docground 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 +217 -0
- package/dist/src/compiler/zod-compiler.d.ts +31 -0
- package/dist/src/compiler/zod-compiler.d.ts.map +1 -0
- package/dist/src/compiler/zod-compiler.js +150 -0
- package/dist/src/compiler/zod-compiler.js.map +1 -0
- package/dist/src/core/extractor.d.ts +17 -0
- package/dist/src/core/extractor.d.ts.map +1 -0
- package/dist/src/core/extractor.js +268 -0
- package/dist/src/core/extractor.js.map +1 -0
- package/dist/src/core/types.d.ts +102 -0
- package/dist/src/core/types.d.ts.map +1 -0
- package/dist/src/core/types.js +2 -0
- package/dist/src/core/types.js.map +1 -0
- package/dist/src/grounding/bbox.d.ts +40 -0
- package/dist/src/grounding/bbox.d.ts.map +1 -0
- package/dist/src/grounding/bbox.js +98 -0
- package/dist/src/grounding/bbox.js.map +1 -0
- package/dist/src/index.d.ts +16 -0
- package/dist/src/index.d.ts.map +1 -0
- package/dist/src/index.js +16 -0
- package/dist/src/index.js.map +1 -0
- package/dist/src/normalizers/currency.d.ts +14 -0
- package/dist/src/normalizers/currency.d.ts.map +1 -0
- package/dist/src/normalizers/currency.js +96 -0
- package/dist/src/normalizers/currency.js.map +1 -0
- package/dist/src/normalizers/tabular.d.ts +16 -0
- package/dist/src/normalizers/tabular.d.ts.map +1 -0
- package/dist/src/normalizers/tabular.js +81 -0
- package/dist/src/normalizers/tabular.js.map +1 -0
- package/dist/src/providers/anthropic-provider.d.ts +13 -0
- package/dist/src/providers/anthropic-provider.d.ts.map +1 -0
- package/dist/src/providers/anthropic-provider.js +62 -0
- package/dist/src/providers/anthropic-provider.js.map +1 -0
- package/dist/src/providers/factory.d.ts +11 -0
- package/dist/src/providers/factory.d.ts.map +1 -0
- package/dist/src/providers/factory.js +32 -0
- package/dist/src/providers/factory.js.map +1 -0
- package/dist/src/providers/gemini-provider.d.ts +13 -0
- package/dist/src/providers/gemini-provider.d.ts.map +1 -0
- package/dist/src/providers/gemini-provider.js +48 -0
- package/dist/src/providers/gemini-provider.js.map +1 -0
- package/dist/src/providers/openai-provider.d.ts +14 -0
- package/dist/src/providers/openai-provider.d.ts.map +1 -0
- package/dist/src/providers/openai-provider.js +48 -0
- package/dist/src/providers/openai-provider.js.map +1 -0
- package/dist/src/providers/types.d.ts +45 -0
- package/dist/src/providers/types.d.ts.map +1 -0
- package/dist/src/providers/types.js +2 -0
- package/dist/src/providers/types.js.map +1 -0
- package/dist/src/stitcher/table-stitcher.d.ts +19 -0
- package/dist/src/stitcher/table-stitcher.d.ts.map +1 -0
- package/dist/src/stitcher/table-stitcher.js +129 -0
- package/dist/src/stitcher/table-stitcher.js.map +1 -0
- package/dist/src/tiling/cutter.d.ts +14 -0
- package/dist/src/tiling/cutter.d.ts.map +1 -0
- package/dist/src/tiling/cutter.js +46 -0
- package/dist/src/tiling/cutter.js.map +1 -0
- package/dist/src/tiling/tiler.d.ts +41 -0
- package/dist/src/tiling/tiler.d.ts.map +1 -0
- package/dist/src/tiling/tiler.js +82 -0
- package/dist/src/tiling/tiler.js.map +1 -0
- package/dist/src/validators/invariants.d.ts +30 -0
- package/dist/src/validators/invariants.d.ts.map +1 -0
- package/dist/src/validators/invariants.js +164 -0
- package/dist/src/validators/invariants.js.map +1 -0
- package/dist/tests/compiler.test.d.ts +2 -0
- package/dist/tests/compiler.test.d.ts.map +1 -0
- package/dist/tests/compiler.test.js +110 -0
- package/dist/tests/compiler.test.js.map +1 -0
- package/dist/tests/extractor.test.d.ts +2 -0
- package/dist/tests/extractor.test.d.ts.map +1 -0
- package/dist/tests/extractor.test.js +195 -0
- package/dist/tests/extractor.test.js.map +1 -0
- package/dist/tests/fixtures.d.ts +9 -0
- package/dist/tests/fixtures.d.ts.map +1 -0
- package/dist/tests/fixtures.js +20 -0
- package/dist/tests/fixtures.js.map +1 -0
- package/dist/tests/grounding.test.d.ts +2 -0
- package/dist/tests/grounding.test.d.ts.map +1 -0
- package/dist/tests/grounding.test.js +82 -0
- package/dist/tests/grounding.test.js.map +1 -0
- package/dist/tests/multi-domain.test.d.ts +2 -0
- package/dist/tests/multi-domain.test.d.ts.map +1 -0
- package/dist/tests/multi-domain.test.js +135 -0
- package/dist/tests/multi-domain.test.js.map +1 -0
- package/dist/tests/normalizer.test.d.ts +2 -0
- package/dist/tests/normalizer.test.d.ts.map +1 -0
- package/dist/tests/normalizer.test.js +70 -0
- package/dist/tests/normalizer.test.js.map +1 -0
- package/dist/tests/providers.test.d.ts +2 -0
- package/dist/tests/providers.test.d.ts.map +1 -0
- package/dist/tests/providers.test.js +110 -0
- package/dist/tests/providers.test.js.map +1 -0
- package/dist/tests/stitcher.test.d.ts +2 -0
- package/dist/tests/stitcher.test.d.ts.map +1 -0
- package/dist/tests/stitcher.test.js +85 -0
- package/dist/tests/stitcher.test.js.map +1 -0
- package/dist/tests/tiling.test.d.ts +2 -0
- package/dist/tests/tiling.test.d.ts.map +1 -0
- package/dist/tests/tiling.test.js +85 -0
- package/dist/tests/tiling.test.js.map +1 -0
- package/dist/tests/validators.test.d.ts +2 -0
- package/dist/tests/validators.test.d.ts.map +1 -0
- package/dist/tests/validators.test.js +125 -0
- package/dist/tests/validators.test.js.map +1 -0
- package/dist/vitest.config.d.ts +3 -0
- package/dist/vitest.config.d.ts.map +1 -0
- package/dist/vitest.config.js +13 -0
- package/dist/vitest.config.js.map +1 -0
- package/package.json +44 -0
|
@@ -0,0 +1,268 @@
|
|
|
1
|
+
import { z } from "zod";
|
|
2
|
+
import { promises as fs } from "fs";
|
|
3
|
+
import { resolveProvider } from "../providers/factory.js";
|
|
4
|
+
import { compileZodToJsonSchema } from "../compiler/zod-compiler.js";
|
|
5
|
+
import { planTiling, needsTiling } from "../tiling/tiler.js";
|
|
6
|
+
import { cropTile, getImageMetadata } from "../tiling/cutter.js";
|
|
7
|
+
import { TableStitcher } from "../stitcher/table-stitcher.js";
|
|
8
|
+
import { InvariantEngine } from "../validators/invariants.js";
|
|
9
|
+
function toBase64String(buffer) {
|
|
10
|
+
return buffer.toString("base64");
|
|
11
|
+
}
|
|
12
|
+
function detectMimeType(buffer) {
|
|
13
|
+
if (buffer[0] === 0x25 && buffer[1] === 0x50 && buffer[2] === 0x44 && buffer[3] === 0x46) {
|
|
14
|
+
return "application/pdf";
|
|
15
|
+
}
|
|
16
|
+
if (buffer[0] === 0x89 && buffer[1] === 0x50 && buffer[2] === 0x4e && buffer[3] === 0x47) {
|
|
17
|
+
return "image/png";
|
|
18
|
+
}
|
|
19
|
+
if (buffer[0] === 0xff && buffer[1] === 0xd8) {
|
|
20
|
+
return "image/jpeg";
|
|
21
|
+
}
|
|
22
|
+
if (buffer[0] === 0x47 && buffer[1] === 0x49 && buffer[2] === 0x46) {
|
|
23
|
+
return "image/gif";
|
|
24
|
+
}
|
|
25
|
+
if (buffer[0] === 0x52 && buffer[1] === 0x49 && buffer[2] === 0x46 && buffer[3] === 0x46) {
|
|
26
|
+
return "image/webp";
|
|
27
|
+
}
|
|
28
|
+
return "image/png";
|
|
29
|
+
}
|
|
30
|
+
const A4_PAGE_SIZE = { width: 595, height: 842 };
|
|
31
|
+
function isDocumentImage(mimeType) {
|
|
32
|
+
return mimeType === "application/pdf";
|
|
33
|
+
}
|
|
34
|
+
async function readInput(input) {
|
|
35
|
+
if (typeof input === "string") {
|
|
36
|
+
const dataUri = input.match(/^data:([^;]+);base64,(.+)$/);
|
|
37
|
+
if (dataUri) {
|
|
38
|
+
return { buffer: Buffer.from(dataUri[2], "base64"), mimeType: dataUri[1] };
|
|
39
|
+
}
|
|
40
|
+
try {
|
|
41
|
+
await fs.access(input);
|
|
42
|
+
const buffer = await fs.readFile(input);
|
|
43
|
+
return { buffer, mimeType: detectMimeType(buffer) };
|
|
44
|
+
}
|
|
45
|
+
catch {
|
|
46
|
+
return { buffer: Buffer.from(input, "base64"), mimeType: "image/png" };
|
|
47
|
+
}
|
|
48
|
+
}
|
|
49
|
+
const buffer = Buffer.isBuffer(input) ? input : Buffer.from(input);
|
|
50
|
+
return { buffer, mimeType: detectMimeType(buffer) };
|
|
51
|
+
}
|
|
52
|
+
function wrapSchema(userSchema) {
|
|
53
|
+
const BBoxSchema = z.array(z.number()).length(4);
|
|
54
|
+
const GroundingSchema = z.object({
|
|
55
|
+
pageIndex: z.number(),
|
|
56
|
+
field: z.string(),
|
|
57
|
+
value: z.unknown(),
|
|
58
|
+
bbox: BBoxSchema,
|
|
59
|
+
confidence: z.number().optional(),
|
|
60
|
+
});
|
|
61
|
+
const CellSpanSchema = z.object({
|
|
62
|
+
rowspan: z.number().optional(),
|
|
63
|
+
colspan: z.number().optional(),
|
|
64
|
+
});
|
|
65
|
+
const TableSchema = z.object({
|
|
66
|
+
pageIndex: z.number(),
|
|
67
|
+
rows: z.array(z.array(z.unknown())),
|
|
68
|
+
header: z.array(z.string()).optional(),
|
|
69
|
+
bbox: BBoxSchema,
|
|
70
|
+
html: z.string().optional(),
|
|
71
|
+
spans: z.array(z.array(CellSpanSchema.optional()).optional()).optional(),
|
|
72
|
+
});
|
|
73
|
+
return z.object({
|
|
74
|
+
data: userSchema,
|
|
75
|
+
groundings: z.array(GroundingSchema).optional(),
|
|
76
|
+
tables: z.array(TableSchema).optional(),
|
|
77
|
+
});
|
|
78
|
+
}
|
|
79
|
+
function buildPrompt(schemaName = "document", options = {}) {
|
|
80
|
+
const lines = [
|
|
81
|
+
`Extract the ${schemaName} from the provided image(s).`,
|
|
82
|
+
"Return valid JSON matching the requested schema exactly.",
|
|
83
|
+
"For each extracted field, include a visual grounding with `bbox` as normalized [ymin, xmin, ymax, xmax] coordinates.",
|
|
84
|
+
"If you detect tables, include them in the `tables` array with rows, header, and bbox.",
|
|
85
|
+
];
|
|
86
|
+
if (options.prompt) {
|
|
87
|
+
lines.push(`Additional instructions: ${options.prompt}`);
|
|
88
|
+
}
|
|
89
|
+
return lines.join("\n");
|
|
90
|
+
}
|
|
91
|
+
function buildRepairPrompt(basePrompt, report) {
|
|
92
|
+
return [
|
|
93
|
+
basePrompt,
|
|
94
|
+
"",
|
|
95
|
+
"The previous extraction failed the following invariant checks; please correct the output and return the full JSON again:",
|
|
96
|
+
report.checks
|
|
97
|
+
.filter((c) => !c.passed)
|
|
98
|
+
.map((c) => `- ${c.id}: ${c.message ?? `diff ${c.diff}`}`)
|
|
99
|
+
.join("\n"),
|
|
100
|
+
].join("\n");
|
|
101
|
+
}
|
|
102
|
+
function collectFields(data, groundings) {
|
|
103
|
+
const fields = {};
|
|
104
|
+
function visit(value, path) {
|
|
105
|
+
if (Array.isArray(value)) {
|
|
106
|
+
value.forEach((v, i) => visit(v, path ? `${path}.${i}` : `${i}`));
|
|
107
|
+
return;
|
|
108
|
+
}
|
|
109
|
+
if (value !== null && typeof value === "object") {
|
|
110
|
+
for (const [k, v] of Object.entries(value)) {
|
|
111
|
+
visit(v, path ? `${path}.${k}` : k);
|
|
112
|
+
}
|
|
113
|
+
return;
|
|
114
|
+
}
|
|
115
|
+
const g = groundings.find((gr) => gr.field === path);
|
|
116
|
+
fields[path] = {
|
|
117
|
+
value,
|
|
118
|
+
confidence: g?.confidence ?? 1,
|
|
119
|
+
pageIndex: g?.pageIndex,
|
|
120
|
+
bbox: g?.bbox,
|
|
121
|
+
};
|
|
122
|
+
}
|
|
123
|
+
visit(data, "");
|
|
124
|
+
return fields;
|
|
125
|
+
}
|
|
126
|
+
function computeGlobalConfidence(invariants, groundings) {
|
|
127
|
+
const invConfidence = invariants?.confidenceScore ?? 1;
|
|
128
|
+
if (groundings.length === 0)
|
|
129
|
+
return invConfidence;
|
|
130
|
+
const groundingAvg = groundings.reduce((sum, g) => sum + (g.confidence ?? 1), 0) / groundings.length;
|
|
131
|
+
return Math.max(0, Math.min(1, invConfidence * groundingAvg));
|
|
132
|
+
}
|
|
133
|
+
export class DocumentExtractor {
|
|
134
|
+
baseConfig;
|
|
135
|
+
constructor(config = {}) {
|
|
136
|
+
this.baseConfig = config;
|
|
137
|
+
}
|
|
138
|
+
resolveProvider(options = {}) {
|
|
139
|
+
if (options.customVLM)
|
|
140
|
+
return options.customVLM;
|
|
141
|
+
if (this.baseConfig.customVLM)
|
|
142
|
+
return this.baseConfig.customVLM;
|
|
143
|
+
return resolveProvider({
|
|
144
|
+
provider: options.provider ?? this.baseConfig.provider,
|
|
145
|
+
apiKey: options.apiKey ?? this.baseConfig.apiKey,
|
|
146
|
+
model: options.model ?? this.baseConfig.model,
|
|
147
|
+
baseUrl: options.baseUrl ?? this.baseConfig.baseUrl,
|
|
148
|
+
});
|
|
149
|
+
}
|
|
150
|
+
async ingestPage(input) {
|
|
151
|
+
const { buffer, mimeType } = await readInput(input);
|
|
152
|
+
const base64 = toBase64String(buffer);
|
|
153
|
+
const image = { mimeType, base64, width: 0, height: 0 };
|
|
154
|
+
if (isDocumentImage(mimeType)) {
|
|
155
|
+
image.width = A4_PAGE_SIZE.width;
|
|
156
|
+
image.height = A4_PAGE_SIZE.height;
|
|
157
|
+
}
|
|
158
|
+
else {
|
|
159
|
+
const meta = await getImageMetadata(image);
|
|
160
|
+
image.width = meta.width;
|
|
161
|
+
image.height = meta.height;
|
|
162
|
+
}
|
|
163
|
+
return {
|
|
164
|
+
index: 0,
|
|
165
|
+
image,
|
|
166
|
+
size: { width: image.width, height: image.height },
|
|
167
|
+
};
|
|
168
|
+
}
|
|
169
|
+
getTileOptions(options) {
|
|
170
|
+
if (options.tiling === "off")
|
|
171
|
+
return undefined;
|
|
172
|
+
if (options.tiling === "auto" || options.tiling === undefined) {
|
|
173
|
+
return { maxTileWidth: 1536, maxTileHeight: 1536, overlap: 0.15 };
|
|
174
|
+
}
|
|
175
|
+
return options.tiling;
|
|
176
|
+
}
|
|
177
|
+
async prepareTiles(page, options) {
|
|
178
|
+
if (isDocumentImage(page.image.mimeType)) {
|
|
179
|
+
return [page.image];
|
|
180
|
+
}
|
|
181
|
+
const tileOpts = this.getTileOptions(options);
|
|
182
|
+
if (!tileOpts || !needsTiling(page.size, tileOpts)) {
|
|
183
|
+
return [page.image];
|
|
184
|
+
}
|
|
185
|
+
const plan = planTiling(page.index, page.size, tileOpts);
|
|
186
|
+
const tileImages = [];
|
|
187
|
+
for (const tile of plan.tiles) {
|
|
188
|
+
const cropped = await cropTile(page.image, tile);
|
|
189
|
+
tileImages.push(cropped);
|
|
190
|
+
options.onProgress?.({ step: "tiling", page: page.index, tile: tile.row * tile.cols + tile.col });
|
|
191
|
+
}
|
|
192
|
+
return tileImages;
|
|
193
|
+
}
|
|
194
|
+
runInvariants(data, tables, options) {
|
|
195
|
+
const engine = new InvariantEngine();
|
|
196
|
+
for (const check of options.invariants ?? []) {
|
|
197
|
+
engine.addCheck(check);
|
|
198
|
+
}
|
|
199
|
+
const context = data !== null && typeof data === "object" ? { ...data, tables } : { value: data, tables };
|
|
200
|
+
return engine.check(context);
|
|
201
|
+
}
|
|
202
|
+
async callProvider(provider, tileImages, prompt, schema, options) {
|
|
203
|
+
options.onProgress?.({ step: "vlm", page: 0 });
|
|
204
|
+
return provider.extract({
|
|
205
|
+
images: tileImages,
|
|
206
|
+
prompt,
|
|
207
|
+
schema,
|
|
208
|
+
responseMimeType: "application/json",
|
|
209
|
+
maxTokens: options.maxTokens,
|
|
210
|
+
temperature: options.temperature ?? 0,
|
|
211
|
+
seed: options.seed,
|
|
212
|
+
});
|
|
213
|
+
}
|
|
214
|
+
async extract(input, schema, options = {}) {
|
|
215
|
+
const start = Date.now();
|
|
216
|
+
const provider = this.resolveProvider(options);
|
|
217
|
+
const page = await this.ingestPage(input);
|
|
218
|
+
const tileImages = await this.prepareTiles(page, options);
|
|
219
|
+
const wrappedSchema = wrapSchema(schema);
|
|
220
|
+
const providerSchema = compileZodToJsonSchema(wrappedSchema);
|
|
221
|
+
const prompt = buildPrompt(schema.constructor.name, options);
|
|
222
|
+
let raw = await this.callProvider(provider, tileImages, prompt, providerSchema, options);
|
|
223
|
+
let parsed;
|
|
224
|
+
try {
|
|
225
|
+
parsed = JSON.parse(raw.text);
|
|
226
|
+
}
|
|
227
|
+
catch {
|
|
228
|
+
throw new Error(`Provider returned invalid JSON: ${raw.text.slice(0, 200)}`);
|
|
229
|
+
}
|
|
230
|
+
let data = wrappedSchema.shape.data.parse(parsed.data);
|
|
231
|
+
let groundings = parsed.groundings ?? [];
|
|
232
|
+
let tables = parsed.tables ?? [];
|
|
233
|
+
let invariantReport;
|
|
234
|
+
let repairs = 0;
|
|
235
|
+
if (options.prove) {
|
|
236
|
+
invariantReport = this.runInvariants(data, tables, options);
|
|
237
|
+
if (!invariantReport.ok && options.autoRepair) {
|
|
238
|
+
const repairPrompt = buildRepairPrompt(prompt, invariantReport);
|
|
239
|
+
const repairRaw = await this.callProvider(provider, tileImages, repairPrompt, providerSchema, options);
|
|
240
|
+
const repairParsed = JSON.parse(repairRaw.text);
|
|
241
|
+
data = wrappedSchema.shape.data.parse(repairParsed.data);
|
|
242
|
+
groundings = repairParsed.groundings ?? [];
|
|
243
|
+
tables = repairParsed.tables ?? [];
|
|
244
|
+
invariantReport = this.runInvariants(data, tables, options);
|
|
245
|
+
repairs = 1;
|
|
246
|
+
}
|
|
247
|
+
}
|
|
248
|
+
const stitchedTables = new TableStitcher().stitch(tables);
|
|
249
|
+
const confidence = computeGlobalConfidence(invariantReport, groundings);
|
|
250
|
+
const fields = collectFields(data, groundings);
|
|
251
|
+
return {
|
|
252
|
+
data,
|
|
253
|
+
confidence,
|
|
254
|
+
fields,
|
|
255
|
+
tables: stitchedTables,
|
|
256
|
+
groundings,
|
|
257
|
+
invariants: invariantReport,
|
|
258
|
+
provenance: {
|
|
259
|
+
provider: provider.name,
|
|
260
|
+
model: provider.model ?? "unknown",
|
|
261
|
+
durationMs: Date.now() - start,
|
|
262
|
+
tilesUsed: tileImages.length,
|
|
263
|
+
repairs,
|
|
264
|
+
},
|
|
265
|
+
};
|
|
266
|
+
}
|
|
267
|
+
}
|
|
268
|
+
//# sourceMappingURL=extractor.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"extractor.js","sourceRoot":"","sources":["../../../src/core/extractor.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,CAAC,EAAE,MAAM,KAAK,CAAC;AACxB,OAAO,EAAE,QAAQ,IAAI,EAAE,EAAE,MAAM,IAAI,CAAC;AACpC,OAAO,EAAE,eAAe,EAA8B,MAAM,yBAAyB,CAAC;AAWtF,OAAO,EAAE,sBAAsB,EAAE,MAAM,6BAA6B,CAAC;AACrE,OAAO,EAAE,UAAU,EAAE,WAAW,EAAE,MAAM,oBAAoB,CAAC;AAC7D,OAAO,EAAE,QAAQ,EAAE,gBAAgB,EAAE,MAAM,qBAAqB,CAAC;AACjE,OAAO,EAAE,aAAa,EAAE,MAAM,+BAA+B,CAAC;AAC9D,OAAO,EAAE,eAAe,EAAwB,MAAM,6BAA6B,CAAC;AAIpF,SAAS,cAAc,CAAC,MAAc;IACpC,OAAO,MAAM,CAAC,QAAQ,CAAC,QAAQ,CAAC,CAAC;AACnC,CAAC;AAED,SAAS,cAAc,CAAC,MAAc;IACpC,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACzF,OAAO,iBAAiB,CAAC;IAC3B,CAAC;IACD,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACzF,OAAO,WAAW,CAAC;IACrB,CAAC;IACD,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QAC7C,OAAO,YAAY,CAAC;IACtB,CAAC;IACD,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACnE,OAAO,WAAW,CAAC;IACrB,CAAC;IACD,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACzF,OAAO,YAAY,CAAC;IACtB,CAAC;IACD,OAAO,WAAW,CAAC;AACrB,CAAC;AAED,MAAM,YAAY,GAAG,EAAE,KAAK,EAAE,GAAG,EAAE,MAAM,EAAE,GAAG,EAAE,CAAC;AAEjD,SAAS,eAAe,CAAC,QAAgB;IACvC,OAAO,QAAQ,KAAK,iBAAiB,CAAC;AACxC,CAAC;AAED,KAAK,UAAU,SAAS,CAAC,KAAmC;IAC1D,IAAI,OAAO,KAAK,KAAK,QAAQ,EAAE,CAAC;QAC9B,MAAM,OAAO,GAAG,KAAK,CAAC,KAAK,CAAC,4BAA4B,CAAC,CAAC;QAC1D,IAAI,OAAO,EAAE,CAAC;YACZ,OAAO,EAAE,MAAM,EAAE,MAAM,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,QAAQ,CAAC,EAAE,QAAQ,EAAE,OAAO,CAAC,CAAC,CAAC,EAAE,CAAC;QAC7E,CAAC;QACD,IAAI,CAAC;YACH,MAAM,EAAE,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC;YACvB,MAAM,MAAM,GAAG,MAAM,EAAE,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC;YACxC,OAAO,EAAE,MAAM,EAAE,QAAQ,EAAE,cAAc,CAAC,MAAM,CAAC,EAAE,CAAC;QACtD,CAAC;QAAC,MAAM,CAAC;YACP,OAAO,EAAE,MAAM,EAAE,MAAM,CAAC,IAAI,CAAC,KAAK,EAAE,QAAQ,CAAC,EAAE,QAAQ,EAAE,WAAW,EAAE,CAAC;QACzE,CAAC;IACH,CAAC;IACD,MAAM,MAAM,GAAG,MAAM,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;IACnE,OAAO,EAAE,MAAM,EAAE,QAAQ,EAAE,cAAc,CAAC,MAAM,CAAC,EAAE,CAAC;AACtD,CAAC;AAED,SAAS,UAAU,CAAI,UAAwB;IAC7C,MAAM,UAAU,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,EAAE,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC;IACjD,MAAM,eAAe,GAAG,CAAC,CAAC,MAAM,CAAC;QAC/B,SAAS,EAAE,CAAC,CAAC,MAAM,EAAE;QACrB,KAAK,EAAE,CAAC,CAAC,MAAM,EAAE;QACjB,KAAK,EAAE,CAAC,CAAC,OAAO,EAAE;QAClB,IAAI,EAAE,UAAU;QAChB,UAAU,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE;KAClC,CAAC,CAAC;IACH,MAAM,cAAc,GAAG,CAAC,CAAC,MAAM,CAAC;QAC9B,OAAO,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE;QAC9B,OAAO,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE;KAC/B,CAAC,CAAC;IACH,MAAM,WAAW,GAAG,CAAC,CAAC,MAAM,CAAC;QAC3B,SAAS,EAAE,CAAC,CAAC,MAAM,EAAE;QACrB,IAAI,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,OAAO,EAAE,CAAC,CAAC;QACnC,MAAM,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,EAAE,CAAC,CAAC,QAAQ,EAAE;QACtC,IAAI,EAAE,UAAU;QAChB,IAAI,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE;QAC3B,KAAK,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,cAAc,CAAC,QAAQ,EAAE,CAAC,CAAC,QAAQ,EAAE,CAAC,CAAC,QAAQ,EAAE;KACzE,CAAC,CAAC;IACH,OAAO,CAAC,CAAC,MAAM,CAAC;QACd,IAAI,EAAE,UAAU;QAChB,UAAU,EAAE,CAAC,CAAC,KAAK,CAAC,eAAe,CAAC,CAAC,QAAQ,EAAE;QAC/C,MAAM,EAAE,CAAC,CAAC,KAAK,CAAC,WAAW,CAAC,CAAC,QAAQ,EAAE;KACxC,CAAC,CAAC;AACL,CAAC;AAED,SAAS,WAAW,CAAC,UAAU,GAAG,UAAU,EAAE,UAA6B,EAAE;IAC3E,MAAM,KAAK,GAAG;QACZ,eAAe,UAAU,8BAA8B;QACvD,0DAA0D;QAC1D,sHAAsH;QACtH,uFAAuF;KACxF,CAAC;IACF,IAAI,OAAO,CAAC,MAAM,EAAE,CAAC;QACnB,KAAK,CAAC,IAAI,CAAC,4BAA4B,OAAO,CAAC,MAAM,EAAE,CAAC,CAAC;IAC3D,CAAC;IACD,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;AAC1B,CAAC;AAED,SAAS,iBAAiB,CAAC,UAAkB,EAAE,MAAuB;IACpE,OAAO;QACL,UAAU;QACV,EAAE;QACF,0HAA0H;QAC1H,MAAM,CAAC,MAAM;aACV,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC;aACxB,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,KAAK,CAAC,CAAC,EAAE,KAAK,CAAC,CAAC,OAAO,IAAI,QAAQ,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC;aACzD,IAAI,CAAC,IAAI,CAAC;KACd,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;AACf,CAAC;AAED,SAAS,aAAa,CAAC,IAAa,EAAE,UAA2B;IAC/D,MAAM,MAAM,GAAgC,EAAE,CAAC;IAE/C,SAAS,KAAK,CAAC,KAAc,EAAE,IAAY;QACzC,IAAI,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC,EAAE,CAAC;YACzB,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC,GAAG,IAAI,IAAI,CAAC,EAAE,CAAC,CAAC,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,CAAC;YAClE,OAAO;QACT,CAAC;QACD,IAAI,KAAK,KAAK,IAAI,IAAI,OAAO,KAAK,KAAK,QAAQ,EAAE,CAAC;YAChD,KAAK,MAAM,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,MAAM,CAAC,OAAO,CAAC,KAAK,CAAC,EAAE,CAAC;gBAC3C,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC,GAAG,IAAI,IAAI,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;YACtC,CAAC;YACD,OAAO;QACT,CAAC;QACD,MAAM,CAAC,GAAG,UAAU,CAAC,IAAI,CAAC,CAAC,EAAE,EAAE,EAAE,CAAC,EAAE,CAAC,KAAK,KAAK,IAAI,CAAC,CAAC;QACrD,MAAM,CAAC,IAAI,CAAC,GAAG;YACb,KAAK;YACL,UAAU,EAAE,CAAC,EAAE,UAAU,IAAI,CAAC;YAC9B,SAAS,EAAE,CAAC,EAAE,SAAS;YACvB,IAAI,EAAE,CAAC,EAAE,IAAI;SACd,CAAC;IACJ,CAAC;IAED,KAAK,CAAC,IAAI,EAAE,EAAE,CAAC,CAAC;IAChB,OAAO,MAAM,CAAC;AAChB,CAAC;AAED,SAAS,uBAAuB,CAAC,UAAuC,EAAE,UAA2B;IACnG,MAAM,aAAa,GAAG,UAAU,EAAE,eAAe,IAAI,CAAC,CAAC;IACvD,IAAI,UAAU,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,aAAa,CAAC;IAClD,MAAM,YAAY,GAChB,UAAU,CAAC,MAAM,CAAC,CAAC,GAAG,EAAE,CAAC,EAAE,EAAE,CAAC,GAAG,GAAG,CAAC,CAAC,CAAC,UAAU,IAAI,CAAC,CAAC,EAAE,CAAC,CAAC,GAAG,UAAU,CAAC,MAAM,CAAC;IAClF,OAAO,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,aAAa,GAAG,YAAY,CAAC,CAAC,CAAC;AAChE,CAAC;AAED,MAAM,OAAO,iBAAiB;IACpB,UAAU,CAAkB;IAEpC,YAAY,SAA0B,EAAE;QACtC,IAAI,CAAC,UAAU,GAAG,MAAM,CAAC;IAC3B,CAAC;IAEO,eAAe,CAAC,UAA6B,EAAE;QACrD,IAAI,OAAO,CAAC,SAAS;YAAE,OAAO,OAAO,CAAC,SAAS,CAAC;QAChD,IAAI,IAAI,CAAC,UAAU,CAAC,SAAS;YAAE,OAAO,IAAI,CAAC,UAAU,CAAC,SAAS,CAAC;QAChE,OAAO,eAAe,CAAC;YACrB,QAAQ,EAAE,OAAO,CAAC,QAAQ,IAAI,IAAI,CAAC,UAAU,CAAC,QAAQ;YACtD,MAAM,EAAE,OAAO,CAAC,MAAM,IAAI,IAAI,CAAC,UAAU,CAAC,MAAM;YAChD,KAAK,EAAE,OAAO,CAAC,KAAK,IAAI,IAAI,CAAC,UAAU,CAAC,KAAK;YAC7C,OAAO,EAAE,OAAO,CAAC,OAAO,IAAI,IAAI,CAAC,UAAU,CAAC,OAAO;SACpD,CAAC,CAAC;IACL,CAAC;IAEO,KAAK,CAAC,UAAU,CAAC,KAAmC;QAC1D,MAAM,EAAE,MAAM,EAAE,QAAQ,EAAE,GAAG,MAAM,SAAS,CAAC,KAAK,CAAC,CAAC;QACpD,MAAM,MAAM,GAAG,cAAc,CAAC,MAAM,CAAC,CAAC;QACtC,MAAM,KAAK,GAAc,EAAE,QAAQ,EAAE,MAAM,EAAE,KAAK,EAAE,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,CAAC;QAEnE,IAAI,eAAe,CAAC,QAAQ,CAAC,EAAE,CAAC;YAC9B,KAAK,CAAC,KAAK,GAAG,YAAY,CAAC,KAAK,CAAC;YACjC,KAAK,CAAC,MAAM,GAAG,YAAY,CAAC,MAAM,CAAC;QACrC,CAAC;aAAM,CAAC;YACN,MAAM,IAAI,GAAG,MAAM,gBAAgB,CAAC,KAAK,CAAC,CAAC;YAC3C,KAAK,CAAC,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC;YACzB,KAAK,CAAC,MAAM,GAAG,IAAI,CAAC,MAAM,CAAC;QAC7B,CAAC;QAED,OAAO;YACL,KAAK,EAAE,CAAC;YACR,KAAK;YACL,IAAI,EAAE,EAAE,KAAK,EAAE,KAAK,CAAC,KAAK,EAAE,MAAM,EAAE,KAAK,CAAC,MAAM,EAAE;SACnD,CAAC;IACJ,CAAC;IAEO,cAAc,CAAC,OAA0B;QAC/C,IAAI,OAAO,CAAC,MAAM,KAAK,KAAK;YAAE,OAAO,SAAS,CAAC;QAC/C,IAAI,OAAO,CAAC,MAAM,KAAK,MAAM,IAAI,OAAO,CAAC,MAAM,KAAK,SAAS,EAAE,CAAC;YAC9D,OAAO,EAAE,YAAY,EAAE,IAAI,EAAE,aAAa,EAAE,IAAI,EAAE,OAAO,EAAE,IAAI,EAAE,CAAC;QACpE,CAAC;QACD,OAAO,OAAO,CAAC,MAAM,CAAC;IACxB,CAAC;IAEO,KAAK,CAAC,YAAY,CAAC,IAAkB,EAAE,OAA0B;QACvE,IAAI,eAAe,CAAC,IAAI,CAAC,KAAK,CAAC,QAAQ,CAAC,EAAE,CAAC;YACzC,OAAO,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACtB,CAAC;QAED,MAAM,QAAQ,GAAG,IAAI,CAAC,cAAc,CAAC,OAAO,CAAC,CAAC;QAC9C,IAAI,CAAC,QAAQ,IAAI,CAAC,WAAW,CAAC,IAAI,CAAC,IAAI,EAAE,QAAQ,CAAC,EAAE,CAAC;YACnD,OAAO,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACtB,CAAC;QAED,MAAM,IAAI,GAAG,UAAU,CAAC,IAAI,CAAC,KAAK,EAAE,IAAI,CAAC,IAAI,EAAE,QAAQ,CAAC,CAAC;QACzD,MAAM,UAAU,GAAgB,EAAE,CAAC;QACnC,KAAK,MAAM,IAAI,IAAI,IAAI,CAAC,KAAK,EAAE,CAAC;YAC9B,MAAM,OAAO,GAAG,MAAM,QAAQ,CAAC,IAAI,CAAC,KAAK,EAAE,IAAI,CAAC,CAAC;YACjD,UAAU,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;YACzB,OAAO,CAAC,UAAU,EAAE,CAAC,EAAE,IAAI,EAAE,QAAQ,EAAE,IAAI,EAAE,IAAI,CAAC,KAAK,EAAE,IAAI,EAAE,IAAI,CAAC,GAAG,GAAG,IAAI,CAAC,IAAI,GAAG,IAAI,CAAC,GAAG,EAAE,CAAC,CAAC;QACpG,CAAC;QACD,OAAO,UAAU,CAAC;IACpB,CAAC;IAEO,aAAa,CACnB,IAAa,EACb,MAAwB,EACxB,OAA0B;QAE1B,MAAM,MAAM,GAAG,IAAI,eAAe,EAAE,CAAC;QACrC,KAAK,MAAM,KAAK,IAAI,OAAO,CAAC,UAAU,IAAI,EAAE,EAAE,CAAC;YAC7C,MAAM,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC;QACzB,CAAC;QACD,MAAM,OAAO,GACX,IAAI,KAAK,IAAI,IAAI,OAAO,IAAI,KAAK,QAAQ,CAAC,CAAC,CAAC,EAAE,GAAG,IAAI,EAAE,MAAM,EAAE,CAAC,CAAC,CAAC,EAAE,KAAK,EAAE,IAAI,EAAE,MAAM,EAAE,CAAC;QAC5F,OAAO,MAAM,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC;IAC/B,CAAC;IAEO,KAAK,CAAC,YAAY,CACxB,QAAqB,EACrB,UAAuB,EACvB,MAAc,EACd,MAA+B,EAC/B,OAA0B;QAE1B,OAAO,CAAC,UAAU,EAAE,CAAC,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,CAAC,EAAE,CAAC,CAAC;QAC/C,OAAO,QAAQ,CAAC,OAAO,CAAC;YACtB,MAAM,EAAE,UAAU;YAClB,MAAM;YACN,MAAM;YACN,gBAAgB,EAAE,kBAAkB;YACpC,SAAS,EAAE,OAAO,CAAC,SAAS;YAC5B,WAAW,EAAE,OAAO,CAAC,WAAW,IAAI,CAAC;YACrC,IAAI,EAAE,OAAO,CAAC,IAAI;SACnB,CAAC,CAAC;IACL,CAAC;IAED,KAAK,CAAC,OAAO,CACX,KAAmC,EACnC,MAAoB,EACpB,UAA6B,EAAE;QAE/B,MAAM,KAAK,GAAG,IAAI,CAAC,GAAG,EAAE,CAAC;QACzB,MAAM,QAAQ,GAAG,IAAI,CAAC,eAAe,CAAC,OAAO,CAAC,CAAC;QAE/C,MAAM,IAAI,GAAG,MAAM,IAAI,CAAC,UAAU,CAAC,KAAK,CAAC,CAAC;QAC1C,MAAM,UAAU,GAAG,MAAM,IAAI,CAAC,YAAY,CAAC,IAAI,EAAE,OAAO,CAAC,CAAC;QAE1D,MAAM,aAAa,GAAG,UAAU,CAAC,MAAM,CAAC,CAAC;QACzC,MAAM,cAAc,GAAG,sBAAsB,CAAC,aAAa,CAAC,CAAC;QAC7D,MAAM,MAAM,GAAG,WAAW,CAAC,MAAM,CAAC,WAAW,CAAC,IAAI,EAAE,OAAO,CAAC,CAAC;QAE7D,IAAI,GAAG,GAAG,MAAM,IAAI,CAAC,YAAY,CAAC,QAAQ,EAAE,UAAU,EAAE,MAAM,EAAE,cAAc,EAAE,OAAO,CAAC,CAAC;QACzF,IAAI,MAAkF,CAAC;QACvF,IAAI,CAAC;YACH,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC;QAChC,CAAC;QAAC,MAAM,CAAC;YACP,MAAM,IAAI,KAAK,CAAC,mCAAmC,GAAG,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC,EAAE,CAAC,CAAC;QAC/E,CAAC;QAED,IAAI,IAAI,GAAM,aAAa,CAAC,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,MAAM,CAAC,IAAI,CAAM,CAAC;QAC/D,IAAI,UAAU,GAAG,MAAM,CAAC,UAAU,IAAI,EAAE,CAAC;QACzC,IAAI,MAAM,GAAG,MAAM,CAAC,MAAM,IAAI,EAAE,CAAC;QACjC,IAAI,eAA4C,CAAC;QACjD,IAAI,OAAO,GAAG,CAAC,CAAC;QAEhB,IAAI,OAAO,CAAC,KAAK,EAAE,CAAC;YAClB,eAAe,GAAG,IAAI,CAAC,aAAa,CAAC,IAAI,EAAE,MAAM,EAAE,OAAO,CAAC,CAAC;YAE5D,IAAI,CAAC,eAAe,CAAC,EAAE,IAAI,OAAO,CAAC,UAAU,EAAE,CAAC;gBAC9C,MAAM,YAAY,GAAG,iBAAiB,CAAC,MAAM,EAAE,eAAe,CAAC,CAAC;gBAChE,MAAM,SAAS,GAAG,MAAM,IAAI,CAAC,YAAY,CACvC,QAAQ,EACR,UAAU,EACV,YAAY,EACZ,cAAc,EACd,OAAO,CACR,CAAC;gBACF,MAAM,YAAY,GAAkB,IAAI,CAAC,KAAK,CAAC,SAAS,CAAC,IAAI,CAAC,CAAC;gBAC/D,IAAI,GAAG,aAAa,CAAC,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,YAAY,CAAC,IAAI,CAAM,CAAC;gBAC9D,UAAU,GAAG,YAAY,CAAC,UAAU,IAAI,EAAE,CAAC;gBAC3C,MAAM,GAAG,YAAY,CAAC,MAAM,IAAI,EAAE,CAAC;gBACnC,eAAe,GAAG,IAAI,CAAC,aAAa,CAAC,IAAI,EAAE,MAAM,EAAE,OAAO,CAAC,CAAC;gBAC5D,OAAO,GAAG,CAAC,CAAC;YACd,CAAC;QACH,CAAC;QAED,MAAM,cAAc,GAAG,IAAI,aAAa,EAAE,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC;QAC1D,MAAM,UAAU,GAAG,uBAAuB,CAAC,eAAe,EAAE,UAAU,CAAC,CAAC;QACxE,MAAM,MAAM,GAAG,aAAa,CAAC,IAAI,EAAE,UAAU,CAAC,CAAC;QAE/C,OAAO;YACL,IAAI;YACJ,UAAU;YACV,MAAM;YACN,MAAM,EAAE,cAAc;YACtB,UAAU;YACV,UAAU,EAAE,eAAe;YAC3B,UAAU,EAAE;gBACV,QAAQ,EAAE,QAAQ,CAAC,IAAI;gBACvB,KAAK,EAAG,QAA+B,CAAC,KAAK,IAAI,SAAS;gBAC1D,UAAU,EAAE,IAAI,CAAC,GAAG,EAAE,GAAG,KAAK;gBAC9B,SAAS,EAAE,UAAU,CAAC,MAAM;gBAC5B,OAAO;aACR;SACF,CAAC;IACJ,CAAC;CACF"}
|
|
@@ -0,0 +1,102 @@
|
|
|
1
|
+
import type { NormalizedBBox } from "../grounding/bbox.js";
|
|
2
|
+
import type { GroundedField, PageImage, VLMProvider } from "../providers/types.js";
|
|
3
|
+
import type { InvariantReport } from "../validators/invariants.js";
|
|
4
|
+
export interface ExtractedTable {
|
|
5
|
+
pageIndex: number;
|
|
6
|
+
rows: unknown[][];
|
|
7
|
+
header?: string[];
|
|
8
|
+
bbox: NormalizedBBox;
|
|
9
|
+
html?: string;
|
|
10
|
+
/** Optional per-cell metadata (colspan/rowspan). */
|
|
11
|
+
spans?: Array<Array<{
|
|
12
|
+
rowspan?: number;
|
|
13
|
+
colspan?: number;
|
|
14
|
+
} | undefined>>;
|
|
15
|
+
}
|
|
16
|
+
export interface FieldResult {
|
|
17
|
+
value: unknown;
|
|
18
|
+
confidence: number;
|
|
19
|
+
pageIndex?: number;
|
|
20
|
+
bbox?: NormalizedBBox;
|
|
21
|
+
}
|
|
22
|
+
export interface ExtractionProvenance {
|
|
23
|
+
provider: string;
|
|
24
|
+
model: string;
|
|
25
|
+
durationMs: number;
|
|
26
|
+
tilesUsed: number;
|
|
27
|
+
repairs: number;
|
|
28
|
+
}
|
|
29
|
+
export interface ExtractionResult<T> {
|
|
30
|
+
/** Typed extraction payload. */
|
|
31
|
+
data: T;
|
|
32
|
+
/** Global confidence score in [0, 1]. */
|
|
33
|
+
confidence: number;
|
|
34
|
+
/** Per-field confidence and grounding (best available). */
|
|
35
|
+
fields: Record<string, FieldResult>;
|
|
36
|
+
/** Extracted and optionally stitched tables. */
|
|
37
|
+
tables: ExtractedTable[];
|
|
38
|
+
/** Visual grounding records for all detected fields. */
|
|
39
|
+
groundings: GroundedField[];
|
|
40
|
+
/** Invariant check report when `prove: true` is used. */
|
|
41
|
+
invariants?: InvariantReport;
|
|
42
|
+
provenance: ExtractionProvenance;
|
|
43
|
+
}
|
|
44
|
+
export interface TileOptions {
|
|
45
|
+
/** Maximum tile width in pixels (default 1536). */
|
|
46
|
+
maxTileWidth?: number;
|
|
47
|
+
/** Maximum tile height in pixels (default 1536). */
|
|
48
|
+
maxTileHeight?: number;
|
|
49
|
+
/** Overlap ratio between tiles in [0, 1] (default 0.15). */
|
|
50
|
+
overlap?: number;
|
|
51
|
+
/** Minimum number of pixels per tile. */
|
|
52
|
+
minPixels?: number;
|
|
53
|
+
/** Maximum number of pixels per tile. */
|
|
54
|
+
maxPixels?: number;
|
|
55
|
+
}
|
|
56
|
+
export interface ExtractionOptions {
|
|
57
|
+
/** Provider to use; defaults to env-variable resolution. */
|
|
58
|
+
provider?: "gemini" | "openai" | "anthropic" | "custom";
|
|
59
|
+
/** Model id; falls back to provider default. */
|
|
60
|
+
model?: string;
|
|
61
|
+
/** API key or falls back to environment variables. */
|
|
62
|
+
apiKey?: string;
|
|
63
|
+
/** Base URL override for OpenAI-compatible endpoints. */
|
|
64
|
+
baseUrl?: string;
|
|
65
|
+
/** Custom VLM provider callable. */
|
|
66
|
+
customVLM?: VLMProvider;
|
|
67
|
+
/** Tiling behaviour: `auto`, `off`, or explicit options. */
|
|
68
|
+
tiling?: "auto" | "off" | TileOptions;
|
|
69
|
+
/** Run invariant checks. */
|
|
70
|
+
prove?: boolean;
|
|
71
|
+
/** Custom invariant checks. */
|
|
72
|
+
invariants?: Array<{
|
|
73
|
+
id: string;
|
|
74
|
+
expression: string;
|
|
75
|
+
tolerance?: number;
|
|
76
|
+
}>;
|
|
77
|
+
/** Re-invoke the VLM with the discrepancy report when invariants fail. */
|
|
78
|
+
autoRepair?: boolean;
|
|
79
|
+
/** Max output tokens. */
|
|
80
|
+
maxTokens?: number;
|
|
81
|
+
/** Sampling temperature. */
|
|
82
|
+
temperature?: number;
|
|
83
|
+
/** Deterministic seed. */
|
|
84
|
+
seed?: number;
|
|
85
|
+
/** Prompt override. */
|
|
86
|
+
prompt?: string;
|
|
87
|
+
/** Optional progress callback. */
|
|
88
|
+
onProgress?: (p: {
|
|
89
|
+
step: string;
|
|
90
|
+
page: number;
|
|
91
|
+
tile?: number;
|
|
92
|
+
}) => void;
|
|
93
|
+
}
|
|
94
|
+
export interface IngestedPage {
|
|
95
|
+
index: number;
|
|
96
|
+
image: PageImage;
|
|
97
|
+
size: {
|
|
98
|
+
width: number;
|
|
99
|
+
height: number;
|
|
100
|
+
};
|
|
101
|
+
}
|
|
102
|
+
//# sourceMappingURL=types.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"types.d.ts","sourceRoot":"","sources":["../../../src/core/types.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,sBAAsB,CAAC;AAC3D,OAAO,KAAK,EAAE,aAAa,EAAE,SAAS,EAAE,WAAW,EAAE,MAAM,uBAAuB,CAAC;AACnF,OAAO,KAAK,EAAE,eAAe,EAAE,MAAM,6BAA6B,CAAC;AAEnE,MAAM,WAAW,cAAc;IAC7B,SAAS,EAAE,MAAM,CAAC;IAClB,IAAI,EAAE,OAAO,EAAE,EAAE,CAAC;IAClB,MAAM,CAAC,EAAE,MAAM,EAAE,CAAC;IAClB,IAAI,EAAE,cAAc,CAAC;IACrB,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,oDAAoD;IACpD,KAAK,CAAC,EAAE,KAAK,CAAC,KAAK,CAAC;QAAE,OAAO,CAAC,EAAE,MAAM,CAAC;QAAC,OAAO,CAAC,EAAE,MAAM,CAAA;KAAE,GAAG,SAAS,CAAC,CAAC,CAAC;CAC1E;AAED,MAAM,WAAW,WAAW;IAC1B,KAAK,EAAE,OAAO,CAAC;IACf,UAAU,EAAE,MAAM,CAAC;IACnB,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,IAAI,CAAC,EAAE,cAAc,CAAC;CACvB;AAED,MAAM,WAAW,oBAAoB;IACnC,QAAQ,EAAE,MAAM,CAAC;IACjB,KAAK,EAAE,MAAM,CAAC;IACd,UAAU,EAAE,MAAM,CAAC;IACnB,SAAS,EAAE,MAAM,CAAC;IAClB,OAAO,EAAE,MAAM,CAAC;CACjB;AAED,MAAM,WAAW,gBAAgB,CAAC,CAAC;IACjC,gCAAgC;IAChC,IAAI,EAAE,CAAC,CAAC;IACR,yCAAyC;IACzC,UAAU,EAAE,MAAM,CAAC;IACnB,2DAA2D;IAC3D,MAAM,EAAE,MAAM,CAAC,MAAM,EAAE,WAAW,CAAC,CAAC;IACpC,gDAAgD;IAChD,MAAM,EAAE,cAAc,EAAE,CAAC;IACzB,wDAAwD;IACxD,UAAU,EAAE,aAAa,EAAE,CAAC;IAC5B,yDAAyD;IACzD,UAAU,CAAC,EAAE,eAAe,CAAC;IAC7B,UAAU,EAAE,oBAAoB,CAAC;CAClC;AAED,MAAM,WAAW,WAAW;IAC1B,mDAAmD;IACnD,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,oDAAoD;IACpD,aAAa,CAAC,EAAE,MAAM,CAAC;IACvB,4DAA4D;IAC5D,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,yCAAyC;IACzC,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,yCAAyC;IACzC,SAAS,CAAC,EAAE,MAAM,CAAC;CACpB;AAED,MAAM,WAAW,iBAAiB;IAChC,4DAA4D;IAC5D,QAAQ,CAAC,EAAE,QAAQ,GAAG,QAAQ,GAAG,WAAW,GAAG,QAAQ,CAAC;IACxD,gDAAgD;IAChD,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,sDAAsD;IACtD,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,yDAAyD;IACzD,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,oCAAoC;IACpC,SAAS,CAAC,EAAE,WAAW,CAAC;IACxB,4DAA4D;IAC5D,MAAM,CAAC,EAAE,MAAM,GAAG,KAAK,GAAG,WAAW,CAAC;IACtC,4BAA4B;IAC5B,KAAK,CAAC,EAAE,OAAO,CAAC;IAChB,+BAA+B;IAC/B,UAAU,CAAC,EAAE,KAAK,CAAC;QAAE,EAAE,EAAE,MAAM,CAAC;QAAC,UAAU,EAAE,MAAM,CAAC;QAAC,SAAS,CAAC,EAAE,MAAM,CAAA;KAAE,CAAC,CAAC;IAC3E,0EAA0E;IAC1E,UAAU,CAAC,EAAE,OAAO,CAAC;IACrB,yBAAyB;IACzB,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,4BAA4B;IAC5B,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,0BAA0B;IAC1B,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,uBAAuB;IACvB,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,kCAAkC;IAClC,UAAU,CAAC,EAAE,CAAC,CAAC,EAAE;QAAE,IAAI,EAAE,MAAM,CAAC;QAAC,IAAI,EAAE,MAAM,CAAC;QAAC,IAAI,CAAC,EAAE,MAAM,CAAA;KAAE,KAAK,IAAI,CAAC;CACzE;AAED,MAAM,WAAW,YAAY;IAC3B,KAAK,EAAE,MAAM,CAAC;IACd,KAAK,EAAE,SAAS,CAAC;IACjB,IAAI,EAAE;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,MAAM,EAAE,MAAM,CAAA;KAAE,CAAC;CACzC"}
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"types.js","sourceRoot":"","sources":["../../../src/core/types.ts"],"names":[],"mappings":""}
|
|
@@ -0,0 +1,40 @@
|
|
|
1
|
+
/**
|
|
2
|
+
* A normalized bounding box in the form `[ymin, xmin, ymax, xmax]`
|
|
3
|
+
* where each value is in the range [0, 1] relative to the page dimensions.
|
|
4
|
+
*/
|
|
5
|
+
export type NormalizedBBox = [ymin: number, xmin: number, ymax: number, xmax: number];
|
|
6
|
+
export interface PointBox {
|
|
7
|
+
x1: number;
|
|
8
|
+
y1: number;
|
|
9
|
+
x2: number;
|
|
10
|
+
y2: number;
|
|
11
|
+
}
|
|
12
|
+
export interface PageSize {
|
|
13
|
+
width: number;
|
|
14
|
+
height: number;
|
|
15
|
+
}
|
|
16
|
+
export interface GroundedEntity<T = unknown> {
|
|
17
|
+
value: T;
|
|
18
|
+
pageIndex: number;
|
|
19
|
+
bbox: NormalizedBBox;
|
|
20
|
+
confidence?: number;
|
|
21
|
+
source: "vlm" | "ocr" | "anchor";
|
|
22
|
+
}
|
|
23
|
+
/**
|
|
24
|
+
* Convert absolute pixel coordinates into a normalized [ymin, xmin, ymax, xmax] box.
|
|
25
|
+
*/
|
|
26
|
+
export declare function normalizeBBox(box: PointBox, pageSize: PageSize): NormalizedBBox;
|
|
27
|
+
/**
|
|
28
|
+
* Convert a normalized box back to absolute pixel coordinates.
|
|
29
|
+
*/
|
|
30
|
+
export declare function denormalizeBBox(bbox: NormalizedBBox, pageSize: PageSize): PointBox;
|
|
31
|
+
/**
|
|
32
|
+
* Compute the Intersection over Union (IoU) of two normalized boxes.
|
|
33
|
+
* Returns a value in [0, 1].
|
|
34
|
+
*/
|
|
35
|
+
export declare function iou(a: NormalizedBBox, b: NormalizedBBox): number;
|
|
36
|
+
/**
|
|
37
|
+
* Greedily merge overlapping boxes whose IoU is at least `iouThreshold`.
|
|
38
|
+
*/
|
|
39
|
+
export declare function mergeBoxes(boxes: NormalizedBBox[], iouThreshold?: number): NormalizedBBox[];
|
|
40
|
+
//# sourceMappingURL=bbox.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"bbox.d.ts","sourceRoot":"","sources":["../../../src/grounding/bbox.ts"],"names":[],"mappings":"AAAA;;;GAGG;AACH,MAAM,MAAM,cAAc,GAAG,CAAC,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,CAAC,CAAC;AAEtF,MAAM,WAAW,QAAQ;IACvB,EAAE,EAAE,MAAM,CAAC;IACX,EAAE,EAAE,MAAM,CAAC;IACX,EAAE,EAAE,MAAM,CAAC;IACX,EAAE,EAAE,MAAM,CAAC;CACZ;AAED,MAAM,WAAW,QAAQ;IACvB,KAAK,EAAE,MAAM,CAAC;IACd,MAAM,EAAE,MAAM,CAAC;CAChB;AAED,MAAM,WAAW,cAAc,CAAC,CAAC,GAAG,OAAO;IACzC,KAAK,EAAE,CAAC,CAAC;IACT,SAAS,EAAE,MAAM,CAAC;IAClB,IAAI,EAAE,cAAc,CAAC;IACrB,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,MAAM,EAAE,KAAK,GAAG,KAAK,GAAG,QAAQ,CAAC;CAClC;AAsBD;;GAEG;AACH,wBAAgB,aAAa,CAAC,GAAG,EAAE,QAAQ,EAAE,QAAQ,EAAE,QAAQ,GAAG,cAAc,CAY/E;AAED;;GAEG;AACH,wBAAgB,eAAe,CAAC,IAAI,EAAE,cAAc,EAAE,QAAQ,EAAE,QAAQ,GAAG,QAAQ,CAQlF;AAED;;;GAGG;AACH,wBAAgB,GAAG,CAAC,CAAC,EAAE,cAAc,EAAE,CAAC,EAAE,cAAc,GAAG,MAAM,CAqBhE;AAED;;GAEG;AACH,wBAAgB,UAAU,CACxB,KAAK,EAAE,cAAc,EAAE,EACvB,YAAY,SAAM,GACjB,cAAc,EAAE,CA4BlB"}
|
|
@@ -0,0 +1,98 @@
|
|
|
1
|
+
function validBox(b) {
|
|
2
|
+
return (Number.isFinite(b[0]) &&
|
|
3
|
+
Number.isFinite(b[1]) &&
|
|
4
|
+
Number.isFinite(b[2]) &&
|
|
5
|
+
Number.isFinite(b[3]) &&
|
|
6
|
+
b[0] <= b[2] &&
|
|
7
|
+
b[1] <= b[3]);
|
|
8
|
+
}
|
|
9
|
+
function safeBox(b) {
|
|
10
|
+
return [
|
|
11
|
+
Math.min(b[0], b[2]),
|
|
12
|
+
Math.min(b[1], b[3]),
|
|
13
|
+
Math.max(b[0], b[2]),
|
|
14
|
+
Math.max(b[1], b[3]),
|
|
15
|
+
];
|
|
16
|
+
}
|
|
17
|
+
/**
|
|
18
|
+
* Convert absolute pixel coordinates into a normalized [ymin, xmin, ymax, xmax] box.
|
|
19
|
+
*/
|
|
20
|
+
export function normalizeBBox(box, pageSize) {
|
|
21
|
+
const minX = Math.min(box.x1, box.x2);
|
|
22
|
+
const maxX = Math.max(box.x1, box.x2);
|
|
23
|
+
const minY = Math.min(box.y1, box.y2);
|
|
24
|
+
const maxY = Math.max(box.y1, box.y2);
|
|
25
|
+
return [
|
|
26
|
+
minY / pageSize.height,
|
|
27
|
+
minX / pageSize.width,
|
|
28
|
+
maxY / pageSize.height,
|
|
29
|
+
maxX / pageSize.width,
|
|
30
|
+
];
|
|
31
|
+
}
|
|
32
|
+
/**
|
|
33
|
+
* Convert a normalized box back to absolute pixel coordinates.
|
|
34
|
+
*/
|
|
35
|
+
export function denormalizeBBox(bbox, pageSize) {
|
|
36
|
+
const [ymin, xmin, ymax, xmax] = safeBox(bbox);
|
|
37
|
+
return {
|
|
38
|
+
x1: xmin * pageSize.width,
|
|
39
|
+
y1: ymin * pageSize.height,
|
|
40
|
+
x2: xmax * pageSize.width,
|
|
41
|
+
y2: ymax * pageSize.height,
|
|
42
|
+
};
|
|
43
|
+
}
|
|
44
|
+
/**
|
|
45
|
+
* Compute the Intersection over Union (IoU) of two normalized boxes.
|
|
46
|
+
* Returns a value in [0, 1].
|
|
47
|
+
*/
|
|
48
|
+
export function iou(a, b) {
|
|
49
|
+
const aBox = safeBox(a);
|
|
50
|
+
const bBox = safeBox(b);
|
|
51
|
+
if (!validBox(aBox) || !validBox(bBox))
|
|
52
|
+
return 0;
|
|
53
|
+
const ymin = Math.max(aBox[0], bBox[0]);
|
|
54
|
+
const xmin = Math.max(aBox[1], bBox[1]);
|
|
55
|
+
const ymax = Math.min(aBox[2], bBox[2]);
|
|
56
|
+
const xmax = Math.min(aBox[3], bBox[3]);
|
|
57
|
+
const interWidth = xmax - xmin;
|
|
58
|
+
const interHeight = ymax - ymin;
|
|
59
|
+
if (interWidth <= 0 || interHeight <= 0)
|
|
60
|
+
return 0;
|
|
61
|
+
const interArea = interWidth * interHeight;
|
|
62
|
+
const aArea = (aBox[2] - aBox[0]) * (aBox[3] - aBox[1]);
|
|
63
|
+
const bArea = (bBox[2] - bBox[0]) * (bBox[3] - bBox[1]);
|
|
64
|
+
const unionArea = aArea + bArea - interArea;
|
|
65
|
+
return unionArea === 0 ? 0 : interArea / unionArea;
|
|
66
|
+
}
|
|
67
|
+
/**
|
|
68
|
+
* Greedily merge overlapping boxes whose IoU is at least `iouThreshold`.
|
|
69
|
+
*/
|
|
70
|
+
export function mergeBoxes(boxes, iouThreshold = 0.5) {
|
|
71
|
+
if (boxes.length === 0)
|
|
72
|
+
return [];
|
|
73
|
+
const working = boxes.map(safeBox);
|
|
74
|
+
let changed = true;
|
|
75
|
+
while (changed) {
|
|
76
|
+
changed = false;
|
|
77
|
+
for (let i = 0; i < working.length; i++) {
|
|
78
|
+
for (let j = i + 1; j < working.length; j++) {
|
|
79
|
+
if (iou(working[i], working[j]) >= iouThreshold) {
|
|
80
|
+
const merged = [
|
|
81
|
+
Math.min(working[i][0], working[j][0]),
|
|
82
|
+
Math.min(working[i][1], working[j][1]),
|
|
83
|
+
Math.max(working[i][2], working[j][2]),
|
|
84
|
+
Math.max(working[i][3], working[j][3]),
|
|
85
|
+
];
|
|
86
|
+
working.splice(j, 1);
|
|
87
|
+
working[i] = merged;
|
|
88
|
+
changed = true;
|
|
89
|
+
break;
|
|
90
|
+
}
|
|
91
|
+
}
|
|
92
|
+
if (changed)
|
|
93
|
+
break;
|
|
94
|
+
}
|
|
95
|
+
}
|
|
96
|
+
return working;
|
|
97
|
+
}
|
|
98
|
+
//# sourceMappingURL=bbox.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"bbox.js","sourceRoot":"","sources":["../../../src/grounding/bbox.ts"],"names":[],"mappings":"AA0BA,SAAS,QAAQ,CAAC,CAAiB;IACjC,OAAO,CACL,MAAM,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;QACrB,MAAM,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;QACrB,MAAM,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;QACrB,MAAM,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;QACrB,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QACZ,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CACb,CAAC;AACJ,CAAC;AAED,SAAS,OAAO,CAAC,CAAiB;IAChC,OAAO;QACL,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC;QACpB,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC;QACpB,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC;QACpB,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC;KACrB,CAAC;AACJ,CAAC;AAED;;GAEG;AACH,MAAM,UAAU,aAAa,CAAC,GAAa,EAAE,QAAkB;IAC7D,MAAM,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,EAAE,EAAE,GAAG,CAAC,EAAE,CAAC,CAAC;IACtC,MAAM,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,EAAE,EAAE,GAAG,CAAC,EAAE,CAAC,CAAC;IACtC,MAAM,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,EAAE,EAAE,GAAG,CAAC,EAAE,CAAC,CAAC;IACtC,MAAM,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,EAAE,EAAE,GAAG,CAAC,EAAE,CAAC,CAAC;IAEtC,OAAO;QACL,IAAI,GAAG,QAAQ,CAAC,MAAM;QACtB,IAAI,GAAG,QAAQ,CAAC,KAAK;QACrB,IAAI,GAAG,QAAQ,CAAC,MAAM;QACtB,IAAI,GAAG,QAAQ,CAAC,KAAK;KACtB,CAAC;AACJ,CAAC;AAED;;GAEG;AACH,MAAM,UAAU,eAAe,CAAC,IAAoB,EAAE,QAAkB;IACtE,MAAM,CAAC,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,CAAC,GAAG,OAAO,CAAC,IAAI,CAAC,CAAC;IAC/C,OAAO;QACL,EAAE,EAAE,IAAI,GAAG,QAAQ,CAAC,KAAK;QACzB,EAAE,EAAE,IAAI,GAAG,QAAQ,CAAC,MAAM;QAC1B,EAAE,EAAE,IAAI,GAAG,QAAQ,CAAC,KAAK;QACzB,EAAE,EAAE,IAAI,GAAG,QAAQ,CAAC,MAAM;KAC3B,CAAC;AACJ,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,GAAG,CAAC,CAAiB,EAAE,CAAiB;IACtD,MAAM,IAAI,GAAG,OAAO,CAAC,CAAC,CAAC,CAAC;IACxB,MAAM,IAAI,GAAG,OAAO,CAAC,CAAC,CAAC,CAAC;IAExB,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC;QAAE,OAAO,CAAC,CAAC;IAEjD,MAAM,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC;IACxC,MAAM,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC;IACxC,MAAM,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC;IACxC,MAAM,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC;IAExC,MAAM,UAAU,GAAG,IAAI,GAAG,IAAI,CAAC;IAC/B,MAAM,WAAW,GAAG,IAAI,GAAG,IAAI,CAAC;IAChC,IAAI,UAAU,IAAI,CAAC,IAAI,WAAW,IAAI,CAAC;QAAE,OAAO,CAAC,CAAC;IAElD,MAAM,SAAS,GAAG,UAAU,GAAG,WAAW,CAAC;IAC3C,MAAM,KAAK,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC;IACxD,MAAM,KAAK,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC;IACxD,MAAM,SAAS,GAAG,KAAK,GAAG,KAAK,GAAG,SAAS,CAAC;IAE5C,OAAO,SAAS,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,SAAS,GAAG,SAAS,CAAC;AACrD,CAAC;AAED;;GAEG;AACH,MAAM,UAAU,UAAU,CACxB,KAAuB,EACvB,YAAY,GAAG,GAAG;IAElB,IAAI,KAAK,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,CAAC;IAElC,MAAM,OAAO,GAAG,KAAK,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC;IACnC,IAAI,OAAO,GAAG,IAAI,CAAC;IAEnB,OAAO,OAAO,EAAE,CAAC;QACf,OAAO,GAAG,KAAK,CAAC;QAChB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,OAAO,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;YACxC,KAAK,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,OAAO,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;gBAC5C,IAAI,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,OAAO,CAAC,CAAC,CAAC,CAAC,IAAI,YAAY,EAAE,CAAC;oBAChD,MAAM,MAAM,GAAmB;wBAC7B,IAAI,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;wBACtC,IAAI,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;wBACtC,IAAI,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;wBACtC,IAAI,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;qBACvC,CAAC;oBACF,OAAO,CAAC,MAAM,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC;oBACrB,OAAO,CAAC,CAAC,CAAC,GAAG,MAAM,CAAC;oBACpB,OAAO,GAAG,IAAI,CAAC;oBACf,MAAM;gBACR,CAAC;YACH,CAAC;YACD,IAAI,OAAO;gBAAE,MAAM;QACrB,CAAC;IACH,CAAC;IAED,OAAO,OAAO,CAAC;AACjB,CAAC"}
|
|
@@ -0,0 +1,16 @@
|
|
|
1
|
+
export * from "./compiler/zod-compiler.js";
|
|
2
|
+
export * from "./grounding/bbox.js";
|
|
3
|
+
export * from "./normalizers/currency.js";
|
|
4
|
+
export * from "./normalizers/tabular.js";
|
|
5
|
+
export * from "./validators/invariants.js";
|
|
6
|
+
export * from "./core/types.js";
|
|
7
|
+
export { DocumentExtractor } from "./core/extractor.js";
|
|
8
|
+
export * from "./providers/types.js";
|
|
9
|
+
export { resolveProvider } from "./providers/factory.js";
|
|
10
|
+
export { GeminiProvider } from "./providers/gemini-provider.js";
|
|
11
|
+
export { OpenAIProvider } from "./providers/openai-provider.js";
|
|
12
|
+
export { AnthropicProvider } from "./providers/anthropic-provider.js";
|
|
13
|
+
export * from "./tiling/tiler.js";
|
|
14
|
+
export { cropTile, getImageMetadata } from "./tiling/cutter.js";
|
|
15
|
+
export { TableStitcher } from "./stitcher/table-stitcher.js";
|
|
16
|
+
//# sourceMappingURL=index.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../src/index.ts"],"names":[],"mappings":"AAAA,cAAc,4BAA4B,CAAC;AAC3C,cAAc,qBAAqB,CAAC;AACpC,cAAc,2BAA2B,CAAC;AAC1C,cAAc,0BAA0B,CAAC;AACzC,cAAc,4BAA4B,CAAC;AAE3C,cAAc,iBAAiB,CAAC;AAChC,OAAO,EAAE,iBAAiB,EAAE,MAAM,qBAAqB,CAAC;AAExD,cAAc,sBAAsB,CAAC;AACrC,OAAO,EAAE,eAAe,EAAE,MAAM,wBAAwB,CAAC;AACzD,OAAO,EAAE,cAAc,EAAE,MAAM,gCAAgC,CAAC;AAChE,OAAO,EAAE,cAAc,EAAE,MAAM,gCAAgC,CAAC;AAChE,OAAO,EAAE,iBAAiB,EAAE,MAAM,mCAAmC,CAAC;AAEtE,cAAc,mBAAmB,CAAC;AAClC,OAAO,EAAE,QAAQ,EAAE,gBAAgB,EAAE,MAAM,oBAAoB,CAAC;AAEhE,OAAO,EAAE,aAAa,EAAE,MAAM,8BAA8B,CAAC"}
|
|
@@ -0,0 +1,16 @@
|
|
|
1
|
+
export * from "./compiler/zod-compiler.js";
|
|
2
|
+
export * from "./grounding/bbox.js";
|
|
3
|
+
export * from "./normalizers/currency.js";
|
|
4
|
+
export * from "./normalizers/tabular.js";
|
|
5
|
+
export * from "./validators/invariants.js";
|
|
6
|
+
export * from "./core/types.js";
|
|
7
|
+
export { DocumentExtractor } from "./core/extractor.js";
|
|
8
|
+
export * from "./providers/types.js";
|
|
9
|
+
export { resolveProvider } from "./providers/factory.js";
|
|
10
|
+
export { GeminiProvider } from "./providers/gemini-provider.js";
|
|
11
|
+
export { OpenAIProvider } from "./providers/openai-provider.js";
|
|
12
|
+
export { AnthropicProvider } from "./providers/anthropic-provider.js";
|
|
13
|
+
export * from "./tiling/tiler.js";
|
|
14
|
+
export { cropTile, getImageMetadata } from "./tiling/cutter.js";
|
|
15
|
+
export { TableStitcher } from "./stitcher/table-stitcher.js";
|
|
16
|
+
//# sourceMappingURL=index.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"index.js","sourceRoot":"","sources":["../../src/index.ts"],"names":[],"mappings":"AAAA,cAAc,4BAA4B,CAAC;AAC3C,cAAc,qBAAqB,CAAC;AACpC,cAAc,2BAA2B,CAAC;AAC1C,cAAc,0BAA0B,CAAC;AACzC,cAAc,4BAA4B,CAAC;AAE3C,cAAc,iBAAiB,CAAC;AAChC,OAAO,EAAE,iBAAiB,EAAE,MAAM,qBAAqB,CAAC;AAExD,cAAc,sBAAsB,CAAC;AACrC,OAAO,EAAE,eAAe,EAAE,MAAM,wBAAwB,CAAC;AACzD,OAAO,EAAE,cAAc,EAAE,MAAM,gCAAgC,CAAC;AAChE,OAAO,EAAE,cAAc,EAAE,MAAM,gCAAgC,CAAC;AAChE,OAAO,EAAE,iBAAiB,EAAE,MAAM,mCAAmC,CAAC;AAEtE,cAAc,mBAAmB,CAAC;AAClC,OAAO,EAAE,QAAQ,EAAE,gBAAgB,EAAE,MAAM,oBAAoB,CAAC;AAEhE,OAAO,EAAE,aAAa,EAAE,MAAM,8BAA8B,CAAC"}
|
|
@@ -0,0 +1,14 @@
|
|
|
1
|
+
/**
|
|
2
|
+
* Parse an accounting number in multi-region formats and return a plain number.
|
|
3
|
+
*
|
|
4
|
+
* Supported inputs:
|
|
5
|
+
* - `$1,250.50` => 1250.5
|
|
6
|
+
* - `1.250,50 €` => 1250.5
|
|
7
|
+
* - `1,250` => 1250
|
|
8
|
+
* - `(450.00)` => -450
|
|
9
|
+
* - `-1.250,50` => -1250.5
|
|
10
|
+
*
|
|
11
|
+
* Returns `null` if the value cannot be parsed.
|
|
12
|
+
*/
|
|
13
|
+
export declare function parseAccountingNumber(val: string | number): number | null;
|
|
14
|
+
//# sourceMappingURL=currency.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"currency.d.ts","sourceRoot":"","sources":["../../../src/normalizers/currency.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;GAWG;AACH,wBAAgB,qBAAqB,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,GAAG,MAAM,GAAG,IAAI,CAmFzE"}
|