@clazic/kordoc 2.2.7 → 2.2.9
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/dist/{chunk-3CNYQD23.js → chunk-FF5M4SDK.js} +2 -2
- package/dist/{chunk-UFIRSH5G.js → chunk-OL2NDK3E.js} +25 -16
- package/dist/chunk-OL2NDK3E.js.map +1 -0
- package/dist/cli.js +13 -6
- package/dist/cli.js.map +1 -1
- package/dist/index.cjs +124 -32
- package/dist/index.cjs.map +1 -1
- package/dist/index.d.cts +7 -0
- package/dist/index.d.ts +7 -0
- package/dist/index.js +124 -32
- package/dist/index.js.map +1 -1
- package/dist/mcp.js +6 -5
- package/dist/mcp.js.map +1 -1
- package/dist/{provider-EPHXUWRL.js → provider-I3XGSVL6.js} +62 -17
- package/dist/provider-I3XGSVL6.js.map +1 -0
- package/dist/{resolve-NYKB5P3U.js → resolve-UFUJEPCJ.js} +10 -4
- package/dist/{resolve-NYKB5P3U.js.map → resolve-UFUJEPCJ.js.map} +1 -1
- package/dist/tesseract-provider-WCVJWBUT.js +56 -0
- package/dist/tesseract-provider-WCVJWBUT.js.map +1 -0
- package/dist/{utils-3EDZ5QEH.js → utils-CU26KLDC.js} +2 -2
- package/dist/{watch-BDL7I557.js → watch-Z6SH4KRB.js} +3 -3
- package/package.json +1 -1
- package/dist/chunk-UFIRSH5G.js.map +0 -1
- package/dist/provider-EPHXUWRL.js.map +0 -1
- package/dist/tesseract-provider-UNJOI25M.js +0 -24
- package/dist/tesseract-provider-UNJOI25M.js.map +0 -1
- /package/dist/{chunk-3CNYQD23.js.map → chunk-FF5M4SDK.js.map} +0 -0
- /package/dist/{utils-3EDZ5QEH.js.map → utils-CU26KLDC.js.map} +0 -0
- /package/dist/{watch-BDL7I557.js.map → watch-Z6SH4KRB.js.map} +0 -0
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
#!/usr/bin/env node
|
|
2
2
|
|
|
3
3
|
// src/utils.ts
|
|
4
|
-
var VERSION = true ? "2.2.
|
|
4
|
+
var VERSION = true ? "2.2.9" : "0.0.0-dev";
|
|
5
5
|
function toArrayBuffer(buf) {
|
|
6
6
|
if (buf.byteOffset === 0 && buf.byteLength === buf.buffer.byteLength) {
|
|
7
7
|
return buf.buffer;
|
|
@@ -90,4 +90,4 @@ export {
|
|
|
90
90
|
sanitizeHref,
|
|
91
91
|
classifyError
|
|
92
92
|
};
|
|
93
|
-
//# sourceMappingURL=chunk-
|
|
93
|
+
//# sourceMappingURL=chunk-FF5M4SDK.js.map
|
|
@@ -6,7 +6,7 @@ import {
|
|
|
6
6
|
precheckZipSize,
|
|
7
7
|
sanitizeHref,
|
|
8
8
|
toArrayBuffer
|
|
9
|
-
} from "./chunk-
|
|
9
|
+
} from "./chunk-FF5M4SDK.js";
|
|
10
10
|
import {
|
|
11
11
|
parsePageRange
|
|
12
12
|
} from "./chunk-MOL7MDBG.js";
|
|
@@ -5406,8 +5406,9 @@ async function parsePdfDocument(buffer, options) {
|
|
|
5406
5406
|
const ocrMode = options?.ocrMode;
|
|
5407
5407
|
if (!ocrProvider && ocrMode && ocrMode !== "off") {
|
|
5408
5408
|
try {
|
|
5409
|
-
const { resolveOcrProvider } = await import("./resolve-
|
|
5410
|
-
|
|
5409
|
+
const { resolveOcrProvider } = await import("./resolve-UFUJEPCJ.js");
|
|
5410
|
+
const concurrency = options?.ocrConcurrency ?? 1;
|
|
5411
|
+
ocrProvider = await resolveOcrProvider(ocrMode, warnings, concurrency);
|
|
5411
5412
|
} catch (resolveErr) {
|
|
5412
5413
|
if (ocrMode !== "auto") {
|
|
5413
5414
|
throw Object.assign(
|
|
@@ -5418,20 +5419,28 @@ async function parsePdfDocument(buffer, options) {
|
|
|
5418
5419
|
}
|
|
5419
5420
|
}
|
|
5420
5421
|
if (ocrProvider) {
|
|
5422
|
+
let ocrBlocks = [];
|
|
5421
5423
|
try {
|
|
5422
|
-
const { ocrPages } = await import("./provider-
|
|
5423
|
-
const
|
|
5424
|
-
|
|
5425
|
-
const ocrMarkdown = blocksToMarkdown(ocrBlocks);
|
|
5426
|
-
return {
|
|
5427
|
-
markdown: ocrMarkdown,
|
|
5428
|
-
blocks: ocrBlocks,
|
|
5429
|
-
metadata,
|
|
5430
|
-
warnings: warnings.length > 0 ? warnings : void 0,
|
|
5431
|
-
isImageBased: true
|
|
5432
|
-
};
|
|
5433
|
-
}
|
|
5424
|
+
const { ocrPages } = await import("./provider-I3XGSVL6.js");
|
|
5425
|
+
const concurrency = options?.ocrConcurrency ?? 1;
|
|
5426
|
+
ocrBlocks = await ocrPages(doc, ocrProvider, pageFilter, effectivePageCount, warnings, concurrency);
|
|
5434
5427
|
} catch {
|
|
5428
|
+
} finally {
|
|
5429
|
+
const terminable = ocrProvider;
|
|
5430
|
+
if (typeof terminable.terminate === "function") {
|
|
5431
|
+
await terminable.terminate().catch(() => {
|
|
5432
|
+
});
|
|
5433
|
+
}
|
|
5434
|
+
}
|
|
5435
|
+
if (ocrBlocks.length > 0) {
|
|
5436
|
+
const ocrMarkdown = blocksToMarkdown(ocrBlocks);
|
|
5437
|
+
return {
|
|
5438
|
+
markdown: ocrMarkdown,
|
|
5439
|
+
blocks: ocrBlocks,
|
|
5440
|
+
metadata,
|
|
5441
|
+
warnings: warnings.length > 0 ? warnings : void 0,
|
|
5442
|
+
isImageBased: true
|
|
5443
|
+
};
|
|
5435
5444
|
}
|
|
5436
5445
|
}
|
|
5437
5446
|
if (ocrMode === "off") {
|
|
@@ -9614,4 +9623,4 @@ export {
|
|
|
9614
9623
|
cfb/cfb.js:
|
|
9615
9624
|
(*! crc32.js (C) 2014-present SheetJS -- http://sheetjs.com *)
|
|
9616
9625
|
*/
|
|
9617
|
-
//# sourceMappingURL=chunk-
|
|
9626
|
+
//# sourceMappingURL=chunk-OL2NDK3E.js.map
|