/detect/reply-header.ts"],"names":[],"mappings":";AAAA;;;;;;;;;;;;;GAaG;;AAmDH,sDAgBC;AA6BD,0DA8EC;AAQD,kDASC;AA7LD,yCAA0C;AAG1C,4EAA4E;AAC5E,MAAM,OAAO,GAAG,CAAC,CAAC;AAElB,qFAAqF;AACrF,MAAM,UAAU,GAAG,CAAC,CAAC;AAErB;;;;;;;;GAQG;AACH,MAAM,gBAAgB,GAAG,6DAA6D,CAAC;AAEvF;;;;;;GAMG;AACH,MAAM,UAAU,GACZ,wfAAwf,CAAC;AAE7f,SAAS,OAAO,CAAC,IAAY;IACzB,MAAM,KAAK,GAAG,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC;IAChC,OAAO,KAAK,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,KAAK,GAAG,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;AAC5D,CAAC;AAUD;;;;;GAKG;AACH,SAAgB,qBAAqB,CAAC,IAAY;IAC9C,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,EAAE,CAAC;IAC5B,IAAI,CAAC,OAAO,IAAI,OAAO,CAAC,MAAM,GAAG,GAAG,EAAE,CAAC;QACnC,OAAO,IAAI,CAAC;IAChB,CAAC;IACD,MAAM,QAAQ,GAAG,IAAA,uBAAY,GAAE,CAAC;IAChC,IAAI,IAAI,GAAoB,IAAI,CAAC;IACjC,KAAK,MAAM,OAAO,IAAI,QAAQ,CAAC,YAAY,EAAE,CAAC;QAC1C,IAAI,CAAC,OAAO,CAAC,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,EAAE,CAAC;YAC/B,SAAS;QACb,CAAC;QACD,IAAI,CAAC,IAAI,IAAI,OAAO,CAAC,QAAQ,GAAG,IAAI,CAAC,QAAQ,EAAE,CAAC;YAC5C,IAAI,GAAG,EAAE,EAAE,EAAE,SAAS,OAAO,CAAC,EAAE,EAAE,EAAE,IAAI,EAAE,cAAc,EAAE,QAAQ,EAAE,OAAO,CAAC,QAAQ,EAAE,CAAC;QAC3F,CAAC;IACL,CAAC;IACD,OAAO,IAAI,CAAC;AAChB,CAAC;AAED,+EAA+E;AAC/E,SAAS,QAAQ,CAAC,KAA0B;IACxC,uFAAuF;IACvF,qEAAqE;IACrE,OAAO,KAAK,CAAC,IAAI,IAAI,UAAU,IAAI,KAAK,CAAC,GAAG,CAAC,YAAY,CAAC,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,MAAM,CAAC,IAAI,KAAK,CAAC,GAAG,CAAC,SAAS,CAAC,CAAC,CAAC;AAC9G,CAAC;AAED;;;;;;;;;;;;;;;;;;;GAmBG;AACH,SAAgB,uBAAuB,CAAC,KAAwB,EAAE,IAAI,GAAG,CAAC;IACtE,MAAM,QAAQ,GAAG,IAAA,uBAAY,GAAE,CAAC;IAEhC,MAAM,QAAQ,GAAa;QACvB,EAAE,EAAE,2BAA2B;QAC/B,IAAI,EAAE,cAAc;QACpB,6EAA6E;QAC7E,QAAQ,EAAE,IAAI;KACjB,CAAC;IAEF,4FAA4F;IAC5F,6FAA6F;IAC7F,sDAAsD;IACtD,IAAI,UAAU,GAAG,CAAC,CAAC,CAAC;IACpB,IAAI,KAAK,GAAG,IAAI,GAAG,EAAU,CAAC;IAC9B,IAAI,IAAI,GAAG,CAAC,CAAC,CAAC;IACd,IAAI,GAAG,GAAG,CAAC,CAAC;IAEZ,KAAK,IAAI,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,EAAE,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QACpD,MAAM,IAAI,GAAG,CAAC,KAAK,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC;QAErC,IAAI,CAAC,IAAI,EAAE,CAAC;YACR,IAAI,UAAU,KAAK,CAAC,CAAC,EAAE,CAAC;gBACpB,UAAU,GAAG,CAAC,CAAC;gBACf,GAAG,GAAG,CAAC,CAAC;gBACR,SAAS;YACb,CAAC;YACD,GAAG,EAAE,CAAC;YACN,IAAI,GAAG,IAAI,OAAO,EAAE,CAAC;gBACjB,SAAS;YACb,CAAC;YACD,qFAAqF;YACrF,wFAAwF;YACxF,sFAAsF;YACtF,yBAAyB;YACzB,IAAI,QAAQ,CAAC,KAAK,CAAC,IAAI,IAAI,IAAI,UAAU,EAAE,CAAC;gBACxC,OAAO,EAAE,SAAS,EAAE,UAAU,EAAE,OAAO,EAAE,IAAI,EAAE,QAAQ,EAAE,CAAC;YAC9D,CAAC;YACD,KAAK,GAAG,IAAI,GAAG,EAAU,CAAC;YAC1B,IAAI,GAAG,CAAC,CAAC,CAAC;YACV,UAAU,GAAG,CAAC,GAAG,OAAO,GAAG,CAAC,CAAC;YAC7B,GAAG,GAAG,OAAO,CAAC;YACd,SAAS;QACb,CAAC;QAED,MAAM,OAAO,GAAG,QAAQ,CAAC,kBAAkB,CAAC,IAAI,CAAC,OAAO,CAAC,EAAE,CAAC,OAAO,CAAC,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC;QACtF,IAAI,CAAC,OAAO,EAAE,CAAC;YACX,kDAAkD;YAClD,IAAI,QAAQ,CAAC,KAAK,CAAC,IAAI,IAAI,IAAI,UAAU,IAAI,UAAU,KAAK,CAAC,CAAC,EAAE,CAAC;gBAC7D,OAAO,EAAE,SAAS,EAAE,UAAU,EAAE,OAAO,EAAE,IAAI,EAAE,QAAQ,EAAE,CAAC;YAC9D,CAAC;YACD,KAAK,GAAG,IAAI,GAAG,EAAU,CAAC;YAC1B,IAAI,GAAG,CAAC,CAAC,CAAC;YACV,UAAU,GAAG,CAAC,CAAC,CAAC;YAChB,GAAG,GAAG,CAAC,CAAC;YACR,SAAS;QACb,CAAC;QAED,IAAI,UAAU,KAAK,CAAC,CAAC,EAAE,CAAC;YACpB,UAAU,GAAG,CAAC,CAAC;QACnB,CAAC;QACD,GAAG,GAAG,CAAC,CAAC;QACR,MAAM,IAAI,GAAG,OAAO,CAAC,IAAI,IAAI,OAAO,CAAC,EAAE,CAAC;QACxC,MAAM,KAAK,GAAG,OAAO,CAAC,IAAI,CAAC,CAAC;QAC5B,sFAAsF;QACtF,sFAAsF;QACtF,0DAA0D;QAC1D,IAAI,IAAI,KAAK,YAAY,CAAC,CAAC,CAAC,gBAAgB,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,KAAK,MAAM,CAAC,CAAC,CAAC,UAAU,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;YACzG,KAAK,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC;QACpB,CAAC;QACD,IAAI,GAAG,CAAC,CAAC;IACb,CAAC;IAED,IAAI,UAAU,KAAK,CAAC,CAAC,IAAI,QAAQ,CAAC,KAAK,CAAC,IAAI,IAAI,IAAI,UAAU,EAAE,CAAC;QAC7D,OAAO,EAAE,SAAS,EAAE,UAAU,EAAE,OAAO,EAAE,IAAI,EAAE,QAAQ,EAAE,CAAC;IAC9D,CAAC;IAED,OAAO,IAAI,CAAC;AAChB,CAAC;AAED;;;;;GAKG;AACH,SAAgB,mBAAmB,CAAC,IAAY;IAC5C,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,EAAE,CAAC;IAC5B,IAAI,UAAU,CAAC,IAAI,CAAC,OAAO,CAAC,EAAE,CAAC;QAC3B,OAAO,EAAE,EAAE,EAAE,+BAA+B,EAAE,IAAI,EAAE,WAAW,EAAE,QAAQ,EAAE,GAAG,EAAE,CAAC;IACrF,CAAC;IACD,IAAI,cAAc,CAAC,IAAI,CAAC,OAAO,CAAC,EAAE,CAAC;QAC/B,OAAO,EAAE,EAAE,EAAE,uBAAuB,EAAE,IAAI,EAAE,WAAW,EAAE,QAAQ,EAAE,IAAI,EAAE,CAAC;IAC9E,CAAC;IACD,OAAO,IAAI,CAAC;AAChB,CAAC"}
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"signature.d.ts","sourceRoot":"","sources":["../../src/detect/signature.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;GAcG;AAGH,OAAO,EAAE,KAAK,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AACjD,OAAO,EAAE,KAAK,aAAa,EAAE,MAAM,kBAAkB,CAAC;
|
|
1
|
+
{"version":3,"file":"signature.d.ts","sourceRoot":"","sources":["../../src/detect/signature.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;GAcG;AAGH,OAAO,EAAE,KAAK,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AACjD,OAAO,EAAE,KAAK,aAAa,EAAE,MAAM,kBAAkB,CAAC;AAatD,wBAAgB,wBAAwB,CAAC,IAAI,EAAE,MAAM,GAAG,QAAQ,GAAG,IAAI,CAQtE;AAED,iFAAiF;AACjF,wBAAgB,sBAAsB,CAAC,KAAK,EAAE,SAAS,MAAM,EAAE,GAAG,QAAQ,EAAE,CAmB3E;AAED;;;;;GAKG;AACH,wBAAgB,oBAAoB,CAAC,QAAQ,EAAE,aAAa,GAAG,QAAQ,GAAG,IAAI,CAc7E"}
|
package/dist/detect/signature.js
CHANGED
|
@@ -21,8 +21,13 @@ exports.detectSignatureShape = detectSignatureShape;
|
|
|
21
21
|
const patterns_1 = require("./patterns");
|
|
22
22
|
/** Signatures beyond this many lines are almost always something else. */
|
|
23
23
|
const MAX_SIGNATURE_LINES = 12;
|
|
24
|
-
/**
|
|
25
|
-
|
|
24
|
+
/**
|
|
25
|
+
* A signature sits after the sender's words, which in a reply with quoted history below can be
|
|
26
|
+
* early by character position, not late. The classifier's walk back from the authored region is
|
|
27
|
+
* what constrains location structurally; this gate only excludes the opening slice, where a
|
|
28
|
+
* contact card is the message rather than a sign-off.
|
|
29
|
+
*/
|
|
30
|
+
const MIN_POSITION = 0.1;
|
|
26
31
|
function detectSignatureDelimiter(line) {
|
|
27
32
|
const patterns = (0, patterns_1.loadPatterns)();
|
|
28
33
|
for (const pattern of patterns.signatureDelimiters) {
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"signature.js","sourceRoot":"","sources":["../../src/detect/signature.ts"],"names":[],"mappings":";AAAA;;;;;;;;;;;;;;GAcG;;
|
|
1
|
+
{"version":3,"file":"signature.js","sourceRoot":"","sources":["../../src/detect/signature.ts"],"names":[],"mappings":";AAAA;;;;;;;;;;;;;;GAcG;;AAiBH,4DAQC;AAGD,wDAmBC;AAQD,oDAcC;AAnED,yCAA0C;AAI1C,0EAA0E;AAC1E,MAAM,mBAAmB,GAAG,EAAE,CAAC;AAE/B;;;;;GAKG;AACH,MAAM,YAAY,GAAG,GAAG,CAAC;AAEzB,SAAgB,wBAAwB,CAAC,IAAY;IACjD,MAAM,QAAQ,GAAG,IAAA,uBAAY,GAAE,CAAC;IAChC,KAAK,MAAM,OAAO,IAAI,QAAQ,CAAC,mBAAmB,EAAE,CAAC;QACjD,IAAI,OAAO,CAAC,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;YAC3B,OAAO,EAAE,EAAE,EAAE,aAAa,OAAO,CAAC,EAAE,EAAE,EAAE,IAAI,EAAE,WAAW,EAAE,QAAQ,EAAE,OAAO,CAAC,QAAQ,EAAE,CAAC;QAC5F,CAAC;IACL,CAAC;IACD,OAAO,IAAI,CAAC;AAChB,CAAC;AAED,iFAAiF;AACjF,SAAgB,sBAAsB,CAAC,KAAwB;IAC3D,MAAM,QAAQ,GAAG,IAAA,uBAAY,GAAE,CAAC;IAChC,MAAM,KAAK,GAAG,IAAI,GAAG,EAAoB,CAAC;IAC1C,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;QACvB,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,EAAE,CAAC;QAC5B,IAAI,CAAC,OAAO,IAAI,OAAO,CAAC,MAAM,GAAG,GAAG,EAAE,CAAC;YACnC,SAAS;QACb,CAAC;QACD,KAAK,MAAM,OAAO,IAAI,QAAQ,CAAC,gBAAgB,EAAE,CAAC;YAC9C,IAAI,OAAO,CAAC,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,EAAE,CAAC;gBAC9B,MAAM,EAAE,GAAG,aAAa,OAAO,CAAC,EAAE,EAAE,CAAC;gBACrC,MAAM,QAAQ,GAAG,KAAK,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;gBAC/B,IAAI,CAAC,QAAQ,IAAI,QAAQ,CAAC,QAAQ,GAAG,OAAO,CAAC,QAAQ,EAAE,CAAC;oBACpD,KAAK,CAAC,GAAG,CAAC,EAAE,EAAE,EAAE,EAAE,EAAE,IAAI,EAAE,WAAW,EAAE,QAAQ,EAAE,OAAO,CAAC,QAAQ,EAAE,CAAC,CAAC;gBACzE,CAAC;YACL,CAAC;QACL,CAAC;IACL,CAAC;IACD,OAAO,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC,CAAC;AAC/B,CAAC;AAED;;;;;GAKG;AACH,SAAgB,oBAAoB,CAAC,QAAuB;IACxD,IAAI,QAAQ,CAAC,UAAU,IAAI,QAAQ,CAAC,QAAQ,GAAG,YAAY,EAAE,CAAC;QAC1D,OAAO,IAAI,CAAC;IAChB,CAAC;IACD,IAAI,QAAQ,CAAC,SAAS,GAAG,mBAAmB,IAAI,QAAQ,CAAC,SAAS,GAAG,EAAE,EAAE,CAAC;QACtE,OAAO,IAAI,CAAC;IAChB,CAAC;IACD,MAAM,UAAU,GAAG,QAAQ,CAAC,eAAe,IAAI,QAAQ,CAAC,cAAc,IAAI,QAAQ,CAAC,MAAM,CAAC;IAC1F,IAAI,CAAC,UAAU,IAAI,QAAQ,CAAC,cAAc,GAAG,GAAG,EAAE,CAAC;QAC/C,OAAO,IAAI,CAAC;IAChB,CAAC;IACD,4FAA4F;IAC5F,2CAA2C;IAC3C,OAAO,EAAE,EAAE,EAAE,iBAAiB,EAAE,IAAI,EAAE,WAAW,EAAE,QAAQ,EAAE,IAAI,EAAE,CAAC;AACxE,CAAC"}
|
package/dist/eval/metrics.d.ts
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"metrics.d.ts","sourceRoot":"","sources":["../../src/eval/metrics.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;GAYG;AAEH,OAAO,EAAU,KAAK,KAAK,EAAE,MAAM,UAAU,CAAC;AAE9C,MAAM,WAAW,aAAa;IAC1B,IAAI,EAAE,MAAM,CAAC;IACb,UAAU,EAAE,MAAM,CAAC;IACnB,SAAS,EAAE,MAAM,CAAC;IAClB,SAAS,EAAE,KAAK,CAAC;IACjB,SAAS,EAAE,KAAK,CAAC;IACjB,uEAAuE;IACvE,MAAM,EAAE,OAAO,CAAC;IAChB,uFAAuF;IACvF,KAAK,CAAC,EAAE,MAAM,CAAC;CAClB;AAED,MAAM,WAAW,YAAY;IACzB,KAAK,EAAE,KAAK,CAAC;IACb,OAAO,EAAE,MAAM,CAAC;IAChB,SAAS,EAAE,MAAM,CAAC;IAClB,aAAa,EAAE,MAAM,CAAC;IACtB,SAAS,EAAE,MAAM,CAAC;IAClB,MAAM,EAAE,MAAM,CAAC;IACf,EAAE,EAAE,MAAM,CAAC;CACd;AAED,MAAM,WAAW,gBAAgB;IAC7B,6DAA6D;IAC7D,eAAe,EAAE,MAAM,CAAC;IACxB,qEAAqE;IACrE,uBAAuB,EAAE,MAAM,CAAC;IAChC,oBAAoB,EAAE,MAAM,CAAC;IAC7B,wEAAwE;IACxE,gBAAgB,EAAE,MAAM,CAAC;IACzB,wBAAwB,EAAE,MAAM,CAAC;IACjC,qBAAqB,EAAE,MAAM,CAAC;IAC9B,qEAAqE;IACrE,WAAW,EAAE,MAAM,CAAC;IACpB,2BAA2B,EAAE,MAAM,CAAC;IACpC,eAAe,EAAE,MAAM,CAAC;IACxB,gEAAgE;IAChE,YAAY,EAAE,MAAM,CAAC;IACrB,wBAAwB,EAAE,MAAM,CAAC;IACjC,wBAAwB,EAAE,MAAM,CAAC;CACpC;AAED,MAAM,WAAW,kBAAkB;IAC/B,sDAAsD;IACtD,oBAAoB,EAAE,MAAM,CAAC;IAC7B,+EAA+E;IAC/E,mBAAmB,EAAE,MAAM,CAAC;IAC5B,wEAAwE;IACxE,oBAAoB,EAAE,MAAM,CAAC;IAC7B,mDAAmD;IACnD,kBAAkB,EAAE,MAAM,CAAC;IAC3B,2DAA2D;IAC3D,kBAAkB,EAAE,MAAM,CAAC;IAC3B,gEAAgE;IAChE,eAAe,EAAE,MAAM,CAAC;IACxB,yFAAyF;IACzF,aAAa,EAAE,MAAM,CAAC;IACtB,QAAQ,EAAE,MAAM,CAAC;CACpB;AAED,MAAM,WAAW,gBAAgB;IAC7B,MAAM,EAAE,MAAM,CAAC;IACf,QAAQ,EAAE,MAAM,CAAC;IACjB,cAAc,EAAE,MAAM,CAAC;IACvB,QAAQ,EAAE,YAAY,EAAE,CAAC;IACzB,SAAS,EAAE,GAAG,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IAC/B,SAAS,EAAE,gBAAgB,CAAC;IAC5B,YAAY,EAAE,kBAAkB,CAAC;IACjC,OAAO,EAAE;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,MAAM,EAAE,MAAM,CAAC;QAAC,SAAS,EAAE,MAAM,CAAC;QAAC,oBAAoB,EAAE,MAAM,CAAA;KAAE,EAAE,CAAC;
|
|
1
|
+
{"version":3,"file":"metrics.d.ts","sourceRoot":"","sources":["../../src/eval/metrics.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;GAYG;AAEH,OAAO,EAAU,KAAK,KAAK,EAAE,MAAM,UAAU,CAAC;AAE9C,MAAM,WAAW,aAAa;IAC1B,IAAI,EAAE,MAAM,CAAC;IACb,UAAU,EAAE,MAAM,CAAC;IACnB,SAAS,EAAE,MAAM,CAAC;IAClB,SAAS,EAAE,KAAK,CAAC;IACjB,SAAS,EAAE,KAAK,CAAC;IACjB,uEAAuE;IACvE,MAAM,EAAE,OAAO,CAAC;IAChB,uFAAuF;IACvF,KAAK,CAAC,EAAE,MAAM,CAAC;CAClB;AAED,MAAM,WAAW,YAAY;IACzB,KAAK,EAAE,KAAK,CAAC;IACb,OAAO,EAAE,MAAM,CAAC;IAChB,SAAS,EAAE,MAAM,CAAC;IAClB,aAAa,EAAE,MAAM,CAAC;IACtB,SAAS,EAAE,MAAM,CAAC;IAClB,MAAM,EAAE,MAAM,CAAC;IACf,EAAE,EAAE,MAAM,CAAC;CACd;AAED,MAAM,WAAW,gBAAgB;IAC7B,6DAA6D;IAC7D,eAAe,EAAE,MAAM,CAAC;IACxB,qEAAqE;IACrE,uBAAuB,EAAE,MAAM,CAAC;IAChC,oBAAoB,EAAE,MAAM,CAAC;IAC7B,wEAAwE;IACxE,gBAAgB,EAAE,MAAM,CAAC;IACzB,wBAAwB,EAAE,MAAM,CAAC;IACjC,qBAAqB,EAAE,MAAM,CAAC;IAC9B,qEAAqE;IACrE,WAAW,EAAE,MAAM,CAAC;IACpB,2BAA2B,EAAE,MAAM,CAAC;IACpC,eAAe,EAAE,MAAM,CAAC;IACxB,gEAAgE;IAChE,YAAY,EAAE,MAAM,CAAC;IACrB,wBAAwB,EAAE,MAAM,CAAC;IACjC,wBAAwB,EAAE,MAAM,CAAC;CACpC;AAED,MAAM,WAAW,kBAAkB;IAC/B,sDAAsD;IACtD,oBAAoB,EAAE,MAAM,CAAC;IAC7B,+EAA+E;IAC/E,mBAAmB,EAAE,MAAM,CAAC;IAC5B,wEAAwE;IACxE,oBAAoB,EAAE,MAAM,CAAC;IAC7B,mDAAmD;IACnD,kBAAkB,EAAE,MAAM,CAAC;IAC3B,2DAA2D;IAC3D,kBAAkB,EAAE,MAAM,CAAC;IAC3B,gEAAgE;IAChE,eAAe,EAAE,MAAM,CAAC;IACxB,yFAAyF;IACzF,aAAa,EAAE,MAAM,CAAC;IACtB,QAAQ,EAAE,MAAM,CAAC;CACpB;AAED,MAAM,WAAW,gBAAgB;IAC7B,MAAM,EAAE,MAAM,CAAC;IACf,QAAQ,EAAE,MAAM,CAAC;IACjB,cAAc,EAAE,MAAM,CAAC;IACvB,QAAQ,EAAE,YAAY,EAAE,CAAC;IACzB,SAAS,EAAE,GAAG,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IAC/B,SAAS,EAAE,gBAAgB,CAAC;IAC5B,YAAY,EAAE,kBAAkB,CAAC;IACjC,OAAO,EAAE;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,MAAM,EAAE,MAAM,CAAC;QAAC,SAAS,EAAE,MAAM,CAAC;QAAC,oBAAoB,EAAE,MAAM,CAAC;QAAC,kBAAkB,EAAE,MAAM,CAAA;KAAE,EAAE,CAAC;CAC7H;AA6ID,wBAAgB,QAAQ,CAAC,IAAI,EAAE,SAAS,aAAa,EAAE,GAAG,gBAAgB,CA2CzE"}
|
package/dist/eval/metrics.js
CHANGED
|
@@ -42,8 +42,6 @@ function labelMetrics(rows, label) {
|
|
|
42
42
|
f1: precision + recall > 0 ? (2 * precision * recall) / (precision + recall) : 0
|
|
43
43
|
};
|
|
44
44
|
}
|
|
45
|
-
/** Labels the reference considers authored by the sender of this message. */
|
|
46
|
-
const AUTHORED = new Set(['NEW_CONTENT', 'SIGNATURE']);
|
|
47
45
|
function retention(rows) {
|
|
48
46
|
let newContentChars = 0;
|
|
49
47
|
let newContentCharsRetained = 0;
|
|
@@ -120,7 +118,12 @@ function catastrophic(rows) {
|
|
|
120
118
|
if (!visible.length && blocks.some(block => block.charCount > 0)) {
|
|
121
119
|
counts.emptyExtraction++;
|
|
122
120
|
}
|
|
123
|
-
|
|
121
|
+
// A footer is a signature or a disclaimer, matching the benchmark's structural check, but
|
|
122
|
+
// only a message that had something else to show can lose its point. An empty reply whose
|
|
123
|
+
// entire authored content is its signature comes out as footer plus hidden history, and
|
|
124
|
+
// that is the correct extraction, not a catastrophe; six such messages in the reference
|
|
125
|
+
// set started tripping this counter the moment history hiding improved.
|
|
126
|
+
if (authored.length && visible.length && visible.every(block => block.reference === 'SIGNATURE' || block.reference === 'DISCLAIMER')) {
|
|
124
127
|
counts.onlyFooterRetained++;
|
|
125
128
|
}
|
|
126
129
|
if (blocks.some(block => block.reference === 'FORWARDED_CONTENT' && block.hidden)) {
|
|
@@ -173,12 +176,16 @@ function evaluate(rows) {
|
|
|
173
176
|
retention: retention(rows),
|
|
174
177
|
catastrophic: catastrophic(rows),
|
|
175
178
|
byGroup: [...groups]
|
|
176
|
-
.map(([group, list]) =>
|
|
177
|
-
|
|
178
|
-
|
|
179
|
-
|
|
180
|
-
|
|
181
|
-
|
|
179
|
+
.map(([group, list]) => {
|
|
180
|
+
const kept = retention(list);
|
|
181
|
+
return {
|
|
182
|
+
group,
|
|
183
|
+
blocks: list.length,
|
|
184
|
+
agreement: list.filter(row => row.reference === row.predicted).length / list.length,
|
|
185
|
+
newContentCharRecall: kept.newContentCharRecall,
|
|
186
|
+
historyHiddenShare: 1 - kept.historyStillVisibleShare
|
|
187
|
+
};
|
|
188
|
+
})
|
|
182
189
|
.sort((a, b) => b.blocks - a.blocks)
|
|
183
190
|
};
|
|
184
191
|
}
|