@webywz/context-diet 0.0.0-stage → 0.1.0
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/LICENSE +21 -0
- package/README.md +170 -2
- package/dist/blocks.d.ts +8 -0
- package/dist/blocks.d.ts.map +1 -0
- package/dist/blocks.js +129 -0
- package/dist/blocks.js.map +1 -0
- package/dist/cli.d.ts +12 -0
- package/dist/cli.d.ts.map +1 -0
- package/dist/cli.js +149 -0
- package/dist/cli.js.map +1 -0
- package/dist/diet.d.ts +12 -0
- package/dist/diet.d.ts.map +1 -0
- package/dist/diet.js +66 -0
- package/dist/diet.js.map +1 -0
- package/dist/heuristics.d.ts +13 -0
- package/dist/heuristics.d.ts.map +1 -0
- package/dist/heuristics.js +238 -0
- package/dist/heuristics.js.map +1 -0
- package/dist/index.d.ts +7 -0
- package/dist/index.d.ts.map +1 -0
- package/dist/index.js +7 -0
- package/dist/index.js.map +1 -0
- package/dist/llm.d.ts +12 -0
- package/dist/llm.d.ts.map +1 -0
- package/dist/llm.js +107 -0
- package/dist/llm.js.map +1 -0
- package/dist/package-info.d.ts +5 -0
- package/dist/package-info.d.ts.map +1 -0
- package/dist/package-info.js +13 -0
- package/dist/package-info.js.map +1 -0
- package/dist/parse-args.d.ts +19 -0
- package/dist/parse-args.d.ts.map +1 -0
- package/dist/parse-args.js +172 -0
- package/dist/parse-args.js.map +1 -0
- package/dist/report.d.ts +24 -0
- package/dist/report.d.ts.map +1 -0
- package/dist/report.js +56 -0
- package/dist/report.js.map +1 -0
- package/dist/tokens.d.ts +6 -0
- package/dist/tokens.d.ts.map +1 -0
- package/dist/tokens.js +73 -0
- package/dist/tokens.js.map +1 -0
- package/dist/truncate.d.ts +4 -0
- package/dist/truncate.d.ts.map +1 -0
- package/dist/truncate.js +136 -0
- package/dist/truncate.js.map +1 -0
- package/dist/types.d.ts +48 -0
- package/dist/types.d.ts.map +1 -0
- package/dist/types.js +18 -0
- package/dist/types.js.map +1 -0
- package/package.json +58 -4
package/dist/report.d.ts
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
1
|
+
import type { CutCounts, DietResult } from "./types.js";
|
|
2
|
+
export interface JsonReport {
|
|
3
|
+
ok: true;
|
|
4
|
+
source: string;
|
|
5
|
+
input: {
|
|
6
|
+
chars: number;
|
|
7
|
+
tokens: number;
|
|
8
|
+
};
|
|
9
|
+
output: {
|
|
10
|
+
chars: number;
|
|
11
|
+
tokens: number;
|
|
12
|
+
};
|
|
13
|
+
saved: {
|
|
14
|
+
chars: number;
|
|
15
|
+
tokens: number;
|
|
16
|
+
charPercent: number;
|
|
17
|
+
tokenPercent: number;
|
|
18
|
+
};
|
|
19
|
+
cuts: CutCounts;
|
|
20
|
+
text: string;
|
|
21
|
+
}
|
|
22
|
+
export declare function toJsonReport(result: DietResult, source: string): JsonReport;
|
|
23
|
+
export declare function formatHumanReport(result: DietResult, source: string, version: string): string;
|
|
24
|
+
//# sourceMappingURL=report.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"report.d.ts","sourceRoot":"","sources":["../src/report.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,SAAS,EAAE,UAAU,EAAE,MAAM,YAAY,CAAC;AAExD,MAAM,WAAW,UAAU;IACzB,EAAE,EAAE,IAAI,CAAC;IACT,MAAM,EAAE,MAAM,CAAC;IACf,KAAK,EAAE;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,MAAM,EAAE,MAAM,CAAA;KAAE,CAAC;IACzC,MAAM,EAAE;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,MAAM,EAAE,MAAM,CAAA;KAAE,CAAC;IAC1C,KAAK,EAAE;QACL,KAAK,EAAE,MAAM,CAAC;QACd,MAAM,EAAE,MAAM,CAAC;QACf,WAAW,EAAE,MAAM,CAAC;QACpB,YAAY,EAAE,MAAM,CAAC;KACtB,CAAC;IACF,IAAI,EAAE,SAAS,CAAC;IAChB,IAAI,EAAE,MAAM,CAAC;CACd;AAWD,wBAAgB,YAAY,CAAC,MAAM,EAAE,UAAU,EAAE,MAAM,EAAE,MAAM,GAAG,UAAU,CAiB3E;AAmBD,wBAAgB,iBAAiB,CAC/B,MAAM,EAAE,UAAU,EAClB,MAAM,EAAE,MAAM,EACd,OAAO,EAAE,MAAM,GACd,MAAM,CAcR"}
|
package/dist/report.js
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
1
|
+
function pct(before, after) {
|
|
2
|
+
if (before <= 0)
|
|
3
|
+
return 0;
|
|
4
|
+
return Math.round(((before - after) / before) * 1000) / 10;
|
|
5
|
+
}
|
|
6
|
+
function commas(n) {
|
|
7
|
+
return n.toLocaleString("en-US");
|
|
8
|
+
}
|
|
9
|
+
export function toJsonReport(result, source) {
|
|
10
|
+
const savedChars = Math.max(0, result.input.chars - result.output.chars);
|
|
11
|
+
const savedTokens = Math.max(0, result.input.tokens - result.output.tokens);
|
|
12
|
+
return {
|
|
13
|
+
ok: true,
|
|
14
|
+
source,
|
|
15
|
+
input: result.input,
|
|
16
|
+
output: result.output,
|
|
17
|
+
saved: {
|
|
18
|
+
chars: savedChars,
|
|
19
|
+
tokens: savedTokens,
|
|
20
|
+
charPercent: pct(result.input.chars, result.output.chars),
|
|
21
|
+
tokenPercent: pct(result.input.tokens, result.output.tokens),
|
|
22
|
+
},
|
|
23
|
+
cuts: result.cuts,
|
|
24
|
+
text: result.text,
|
|
25
|
+
};
|
|
26
|
+
}
|
|
27
|
+
function cutLines(cuts) {
|
|
28
|
+
const rows = [
|
|
29
|
+
["ansi sequences", String(cuts.ansiSequences)],
|
|
30
|
+
["boilerplate lines", String(cuts.boilerplateLines)],
|
|
31
|
+
["duplicate paragraphs", String(cuts.duplicateParagraphs)],
|
|
32
|
+
["whitespace runs", String(cuts.whitespaceRuns)],
|
|
33
|
+
["omitted heavy blobs", String(cuts.omittedHeavyBlobs)],
|
|
34
|
+
["truncated blocks", String(cuts.truncatedBlocks)],
|
|
35
|
+
["truncated tokens", `~${commas(cuts.truncatedTokens)}`],
|
|
36
|
+
["llm deepen", cuts.deepened ? "yes" : "no"],
|
|
37
|
+
];
|
|
38
|
+
const width = Math.max(...rows.map(([k]) => k.length));
|
|
39
|
+
return rows.map(([k, v]) => ` ${k.padEnd(width)} ${v}`);
|
|
40
|
+
}
|
|
41
|
+
export function formatHumanReport(result, source, version) {
|
|
42
|
+
const { input, output } = result;
|
|
43
|
+
const charPct = pct(input.chars, output.chars);
|
|
44
|
+
const tokPct = pct(input.tokens, output.tokens);
|
|
45
|
+
const sign = (n) => (n > 0 ? `−${n}%` : n < 0 ? `+${Math.abs(n)}%` : "0%");
|
|
46
|
+
const lines = [
|
|
47
|
+
`context-diet ${version}`,
|
|
48
|
+
`source: ${source}`,
|
|
49
|
+
`chars: ${commas(input.chars)} → ${commas(output.chars)} (${sign(charPct)})`,
|
|
50
|
+
`tokens: ~${commas(input.tokens)} → ~${commas(output.tokens)} (${sign(tokPct)})`,
|
|
51
|
+
`cut:`,
|
|
52
|
+
...cutLines(result.cuts),
|
|
53
|
+
];
|
|
54
|
+
return lines.join("\n") + "\n";
|
|
55
|
+
}
|
|
56
|
+
//# sourceMappingURL=report.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"report.js","sourceRoot":"","sources":["../src/report.ts"],"names":[],"mappings":"AAiBA,SAAS,GAAG,CAAC,MAAc,EAAE,KAAa;IACxC,IAAI,MAAM,IAAI,CAAC;QAAE,OAAO,CAAC,CAAC;IAC1B,OAAO,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,GAAG,KAAK,CAAC,GAAG,MAAM,CAAC,GAAG,IAAI,CAAC,GAAG,EAAE,CAAC;AAC7D,CAAC;AAED,SAAS,MAAM,CAAC,CAAS;IACvB,OAAO,CAAC,CAAC,cAAc,CAAC,OAAO,CAAC,CAAC;AACnC,CAAC;AAED,MAAM,UAAU,YAAY,CAAC,MAAkB,EAAE,MAAc;IAC7D,MAAM,UAAU,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,MAAM,CAAC,KAAK,CAAC,KAAK,GAAG,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC;IACzE,MAAM,WAAW,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,MAAM,CAAC,KAAK,CAAC,MAAM,GAAG,MAAM,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC;IAC5E,OAAO;QACL,EAAE,EAAE,IAAI;QACR,MAAM;QACN,KAAK,EAAE,MAAM,CAAC,KAAK;QACnB,MAAM,EAAE,MAAM,CAAC,MAAM;QACrB,KAAK,EAAE;YACL,KAAK,EAAE,UAAU;YACjB,MAAM,EAAE,WAAW;YACnB,WAAW,EAAE,GAAG,CAAC,MAAM,CAAC,KAAK,CAAC,KAAK,EAAE,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC;YACzD,YAAY,EAAE,GAAG,CAAC,MAAM,CAAC,KAAK,CAAC,MAAM,EAAE,MAAM,CAAC,MAAM,CAAC,MAAM,CAAC;SAC7D;QACD,IAAI,EAAE,MAAM,CAAC,IAAI;QACjB,IAAI,EAAE,MAAM,CAAC,IAAI;KAClB,CAAC;AACJ,CAAC;AAED,SAAS,QAAQ,CAAC,IAAe;IAC/B,MAAM,IAAI,GAA4B;QACpC,CAAC,gBAAgB,EAAE,MAAM,CAAC,IAAI,CAAC,aAAa,CAAC,CAAC;QAC9C,CAAC,mBAAmB,EAAE,MAAM,CAAC,IAAI,CAAC,gBAAgB,CAAC,CAAC;QACpD,CAAC,sBAAsB,EAAE,MAAM,CAAC,IAAI,CAAC,mBAAmB,CAAC,CAAC;QAC1D,CAAC,iBAAiB,EAAE,MAAM,CAAC,IAAI,CAAC,cAAc,CAAC,CAAC;QAChD,CAAC,qBAAqB,EAAE,MAAM,CAAC,IAAI,CAAC,iBAAiB,CAAC,CAAC;QACvD,CAAC,kBAAkB,EAAE,MAAM,CAAC,IAAI,CAAC,eAAe,CAAC,CAAC;QAClD,CAAC,kBAAkB,EAAE,IAAI,MAAM,CAAC,IAAI,CAAC,eAAe,CAAC,EAAE,CAAC;QACxD,CAAC,YAAY,EAAE,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,IAAI,CAAC;KAC7C,CAAC;IACF,MAAM,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC;IACvD,OAAO,IAAI,CAAC,GAAG,CACb,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,EAAE,EAAE,CAAC,KAAK,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC,KAAK,CAAC,EAAE,CACzC,CAAC;AACJ,CAAC;AAED,MAAM,UAAU,iBAAiB,CAC/B,MAAkB,EAClB,MAAc,EACd,OAAe;IAEf,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,GAAG,MAAM,CAAC;IACjC,MAAM,OAAO,GAAG,GAAG,CAAC,KAAK,CAAC,KAAK,EAAE,MAAM,CAAC,KAAK,CAAC,CAAC;IAC/C,MAAM,MAAM,GAAG,GAAG,CAAC,KAAK,CAAC,MAAM,EAAE,MAAM,CAAC,MAAM,CAAC,CAAC;IAChD,MAAM,IAAI,GAAG,CAAC,CAAS,EAAE,EAAE,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;IACnF,MAAM,KAAK,GAAG;QACZ,gBAAgB,OAAO,EAAE;QACzB,aAAa,MAAM,EAAE;QACrB,aAAa,MAAM,CAAC,KAAK,CAAC,KAAK,CAAC,MAAM,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,MAAM,IAAI,CAAC,OAAO,CAAC,GAAG;QAChF,cAAc,MAAM,CAAC,KAAK,CAAC,MAAM,CAAC,OAAO,MAAM,CAAC,MAAM,CAAC,MAAM,CAAC,MAAM,IAAI,CAAC,MAAM,CAAC,GAAG;QACnF,MAAM;QACN,GAAG,QAAQ,CAAC,MAAM,CAAC,IAAI,CAAC;KACzB,CAAC;IACF,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,GAAG,IAAI,CAAC;AACjC,CAAC"}
|
package/dist/tokens.d.ts
ADDED
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"tokens.d.ts","sourceRoot":"","sources":["../src/tokens.ts"],"names":[],"mappings":"AA4BA,wBAAgB,cAAc,CAAC,IAAI,EAAE,MAAM,GAAG,MAAM,CAoDnD;AAED,wBAAgB,QAAQ,CAAC,IAAI,EAAE,MAAM,GAAG;IAAE,KAAK,EAAE,MAAM,CAAC;IAAC,MAAM,EAAE,MAAM,CAAA;CAAE,CAExE"}
|
package/dist/tokens.js
ADDED
|
@@ -0,0 +1,73 @@
|
|
|
1
|
+
/**
|
|
2
|
+
* Approximate a cl100k_base (GPT-4 family) token count without shipping a BPE
|
|
3
|
+
* table. Tuned for English prose, source code, and CJK. Typical error vs
|
|
4
|
+
* tiktoken is within ~20% on mixed technical text — enough to budget a prompt,
|
|
5
|
+
* not enough to bill on.
|
|
6
|
+
*/
|
|
7
|
+
const CJK_CODE_POINT = (code) => (code >= 0x4e00 && code <= 0x9fff) ||
|
|
8
|
+
(code >= 0x3400 && code <= 0x4dbf) ||
|
|
9
|
+
(code >= 0xf900 && code <= 0xfaff) ||
|
|
10
|
+
(code >= 0x3040 && code <= 0x30ff) ||
|
|
11
|
+
(code >= 0x31f0 && code <= 0x31ff) ||
|
|
12
|
+
(code >= 0xac00 && code <= 0xd7af) ||
|
|
13
|
+
(code >= 0x1100 && code <= 0x11ff);
|
|
14
|
+
function isLetter(ch) {
|
|
15
|
+
const code = ch.charCodeAt(0);
|
|
16
|
+
return ((code >= 65 && code <= 90) ||
|
|
17
|
+
(code >= 97 && code <= 122) ||
|
|
18
|
+
(code >= 0xc0 && code <= 0x24f && ch.toLowerCase() !== ch.toUpperCase()));
|
|
19
|
+
}
|
|
20
|
+
function isDigit(ch) {
|
|
21
|
+
return ch >= "0" && ch <= "9";
|
|
22
|
+
}
|
|
23
|
+
export function estimateTokens(text) {
|
|
24
|
+
if (!text)
|
|
25
|
+
return 0;
|
|
26
|
+
let tokens = 0;
|
|
27
|
+
let i = 0;
|
|
28
|
+
const len = text.length;
|
|
29
|
+
while (i < len) {
|
|
30
|
+
const ch = text[i];
|
|
31
|
+
const code = text.charCodeAt(i);
|
|
32
|
+
if (ch === " " || ch === "\t" || ch === "\r") {
|
|
33
|
+
i++;
|
|
34
|
+
continue;
|
|
35
|
+
}
|
|
36
|
+
if (ch === "\n") {
|
|
37
|
+
tokens += 0.25;
|
|
38
|
+
i++;
|
|
39
|
+
continue;
|
|
40
|
+
}
|
|
41
|
+
if (CJK_CODE_POINT(code)) {
|
|
42
|
+
tokens += 1;
|
|
43
|
+
i++;
|
|
44
|
+
continue;
|
|
45
|
+
}
|
|
46
|
+
if (isLetter(ch)) {
|
|
47
|
+
let j = i + 1;
|
|
48
|
+
while (j < len && isLetter(text[j]))
|
|
49
|
+
j++;
|
|
50
|
+
const wordLen = j - i;
|
|
51
|
+
// Short function words / identifiers are usually 1 piece; longer words
|
|
52
|
+
// split roughly every 4 characters, matching cl100k's byte-pair grain.
|
|
53
|
+
tokens += wordLen <= 5 ? 1 : Math.ceil(wordLen / 4);
|
|
54
|
+
i = j;
|
|
55
|
+
continue;
|
|
56
|
+
}
|
|
57
|
+
if (isDigit(ch)) {
|
|
58
|
+
let j = i + 1;
|
|
59
|
+
while (j < len && isDigit(text[j]))
|
|
60
|
+
j++;
|
|
61
|
+
tokens += Math.max(1, Math.ceil((j - i) / 3));
|
|
62
|
+
i = j;
|
|
63
|
+
continue;
|
|
64
|
+
}
|
|
65
|
+
tokens += 1;
|
|
66
|
+
i++;
|
|
67
|
+
}
|
|
68
|
+
return Math.max(text.length === 0 ? 0 : 1, Math.round(tokens));
|
|
69
|
+
}
|
|
70
|
+
export function snapshot(text) {
|
|
71
|
+
return { chars: text.length, tokens: estimateTokens(text) };
|
|
72
|
+
}
|
|
73
|
+
//# sourceMappingURL=tokens.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"tokens.js","sourceRoot":"","sources":["../src/tokens.ts"],"names":[],"mappings":"AAAA;;;;;GAKG;AACH,MAAM,cAAc,GAAG,CAAC,IAAY,EAAW,EAAE,CAC/C,CAAC,IAAI,IAAI,MAAM,IAAI,IAAI,IAAI,MAAM,CAAC;IAClC,CAAC,IAAI,IAAI,MAAM,IAAI,IAAI,IAAI,MAAM,CAAC;IAClC,CAAC,IAAI,IAAI,MAAM,IAAI,IAAI,IAAI,MAAM,CAAC;IAClC,CAAC,IAAI,IAAI,MAAM,IAAI,IAAI,IAAI,MAAM,CAAC;IAClC,CAAC,IAAI,IAAI,MAAM,IAAI,IAAI,IAAI,MAAM,CAAC;IAClC,CAAC,IAAI,IAAI,MAAM,IAAI,IAAI,IAAI,MAAM,CAAC;IAClC,CAAC,IAAI,IAAI,MAAM,IAAI,IAAI,IAAI,MAAM,CAAC,CAAC;AAErC,SAAS,QAAQ,CAAC,EAAU;IAC1B,MAAM,IAAI,GAAG,EAAE,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC;IAC9B,OAAO,CACL,CAAC,IAAI,IAAI,EAAE,IAAI,IAAI,IAAI,EAAE,CAAC;QAC1B,CAAC,IAAI,IAAI,EAAE,IAAI,IAAI,IAAI,GAAG,CAAC;QAC3B,CAAC,IAAI,IAAI,IAAI,IAAI,IAAI,IAAI,KAAK,IAAI,EAAE,CAAC,WAAW,EAAE,KAAK,EAAE,CAAC,WAAW,EAAE,CAAC,CACzE,CAAC;AACJ,CAAC;AAED,SAAS,OAAO,CAAC,EAAU;IACzB,OAAO,EAAE,IAAI,GAAG,IAAI,EAAE,IAAI,GAAG,CAAC;AAChC,CAAC;AAED,MAAM,UAAU,cAAc,CAAC,IAAY;IACzC,IAAI,CAAC,IAAI;QAAE,OAAO,CAAC,CAAC;IAEpB,IAAI,MAAM,GAAG,CAAC,CAAC;IACf,IAAI,CAAC,GAAG,CAAC,CAAC;IACV,MAAM,GAAG,GAAG,IAAI,CAAC,MAAM,CAAC;IAExB,OAAO,CAAC,GAAG,GAAG,EAAE,CAAC;QACf,MAAM,EAAE,GAAG,IAAI,CAAC,CAAC,CAAE,CAAC;QACpB,MAAM,IAAI,GAAG,IAAI,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC;QAEhC,IAAI,EAAE,KAAK,GAAG,IAAI,EAAE,KAAK,IAAI,IAAI,EAAE,KAAK,IAAI,EAAE,CAAC;YAC7C,CAAC,EAAE,CAAC;YACJ,SAAS;QACX,CAAC;QAED,IAAI,EAAE,KAAK,IAAI,EAAE,CAAC;YAChB,MAAM,IAAI,IAAI,CAAC;YACf,CAAC,EAAE,CAAC;YACJ,SAAS;QACX,CAAC;QAED,IAAI,cAAc,CAAC,IAAI,CAAC,EAAE,CAAC;YACzB,MAAM,IAAI,CAAC,CAAC;YACZ,CAAC,EAAE,CAAC;YACJ,SAAS;QACX,CAAC;QAED,IAAI,QAAQ,CAAC,EAAE,CAAC,EAAE,CAAC;YACjB,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;YACd,OAAO,CAAC,GAAG,GAAG,IAAI,QAAQ,CAAC,IAAI,CAAC,CAAC,CAAE,CAAC;gBAAE,CAAC,EAAE,CAAC;YAC1C,MAAM,OAAO,GAAG,CAAC,GAAG,CAAC,CAAC;YACtB,uEAAuE;YACvE,uEAAuE;YACvE,MAAM,IAAI,OAAO,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,OAAO,GAAG,CAAC,CAAC,CAAC;YACpD,CAAC,GAAG,CAAC,CAAC;YACN,SAAS;QACX,CAAC;QAED,IAAI,OAAO,CAAC,EAAE,CAAC,EAAE,CAAC;YAChB,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;YACd,OAAO,CAAC,GAAG,GAAG,IAAI,OAAO,CAAC,IAAI,CAAC,CAAC,CAAE,CAAC;gBAAE,CAAC,EAAE,CAAC;YACzC,MAAM,IAAI,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC;YAC9C,CAAC,GAAG,CAAC,CAAC;YACN,SAAS;QACX,CAAC;QAED,MAAM,IAAI,CAAC,CAAC;QACZ,CAAC,EAAE,CAAC;IACN,CAAC;IAED,OAAO,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,MAAM,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,IAAI,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC;AACjE,CAAC;AAED,MAAM,UAAU,QAAQ,CAAC,IAAY;IACnC,OAAO,EAAE,KAAK,EAAE,IAAI,CAAC,MAAM,EAAE,MAAM,EAAE,cAAc,CAAC,IAAI,CAAC,EAAE,CAAC;AAC9D,CAAC"}
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"truncate.d.ts","sourceRoot":"","sources":["../src/truncate.ts"],"names":[],"mappings":"AAAA,OAAO,EAAc,KAAK,KAAK,EAAE,MAAM,aAAa,CAAC;AAErD,OAAO,KAAK,EAAE,SAAS,EAAE,YAAY,EAAE,MAAM,YAAY,CAAC;AA0D1D,wBAAgB,gBAAgB,CAC9B,MAAM,EAAE,KAAK,EAAE,EACf,SAAS,EAAE,MAAM,EACjB,IAAI,EAAE,YAAY,EAAE,EACpB,IAAI,EAAE,SAAS,GACd,KAAK,EAAE,CA4ET"}
|
package/dist/truncate.js
ADDED
|
@@ -0,0 +1,136 @@
|
|
|
1
|
+
import { joinBlocks } from "./blocks.js";
|
|
2
|
+
import { estimateTokens } from "./tokens.js";
|
|
3
|
+
function scoreBlock(block, keep) {
|
|
4
|
+
let score = 1;
|
|
5
|
+
switch (block.type) {
|
|
6
|
+
case "heading":
|
|
7
|
+
score = 4;
|
|
8
|
+
if (keep.has("headings"))
|
|
9
|
+
score += 24;
|
|
10
|
+
break;
|
|
11
|
+
case "code":
|
|
12
|
+
score = 4;
|
|
13
|
+
if (keep.has("code"))
|
|
14
|
+
score += 24;
|
|
15
|
+
break;
|
|
16
|
+
case "list":
|
|
17
|
+
score = 3;
|
|
18
|
+
if (keep.has("lists"))
|
|
19
|
+
score += 24;
|
|
20
|
+
break;
|
|
21
|
+
case "paragraph":
|
|
22
|
+
score = 2;
|
|
23
|
+
break;
|
|
24
|
+
case "rule":
|
|
25
|
+
score = 0;
|
|
26
|
+
break;
|
|
27
|
+
case "blank":
|
|
28
|
+
score = 0;
|
|
29
|
+
break;
|
|
30
|
+
default:
|
|
31
|
+
score = 1;
|
|
32
|
+
}
|
|
33
|
+
// Prefer earlier context when scores tie.
|
|
34
|
+
return score;
|
|
35
|
+
}
|
|
36
|
+
function trimBlockToTokens(raw, maxTokens) {
|
|
37
|
+
if (estimateTokens(raw) <= maxTokens)
|
|
38
|
+
return raw;
|
|
39
|
+
const lines = raw.split("\n");
|
|
40
|
+
if (lines.length > 1) {
|
|
41
|
+
const kept = [];
|
|
42
|
+
for (const line of lines) {
|
|
43
|
+
const trial = [...kept, line].join("\n");
|
|
44
|
+
if (estimateTokens(trial + "\n…") > maxTokens)
|
|
45
|
+
break;
|
|
46
|
+
kept.push(line);
|
|
47
|
+
}
|
|
48
|
+
if (kept.length > 0)
|
|
49
|
+
return `${kept.join("\n")}\n…`;
|
|
50
|
+
}
|
|
51
|
+
// Character walk as a last resort.
|
|
52
|
+
let lo = 0;
|
|
53
|
+
let hi = raw.length;
|
|
54
|
+
while (lo < hi) {
|
|
55
|
+
const mid = Math.ceil((lo + hi) / 2);
|
|
56
|
+
const slice = `${raw.slice(0, mid).trimEnd()}\n…`;
|
|
57
|
+
if (estimateTokens(slice) <= maxTokens)
|
|
58
|
+
lo = mid;
|
|
59
|
+
else
|
|
60
|
+
hi = mid - 1;
|
|
61
|
+
}
|
|
62
|
+
const cut = raw.slice(0, lo).trimEnd();
|
|
63
|
+
return cut.length === 0 ? "…" : `${cut}\n…`;
|
|
64
|
+
}
|
|
65
|
+
export function truncateToBudget(blocks, maxTokens, keep, cuts) {
|
|
66
|
+
if (!(maxTokens > 0))
|
|
67
|
+
return blocks;
|
|
68
|
+
const keepSet = new Set(keep);
|
|
69
|
+
const tokens = blocks.map((b) => estimateTokens(b.raw));
|
|
70
|
+
let total = tokens.reduce((a, b) => a + b, 0);
|
|
71
|
+
if (total <= maxTokens)
|
|
72
|
+
return blocks;
|
|
73
|
+
const scores = blocks.map((b) => scoreBlock(b, keepSet));
|
|
74
|
+
const order = blocks
|
|
75
|
+
.map((_, i) => i)
|
|
76
|
+
.sort((a, b) => {
|
|
77
|
+
const ds = scores[a] - scores[b];
|
|
78
|
+
if (ds !== 0)
|
|
79
|
+
return ds;
|
|
80
|
+
return b - a; // drop later ties first
|
|
81
|
+
});
|
|
82
|
+
const dropped = new Set();
|
|
83
|
+
for (const i of order) {
|
|
84
|
+
if (total <= maxTokens)
|
|
85
|
+
break;
|
|
86
|
+
const block = blocks[i];
|
|
87
|
+
if (block.type === "blank")
|
|
88
|
+
continue;
|
|
89
|
+
// Never drop the last remaining non-blank block; trim it instead.
|
|
90
|
+
const remainingNonBlank = blocks.filter((b, idx) => !dropped.has(idx) && b.type !== "blank").length;
|
|
91
|
+
if (remainingNonBlank <= 1)
|
|
92
|
+
continue;
|
|
93
|
+
dropped.add(i);
|
|
94
|
+
total -= tokens[i];
|
|
95
|
+
cuts.truncatedBlocks += 1;
|
|
96
|
+
cuts.truncatedTokens += tokens[i];
|
|
97
|
+
}
|
|
98
|
+
let next = blocks.filter((_, i) => !dropped.has(i));
|
|
99
|
+
next = next.filter((b, i, arr) => {
|
|
100
|
+
if (b.type !== "blank")
|
|
101
|
+
return true;
|
|
102
|
+
if (i === 0 || i === arr.length - 1)
|
|
103
|
+
return false;
|
|
104
|
+
return arr[i - 1]?.type !== "blank";
|
|
105
|
+
});
|
|
106
|
+
total = next.reduce((a, b) => a + estimateTokens(b.raw), 0);
|
|
107
|
+
if (total <= maxTokens)
|
|
108
|
+
return next;
|
|
109
|
+
// Still over: trim the lowest-score, largest remaining block from the end.
|
|
110
|
+
const trimOrder = next
|
|
111
|
+
.map((b, i) => ({ i, b, s: scoreBlock(b, keepSet), t: estimateTokens(b.raw) }))
|
|
112
|
+
.filter((x) => x.b.type !== "blank")
|
|
113
|
+
.sort((a, b) => a.s - b.s || b.t - a.t || b.i - a.i);
|
|
114
|
+
for (const item of trimOrder) {
|
|
115
|
+
total = next.reduce((a, b) => a + estimateTokens(b.raw), 0);
|
|
116
|
+
if (total <= maxTokens)
|
|
117
|
+
break;
|
|
118
|
+
const others = total - item.t;
|
|
119
|
+
const budget = Math.max(8, maxTokens - others);
|
|
120
|
+
const trimmed = trimBlockToTokens(item.b.raw, budget);
|
|
121
|
+
if (trimmed !== item.b.raw) {
|
|
122
|
+
cuts.truncatedBlocks += 1;
|
|
123
|
+
cuts.truncatedTokens += Math.max(0, item.t - estimateTokens(trimmed));
|
|
124
|
+
next = next.map((b, i) => i === item.i ? { ...b, raw: trimmed } : b);
|
|
125
|
+
}
|
|
126
|
+
}
|
|
127
|
+
const joined = joinBlocks(next);
|
|
128
|
+
if (estimateTokens(joined) <= maxTokens)
|
|
129
|
+
return next;
|
|
130
|
+
// Absolute fallback: hard-cut the joined document.
|
|
131
|
+
const hard = trimBlockToTokens(joined, maxTokens);
|
|
132
|
+
cuts.truncatedBlocks += 1;
|
|
133
|
+
cuts.truncatedTokens += Math.max(0, estimateTokens(joined) - estimateTokens(hard));
|
|
134
|
+
return [{ type: "paragraph", raw: hard }];
|
|
135
|
+
}
|
|
136
|
+
//# sourceMappingURL=truncate.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"truncate.js","sourceRoot":"","sources":["../src/truncate.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,UAAU,EAAc,MAAM,aAAa,CAAC;AACrD,OAAO,EAAE,cAAc,EAAE,MAAM,aAAa,CAAC;AAG7C,SAAS,UAAU,CAAC,KAAY,EAAE,IAAuB;IACvD,IAAI,KAAK,GAAG,CAAC,CAAC;IACd,QAAQ,KAAK,CAAC,IAAI,EAAE,CAAC;QACnB,KAAK,SAAS;YACZ,KAAK,GAAG,CAAC,CAAC;YACV,IAAI,IAAI,CAAC,GAAG,CAAC,UAAU,CAAC;gBAAE,KAAK,IAAI,EAAE,CAAC;YACtC,MAAM;QACR,KAAK,MAAM;YACT,KAAK,GAAG,CAAC,CAAC;YACV,IAAI,IAAI,CAAC,GAAG,CAAC,MAAM,CAAC;gBAAE,KAAK,IAAI,EAAE,CAAC;YAClC,MAAM;QACR,KAAK,MAAM;YACT,KAAK,GAAG,CAAC,CAAC;YACV,IAAI,IAAI,CAAC,GAAG,CAAC,OAAO,CAAC;gBAAE,KAAK,IAAI,EAAE,CAAC;YACnC,MAAM;QACR,KAAK,WAAW;YACd,KAAK,GAAG,CAAC,CAAC;YACV,MAAM;QACR,KAAK,MAAM;YACT,KAAK,GAAG,CAAC,CAAC;YACV,MAAM;QACR,KAAK,OAAO;YACV,KAAK,GAAG,CAAC,CAAC;YACV,MAAM;QACR;YACE,KAAK,GAAG,CAAC,CAAC;IACd,CAAC;IACD,0CAA0C;IAC1C,OAAO,KAAK,CAAC;AACf,CAAC;AAED,SAAS,iBAAiB,CAAC,GAAW,EAAE,SAAiB;IACvD,IAAI,cAAc,CAAC,GAAG,CAAC,IAAI,SAAS;QAAE,OAAO,GAAG,CAAC;IACjD,MAAM,KAAK,GAAG,GAAG,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;IAC9B,IAAI,KAAK,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QACrB,MAAM,IAAI,GAAa,EAAE,CAAC;QAC1B,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;YACzB,MAAM,KAAK,GAAG,CAAC,GAAG,IAAI,EAAE,IAAI,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;YACzC,IAAI,cAAc,CAAC,KAAK,GAAG,KAAK,CAAC,GAAG,SAAS;gBAAE,MAAM;YACrD,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QAClB,CAAC;QACD,IAAI,IAAI,CAAC,MAAM,GAAG,CAAC;YAAE,OAAO,GAAG,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC;IACtD,CAAC;IACD,mCAAmC;IACnC,IAAI,EAAE,GAAG,CAAC,CAAC;IACX,IAAI,EAAE,GAAG,GAAG,CAAC,MAAM,CAAC;IACpB,OAAO,EAAE,GAAG,EAAE,EAAE,CAAC;QACf,MAAM,GAAG,GAAG,IAAI,CAAC,IAAI,CAAC,CAAC,EAAE,GAAG,EAAE,CAAC,GAAG,CAAC,CAAC,CAAC;QACrC,MAAM,KAAK,GAAG,GAAG,GAAG,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC,CAAC,OAAO,EAAE,KAAK,CAAC;QAClD,IAAI,cAAc,CAAC,KAAK,CAAC,IAAI,SAAS;YAAE,EAAE,GAAG,GAAG,CAAC;;YAC5C,EAAE,GAAG,GAAG,GAAG,CAAC,CAAC;IACpB,CAAC;IACD,MAAM,GAAG,GAAG,GAAG,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,OAAO,EAAE,CAAC;IACvC,OAAO,GAAG,CAAC,MAAM,KAAK,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,GAAG,KAAK,CAAC;AAC9C,CAAC;AAED,MAAM,UAAU,gBAAgB,CAC9B,MAAe,EACf,SAAiB,EACjB,IAAoB,EACpB,IAAe;IAEf,IAAI,CAAC,CAAC,SAAS,GAAG,CAAC,CAAC;QAAE,OAAO,MAAM,CAAC;IACpC,MAAM,OAAO,GAAG,IAAI,GAAG,CAAC,IAAI,CAAC,CAAC;IAC9B,MAAM,MAAM,GAAG,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,cAAc,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;IACxD,IAAI,KAAK,GAAG,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,CAAC;IAC9C,IAAI,KAAK,IAAI,SAAS;QAAE,OAAO,MAAM,CAAC;IAEtC,MAAM,MAAM,GAAG,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,UAAU,CAAC,CAAC,EAAE,OAAO,CAAC,CAAC,CAAC;IACzD,MAAM,KAAK,GAAG,MAAM;SACjB,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC;SAChB,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE;QACb,MAAM,EAAE,GAAG,MAAM,CAAC,CAAC,CAAE,GAAG,MAAM,CAAC,CAAC,CAAE,CAAC;QACnC,IAAI,EAAE,KAAK,CAAC;YAAE,OAAO,EAAE,CAAC;QACxB,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,wBAAwB;IACxC,CAAC,CAAC,CAAC;IAEL,MAAM,OAAO,GAAG,IAAI,GAAG,EAAU,CAAC;IAClC,KAAK,MAAM,CAAC,IAAI,KAAK,EAAE,CAAC;QACtB,IAAI,KAAK,IAAI,SAAS;YAAE,MAAM;QAC9B,MAAM,KAAK,GAAG,MAAM,CAAC,CAAC,CAAE,CAAC;QACzB,IAAI,KAAK,CAAC,IAAI,KAAK,OAAO;YAAE,SAAS;QACrC,kEAAkE;QAClE,MAAM,iBAAiB,GACrB,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,GAAG,EAAE,EAAE,CAAC,CAAC,OAAO,CAAC,GAAG,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC,IAAI,KAAK,OAAO,CAAC,CAAC,MAAM,CAAC;QAC5E,IAAI,iBAAiB,IAAI,CAAC;YAAE,SAAS;QACrC,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC;QACf,KAAK,IAAI,MAAM,CAAC,CAAC,CAAE,CAAC;QACpB,IAAI,CAAC,eAAe,IAAI,CAAC,CAAC;QAC1B,IAAI,CAAC,eAAe,IAAI,MAAM,CAAC,CAAC,CAAE,CAAC;IACrC,CAAC;IAED,IAAI,IAAI,GAAY,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;IAC7D,IAAI,GAAG,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,GAAG,EAAE,EAAE;QAC/B,IAAI,CAAC,CAAC,IAAI,KAAK,OAAO;YAAE,OAAO,IAAI,CAAC;QACpC,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,KAAK,GAAG,CAAC,MAAM,GAAG,CAAC;YAAE,OAAO,KAAK,CAAC;QAClD,OAAO,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,KAAK,OAAO,CAAC;IACtC,CAAC,CAAC,CAAC;IAEH,KAAK,GAAG,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,GAAG,cAAc,CAAC,CAAC,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,CAAC;IAC5D,IAAI,KAAK,IAAI,SAAS;QAAE,OAAO,IAAI,CAAC;IAEpC,2EAA2E;IAC3E,MAAM,SAAS,GAAG,IAAI;SACnB,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,EAAE,CAAC,EAAE,CAAC,EAAE,CAAC,EAAE,UAAU,CAAC,CAAC,EAAE,OAAO,CAAC,EAAE,CAAC,EAAE,cAAc,CAAC,CAAC,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;SAC9E,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,KAAK,OAAO,CAAC;SACnC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;IAEvD,KAAK,MAAM,IAAI,IAAI,SAAS,EAAE,CAAC;QAC7B,KAAK,GAAG,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,GAAG,cAAc,CAAC,CAAC,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,CAAC;QAC5D,IAAI,KAAK,IAAI,SAAS;YAAE,MAAM;QAC9B,MAAM,MAAM,GAAG,KAAK,GAAG,IAAI,CAAC,CAAC,CAAC;QAC9B,MAAM,MAAM,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,SAAS,GAAG,MAAM,CAAC,CAAC;QAC/C,MAAM,OAAO,GAAG,iBAAiB,CAAC,IAAI,CAAC,CAAC,CAAC,GAAG,EAAE,MAAM,CAAC,CAAC;QACtD,IAAI,OAAO,KAAK,IAAI,CAAC,CAAC,CAAC,GAAG,EAAE,CAAC;YAC3B,IAAI,CAAC,eAAe,IAAI,CAAC,CAAC;YAC1B,IAAI,CAAC,eAAe,IAAI,IAAI,CAAC,GAAG,CAC9B,CAAC,EACD,IAAI,CAAC,CAAC,GAAG,cAAc,CAAC,OAAO,CAAC,CACjC,CAAC;YACF,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CACvB,CAAC,KAAK,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,GAAG,CAAC,EAAE,GAAG,EAAE,OAAO,EAAE,CAAC,CAAC,CAAC,CAAC,CAC1C,CAAC;QACJ,CAAC;IACH,CAAC;IAED,MAAM,MAAM,GAAG,UAAU,CAAC,IAAI,CAAC,CAAC;IAChC,IAAI,cAAc,CAAC,MAAM,CAAC,IAAI,SAAS;QAAE,OAAO,IAAI,CAAC;IAErD,mDAAmD;IACnD,MAAM,IAAI,GAAG,iBAAiB,CAAC,MAAM,EAAE,SAAS,CAAC,CAAC;IAClD,IAAI,CAAC,eAAe,IAAI,CAAC,CAAC;IAC1B,IAAI,CAAC,eAAe,IAAI,IAAI,CAAC,GAAG,CAC9B,CAAC,EACD,cAAc,CAAC,MAAM,CAAC,GAAG,cAAc,CAAC,IAAI,CAAC,CAC9C,CAAC;IACF,OAAO,CAAC,EAAE,IAAI,EAAE,WAAW,EAAE,GAAG,EAAE,IAAI,EAAE,CAAC,CAAC;AAC5C,CAAC"}
|
package/dist/types.d.ts
ADDED
|
@@ -0,0 +1,48 @@
|
|
|
1
|
+
export type KeepCategory = "headings" | "code" | "lists";
|
|
2
|
+
export type LlmProvider = "openai" | "anthropic";
|
|
3
|
+
export interface CutCounts {
|
|
4
|
+
/** ANSI escape sequences removed. */
|
|
5
|
+
ansiSequences: number;
|
|
6
|
+
/** Repeated / chrome lines dropped (logs, banners, signatures). */
|
|
7
|
+
boilerplateLines: number;
|
|
8
|
+
/** Near-identical paragraphs dropped. */
|
|
9
|
+
duplicateParagraphs: number;
|
|
10
|
+
/** Extra blank-line runs or trailing spaces collapsed. */
|
|
11
|
+
whitespaceRuns: number;
|
|
12
|
+
/** Long base64 / data-URI blobs replaced with placeholders. */
|
|
13
|
+
omittedHeavyBlobs: number;
|
|
14
|
+
/** Blocks dropped or trimmed to fit `--max-tokens`. */
|
|
15
|
+
truncatedBlocks: number;
|
|
16
|
+
/** Estimated tokens removed by truncation (not other steps). */
|
|
17
|
+
truncatedTokens: number;
|
|
18
|
+
/** True when an optional LLM pass rewrote the text. */
|
|
19
|
+
deepened: boolean;
|
|
20
|
+
}
|
|
21
|
+
export interface DietOptions {
|
|
22
|
+
/** Soft cap on estimated output tokens. */
|
|
23
|
+
maxTokens?: number;
|
|
24
|
+
/** Prefer these block types when truncating to a token budget. */
|
|
25
|
+
keep?: KeepCategory[];
|
|
26
|
+
/**
|
|
27
|
+
* Optional second-pass compression via OpenAI or Anthropic.
|
|
28
|
+
* Requires a key in the environment; never implied.
|
|
29
|
+
*/
|
|
30
|
+
deepen?: boolean;
|
|
31
|
+
provider?: LlmProvider;
|
|
32
|
+
/** Injected fetch (tests). Defaults to global fetch. */
|
|
33
|
+
fetch?: typeof fetch;
|
|
34
|
+
env?: NodeJS.ProcessEnv;
|
|
35
|
+
}
|
|
36
|
+
export interface SizeSnapshot {
|
|
37
|
+
chars: number;
|
|
38
|
+
tokens: number;
|
|
39
|
+
}
|
|
40
|
+
export interface DietResult {
|
|
41
|
+
text: string;
|
|
42
|
+
input: SizeSnapshot;
|
|
43
|
+
output: SizeSnapshot;
|
|
44
|
+
cuts: CutCounts;
|
|
45
|
+
}
|
|
46
|
+
export declare const KEEP_CATEGORIES: readonly KeepCategory[];
|
|
47
|
+
export declare function emptyCuts(): CutCounts;
|
|
48
|
+
//# sourceMappingURL=types.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"types.d.ts","sourceRoot":"","sources":["../src/types.ts"],"names":[],"mappings":"AAAA,MAAM,MAAM,YAAY,GAAG,UAAU,GAAG,MAAM,GAAG,OAAO,CAAC;AAEzD,MAAM,MAAM,WAAW,GAAG,QAAQ,GAAG,WAAW,CAAC;AAEjD,MAAM,WAAW,SAAS;IACxB,qCAAqC;IACrC,aAAa,EAAE,MAAM,CAAC;IACtB,mEAAmE;IACnE,gBAAgB,EAAE,MAAM,CAAC;IACzB,yCAAyC;IACzC,mBAAmB,EAAE,MAAM,CAAC;IAC5B,0DAA0D;IAC1D,cAAc,EAAE,MAAM,CAAC;IACvB,+DAA+D;IAC/D,iBAAiB,EAAE,MAAM,CAAC;IAC1B,uDAAuD;IACvD,eAAe,EAAE,MAAM,CAAC;IACxB,gEAAgE;IAChE,eAAe,EAAE,MAAM,CAAC;IACxB,uDAAuD;IACvD,QAAQ,EAAE,OAAO,CAAC;CACnB;AAED,MAAM,WAAW,WAAW;IAC1B,2CAA2C;IAC3C,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,kEAAkE;IAClE,IAAI,CAAC,EAAE,YAAY,EAAE,CAAC;IACtB;;;OAGG;IACH,MAAM,CAAC,EAAE,OAAO,CAAC;IACjB,QAAQ,CAAC,EAAE,WAAW,CAAC;IACvB,wDAAwD;IACxD,KAAK,CAAC,EAAE,OAAO,KAAK,CAAC;IACrB,GAAG,CAAC,EAAE,MAAM,CAAC,UAAU,CAAC;CACzB;AAED,MAAM,WAAW,YAAY;IAC3B,KAAK,EAAE,MAAM,CAAC;IACd,MAAM,EAAE,MAAM,CAAC;CAChB;AAED,MAAM,WAAW,UAAU;IACzB,IAAI,EAAE,MAAM,CAAC;IACb,KAAK,EAAE,YAAY,CAAC;IACpB,MAAM,EAAE,YAAY,CAAC;IACrB,IAAI,EAAE,SAAS,CAAC;CACjB;AAED,eAAO,MAAM,eAAe,EAAE,SAAS,YAAY,EAIlD,CAAC;AAEF,wBAAgB,SAAS,IAAI,SAAS,CAWrC"}
|
package/dist/types.js
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
1
|
+
export const KEEP_CATEGORIES = [
|
|
2
|
+
"headings",
|
|
3
|
+
"code",
|
|
4
|
+
"lists",
|
|
5
|
+
];
|
|
6
|
+
export function emptyCuts() {
|
|
7
|
+
return {
|
|
8
|
+
ansiSequences: 0,
|
|
9
|
+
boilerplateLines: 0,
|
|
10
|
+
duplicateParagraphs: 0,
|
|
11
|
+
whitespaceRuns: 0,
|
|
12
|
+
omittedHeavyBlobs: 0,
|
|
13
|
+
truncatedBlocks: 0,
|
|
14
|
+
truncatedTokens: 0,
|
|
15
|
+
deepened: false,
|
|
16
|
+
};
|
|
17
|
+
}
|
|
18
|
+
//# sourceMappingURL=types.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"types.js","sourceRoot":"","sources":["../src/types.ts"],"names":[],"mappings":"AAmDA,MAAM,CAAC,MAAM,eAAe,GAA4B;IACtD,UAAU;IACV,MAAM;IACN,OAAO;CACR,CAAC;AAEF,MAAM,UAAU,SAAS;IACvB,OAAO;QACL,aAAa,EAAE,CAAC;QAChB,gBAAgB,EAAE,CAAC;QACnB,mBAAmB,EAAE,CAAC;QACtB,cAAc,EAAE,CAAC;QACjB,iBAAiB,EAAE,CAAC;QACpB,eAAe,EAAE,CAAC;QAClB,eAAe,EAAE,CAAC;QAClB,QAAQ,EAAE,KAAK;KAChB,CAAC;AACJ,CAAC"}
|
package/package.json
CHANGED
|
@@ -1,6 +1,60 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "@webywz/context-diet",
|
|
3
|
-
"version": "0.
|
|
4
|
-
"
|
|
5
|
-
"
|
|
6
|
-
|
|
3
|
+
"version": "0.1.0",
|
|
4
|
+
"description": "Slim LLM context before you send it — drop noise, keep the signal, estimate tokens.",
|
|
5
|
+
"keywords": [
|
|
6
|
+
"llm",
|
|
7
|
+
"context",
|
|
8
|
+
"tokens",
|
|
9
|
+
"prompt",
|
|
10
|
+
"cli",
|
|
11
|
+
"openai",
|
|
12
|
+
"anthropic",
|
|
13
|
+
"tiktoken"
|
|
14
|
+
],
|
|
15
|
+
"homepage": "https://github.com/webywz/context-diet#readme",
|
|
16
|
+
"bugs": {
|
|
17
|
+
"url": "https://github.com/webywz/context-diet/issues"
|
|
18
|
+
},
|
|
19
|
+
"license": "MIT",
|
|
20
|
+
"author": "webywz",
|
|
21
|
+
"repository": {
|
|
22
|
+
"type": "git",
|
|
23
|
+
"url": "git+https://github.com/webywz/context-diet.git"
|
|
24
|
+
},
|
|
25
|
+
"type": "module",
|
|
26
|
+
"main": "./dist/index.js",
|
|
27
|
+
"types": "./dist/index.d.ts",
|
|
28
|
+
"exports": {
|
|
29
|
+
".": {
|
|
30
|
+
"types": "./dist/index.d.ts",
|
|
31
|
+
"import": "./dist/index.js"
|
|
32
|
+
}
|
|
33
|
+
},
|
|
34
|
+
"bin": {
|
|
35
|
+
"context-diet": "./dist/cli.js"
|
|
36
|
+
},
|
|
37
|
+
"files": [
|
|
38
|
+
"dist",
|
|
39
|
+
"LICENSE",
|
|
40
|
+
"README.md"
|
|
41
|
+
],
|
|
42
|
+
"engines": {
|
|
43
|
+
"node": ">=18"
|
|
44
|
+
},
|
|
45
|
+
"publishConfig": {
|
|
46
|
+
"access": "public"
|
|
47
|
+
},
|
|
48
|
+
"scripts": {
|
|
49
|
+
"build": "tsc",
|
|
50
|
+
"test": "tsx --test test/*.test.ts",
|
|
51
|
+
"typecheck": "tsc --noEmit",
|
|
52
|
+
"dev": "tsx src/cli.ts",
|
|
53
|
+
"prepublishOnly": "npm test && npm run build"
|
|
54
|
+
},
|
|
55
|
+
"devDependencies": {
|
|
56
|
+
"@types/node": "^22.15.29",
|
|
57
|
+
"tsx": "^4.23.15",
|
|
58
|
+
"typescript": "^5.8.3"
|
|
59
|
+
}
|
|
60
|
+
}
|