firecrawl 4.34.2 → 4.35.0
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/dist/{chunk-PFIWJELS.js → chunk-666CR5HH.js} +1 -1
- package/dist/index.cjs +1 -1
- package/dist/index.d.cts +8 -0
- package/dist/index.d.ts +8 -0
- package/dist/index.js +2 -2
- package/dist/{package-H4IIQQQK.js → package-ZOVFJE7Z.js} +1 -1
- package/package.json +1 -1
- package/src/__tests__/unit/v2/validation.test.ts +3 -2
- package/src/v2/types.ts +8 -0
|
@@ -12,7 +12,7 @@ var require_package = __commonJS({
|
|
|
12
12
|
"package.json"(exports, module) {
|
|
13
13
|
module.exports = {
|
|
14
14
|
name: "@mendable/firecrawl-js",
|
|
15
|
-
version: "4.
|
|
15
|
+
version: "4.35.0",
|
|
16
16
|
description: "JavaScript SDK for the Firecrawl API: web scraping, crawling, web search, and scientific literature search over a research paper index of PubMed, bioRxiv, medRxiv and arXiv abstracts",
|
|
17
17
|
main: "dist/index.js",
|
|
18
18
|
types: "dist/index.d.ts",
|
package/dist/index.cjs
CHANGED
|
@@ -39,7 +39,7 @@ var require_package = __commonJS({
|
|
|
39
39
|
"package.json"(exports2, module2) {
|
|
40
40
|
module2.exports = {
|
|
41
41
|
name: "@mendable/firecrawl-js",
|
|
42
|
-
version: "4.
|
|
42
|
+
version: "4.35.0",
|
|
43
43
|
description: "JavaScript SDK for the Firecrawl API: web scraping, crawling, web search, and scientific literature search over a research paper index of PubMed, bioRxiv, medRxiv and arXiv abstracts",
|
|
44
44
|
main: "dist/index.js",
|
|
45
45
|
types: "dist/index.d.ts",
|
package/dist/index.d.cts
CHANGED
|
@@ -129,6 +129,14 @@ type PDFParser = {
|
|
|
129
129
|
pageMarkdown?: boolean;
|
|
130
130
|
/** Include per-page typed layout blocks (bounding boxes, block types, reading order). */
|
|
131
131
|
blocks?: boolean;
|
|
132
|
+
/**
|
|
133
|
+
* Join PDF pages in `document.markdown` with `\n\n---\n\n<!-- page N -->\n\n`,
|
|
134
|
+
* where N is the 1-based physical page of the content that follows. Markers
|
|
135
|
+
* appear between pages only (no leading marker for page 1), and numbering may
|
|
136
|
+
* skip pages merged by cross-page stitching — use `pages: true` when every
|
|
137
|
+
* physical page is needed. No new response field.
|
|
138
|
+
*/
|
|
139
|
+
pageMarkers?: boolean;
|
|
132
140
|
};
|
|
133
141
|
interface PdfPage {
|
|
134
142
|
pageNumber: number;
|
package/dist/index.d.ts
CHANGED
|
@@ -129,6 +129,14 @@ type PDFParser = {
|
|
|
129
129
|
pageMarkdown?: boolean;
|
|
130
130
|
/** Include per-page typed layout blocks (bounding boxes, block types, reading order). */
|
|
131
131
|
blocks?: boolean;
|
|
132
|
+
/**
|
|
133
|
+
* Join PDF pages in `document.markdown` with `\n\n---\n\n<!-- page N -->\n\n`,
|
|
134
|
+
* where N is the 1-based physical page of the content that follows. Markers
|
|
135
|
+
* appear between pages only (no leading marker for page 1), and numbering may
|
|
136
|
+
* skip pages merged by cross-page stitching — use `pages: true` when every
|
|
137
|
+
* physical page is needed. No new response field.
|
|
138
|
+
*/
|
|
139
|
+
pageMarkers?: boolean;
|
|
132
140
|
};
|
|
133
141
|
interface PdfPage {
|
|
134
142
|
pageNumber: number;
|
package/dist/index.js
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
import {
|
|
2
2
|
require_package
|
|
3
|
-
} from "./chunk-
|
|
3
|
+
} from "./chunk-666CR5HH.js";
|
|
4
4
|
|
|
5
5
|
// src/v2/utils/httpClient.ts
|
|
6
6
|
import axios from "axios";
|
|
@@ -2381,7 +2381,7 @@ var FirecrawlApp = class {
|
|
|
2381
2381
|
if (typeof process !== "undefined" && process.env && process.env.npm_package_version) {
|
|
2382
2382
|
return process.env.npm_package_version;
|
|
2383
2383
|
}
|
|
2384
|
-
const packageJson = await import("./package-
|
|
2384
|
+
const packageJson = await import("./package-ZOVFJE7Z.js");
|
|
2385
2385
|
return packageJson.default.version;
|
|
2386
2386
|
} catch (error) {
|
|
2387
2387
|
const isTest = typeof process !== "undefined" && (process.env.JEST_WORKER_ID != null || false);
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "firecrawl",
|
|
3
|
-
"version": "4.
|
|
3
|
+
"version": "4.35.0",
|
|
4
4
|
"description": "JavaScript SDK for the Firecrawl API: web scraping, crawling, web search, and scientific literature search over a research paper index of PubMed, bioRxiv, medRxiv and arXiv abstracts",
|
|
5
5
|
"main": "dist/index.js",
|
|
6
6
|
"types": "dist/index.d.ts",
|
|
@@ -89,9 +89,9 @@ describe("v2 utils: validation", () => {
|
|
|
89
89
|
expect(options.parsers).toEqual(before);
|
|
90
90
|
});
|
|
91
91
|
|
|
92
|
-
test("ensureValidScrapeOptions: leaves PDF parser pages and
|
|
92
|
+
test("ensureValidScrapeOptions: leaves PDF parser pages, blocks, and pageMarkers options untouched", () => {
|
|
93
93
|
const options = {
|
|
94
|
-
parsers: [{ type: "pdf", mode: "auto", blocks: true, pages: true }],
|
|
94
|
+
parsers: [{ type: "pdf", mode: "auto", blocks: true, pages: true, pageMarkers: true }],
|
|
95
95
|
} as any;
|
|
96
96
|
expect(() => ensureValidScrapeOptions(options)).not.toThrow();
|
|
97
97
|
expect(options.parsers[0]).toEqual({
|
|
@@ -99,6 +99,7 @@ describe("v2 utils: validation", () => {
|
|
|
99
99
|
mode: "auto",
|
|
100
100
|
blocks: true,
|
|
101
101
|
pages: true,
|
|
102
|
+
pageMarkers: true,
|
|
102
103
|
});
|
|
103
104
|
});
|
|
104
105
|
|
package/src/v2/types.ts
CHANGED
|
@@ -199,6 +199,14 @@ export type PDFParser = {
|
|
|
199
199
|
pageMarkdown?: boolean;
|
|
200
200
|
/** Include per-page typed layout blocks (bounding boxes, block types, reading order). */
|
|
201
201
|
blocks?: boolean;
|
|
202
|
+
/**
|
|
203
|
+
* Join PDF pages in `document.markdown` with `\n\n---\n\n<!-- page N -->\n\n`,
|
|
204
|
+
* where N is the 1-based physical page of the content that follows. Markers
|
|
205
|
+
* appear between pages only (no leading marker for page 1), and numbering may
|
|
206
|
+
* skip pages merged by cross-page stitching — use `pages: true` when every
|
|
207
|
+
* physical page is needed. No new response field.
|
|
208
|
+
*/
|
|
209
|
+
pageMarkers?: boolean;
|
|
202
210
|
};
|
|
203
211
|
|
|
204
212
|
export interface PdfPage {
|