@t4h.framework/fs 0.4.0 → 0.6.0
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/.ai/skills/framework-fs/SKILL.md +72 -115
- package/.ai/skills/framework-fs-split/SKILL.md +255 -0
- package/README.md +184 -3
- package/dist/FileSystemClaim.d.ts +4 -0
- package/dist/FileSystemClaim.d.ts.map +1 -1
- package/dist/FileSystemClaim.js.map +1 -1
- package/dist/activities/FileJoinActivity.d.ts +39 -0
- package/dist/activities/FileJoinActivity.d.ts.map +1 -0
- package/dist/activities/FileJoinActivity.js +87 -0
- package/dist/activities/FileJoinActivity.js.map +1 -0
- package/dist/activities/FileSplitActivity.d.ts +38 -0
- package/dist/activities/FileSplitActivity.d.ts.map +1 -0
- package/dist/activities/FileSplitActivity.js +51 -0
- package/dist/activities/FileSplitActivity.js.map +1 -0
- package/dist/activities/__tests__/FileJoinActivity.spec.d.ts +2 -0
- package/dist/activities/__tests__/FileJoinActivity.spec.d.ts.map +1 -0
- package/dist/activities/__tests__/FileJoinActivity.spec.js +160 -0
- package/dist/activities/__tests__/FileJoinActivity.spec.js.map +1 -0
- package/dist/activities/__tests__/FileSplitActivity.spec.d.ts +2 -0
- package/dist/activities/__tests__/FileSplitActivity.spec.d.ts.map +1 -0
- package/dist/activities/__tests__/FileSplitActivity.spec.js +149 -0
- package/dist/activities/__tests__/FileSplitActivity.spec.js.map +1 -0
- package/dist/fs.d.ts +3 -0
- package/dist/fs.d.ts.map +1 -1
- package/dist/fs.js +3 -0
- package/dist/fs.js.map +1 -1
- package/dist/helpers/__tests__/byte-delimiter-scanner.spec.d.ts +2 -0
- package/dist/helpers/__tests__/byte-delimiter-scanner.spec.d.ts.map +1 -0
- package/dist/helpers/__tests__/byte-delimiter-scanner.spec.js +127 -0
- package/dist/helpers/__tests__/byte-delimiter-scanner.spec.js.map +1 -0
- package/dist/helpers/__tests__/csv-reader.spec.d.ts +2 -0
- package/dist/helpers/__tests__/csv-reader.spec.d.ts.map +1 -0
- package/dist/helpers/__tests__/csv-reader.spec.js +60 -0
- package/dist/helpers/__tests__/csv-reader.spec.js.map +1 -0
- package/dist/helpers/byte-delimiter-scanner.d.ts +36 -0
- package/dist/helpers/byte-delimiter-scanner.d.ts.map +1 -0
- package/dist/helpers/byte-delimiter-scanner.js +61 -0
- package/dist/helpers/byte-delimiter-scanner.js.map +1 -0
- package/dist/helpers/csv-reader.d.ts +8 -0
- package/dist/helpers/csv-reader.d.ts.map +1 -0
- package/dist/helpers/csv-reader.js +14 -0
- package/dist/helpers/csv-reader.js.map +1 -0
- package/dist/helpers/parsers.d.ts +9 -0
- package/dist/helpers/parsers.d.ts.map +1 -0
- package/dist/helpers/parsers.js +14 -0
- package/dist/helpers/parsers.js.map +1 -0
- package/dist/testing/index.d.ts +3 -0
- package/dist/testing/index.d.ts.map +1 -0
- package/dist/testing/index.js +32 -0
- package/dist/testing/index.js.map +1 -0
- package/package.json +8 -5
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"FileSplitActivity.spec.d.ts","sourceRoot":"","sources":["../../../src/activities/__tests__/FileSplitActivity.spec.ts"],"names":[],"mappings":""}
|
|
@@ -0,0 +1,149 @@
|
|
|
1
|
+
import { Readable } from 'node:stream';
|
|
2
|
+
import { buffer } from 'node:stream/consumers';
|
|
3
|
+
import { AsyncActivityResultFulfilled, AsyncActivityResultPending, Claim, WorkflowBatchRef, WorkflowClaim, } from '@t4h.framework/core';
|
|
4
|
+
import { describe, expect, it, vi } from 'vitest';
|
|
5
|
+
import { FileSystemClaim } from '../../FileSystemClaim.js';
|
|
6
|
+
import { FileSplitActivity, Split, } from '../FileSplitActivity.js';
|
|
7
|
+
function makeBinary(content) {
|
|
8
|
+
return {
|
|
9
|
+
contentType: 'text/plain',
|
|
10
|
+
contentLength: Buffer.byteLength(content),
|
|
11
|
+
stream: () => Readable.from([Buffer.from(content)]),
|
|
12
|
+
};
|
|
13
|
+
}
|
|
14
|
+
const childWorkflow = { id: 'child-workflow' };
|
|
15
|
+
async function bootstrapActivity(input) {
|
|
16
|
+
const durable = { id: 'durable-copy' };
|
|
17
|
+
const written = [];
|
|
18
|
+
const write = vi.fn(async (_path, readable) => {
|
|
19
|
+
written.push(await buffer(readable));
|
|
20
|
+
return durable;
|
|
21
|
+
});
|
|
22
|
+
const slice = vi.fn((binary, range, encoding) => ({ binary, range, encoding }));
|
|
23
|
+
const startMany = vi.fn(async (_workflow, inputs) => new WorkflowBatchRef('batch-1', inputs.length));
|
|
24
|
+
const result = await Claim.run([
|
|
25
|
+
{
|
|
26
|
+
provide: FileSystemClaim,
|
|
27
|
+
value: { write, slice },
|
|
28
|
+
},
|
|
29
|
+
{
|
|
30
|
+
provide: WorkflowClaim,
|
|
31
|
+
value: { startMany },
|
|
32
|
+
},
|
|
33
|
+
], () => {
|
|
34
|
+
const activity = new FileSplitActivity();
|
|
35
|
+
activity.id = 'activity-1';
|
|
36
|
+
return activity.bootstrap(input);
|
|
37
|
+
});
|
|
38
|
+
return { result, durable, write, slice, startMany, written };
|
|
39
|
+
}
|
|
40
|
+
describe('FileSplitActivity', () => {
|
|
41
|
+
describe('toInput', () => {
|
|
42
|
+
it('stores the workflow by id alongside the binary and options', () => {
|
|
43
|
+
const activity = new FileSplitActivity();
|
|
44
|
+
const binary = makeBinary('');
|
|
45
|
+
const options = { delimiter: Split.line() };
|
|
46
|
+
expect(activity.toInput(binary, childWorkflow, options)).toEqual({
|
|
47
|
+
binary,
|
|
48
|
+
workflow: 'child-workflow',
|
|
49
|
+
options,
|
|
50
|
+
});
|
|
51
|
+
});
|
|
52
|
+
it('throws when the workflow has no id', () => {
|
|
53
|
+
const activity = new FileSplitActivity();
|
|
54
|
+
expect(() => activity.toInput(makeBinary(''), {}, {
|
|
55
|
+
delimiter: Split.line(),
|
|
56
|
+
})).toThrow(/without an id/);
|
|
57
|
+
});
|
|
58
|
+
});
|
|
59
|
+
describe('bootstrap', () => {
|
|
60
|
+
it('persists the source once, slices by ranges, starts the batch and suspends', async () => {
|
|
61
|
+
const { result, durable, write, slice, startMany, written } = await bootstrapActivity({
|
|
62
|
+
binary: makeBinary('foo\nbar\nbaz'),
|
|
63
|
+
workflow: 'child-workflow',
|
|
64
|
+
options: { delimiter: Split.line() },
|
|
65
|
+
});
|
|
66
|
+
expect(write).toHaveBeenCalledExactlyOnceWith('split/activity-1', expect.anything());
|
|
67
|
+
expect(written[0]?.toString()).toBe('foo\nbar\nbaz');
|
|
68
|
+
expect(slice).toHaveBeenCalledTimes(3);
|
|
69
|
+
expect(slice).toHaveBeenNthCalledWith(1, durable, { start: 0, end: 3 }, 'utf8');
|
|
70
|
+
expect(slice).toHaveBeenNthCalledWith(2, durable, { start: 4, end: 7 }, 'utf8');
|
|
71
|
+
expect(slice).toHaveBeenNthCalledWith(3, durable, { start: 8, end: 11 }, 'utf8');
|
|
72
|
+
expect(startMany).toHaveBeenCalledExactlyOnceWith('child-workflow', [
|
|
73
|
+
slice.mock.results[0]?.value,
|
|
74
|
+
slice.mock.results[1]?.value,
|
|
75
|
+
slice.mock.results[2]?.value,
|
|
76
|
+
]);
|
|
77
|
+
expect(result).toBeInstanceOf(AsyncActivityResultPending);
|
|
78
|
+
expect(result.schema).toBeDefined();
|
|
79
|
+
});
|
|
80
|
+
it('splits positional records with the record-start delimiter', async () => {
|
|
81
|
+
const { slice, result } = await bootstrapActivity({
|
|
82
|
+
binary: makeBinary('10A;x\n10B;y\n10C;z'),
|
|
83
|
+
workflow: 'child-workflow',
|
|
84
|
+
options: { delimiter: Split.char('\n10', { next: 2 }) },
|
|
85
|
+
});
|
|
86
|
+
expect(slice.mock.calls.map(([, range]) => range)).toEqual([
|
|
87
|
+
{ start: 0, end: 5 },
|
|
88
|
+
{ start: 6, end: 11 },
|
|
89
|
+
{ start: 12, end: 17 },
|
|
90
|
+
]);
|
|
91
|
+
expect(result).toBeInstanceOf(AsyncActivityResultPending);
|
|
92
|
+
});
|
|
93
|
+
it('resolves immediately for a file with no usable slices', async () => {
|
|
94
|
+
const { result, slice, startMany } = await bootstrapActivity({
|
|
95
|
+
binary: makeBinary('\n\n'),
|
|
96
|
+
workflow: 'child-workflow',
|
|
97
|
+
options: { delimiter: Split.line() },
|
|
98
|
+
});
|
|
99
|
+
expect(slice).not.toHaveBeenCalled();
|
|
100
|
+
expect(startMany).toHaveBeenCalledExactlyOnceWith('child-workflow', []);
|
|
101
|
+
expect(result).toEqual(new AsyncActivityResultFulfilled({ batch: 'batch-1', count: 0 }));
|
|
102
|
+
});
|
|
103
|
+
it('schedules the timeout when awaiting the batch', async () => {
|
|
104
|
+
const before = Date.now();
|
|
105
|
+
const { result } = await bootstrapActivity({
|
|
106
|
+
binary: makeBinary('foo\nbar'),
|
|
107
|
+
workflow: 'child-workflow',
|
|
108
|
+
options: { delimiter: Split.line(), timeoutInMs: 60_000 },
|
|
109
|
+
});
|
|
110
|
+
const pending = result;
|
|
111
|
+
expect(pending).toBeInstanceOf(AsyncActivityResultPending);
|
|
112
|
+
expect(pending.timeoutAt?.getTime()).toBeGreaterThanOrEqual(before + 60_000);
|
|
113
|
+
expect(pending.timeoutAt?.getTime()).toBeLessThanOrEqual(Date.now() + 60_000);
|
|
114
|
+
});
|
|
115
|
+
it('suspends without a timeout when none is configured', async () => {
|
|
116
|
+
const { result } = await bootstrapActivity({
|
|
117
|
+
binary: makeBinary('foo\nbar'),
|
|
118
|
+
workflow: 'child-workflow',
|
|
119
|
+
options: { delimiter: Split.line() },
|
|
120
|
+
});
|
|
121
|
+
expect(result.timeoutAt).toBeUndefined();
|
|
122
|
+
});
|
|
123
|
+
it('forwards the configured encoding to the scanner and the slices', async () => {
|
|
124
|
+
const { slice } = await bootstrapActivity({
|
|
125
|
+
binary: makeBinary('a;b'),
|
|
126
|
+
workflow: 'child-workflow',
|
|
127
|
+
options: { delimiter: Split.char(';'), encoding: 'latin1' },
|
|
128
|
+
});
|
|
129
|
+
expect(slice).toHaveBeenCalledWith(expect.anything(), { start: 0, end: 1 }, 'latin1');
|
|
130
|
+
});
|
|
131
|
+
});
|
|
132
|
+
describe('advance', () => {
|
|
133
|
+
it('resolves with the batch reference payload once the batch is terminal', () => {
|
|
134
|
+
const activity = new FileSplitActivity();
|
|
135
|
+
const payload = { batch: 'batch-1', count: 3 };
|
|
136
|
+
expect(activity.advance(payload)).toEqual(new AsyncActivityResultFulfilled(payload));
|
|
137
|
+
});
|
|
138
|
+
});
|
|
139
|
+
describe('toOutput', () => {
|
|
140
|
+
it('maps the payload to a WorkflowBatchRef', () => {
|
|
141
|
+
const activity = new FileSplitActivity();
|
|
142
|
+
const result = activity.toOutput({ batch: 'batch-1', count: 7 });
|
|
143
|
+
expect(result).toBeInstanceOf(WorkflowBatchRef);
|
|
144
|
+
expect(result.id).toBe('batch-1');
|
|
145
|
+
expect(result.count).toBe(7);
|
|
146
|
+
});
|
|
147
|
+
});
|
|
148
|
+
});
|
|
149
|
+
//# sourceMappingURL=FileSplitActivity.spec.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"FileSplitActivity.spec.js","sourceRoot":"","sources":["../../../src/activities/__tests__/FileSplitActivity.spec.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,aAAa,CAAA;AACtC,OAAO,EAAE,MAAM,EAAE,MAAM,uBAAuB,CAAA;AAE9C,OAAO,EACL,4BAA4B,EAC5B,0BAA0B,EAE1B,KAAK,EAEL,gBAAgB,EAChB,aAAa,GACd,MAAM,qBAAqB,CAAA;AAC5B,OAAO,EAAE,QAAQ,EAAE,MAAM,EAAE,EAAE,EAAE,EAAE,EAAE,MAAM,QAAQ,CAAA;AAEjD,OAAO,EAAE,eAAe,EAAE,MAAM,0BAA0B,CAAA;AAC1D,OAAO,EACL,iBAAiB,EAEjB,KAAK,GACN,MAAM,yBAAyB,CAAA;AAEhC,SAAS,UAAU,CAAC,OAAe;IACjC,OAAO;QACL,WAAW,EAAE,YAAY;QACzB,aAAa,EAAE,MAAM,CAAC,UAAU,CAAC,OAAO,CAAC;QACzC,MAAM,EAAE,GAAG,EAAE,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAC,MAAM,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,CAAC;KAC/B,CAAA;AACxB,CAAC;AAED,MAAM,aAAa,GAAG,EAAE,EAAE,EAAE,gBAAgB,EAAmC,CAAA;AAE/E,KAAK,UAAU,iBAAiB,CAAC,KAA6B;IAC5D,MAAM,OAAO,GAAG,EAAE,EAAE,EAAE,cAAc,EAAuB,CAAA;IAE3D,MAAM,OAAO,GAAa,EAAE,CAAA;IAE5B,MAAM,KAAK,GAAG,EAAE,CAAC,EAAE,CAAC,KAAK,EAAE,KAAa,EAAE,QAAkB,EAAE,EAAE;QAC9D,OAAO,CAAC,IAAI,CAAC,MAAM,MAAM,CAAC,QAAQ,CAAC,CAAC,CAAA;QAEpC,OAAO,OAAO,CAAA;IAChB,CAAC,CAAC,CAAA;IAEF,MAAM,KAAK,GAAG,EAAE,CAAC,EAAE,CACjB,CAAC,MAAc,EAAE,KAAqC,EAAE,QAAQ,EAAE,EAAE,CAClE,CAAC,EAAE,MAAM,EAAE,KAAK,EAAE,QAAQ,EAAE,CAAsB,CACrD,CAAA;IAED,MAAM,SAAS,GAAG,EAAE,CAAC,EAAE,CACrB,KAAK,EAAE,SAAS,EAAE,MAAsB,EAAE,EAAE,CAC1C,IAAI,gBAAgB,CAAC,SAAS,EAAE,MAAM,CAAC,MAAM,CAAC,CACjD,CAAA;IAED,MAAM,MAAM,GAAG,MAAM,KAAK,CAAC,GAAG,CAC5B;QACE;YACE,OAAO,EAAE,eAAe;YACxB,KAAK,EAAE,EAAE,KAAK,EAAE,KAAK,EAAgC;SACtD;QACD;YACE,OAAO,EAAE,aAAa;YACtB,KAAK,EAAE,EAAE,SAAS,EAA8B;SACjD;KACF,EACD,GAAG,EAAE;QACH,MAAM,QAAQ,GAAG,IAAI,iBAAiB,EAAE,CAAA;QAExC,QAAQ,CAAC,EAAE,GAAG,YAAY,CAAA;QAE1B,OAAO,QAAQ,CAAC,SAAS,CAAC,KAAK,CAAC,CAAA;IAClC,CAAC,CACF,CAAA;IAED,OAAO,EAAE,MAAM,EAAE,OAAO,EAAE,KAAK,EAAE,KAAK,EAAE,SAAS,EAAE,OAAO,EAAE,CAAA;AAC9D,CAAC;AAED,QAAQ,CAAC,mBAAmB,EAAE,GAAG,EAAE;IACjC,QAAQ,CAAC,SAAS,EAAE,GAAG,EAAE;QACvB,EAAE,CAAC,4DAA4D,EAAE,GAAG,EAAE;YACpE,MAAM,QAAQ,GAAG,IAAI,iBAAiB,EAAE,CAAA;YACxC,MAAM,MAAM,GAAG,UAAU,CAAC,EAAE,CAAC,CAAA;YAC7B,MAAM,OAAO,GAAG,EAAE,SAAS,EAAE,KAAK,CAAC,IAAI,EAAE,EAAE,CAAA;YAE3C,MAAM,CAAC,QAAQ,CAAC,OAAO,CAAC,MAAM,EAAE,aAAa,EAAE,OAAO,CAAC,CAAC,CAAC,OAAO,CAAC;gBAC/D,MAAM;gBACN,QAAQ,EAAE,gBAAgB;gBAC1B,OAAO;aACR,CAAC,CAAA;QACJ,CAAC,CAAC,CAAA;QAEF,EAAE,CAAC,oCAAoC,EAAE,GAAG,EAAE;YAC5C,MAAM,QAAQ,GAAG,IAAI,iBAAiB,EAAE,CAAA;YAExC,MAAM,CAAC,GAAG,EAAE,CACV,QAAQ,CAAC,OAAO,CAAC,UAAU,CAAC,EAAE,CAAC,EAAE,EAAwB,EAAE;gBACzD,SAAS,EAAE,KAAK,CAAC,IAAI,EAAE;aACxB,CAAC,CACH,CAAC,OAAO,CAAC,eAAe,CAAC,CAAA;QAC5B,CAAC,CAAC,CAAA;IACJ,CAAC,CAAC,CAAA;IAEF,QAAQ,CAAC,WAAW,EAAE,GAAG,EAAE;QACzB,EAAE,CAAC,2EAA2E,EAAE,KAAK,IAAI,EAAE;YACzF,MAAM,EAAE,MAAM,EAAE,OAAO,EAAE,KAAK,EAAE,KAAK,EAAE,SAAS,EAAE,OAAO,EAAE,GACzD,MAAM,iBAAiB,CAAC;gBACtB,MAAM,EAAE,UAAU,CAAC,eAAe,CAAC;gBACnC,QAAQ,EAAE,gBAAgB;gBAC1B,OAAO,EAAE,EAAE,SAAS,EAAE,KAAK,CAAC,IAAI,EAAE,EAAE;aACrC,CAAC,CAAA;YAEJ,MAAM,CAAC,KAAK,CAAC,CAAC,+BAA+B,CAC3C,kBAAkB,EAClB,MAAM,CAAC,QAAQ,EAAE,CAClB,CAAA;YACD,MAAM,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,QAAQ,EAAE,CAAC,CAAC,IAAI,CAAC,eAAe,CAAC,CAAA;YAEpD,MAAM,CAAC,KAAK,CAAC,CAAC,qBAAqB,CAAC,CAAC,CAAC,CAAA;YACtC,MAAM,CAAC,KAAK,CAAC,CAAC,uBAAuB,CACnC,CAAC,EACD,OAAO,EACP,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE,EACpB,MAAM,CACP,CAAA;YACD,MAAM,CAAC,KAAK,CAAC,CAAC,uBAAuB,CACnC,CAAC,EACD,OAAO,EACP,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE,EACpB,MAAM,CACP,CAAA;YACD,MAAM,CAAC,KAAK,CAAC,CAAC,uBAAuB,CACnC,CAAC,EACD,OAAO,EACP,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,EAAE,EAAE,EACrB,MAAM,CACP,CAAA;YAED,MAAM,CAAC,SAAS,CAAC,CAAC,+BAA+B,CAAC,gBAAgB,EAAE;gBAClE,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,KAAK;gBAC5B,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,KAAK;gBAC5B,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,KAAK;aAC7B,CAAC,CAAA;YAEF,MAAM,CAAC,MAAM,CAAC,CAAC,cAAc,CAAC,0BAA0B,CAAC,CAAA;YACzD,MAAM,CAAE,MAAqC,CAAC,MAAM,CAAC,CAAC,WAAW,EAAE,CAAA;QACrE,CAAC,CAAC,CAAA;QAEF,EAAE,CAAC,2DAA2D,EAAE,KAAK,IAAI,EAAE;YACzE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,GAAG,MAAM,iBAAiB,CAAC;gBAChD,MAAM,EAAE,UAAU,CAAC,qBAAqB,CAAC;gBACzC,QAAQ,EAAE,gBAAgB;gBAC1B,OAAO,EAAE,EAAE,SAAS,EAAE,KAAK,CAAC,IAAI,CAAC,MAAM,EAAE,EAAE,IAAI,EAAE,CAAC,EAAE,CAAC,EAAE;aACxD,CAAC,CAAA;YAEF,MAAM,CAAC,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,KAAK,CAAC,EAAE,EAAE,CAAC,KAAK,CAAC,CAAC,CAAC,OAAO,CAAC;gBACzD,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;gBACpB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,EAAE,EAAE;gBACrB,EAAE,KAAK,EAAE,EAAE,EAAE,GAAG,EAAE,EAAE,EAAE;aACvB,CAAC,CAAA;YAEF,MAAM,CAAC,MAAM,CAAC,CAAC,cAAc,CAAC,0BAA0B,CAAC,CAAA;QAC3D,CAAC,CAAC,CAAA;QAEF,EAAE,CAAC,uDAAuD,EAAE,KAAK,IAAI,EAAE;YACrE,MAAM,EAAE,MAAM,EAAE,KAAK,EAAE,SAAS,EAAE,GAAG,MAAM,iBAAiB,CAAC;gBAC3D,MAAM,EAAE,UAAU,CAAC,MAAM,CAAC;gBAC1B,QAAQ,EAAE,gBAAgB;gBAC1B,OAAO,EAAE,EAAE,SAAS,EAAE,KAAK,CAAC,IAAI,EAAE,EAAE;aACrC,CAAC,CAAA;YAEF,MAAM,CAAC,KAAK,CAAC,CAAC,GAAG,CAAC,gBAAgB,EAAE,CAAA;YACpC,MAAM,CAAC,SAAS,CAAC,CAAC,+BAA+B,CAAC,gBAAgB,EAAE,EAAE,CAAC,CAAA;YACvE,MAAM,CAAC,MAAM,CAAC,CAAC,OAAO,CACpB,IAAI,4BAA4B,CAAC,EAAE,KAAK,EAAE,SAAS,EAAE,KAAK,EAAE,CAAC,EAAE,CAAC,CACjE,CAAA;QACH,CAAC,CAAC,CAAA;QAEF,EAAE,CAAC,+CAA+C,EAAE,KAAK,IAAI,EAAE;YAC7D,MAAM,MAAM,GAAG,IAAI,CAAC,GAAG,EAAE,CAAA;YAEzB,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,iBAAiB,CAAC;gBACzC,MAAM,EAAE,UAAU,CAAC,UAAU,CAAC;gBAC9B,QAAQ,EAAE,gBAAgB;gBAC1B,OAAO,EAAE,EAAE,SAAS,EAAE,KAAK,CAAC,IAAI,EAAE,EAAE,WAAW,EAAE,MAAM,EAAE;aAC1D,CAAC,CAAA;YAEF,MAAM,OAAO,GAAG,MAAoC,CAAA;YAEpD,MAAM,CAAC,OAAO,CAAC,CAAC,cAAc,CAAC,0BAA0B,CAAC,CAAA;YAC1D,MAAM,CAAC,OAAO,CAAC,SAAS,EAAE,OAAO,EAAE,CAAC,CAAC,sBAAsB,CACzD,MAAM,GAAG,MAAM,CAChB,CAAA;YACD,MAAM,CAAC,OAAO,CAAC,SAAS,EAAE,OAAO,EAAE,CAAC,CAAC,mBAAmB,CACtD,IAAI,CAAC,GAAG,EAAE,GAAG,MAAM,CACpB,CAAA;QACH,CAAC,CAAC,CAAA;QAEF,EAAE,CAAC,oDAAoD,EAAE,KAAK,IAAI,EAAE;YAClE,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,iBAAiB,CAAC;gBACzC,MAAM,EAAE,UAAU,CAAC,UAAU,CAAC;gBAC9B,QAAQ,EAAE,gBAAgB;gBAC1B,OAAO,EAAE,EAAE,SAAS,EAAE,KAAK,CAAC,IAAI,EAAE,EAAE;aACrC,CAAC,CAAA;YAEF,MAAM,CAAE,MAAqC,CAAC,SAAS,CAAC,CAAC,aAAa,EAAE,CAAA;QAC1E,CAAC,CAAC,CAAA;QAEF,EAAE,CAAC,gEAAgE,EAAE,KAAK,IAAI,EAAE;YAC9E,MAAM,EAAE,KAAK,EAAE,GAAG,MAAM,iBAAiB,CAAC;gBACxC,MAAM,EAAE,UAAU,CAAC,KAAK,CAAC;gBACzB,QAAQ,EAAE,gBAAgB;gBAC1B,OAAO,EAAE,EAAE,SAAS,EAAE,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,QAAQ,EAAE,QAAQ,EAAE;aAC5D,CAAC,CAAA;YAEF,MAAM,CAAC,KAAK,CAAC,CAAC,oBAAoB,CAChC,MAAM,CAAC,QAAQ,EAAE,EACjB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE,EACpB,QAAQ,CACT,CAAA;QACH,CAAC,CAAC,CAAA;IACJ,CAAC,CAAC,CAAA;IAEF,QAAQ,CAAC,SAAS,EAAE,GAAG,EAAE;QACvB,EAAE,CAAC,sEAAsE,EAAE,GAAG,EAAE;YAC9E,MAAM,QAAQ,GAAG,IAAI,iBAAiB,EAAE,CAAA;YACxC,MAAM,OAAO,GAAG,EAAE,KAAK,EAAE,SAAS,EAAE,KAAK,EAAE,CAAC,EAAE,CAAA;YAE9C,MAAM,CAAC,QAAQ,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC,CAAC,OAAO,CACvC,IAAI,4BAA4B,CAAC,OAAO,CAAC,CAC1C,CAAA;QACH,CAAC,CAAC,CAAA;IACJ,CAAC,CAAC,CAAA;IAEF,QAAQ,CAAC,UAAU,EAAE,GAAG,EAAE;QACxB,EAAE,CAAC,wCAAwC,EAAE,GAAG,EAAE;YAChD,MAAM,QAAQ,GAAG,IAAI,iBAAiB,EAAE,CAAA;YACxC,MAAM,MAAM,GAAG,QAAQ,CAAC,QAAQ,CAAC,EAAE,KAAK,EAAE,SAAS,EAAE,KAAK,EAAE,CAAC,EAAE,CAAC,CAAA;YAEhE,MAAM,CAAC,MAAM,CAAC,CAAC,cAAc,CAAC,gBAAgB,CAAC,CAAA;YAC/C,MAAM,CAAC,MAAM,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,CAAA;YACjC,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QAC9B,CAAC,CAAC,CAAA;IACJ,CAAC,CAAC,CAAA;AACJ,CAAC,CAAC,CAAA"}
|
package/dist/fs.d.ts
CHANGED
package/dist/fs.d.ts.map
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"fs.d.ts","sourceRoot":"","sources":["../src/fs.ts"],"names":[],"mappings":"AAAA,cAAc,sBAAsB,CAAA"}
|
|
1
|
+
{"version":3,"file":"fs.d.ts","sourceRoot":"","sources":["../src/fs.ts"],"names":[],"mappings":"AAAA,cAAc,sBAAsB,CAAA;AACpC,cAAc,kCAAkC,CAAA;AAChD,cAAc,mCAAmC,CAAA;AACjD,cAAc,qCAAqC,CAAA"}
|
package/dist/fs.js
CHANGED
package/dist/fs.js.map
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"fs.js","sourceRoot":"","sources":["../src/fs.ts"],"names":[],"mappings":"AAAA,cAAc,sBAAsB,CAAA"}
|
|
1
|
+
{"version":3,"file":"fs.js","sourceRoot":"","sources":["../src/fs.ts"],"names":[],"mappings":"AAAA,cAAc,sBAAsB,CAAA;AACpC,cAAc,kCAAkC,CAAA;AAChD,cAAc,mCAAmC,CAAA;AACjD,cAAc,qCAAqC,CAAA"}
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"byte-delimiter-scanner.spec.d.ts","sourceRoot":"","sources":["../../../src/helpers/__tests__/byte-delimiter-scanner.spec.ts"],"names":[],"mappings":""}
|
|
@@ -0,0 +1,127 @@
|
|
|
1
|
+
import { Readable } from 'node:stream';
|
|
2
|
+
import { buffer } from 'node:stream/consumers';
|
|
3
|
+
import { describe, expect, it } from 'vitest';
|
|
4
|
+
import { ByteDelimiterScanner, } from '../byte-delimiter-scanner.js';
|
|
5
|
+
async function scan(chunks, options) {
|
|
6
|
+
const scanner = new ByteDelimiterScanner(options);
|
|
7
|
+
const output = await buffer(Readable.from(chunks.map(chunk => Buffer.isBuffer(chunk) ? chunk : Buffer.from(chunk))).pipe(scanner));
|
|
8
|
+
return { ranges: scanner.ranges, output };
|
|
9
|
+
}
|
|
10
|
+
describe('ByteDelimiterScanner', () => {
|
|
11
|
+
it('cuts at every delimiter, discarding it by default', async () => {
|
|
12
|
+
const { ranges } = await scan(['foo\nbar\nbaz'], {
|
|
13
|
+
delimiter: { value: '\n' },
|
|
14
|
+
});
|
|
15
|
+
expect(ranges).toEqual([
|
|
16
|
+
{ start: 0, end: 3 },
|
|
17
|
+
{ start: 4, end: 7 },
|
|
18
|
+
{ start: 8, end: 11 },
|
|
19
|
+
]);
|
|
20
|
+
});
|
|
21
|
+
it('skips the empty final slice when the file ends with a delimiter', async () => {
|
|
22
|
+
const { ranges } = await scan(['foo\nbar\n'], {
|
|
23
|
+
delimiter: { value: '\n' },
|
|
24
|
+
});
|
|
25
|
+
expect(ranges).toEqual([
|
|
26
|
+
{ start: 0, end: 3 },
|
|
27
|
+
{ start: 4, end: 7 },
|
|
28
|
+
]);
|
|
29
|
+
});
|
|
30
|
+
it('keeps empty slices when skipEmpty is off', async () => {
|
|
31
|
+
const { ranges } = await scan(['foo\n\nbar'], {
|
|
32
|
+
delimiter: { value: '\n' },
|
|
33
|
+
skipEmpty: false,
|
|
34
|
+
});
|
|
35
|
+
expect(ranges).toEqual([
|
|
36
|
+
{ start: 0, end: 3 },
|
|
37
|
+
{ start: 4, end: 4 },
|
|
38
|
+
{ start: 5, end: 8 },
|
|
39
|
+
]);
|
|
40
|
+
});
|
|
41
|
+
it('passes the bytes through unchanged', async () => {
|
|
42
|
+
const content = 'foo\nbar\nbaz\n';
|
|
43
|
+
const { output } = await scan([content], { delimiter: { value: '\n' } });
|
|
44
|
+
expect(output.toString()).toBe(content);
|
|
45
|
+
});
|
|
46
|
+
it('finds a delimiter split across chunk boundaries', async () => {
|
|
47
|
+
const { ranges } = await scan(['foo\r', '\nbar'], {
|
|
48
|
+
delimiter: { value: '\r\n' },
|
|
49
|
+
});
|
|
50
|
+
expect(ranges).toEqual([
|
|
51
|
+
{ start: 0, end: 3 },
|
|
52
|
+
{ start: 5, end: 8 },
|
|
53
|
+
]);
|
|
54
|
+
});
|
|
55
|
+
it('does not double count a match ending inside the kept tail', async () => {
|
|
56
|
+
// '\n' is both a full match and a possible prefix of the 2-byte tail kept
|
|
57
|
+
// between chunks; a rescan would produce a phantom empty slice.
|
|
58
|
+
const { ranges } = await scan(['a\n10b', '\n10c'], {
|
|
59
|
+
delimiter: { value: '\n10', keep: { next: 2 } },
|
|
60
|
+
});
|
|
61
|
+
expect(ranges).toEqual([
|
|
62
|
+
{ start: 0, end: 1 },
|
|
63
|
+
{ start: 2, end: 5 },
|
|
64
|
+
{ start: 6, end: 9 },
|
|
65
|
+
]);
|
|
66
|
+
});
|
|
67
|
+
it('supports the positional record-start cut: newline dies, marker stays with the next slice', async () => {
|
|
68
|
+
const content = '10A;x\n10B;y\n10C;z';
|
|
69
|
+
const { ranges } = await scan([content], {
|
|
70
|
+
delimiter: { value: '\n10', keep: { next: 2 } },
|
|
71
|
+
});
|
|
72
|
+
expect(ranges).toEqual([
|
|
73
|
+
{ start: 0, end: 5 },
|
|
74
|
+
{ start: 6, end: 11 },
|
|
75
|
+
{ start: 12, end: 17 },
|
|
76
|
+
]);
|
|
77
|
+
expect(content.slice(6, 11)).toBe('10B;y');
|
|
78
|
+
});
|
|
79
|
+
it('supports keeping delimiter bytes with the previous slice', async () => {
|
|
80
|
+
const { ranges } = await scan(['a;b;c'], {
|
|
81
|
+
delimiter: { value: ';', keep: { previous: 1 } },
|
|
82
|
+
});
|
|
83
|
+
expect(ranges).toEqual([
|
|
84
|
+
{ start: 0, end: 2 },
|
|
85
|
+
{ start: 2, end: 4 },
|
|
86
|
+
{ start: 4, end: 5 },
|
|
87
|
+
]);
|
|
88
|
+
});
|
|
89
|
+
it('cuts only every N occurrences to group records', async () => {
|
|
90
|
+
const { ranges } = await scan(['a\nb\nc\nd'], {
|
|
91
|
+
delimiter: { value: '\n' },
|
|
92
|
+
every: 2,
|
|
93
|
+
});
|
|
94
|
+
expect(ranges).toEqual([
|
|
95
|
+
{ start: 0, end: 3 },
|
|
96
|
+
{ start: 4, end: 7 },
|
|
97
|
+
]);
|
|
98
|
+
});
|
|
99
|
+
it('yields the whole file as one slice when the delimiter never occurs', async () => {
|
|
100
|
+
const { ranges } = await scan(['no-delimiter-here'], {
|
|
101
|
+
delimiter: { value: '\n' },
|
|
102
|
+
});
|
|
103
|
+
expect(ranges).toEqual([{ start: 0, end: 17 }]);
|
|
104
|
+
});
|
|
105
|
+
it('converts the delimiter to bytes using the configured encoding', async () => {
|
|
106
|
+
const { ranges } = await scan([Buffer.from('aébé', 'latin1')], {
|
|
107
|
+
delimiter: { value: 'é' },
|
|
108
|
+
encoding: 'latin1',
|
|
109
|
+
});
|
|
110
|
+
expect(ranges).toEqual([
|
|
111
|
+
{ start: 0, end: 1 },
|
|
112
|
+
{ start: 2, end: 3 },
|
|
113
|
+
]);
|
|
114
|
+
});
|
|
115
|
+
it('rejects an empty delimiter', () => {
|
|
116
|
+
expect(() => new ByteDelimiterScanner({ delimiter: { value: '' } })).toThrow(/must not be empty/);
|
|
117
|
+
});
|
|
118
|
+
it('rejects keep sizes exceeding the delimiter length', () => {
|
|
119
|
+
expect(() => new ByteDelimiterScanner({
|
|
120
|
+
delimiter: { value: ';', keep: { previous: 1, next: 1 } },
|
|
121
|
+
})).toThrow(/must not exceed/);
|
|
122
|
+
});
|
|
123
|
+
it('rejects a non-positive every', () => {
|
|
124
|
+
expect(() => new ByteDelimiterScanner({ delimiter: { value: ';' }, every: 0 })).toThrow(/positive integer/);
|
|
125
|
+
});
|
|
126
|
+
});
|
|
127
|
+
//# sourceMappingURL=byte-delimiter-scanner.spec.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"byte-delimiter-scanner.spec.js","sourceRoot":"","sources":["../../../src/helpers/__tests__/byte-delimiter-scanner.spec.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,aAAa,CAAA;AACtC,OAAO,EAAE,MAAM,EAAE,MAAM,uBAAuB,CAAA;AAE9C,OAAO,EAAE,QAAQ,EAAE,MAAM,EAAE,EAAE,EAAE,MAAM,QAAQ,CAAA;AAE7C,OAAO,EACL,oBAAoB,GAGrB,MAAM,8BAA8B,CAAA;AAErC,KAAK,UAAU,IAAI,CACjB,MAA2B,EAC3B,OAAoC;IAEpC,MAAM,OAAO,GAAG,IAAI,oBAAoB,CAAC,OAAO,CAAC,CAAA;IAEjD,MAAM,MAAM,GAAG,MAAM,MAAM,CACzB,QAAQ,CAAC,IAAI,CACX,MAAM,CAAC,GAAG,CAAC,KAAK,CAAC,EAAE,CACjB,MAAM,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CACpD,CACF,CAAC,IAAI,CAAC,OAAO,CAAC,CAChB,CAAA;IAED,OAAO,EAAE,MAAM,EAAE,OAAO,CAAC,MAAM,EAAE,MAAM,EAAE,CAAA;AAC3C,CAAC;AAED,QAAQ,CAAC,sBAAsB,EAAE,GAAG,EAAE;IACpC,EAAE,CAAC,mDAAmD,EAAE,KAAK,IAAI,EAAE;QACjE,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,IAAI,CAAC,CAAC,eAAe,CAAC,EAAE;YAC/C,SAAS,EAAE,EAAE,KAAK,EAAE,IAAI,EAAE;SAC3B,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,CAAC,OAAO,CAAC;YACrB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;YACpB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;YACpB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,EAAE,EAAE;SACtB,CAAC,CAAA;IACJ,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,iEAAiE,EAAE,KAAK,IAAI,EAAE;QAC/E,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,IAAI,CAAC,CAAC,YAAY,CAAC,EAAE;YAC5C,SAAS,EAAE,EAAE,KAAK,EAAE,IAAI,EAAE;SAC3B,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,CAAC,OAAO,CAAC;YACrB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;YACpB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;SACrB,CAAC,CAAA;IACJ,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,0CAA0C,EAAE,KAAK,IAAI,EAAE;QACxD,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,IAAI,CAAC,CAAC,YAAY,CAAC,EAAE;YAC5C,SAAS,EAAE,EAAE,KAAK,EAAE,IAAI,EAAE;YAC1B,SAAS,EAAE,KAAK;SACjB,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,CAAC,OAAO,CAAC;YACrB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;YACpB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;YACpB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;SACrB,CAAC,CAAA;IACJ,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,oCAAoC,EAAE,KAAK,IAAI,EAAE;QAClD,MAAM,OAAO,GAAG,iBAAiB,CAAA;QAEjC,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,IAAI,CAAC,CAAC,OAAO,CAAC,EAAE,EAAE,SAAS,EAAE,EAAE,KAAK,EAAE,IAAI,EAAE,EAAE,CAAC,CAAA;QAExE,MAAM,CAAC,MAAM,CAAC,QAAQ,EAAE,CAAC,CAAC,IAAI,CAAC,OAAO,CAAC,CAAA;IACzC,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,iDAAiD,EAAE,KAAK,IAAI,EAAE;QAC/D,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,IAAI,CAAC,CAAC,OAAO,EAAE,OAAO,CAAC,EAAE;YAChD,SAAS,EAAE,EAAE,KAAK,EAAE,MAAM,EAAE;SAC7B,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,CAAC,OAAO,CAAC;YACrB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;YACpB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;SACrB,CAAC,CAAA;IACJ,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,2DAA2D,EAAE,KAAK,IAAI,EAAE;QACzE,0EAA0E;QAC1E,gEAAgE;QAChE,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,IAAI,CAAC,CAAC,QAAQ,EAAE,OAAO,CAAC,EAAE;YACjD,SAAS,EAAE,EAAE,KAAK,EAAE,MAAM,EAAE,IAAI,EAAE,EAAE,IAAI,EAAE,CAAC,EAAE,EAAE;SAChD,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,CAAC,OAAO,CAAC;YACrB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;YACpB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;YACpB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;SACrB,CAAC,CAAA;IACJ,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,0FAA0F,EAAE,KAAK,IAAI,EAAE;QACxG,MAAM,OAAO,GAAG,qBAAqB,CAAA;QAErC,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,IAAI,CAAC,CAAC,OAAO,CAAC,EAAE;YACvC,SAAS,EAAE,EAAE,KAAK,EAAE,MAAM,EAAE,IAAI,EAAE,EAAE,IAAI,EAAE,CAAC,EAAE,EAAE;SAChD,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,CAAC,OAAO,CAAC;YACrB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;YACpB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,EAAE,EAAE;YACrB,EAAE,KAAK,EAAE,EAAE,EAAE,GAAG,EAAE,EAAE,EAAE;SACvB,CAAC,CAAA;QAEF,MAAM,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,OAAO,CAAC,CAAA;IAC5C,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,0DAA0D,EAAE,KAAK,IAAI,EAAE;QACxE,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,IAAI,CAAC,CAAC,OAAO,CAAC,EAAE;YACvC,SAAS,EAAE,EAAE,KAAK,EAAE,GAAG,EAAE,IAAI,EAAE,EAAE,QAAQ,EAAE,CAAC,EAAE,EAAE;SACjD,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,CAAC,OAAO,CAAC;YACrB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;YACpB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;YACpB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;SACrB,CAAC,CAAA;IACJ,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,gDAAgD,EAAE,KAAK,IAAI,EAAE;QAC9D,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,IAAI,CAAC,CAAC,YAAY,CAAC,EAAE;YAC5C,SAAS,EAAE,EAAE,KAAK,EAAE,IAAI,EAAE;YAC1B,KAAK,EAAE,CAAC;SACT,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,CAAC,OAAO,CAAC;YACrB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;YACpB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;SACrB,CAAC,CAAA;IACJ,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,oEAAoE,EAAE,KAAK,IAAI,EAAE;QAClF,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,IAAI,CAAC,CAAC,mBAAmB,CAAC,EAAE;YACnD,SAAS,EAAE,EAAE,KAAK,EAAE,IAAI,EAAE;SAC3B,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,CAAC,OAAO,CAAC,CAAC,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,EAAE,EAAE,CAAC,CAAC,CAAA;IACjD,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,+DAA+D,EAAE,KAAK,IAAI,EAAE;QAC7E,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,IAAI,CAAC,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,EAAE,QAAQ,CAAC,CAAC,EAAE;YAC7D,SAAS,EAAE,EAAE,KAAK,EAAE,GAAG,EAAE;YACzB,QAAQ,EAAE,QAAQ;SACnB,CAAC,CAAA;QAEF,MAAM,CAAC,MAAM,CAAC,CAAC,OAAO,CAAC;YACrB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;YACpB,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE;SACrB,CAAC,CAAA;IACJ,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,4BAA4B,EAAE,GAAG,EAAE;QACpC,MAAM,CACJ,GAAG,EAAE,CAAC,IAAI,oBAAoB,CAAC,EAAE,SAAS,EAAE,EAAE,KAAK,EAAE,EAAE,EAAE,EAAE,CAAC,CAC7D,CAAC,OAAO,CAAC,mBAAmB,CAAC,CAAA;IAChC,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,mDAAmD,EAAE,GAAG,EAAE;QAC3D,MAAM,CACJ,GAAG,EAAE,CACH,IAAI,oBAAoB,CAAC;YACvB,SAAS,EAAE,EAAE,KAAK,EAAE,GAAG,EAAE,IAAI,EAAE,EAAE,QAAQ,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,EAAE;SAC1D,CAAC,CACL,CAAC,OAAO,CAAC,iBAAiB,CAAC,CAAA;IAC9B,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,8BAA8B,EAAE,GAAG,EAAE;QACtC,MAAM,CACJ,GAAG,EAAE,CAAC,IAAI,oBAAoB,CAAC,EAAE,SAAS,EAAE,EAAE,KAAK,EAAE,GAAG,EAAE,EAAE,KAAK,EAAE,CAAC,EAAE,CAAC,CACxE,CAAC,OAAO,CAAC,kBAAkB,CAAC,CAAA;IAC/B,CAAC,CAAC,CAAA;AACJ,CAAC,CAAC,CAAA"}
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"csv-reader.spec.d.ts","sourceRoot":"","sources":["../../../src/helpers/__tests__/csv-reader.spec.ts"],"names":[],"mappings":""}
|
|
@@ -0,0 +1,60 @@
|
|
|
1
|
+
import { Readable } from 'node:stream';
|
|
2
|
+
import { describe, expect, it, vi } from 'vitest';
|
|
3
|
+
import { csvReader } from '../csv-reader.js';
|
|
4
|
+
async function collect(iterable) {
|
|
5
|
+
const records = [];
|
|
6
|
+
for await (const record of iterable)
|
|
7
|
+
records.push(record);
|
|
8
|
+
return records;
|
|
9
|
+
}
|
|
10
|
+
describe('csvReader', () => {
|
|
11
|
+
it('parses rows into objects keyed by the header row', async () => {
|
|
12
|
+
const stream = Readable.from(['name,age\nalice,30\nbob,25\n']);
|
|
13
|
+
const records = await collect(csvReader(stream));
|
|
14
|
+
expect(records).toEqual([
|
|
15
|
+
JSON.stringify({ name: 'alice', age: '30' }),
|
|
16
|
+
JSON.stringify({ name: 'bob', age: '25' }),
|
|
17
|
+
]);
|
|
18
|
+
});
|
|
19
|
+
it('uses comma as the default delimiter', async () => {
|
|
20
|
+
const stream = Readable.from(['a,b\n1,2\n']);
|
|
21
|
+
const records = await collect(csvReader(stream));
|
|
22
|
+
expect(records).toEqual([JSON.stringify({ a: '1', b: '2' })]);
|
|
23
|
+
});
|
|
24
|
+
it('respects a custom delimiter', async () => {
|
|
25
|
+
const stream = Readable.from(['a;b\n1;2\n']);
|
|
26
|
+
const records = await collect(csvReader(stream, { delimiter: ';' }));
|
|
27
|
+
expect(records).toEqual([JSON.stringify({ a: '1', b: '2' })]);
|
|
28
|
+
});
|
|
29
|
+
it('tolerates rows with a mismatched column count (relax_column_count)', async () => {
|
|
30
|
+
const stream = Readable.from(['a,b\n1\n2,3,4\n']);
|
|
31
|
+
const records = await collect(csvReader(stream, { delimiter: ',' }));
|
|
32
|
+
expect(records).toEqual([
|
|
33
|
+
JSON.stringify({ a: '1' }),
|
|
34
|
+
JSON.stringify({ a: '2', b: '3' }),
|
|
35
|
+
]);
|
|
36
|
+
});
|
|
37
|
+
it('yields nothing for an empty stream', async () => {
|
|
38
|
+
const stream = Readable.from(['']);
|
|
39
|
+
const records = await collect(csvReader(stream));
|
|
40
|
+
expect(records).toEqual([]);
|
|
41
|
+
});
|
|
42
|
+
it('yields nothing when only the header row is present', async () => {
|
|
43
|
+
const stream = Readable.from(['a,b\n']);
|
|
44
|
+
const records = await collect(csvReader(stream));
|
|
45
|
+
expect(records).toEqual([]);
|
|
46
|
+
});
|
|
47
|
+
it('sets the stream encoding when the encoding option is provided', async () => {
|
|
48
|
+
const stream = Readable.from(['a,b\n1,2\n']);
|
|
49
|
+
const setEncoding = vi.spyOn(stream, 'setEncoding');
|
|
50
|
+
await collect(csvReader(stream, { encoding: 'latin1' }));
|
|
51
|
+
expect(setEncoding).toHaveBeenCalledWith('latin1');
|
|
52
|
+
});
|
|
53
|
+
it('does not touch the stream encoding when no encoding option is provided', async () => {
|
|
54
|
+
const stream = Readable.from(['a,b\n1,2\n']);
|
|
55
|
+
const setEncoding = vi.spyOn(stream, 'setEncoding');
|
|
56
|
+
await collect(csvReader(stream));
|
|
57
|
+
expect(setEncoding).not.toHaveBeenCalled();
|
|
58
|
+
});
|
|
59
|
+
});
|
|
60
|
+
//# sourceMappingURL=csv-reader.spec.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"csv-reader.spec.js","sourceRoot":"","sources":["../../../src/helpers/__tests__/csv-reader.spec.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,aAAa,CAAA;AAEtC,OAAO,EAAE,QAAQ,EAAE,MAAM,EAAE,EAAE,EAAE,EAAE,EAAE,MAAM,QAAQ,CAAA;AAEjD,OAAO,EAAE,SAAS,EAAE,MAAM,kBAAkB,CAAA;AAE5C,KAAK,UAAU,OAAO,CAAI,QAA0B;IAClD,MAAM,OAAO,GAAQ,EAAE,CAAA;IAEvB,IAAI,KAAK,EAAE,MAAM,MAAM,IAAI,QAAQ;QAAE,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,CAAA;IAEzD,OAAO,OAAO,CAAA;AAChB,CAAC;AAED,QAAQ,CAAC,WAAW,EAAE,GAAG,EAAE;IACzB,EAAE,CAAC,kDAAkD,EAAE,KAAK,IAAI,EAAE;QAChE,MAAM,MAAM,GAAG,QAAQ,CAAC,IAAI,CAAC,CAAC,8BAA8B,CAAC,CAAC,CAAA;QAE9D,MAAM,OAAO,GAAG,MAAM,OAAO,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC,CAAA;QAEhD,MAAM,CAAC,OAAO,CAAC,CAAC,OAAO,CAAC;YACtB,IAAI,CAAC,SAAS,CAAC,EAAE,IAAI,EAAE,OAAO,EAAE,GAAG,EAAE,IAAI,EAAE,CAAC;YAC5C,IAAI,CAAC,SAAS,CAAC,EAAE,IAAI,EAAE,KAAK,EAAE,GAAG,EAAE,IAAI,EAAE,CAAC;SAC3C,CAAC,CAAA;IACJ,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,qCAAqC,EAAE,KAAK,IAAI,EAAE;QACnD,MAAM,MAAM,GAAG,QAAQ,CAAC,IAAI,CAAC,CAAC,YAAY,CAAC,CAAC,CAAA;QAE5C,MAAM,OAAO,GAAG,MAAM,OAAO,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC,CAAA;QAEhD,MAAM,CAAC,OAAO,CAAC,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,CAAC,CAAC,CAAA;IAC/D,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,6BAA6B,EAAE,KAAK,IAAI,EAAE;QAC3C,MAAM,MAAM,GAAG,QAAQ,CAAC,IAAI,CAAC,CAAC,YAAY,CAAC,CAAC,CAAA;QAE5C,MAAM,OAAO,GAAG,MAAM,OAAO,CAAC,SAAS,CAAC,MAAM,EAAE,EAAE,SAAS,EAAE,GAAG,EAAE,CAAC,CAAC,CAAA;QAEpE,MAAM,CAAC,OAAO,CAAC,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,CAAC,CAAC,CAAA;IAC/D,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,oEAAoE,EAAE,KAAK,IAAI,EAAE;QAClF,MAAM,MAAM,GAAG,QAAQ,CAAC,IAAI,CAAC,CAAC,iBAAiB,CAAC,CAAC,CAAA;QAEjD,MAAM,OAAO,GAAG,MAAM,OAAO,CAAC,SAAS,CAAC,MAAM,EAAE,EAAE,SAAS,EAAE,GAAG,EAAE,CAAC,CAAC,CAAA;QAEpE,MAAM,CAAC,OAAO,CAAC,CAAC,OAAO,CAAC;YACtB,IAAI,CAAC,SAAS,CAAC,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC;YAC1B,IAAI,CAAC,SAAS,CAAC,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC;SACnC,CAAC,CAAA;IACJ,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,oCAAoC,EAAE,KAAK,IAAI,EAAE;QAClD,MAAM,MAAM,GAAG,QAAQ,CAAC,IAAI,CAAC,CAAC,EAAE,CAAC,CAAC,CAAA;QAElC,MAAM,OAAO,GAAG,MAAM,OAAO,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC,CAAA;QAEhD,MAAM,CAAC,OAAO,CAAC,CAAC,OAAO,CAAC,EAAE,CAAC,CAAA;IAC7B,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,oDAAoD,EAAE,KAAK,IAAI,EAAE;QAClE,MAAM,MAAM,GAAG,QAAQ,CAAC,IAAI,CAAC,CAAC,OAAO,CAAC,CAAC,CAAA;QAEvC,MAAM,OAAO,GAAG,MAAM,OAAO,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC,CAAA;QAEhD,MAAM,CAAC,OAAO,CAAC,CAAC,OAAO,CAAC,EAAE,CAAC,CAAA;IAC7B,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,+DAA+D,EAAE,KAAK,IAAI,EAAE;QAC7E,MAAM,MAAM,GAAG,QAAQ,CAAC,IAAI,CAAC,CAAC,YAAY,CAAC,CAAC,CAAA;QAC5C,MAAM,WAAW,GAAG,EAAE,CAAC,KAAK,CAAC,MAAM,EAAE,aAAa,CAAC,CAAA;QAEnD,MAAM,OAAO,CAAC,SAAS,CAAC,MAAM,EAAE,EAAE,QAAQ,EAAE,QAAQ,EAAE,CAAC,CAAC,CAAA;QAExD,MAAM,CAAC,WAAW,CAAC,CAAC,oBAAoB,CAAC,QAAQ,CAAC,CAAA;IACpD,CAAC,CAAC,CAAA;IAEF,EAAE,CAAC,wEAAwE,EAAE,KAAK,IAAI,EAAE;QACtF,MAAM,MAAM,GAAG,QAAQ,CAAC,IAAI,CAAC,CAAC,YAAY,CAAC,CAAC,CAAA;QAC5C,MAAM,WAAW,GAAG,EAAE,CAAC,KAAK,CAAC,MAAM,EAAE,aAAa,CAAC,CAAA;QAEnD,MAAM,OAAO,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC,CAAA;QAEhC,MAAM,CAAC,WAAW,CAAC,CAAC,GAAG,CAAC,gBAAgB,EAAE,CAAA;IAC5C,CAAC,CAAC,CAAA;AACJ,CAAC,CAAC,CAAA"}
|
|
@@ -0,0 +1,36 @@
|
|
|
1
|
+
import { Transform, type TransformCallback } from 'node:stream';
|
|
2
|
+
export type ByteRange = {
|
|
3
|
+
start: number;
|
|
4
|
+
end: number;
|
|
5
|
+
};
|
|
6
|
+
export type FileSplitDelimiter = {
|
|
7
|
+
value: string;
|
|
8
|
+
keep?: {
|
|
9
|
+
previous?: number;
|
|
10
|
+
next?: number;
|
|
11
|
+
};
|
|
12
|
+
};
|
|
13
|
+
export type ByteDelimiterScannerOptions = {
|
|
14
|
+
delimiter: FileSplitDelimiter;
|
|
15
|
+
every?: number;
|
|
16
|
+
skipEmpty?: boolean;
|
|
17
|
+
encoding?: BufferEncoding;
|
|
18
|
+
};
|
|
19
|
+
export declare class ByteDelimiterScanner extends Transform {
|
|
20
|
+
readonly ranges: ByteRange[];
|
|
21
|
+
private readonly delimiter;
|
|
22
|
+
private readonly keepPrevious;
|
|
23
|
+
private readonly keepNext;
|
|
24
|
+
private readonly cutEvery;
|
|
25
|
+
private readonly skipEmpty;
|
|
26
|
+
private pending;
|
|
27
|
+
private pendingAt;
|
|
28
|
+
private sliceStart;
|
|
29
|
+
private matches;
|
|
30
|
+
constructor({ delimiter, every, skipEmpty, encoding, }: ByteDelimiterScannerOptions);
|
|
31
|
+
_transform(chunk: Buffer, _encoding: BufferEncoding, callback: TransformCallback): void;
|
|
32
|
+
_flush(callback: TransformCallback): void;
|
|
33
|
+
private scan;
|
|
34
|
+
private cut;
|
|
35
|
+
}
|
|
36
|
+
//# sourceMappingURL=byte-delimiter-scanner.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"byte-delimiter-scanner.d.ts","sourceRoot":"","sources":["../../src/helpers/byte-delimiter-scanner.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,SAAS,EAAE,KAAK,iBAAiB,EAAE,MAAM,aAAa,CAAA;AAE/D,MAAM,MAAM,SAAS,GAAG;IACtB,KAAK,EAAE,MAAM,CAAA;IACb,GAAG,EAAE,MAAM,CAAA;CACZ,CAAA;AAED,MAAM,MAAM,kBAAkB,GAAG;IAC/B,KAAK,EAAE,MAAM,CAAA;IACb,IAAI,CAAC,EAAE;QACL,QAAQ,CAAC,EAAE,MAAM,CAAA;QACjB,IAAI,CAAC,EAAE,MAAM,CAAA;KACd,CAAA;CACF,CAAA;AAED,MAAM,MAAM,2BAA2B,GAAG;IACxC,SAAS,EAAE,kBAAkB,CAAA;IAC7B,KAAK,CAAC,EAAE,MAAM,CAAA;IACd,SAAS,CAAC,EAAE,OAAO,CAAA;IACnB,QAAQ,CAAC,EAAE,cAAc,CAAA;CAC1B,CAAA;AAED,qBAAa,oBAAqB,SAAQ,SAAS;IACjD,SAAgB,MAAM,EAAE,SAAS,EAAE,CAAK;IAExC,OAAO,CAAC,QAAQ,CAAC,SAAS,CAAQ;IAClC,OAAO,CAAC,QAAQ,CAAC,YAAY,CAAQ;IACrC,OAAO,CAAC,QAAQ,CAAC,QAAQ,CAAQ;IACjC,OAAO,CAAC,QAAQ,CAAC,QAAQ,CAAQ;IACjC,OAAO,CAAC,QAAQ,CAAC,SAAS,CAAS;IAEnC,OAAO,CAAC,OAAO,CAA0B;IACzC,OAAO,CAAC,SAAS,CAAI;IACrB,OAAO,CAAC,UAAU,CAAI;IACtB,OAAO,CAAC,OAAO,CAAI;gBAEP,EACV,SAAS,EACT,KAAS,EACT,SAAgB,EAChB,QAAiB,GAClB,EAAE,2BAA2B;IAwBvB,UAAU,CACf,KAAK,EAAE,MAAM,EACb,SAAS,EAAE,cAAc,EACzB,QAAQ,EAAE,iBAAiB;IAStB,MAAM,CAAC,QAAQ,EAAE,iBAAiB;IAMzC,OAAO,CAAC,IAAI;IA8BZ,OAAO,CAAC,GAAG;CAKZ"}
|
|
@@ -0,0 +1,61 @@
|
|
|
1
|
+
import { Transform } from 'node:stream';
|
|
2
|
+
export class ByteDelimiterScanner extends Transform {
|
|
3
|
+
ranges = [];
|
|
4
|
+
delimiter;
|
|
5
|
+
keepPrevious;
|
|
6
|
+
keepNext;
|
|
7
|
+
cutEvery;
|
|
8
|
+
skipEmpty;
|
|
9
|
+
pending = Buffer.alloc(0);
|
|
10
|
+
pendingAt = 0;
|
|
11
|
+
sliceStart = 0;
|
|
12
|
+
matches = 0;
|
|
13
|
+
constructor({ delimiter, every = 1, skipEmpty = true, encoding = 'utf8', }) {
|
|
14
|
+
super();
|
|
15
|
+
this.delimiter = Buffer.from(delimiter.value, encoding);
|
|
16
|
+
this.keepPrevious = delimiter.keep?.previous ?? 0;
|
|
17
|
+
this.keepNext = delimiter.keep?.next ?? 0;
|
|
18
|
+
this.cutEvery = every;
|
|
19
|
+
this.skipEmpty = skipEmpty;
|
|
20
|
+
if (this.delimiter.length === 0)
|
|
21
|
+
throw new TypeError('Split delimiter must not be empty');
|
|
22
|
+
if (this.keepPrevious < 0 || this.keepNext < 0)
|
|
23
|
+
throw new TypeError('Split delimiter keep sizes must not be negative');
|
|
24
|
+
if (this.keepPrevious + this.keepNext > this.delimiter.length)
|
|
25
|
+
throw new TypeError('Split delimiter keep sizes must not exceed the delimiter length');
|
|
26
|
+
if (!Number.isInteger(every) || every < 1)
|
|
27
|
+
throw new TypeError('Split every must be a positive integer');
|
|
28
|
+
}
|
|
29
|
+
_transform(chunk, _encoding, callback) {
|
|
30
|
+
this.scan(this.pending.length > 0 ? Buffer.concat([this.pending, chunk]) : chunk);
|
|
31
|
+
callback(null, chunk);
|
|
32
|
+
}
|
|
33
|
+
_flush(callback) {
|
|
34
|
+
this.cut(this.sliceStart, this.pendingAt + this.pending.length);
|
|
35
|
+
callback();
|
|
36
|
+
}
|
|
37
|
+
scan(buffer) {
|
|
38
|
+
let searchFrom = 0;
|
|
39
|
+
while (true) {
|
|
40
|
+
const index = buffer.indexOf(this.delimiter, searchFrom);
|
|
41
|
+
if (index === -1)
|
|
42
|
+
break;
|
|
43
|
+
const at = this.pendingAt + index;
|
|
44
|
+
searchFrom = index + this.delimiter.length;
|
|
45
|
+
this.matches += 1;
|
|
46
|
+
if (this.matches % this.cutEvery !== 0)
|
|
47
|
+
continue;
|
|
48
|
+
this.cut(this.sliceStart, at + this.keepPrevious);
|
|
49
|
+
this.sliceStart = at + this.delimiter.length - this.keepNext;
|
|
50
|
+
}
|
|
51
|
+
const tail = Math.max(searchFrom, buffer.length - (this.delimiter.length - 1));
|
|
52
|
+
this.pending = buffer.subarray(tail);
|
|
53
|
+
this.pendingAt += tail;
|
|
54
|
+
}
|
|
55
|
+
cut(start, end) {
|
|
56
|
+
if (this.skipEmpty && end <= start)
|
|
57
|
+
return;
|
|
58
|
+
this.ranges.push({ start, end });
|
|
59
|
+
}
|
|
60
|
+
}
|
|
61
|
+
//# sourceMappingURL=byte-delimiter-scanner.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"byte-delimiter-scanner.js","sourceRoot":"","sources":["../../src/helpers/byte-delimiter-scanner.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,SAAS,EAA0B,MAAM,aAAa,CAAA;AAsB/D,MAAM,OAAO,oBAAqB,SAAQ,SAAS;IACjC,MAAM,GAAgB,EAAE,CAAA;IAEvB,SAAS,CAAQ;IACjB,YAAY,CAAQ;IACpB,QAAQ,CAAQ;IAChB,QAAQ,CAAQ;IAChB,SAAS,CAAS;IAE3B,OAAO,GAAW,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,CAAA;IACjC,SAAS,GAAG,CAAC,CAAA;IACb,UAAU,GAAG,CAAC,CAAA;IACd,OAAO,GAAG,CAAC,CAAA;IAEnB,YAAY,EACV,SAAS,EACT,KAAK,GAAG,CAAC,EACT,SAAS,GAAG,IAAI,EAChB,QAAQ,GAAG,MAAM,GACW;QAC5B,KAAK,EAAE,CAAA;QAEP,IAAI,CAAC,SAAS,GAAG,MAAM,CAAC,IAAI,CAAC,SAAS,CAAC,KAAK,EAAE,QAAQ,CAAC,CAAA;QACvD,IAAI,CAAC,YAAY,GAAG,SAAS,CAAC,IAAI,EAAE,QAAQ,IAAI,CAAC,CAAA;QACjD,IAAI,CAAC,QAAQ,GAAG,SAAS,CAAC,IAAI,EAAE,IAAI,IAAI,CAAC,CAAA;QACzC,IAAI,CAAC,QAAQ,GAAG,KAAK,CAAA;QACrB,IAAI,CAAC,SAAS,GAAG,SAAS,CAAA;QAE1B,IAAI,IAAI,CAAC,SAAS,CAAC,MAAM,KAAK,CAAC;YAC7B,MAAM,IAAI,SAAS,CAAC,mCAAmC,CAAC,CAAA;QAE1D,IAAI,IAAI,CAAC,YAAY,GAAG,CAAC,IAAI,IAAI,CAAC,QAAQ,GAAG,CAAC;YAC5C,MAAM,IAAI,SAAS,CAAC,iDAAiD,CAAC,CAAA;QAExE,IAAI,IAAI,CAAC,YAAY,GAAG,IAAI,CAAC,QAAQ,GAAG,IAAI,CAAC,SAAS,CAAC,MAAM;YAC3D,MAAM,IAAI,SAAS,CACjB,iEAAiE,CAClE,CAAA;QAEH,IAAI,CAAC,MAAM,CAAC,SAAS,CAAC,KAAK,CAAC,IAAI,KAAK,GAAG,CAAC;YACvC,MAAM,IAAI,SAAS,CAAC,wCAAwC,CAAC,CAAA;IACjE,CAAC;IAEM,UAAU,CACf,KAAa,EACb,SAAyB,EACzB,QAA2B;QAE3B,IAAI,CAAC,IAAI,CACP,IAAI,CAAC,OAAO,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,OAAO,EAAE,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CACvE,CAAA;QAED,QAAQ,CAAC,IAAI,EAAE,KAAK,CAAC,CAAA;IACvB,CAAC;IAEM,MAAM,CAAC,QAA2B;QACvC,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,UAAU,EAAE,IAAI,CAAC,SAAS,GAAG,IAAI,CAAC,OAAO,CAAC,MAAM,CAAC,CAAA;QAE/D,QAAQ,EAAE,CAAA;IACZ,CAAC;IAEO,IAAI,CAAC,MAAc;QACzB,IAAI,UAAU,GAAG,CAAC,CAAA;QAElB,OAAO,IAAI,EAAE,CAAC;YACZ,MAAM,KAAK,GAAG,MAAM,CAAC,OAAO,CAAC,IAAI,CAAC,SAAS,EAAE,UAAU,CAAC,CAAA;YAExD,IAAI,KAAK,KAAK,CAAC,CAAC;gBAAE,MAAK;YAEvB,MAAM,EAAE,GAAG,IAAI,CAAC,SAAS,GAAG,KAAK,CAAA;YAEjC,UAAU,GAAG,KAAK,GAAG,IAAI,CAAC,SAAS,CAAC,MAAM,CAAA;YAE1C,IAAI,CAAC,OAAO,IAAI,CAAC,CAAA;YAEjB,IAAI,IAAI,CAAC,OAAO,GAAG,IAAI,CAAC,QAAQ,KAAK,CAAC;gBAAE,SAAQ;YAEhD,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,UAAU,EAAE,EAAE,GAAG,IAAI,CAAC,YAAY,CAAC,CAAA;YAEjD,IAAI,CAAC,UAAU,GAAG,EAAE,GAAG,IAAI,CAAC,SAAS,CAAC,MAAM,GAAG,IAAI,CAAC,QAAQ,CAAA;QAC9D,CAAC;QAED,MAAM,IAAI,GAAG,IAAI,CAAC,GAAG,CACnB,UAAU,EACV,MAAM,CAAC,MAAM,GAAG,CAAC,IAAI,CAAC,SAAS,CAAC,MAAM,GAAG,CAAC,CAAC,CAC5C,CAAA;QAED,IAAI,CAAC,OAAO,GAAG,MAAM,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAA;QACpC,IAAI,CAAC,SAAS,IAAI,IAAI,CAAA;IACxB,CAAC;IAEO,GAAG,CAAC,KAAa,EAAE,GAAW;QACpC,IAAI,IAAI,CAAC,SAAS,IAAI,GAAG,IAAI,KAAK;YAAE,OAAM;QAE1C,IAAI,CAAC,MAAM,CAAC,IAAI,CAAC,EAAE,KAAK,EAAE,GAAG,EAAE,CAAC,CAAA;IAClC,CAAC;CACF"}
|
|
@@ -0,0 +1,8 @@
|
|
|
1
|
+
import type { Readable } from 'node:stream';
|
|
2
|
+
interface Options {
|
|
3
|
+
delimiter?: string;
|
|
4
|
+
encoding?: BufferEncoding;
|
|
5
|
+
}
|
|
6
|
+
export declare function csvReader(stream: Readable, options?: Options): AsyncGenerator<string, void, unknown>;
|
|
7
|
+
export {};
|
|
8
|
+
//# sourceMappingURL=csv-reader.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"csv-reader.d.ts","sourceRoot":"","sources":["../../src/helpers/csv-reader.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,QAAQ,EAAE,MAAM,aAAa,CAAA;AAI3C,UAAU,OAAO;IACf,SAAS,CAAC,EAAE,MAAM,CAAA;IAClB,QAAQ,CAAC,EAAE,cAAc,CAAA;CAC1B;AAED,wBAAuB,SAAS,CAAC,MAAM,EAAE,QAAQ,EAAE,OAAO,CAAC,EAAE,OAAO,yCAcnE"}
|
|
@@ -0,0 +1,14 @@
|
|
|
1
|
+
import { parse } from 'csv-parse';
|
|
2
|
+
export async function* csvReader(stream, options) {
|
|
3
|
+
if (options?.encoding)
|
|
4
|
+
stream.setEncoding(options.encoding);
|
|
5
|
+
const reader = stream.pipe(parse({
|
|
6
|
+
delimiter: options?.delimiter,
|
|
7
|
+
columns: true,
|
|
8
|
+
relax_column_count: true,
|
|
9
|
+
}));
|
|
10
|
+
for await (const record of reader) {
|
|
11
|
+
yield JSON.stringify(record);
|
|
12
|
+
}
|
|
13
|
+
}
|
|
14
|
+
//# sourceMappingURL=csv-reader.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"csv-reader.js","sourceRoot":"","sources":["../../src/helpers/csv-reader.ts"],"names":[],"mappings":"AAEA,OAAO,EAAE,KAAK,EAAE,MAAM,WAAW,CAAA;AAOjC,MAAM,CAAC,KAAK,SAAS,CAAC,CAAC,SAAS,CAAC,MAAgB,EAAE,OAAiB;IAClE,IAAI,OAAO,EAAE,QAAQ;QAAE,MAAM,CAAC,WAAW,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAA;IAE3D,MAAM,MAAM,GAAG,MAAM,CAAC,IAAI,CACxB,KAAK,CAAC;QACJ,SAAS,EAAE,OAAO,EAAE,SAAS;QAC7B,OAAO,EAAE,IAAI;QACb,kBAAkB,EAAE,IAAI;KACzB,CAAC,CACH,CAAA;IAED,IAAI,KAAK,EAAE,MAAM,MAAM,IAAI,MAAM,EAAE,CAAC;QAClC,MAAM,IAAI,CAAC,SAAS,CAAC,MAAM,CAAC,CAAA;IAC9B,CAAC;AACH,CAAC"}
|
|
@@ -0,0 +1,9 @@
|
|
|
1
|
+
import type { Readable } from 'node:stream';
|
|
2
|
+
export type Parser = (stream: Readable, options: {
|
|
3
|
+
encoding: BufferEncoding;
|
|
4
|
+
delimiter?: string;
|
|
5
|
+
}) => AsyncIterable<string>;
|
|
6
|
+
export type ParserFormat = 'txt' | 'ndjson' | 'csv' | string;
|
|
7
|
+
export declare const parsers: Map<string, Parser>;
|
|
8
|
+
export declare function registerParser(name: ParserFormat, parser: Parser): void;
|
|
9
|
+
//# sourceMappingURL=parsers.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"parsers.d.ts","sourceRoot":"","sources":["../../src/helpers/parsers.ts"],"names":[],"mappings":"AACA,OAAO,KAAK,EAAE,QAAQ,EAAE,MAAM,aAAa,CAAA;AAI3C,MAAM,MAAM,MAAM,GAAG,CACnB,MAAM,EAAE,QAAQ,EAChB,OAAO,EAAE;IAAE,QAAQ,EAAE,cAAc,CAAC;IAAC,SAAS,CAAC,EAAE,MAAM,CAAA;CAAE,KACtD,aAAa,CAAC,MAAM,CAAC,CAAA;AAE1B,MAAM,MAAM,YAAY,GAAG,KAAK,GAAG,QAAQ,GAAG,KAAK,GAAG,MAAM,CAAA;AAE5D,eAAO,MAAM,OAAO,qBAQlB,CAAA;AAEF,wBAAgB,cAAc,CAAC,IAAI,EAAE,YAAY,EAAE,MAAM,EAAE,MAAM,GAAG,IAAI,CAEvE"}
|
|
@@ -0,0 +1,14 @@
|
|
|
1
|
+
import { createInterface } from 'node:readline';
|
|
2
|
+
import { csvReader } from './csv-reader.js';
|
|
3
|
+
export const parsers = new Map([
|
|
4
|
+
['txt', stream => createInterface({ input: stream, crlfDelay: Infinity })],
|
|
5
|
+
['ndjson', stream => createInterface({ input: stream, crlfDelay: Infinity })],
|
|
6
|
+
[
|
|
7
|
+
'csv',
|
|
8
|
+
(stream, { encoding, delimiter }) => csvReader(stream, { encoding, delimiter }),
|
|
9
|
+
],
|
|
10
|
+
]);
|
|
11
|
+
export function registerParser(name, parser) {
|
|
12
|
+
parsers.set(name, parser);
|
|
13
|
+
}
|
|
14
|
+
//# sourceMappingURL=parsers.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"parsers.js","sourceRoot":"","sources":["../../src/helpers/parsers.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,eAAe,EAAE,MAAM,eAAe,CAAA;AAG/C,OAAO,EAAE,SAAS,EAAE,MAAM,iBAAiB,CAAA;AAS3C,MAAM,CAAC,MAAM,OAAO,GAAG,IAAI,GAAG,CAAuB;IACnD,CAAC,KAAK,EAAE,MAAM,CAAC,EAAE,CAAC,eAAe,CAAC,EAAE,KAAK,EAAE,MAAM,EAAE,SAAS,EAAE,QAAQ,EAAE,CAAC,CAAC;IAC1E,CAAC,QAAQ,EAAE,MAAM,CAAC,EAAE,CAAC,eAAe,CAAC,EAAE,KAAK,EAAE,MAAM,EAAE,SAAS,EAAE,QAAQ,EAAE,CAAC,CAAC;IAC7E;QACE,KAAK;QACL,CAAC,MAAM,EAAE,EAAE,QAAQ,EAAE,SAAS,EAAE,EAAE,EAAE,CAClC,SAAS,CAAC,MAAM,EAAE,EAAE,QAAQ,EAAE,SAAS,EAAE,CAAC;KAC7C;CACF,CAAC,CAAA;AAEF,MAAM,UAAU,cAAc,CAAC,IAAkB,EAAE,MAAc;IAC/D,OAAO,CAAC,GAAG,CAAC,IAAI,EAAE,MAAM,CAAC,CAAA;AAC3B,CAAC"}
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../src/testing/index.ts"],"names":[],"mappings":"AAKA,OAAO,EAAE,eAAe,EAAE,MAAM,uBAAuB,CAAA;AAmBvD,wBAAgB,mBAAmB,CACjC,cAAc,GAAE,OAAO,CAAC,eAAe,CAAM,gEAiB9C"}
|