officeparser 6.0.3 → 6.0.6
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +2 -2
- package/dist/OfficeParser.d.ts +0 -1
- package/dist/OfficeParser.js +20 -9
- package/dist/officeparser.browser.js +86 -85
- package/dist/officeparser.browser.js.map +4 -4
- package/dist/parsers/ExcelParser.d.ts +0 -1
- package/dist/parsers/ExcelParser.js +2 -2
- package/dist/parsers/OpenOfficeParser.d.ts +0 -1
- package/dist/parsers/OpenOfficeParser.js +5 -5
- package/dist/parsers/PdfParser.d.ts +0 -1
- package/dist/parsers/PdfParser.js +4 -44
- package/dist/parsers/PowerPointParser.d.ts +0 -1
- package/dist/parsers/PowerPointParser.js +2 -2
- package/dist/parsers/RtfParser.d.ts +0 -1
- package/dist/parsers/RtfParser.js +28 -28
- package/dist/parsers/WordParser.d.ts +0 -1
- package/dist/types.d.ts +1 -1
- package/dist/utils/chartUtils.d.ts +0 -1
- package/dist/utils/errorUtils.js +1 -1
- package/dist/utils/imageUtils.d.ts +0 -1
- package/dist/utils/imageUtils.js +4 -4
- package/dist/utils/moduleLoader.d.ts +17 -0
- package/dist/utils/moduleLoader.js +84 -0
- package/dist/utils/ocrUtils.d.ts +0 -1
- package/dist/utils/zipUtils.d.ts +0 -1
- package/package.json +16 -13
package/README.md
CHANGED
|
@@ -471,7 +471,7 @@ The browser bundle exposes the `officeParser` namespace. Include the bundle file
|
|
|
471
471
|
```
|
|
472
472
|
|
|
473
473
|
### PDF Worker Configuration in Browser
|
|
474
|
-
When using `officeparser` in a browser environment to parse PDF files, you may provide the `pdfWorkerSrc` configuration option. If not provided, it defaults to a CDN link for `pdfjs-dist@5.
|
|
474
|
+
When using `officeparser` in a browser environment to parse PDF files, you may provide the `pdfWorkerSrc` configuration option. If not provided, it defaults to a CDN link for `pdfjs-dist@5.5.207`.
|
|
475
475
|
|
|
476
476
|
```javascript
|
|
477
477
|
const file = ...; // File object or ArrayBuffer
|
|
@@ -481,7 +481,7 @@ const ast = await officeParser.parseOffice(file);
|
|
|
481
481
|
|
|
482
482
|
// Or override it with your own path or a different version:
|
|
483
483
|
const ast2 = await officeParser.parseOffice(file, {
|
|
484
|
-
pdfWorkerSrc: "https://unpkg.com/pdfjs-dist@5.
|
|
484
|
+
pdfWorkerSrc: "https://unpkg.com/pdfjs-dist@5.5.207/build/pdf.worker.min.mjs"
|
|
485
485
|
});
|
|
486
486
|
```
|
|
487
487
|
|
package/dist/OfficeParser.d.ts
CHANGED
package/dist/OfficeParser.js
CHANGED
|
@@ -49,16 +49,25 @@ var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (
|
|
|
49
49
|
}) : function(o, v) {
|
|
50
50
|
o["default"] = v;
|
|
51
51
|
});
|
|
52
|
-
var __importStar = (this && this.__importStar) || function (
|
|
53
|
-
|
|
54
|
-
|
|
55
|
-
|
|
56
|
-
|
|
57
|
-
|
|
58
|
-
};
|
|
52
|
+
var __importStar = (this && this.__importStar) || (function () {
|
|
53
|
+
var ownKeys = function(o) {
|
|
54
|
+
ownKeys = Object.getOwnPropertyNames || function (o) {
|
|
55
|
+
var ar = [];
|
|
56
|
+
for (var k in o) if (Object.prototype.hasOwnProperty.call(o, k)) ar[ar.length] = k;
|
|
57
|
+
return ar;
|
|
58
|
+
};
|
|
59
|
+
return ownKeys(o);
|
|
60
|
+
};
|
|
61
|
+
return function (mod) {
|
|
62
|
+
if (mod && mod.__esModule) return mod;
|
|
63
|
+
var result = {};
|
|
64
|
+
if (mod != null) for (var k = ownKeys(mod), i = 0; i < k.length; i++) if (k[i] !== "default") __createBinding(result, mod, k[i]);
|
|
65
|
+
__setModuleDefault(result, mod);
|
|
66
|
+
return result;
|
|
67
|
+
};
|
|
68
|
+
})();
|
|
59
69
|
Object.defineProperty(exports, "__esModule", { value: true });
|
|
60
70
|
exports.OfficeParser = void 0;
|
|
61
|
-
const fileType = __importStar(require("file-type"));
|
|
62
71
|
const fs = __importStar(require("fs"));
|
|
63
72
|
const ExcelParser_1 = require("./parsers/ExcelParser");
|
|
64
73
|
const OpenOfficeParser_1 = require("./parsers/OpenOfficeParser");
|
|
@@ -67,6 +76,7 @@ const PowerPointParser_1 = require("./parsers/PowerPointParser");
|
|
|
67
76
|
const RtfParser_1 = require("./parsers/RtfParser");
|
|
68
77
|
const WordParser_1 = require("./parsers/WordParser");
|
|
69
78
|
const errorUtils_1 = require("./utils/errorUtils");
|
|
79
|
+
const moduleLoader_1 = require("./utils/moduleLoader");
|
|
70
80
|
/**
|
|
71
81
|
* Main parser class providing office document parsing functionality.
|
|
72
82
|
*
|
|
@@ -169,7 +179,8 @@ class OfficeParser {
|
|
|
169
179
|
throw (0, errorUtils_1.getOfficeError)(errorUtils_1.OfficeErrorType.INVALID_INPUT, internalConfig);
|
|
170
180
|
}
|
|
171
181
|
if (!ext) {
|
|
172
|
-
const
|
|
182
|
+
const { fileTypeFromBuffer } = await (0, moduleLoader_1.loadFileType)();
|
|
183
|
+
const type = await fileTypeFromBuffer(buffer);
|
|
173
184
|
if (type) {
|
|
174
185
|
ext = type.ext.toLowerCase();
|
|
175
186
|
}
|