@crawlee/utils 4.0.0-beta.222 → 4.0.0-beta.224
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/index.d.ts
CHANGED
|
@@ -5,7 +5,7 @@ export type { DownloadListOfUrlsOptions, ExtractUrlsOptions } from './internals/
|
|
|
5
5
|
export { sleep } from './internals/general.js';
|
|
6
6
|
export * as social from './internals/social.js';
|
|
7
7
|
export * from './internals/extract-microdata.js';
|
|
8
|
-
export * from './internals/
|
|
8
|
+
export * from './internals/open-graph-parser.js';
|
|
9
9
|
export * from './internals/robots.js';
|
|
10
10
|
export { discoverValidSitemaps, Sitemap } from './internals/sitemap.js';
|
|
11
11
|
export type { ParseSitemapOptions } from './internals/sitemap.js';
|
package/index.js
CHANGED
|
@@ -4,7 +4,7 @@ export { EnqueueStrategy } from './internals/url.js';
|
|
|
4
4
|
export { sleep } from './internals/general.js';
|
|
5
5
|
export * as social from './internals/social.js';
|
|
6
6
|
export * from './internals/extract-microdata.js';
|
|
7
|
-
export * from './internals/
|
|
7
|
+
export * from './internals/open-graph-parser.js';
|
|
8
8
|
export * from './internals/robots.js';
|
|
9
9
|
export { discoverValidSitemaps, Sitemap } from './internals/sitemap.js';
|
|
10
10
|
export { ArgumentValidationError } from './internals/validation.js';
|
package/internals/url.d.ts
CHANGED
|
@@ -56,8 +56,8 @@ export declare const UNSUPPORTED_SCHEME_MESSAGE = "unsupported URL scheme (only
|
|
|
56
56
|
* Check whether `target` matches `origin` under the given enqueue `strategy`. The URL scheme is not
|
|
57
57
|
* considered here (use {@link filterUrl} for the combined scheme + strategy check).
|
|
58
58
|
*
|
|
59
|
-
* The `enqueueLinks` implementation in `@crawlee/
|
|
60
|
-
* (see `packages/
|
|
59
|
+
* The `enqueueLinks` implementation in `@crawlee/basic` matches the same strategies via glob patterns
|
|
60
|
+
* (see `packages/basic-crawler/src/internals/enqueue-links/enqueue-links.ts`) — keep the two in sync when changing either.
|
|
61
61
|
*/
|
|
62
62
|
export declare function matchesEnqueueStrategy(strategy: EnqueueStrategy | `${EnqueueStrategy}`, target: URL, origin: URL): boolean;
|
|
63
63
|
/**
|
package/internals/url.js
CHANGED
|
@@ -73,8 +73,8 @@ function normalizeHostname(hostname) {
|
|
|
73
73
|
* Check whether `target` matches `origin` under the given enqueue `strategy`. The URL scheme is not
|
|
74
74
|
* considered here (use {@link filterUrl} for the combined scheme + strategy check).
|
|
75
75
|
*
|
|
76
|
-
* The `enqueueLinks` implementation in `@crawlee/
|
|
77
|
-
* (see `packages/
|
|
76
|
+
* The `enqueueLinks` implementation in `@crawlee/basic` matches the same strategies via glob patterns
|
|
77
|
+
* (see `packages/basic-crawler/src/internals/enqueue-links/enqueue-links.ts`) — keep the two in sync when changing either.
|
|
78
78
|
*/
|
|
79
79
|
export function matchesEnqueueStrategy(strategy, target, origin) {
|
|
80
80
|
switch (strategy) {
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "@crawlee/utils",
|
|
3
|
-
"version": "4.0.0-beta.
|
|
3
|
+
"version": "4.0.0-beta.224",
|
|
4
4
|
"description": "A set of shared utilities that can be used by crawlers",
|
|
5
5
|
"engines": {
|
|
6
6
|
"node": ">=22.13.0"
|
|
@@ -43,8 +43,8 @@
|
|
|
43
43
|
},
|
|
44
44
|
"dependencies": {
|
|
45
45
|
"@apify/validations": "^1.1.0",
|
|
46
|
-
"@crawlee/http-client": "4.0.0-beta.
|
|
47
|
-
"@crawlee/types": "4.0.0-beta.
|
|
46
|
+
"@crawlee/http-client": "4.0.0-beta.224",
|
|
47
|
+
"@crawlee/types": "4.0.0-beta.224",
|
|
48
48
|
"@types/sax": "^1.2.7",
|
|
49
49
|
"cheerio": "^1.0.0",
|
|
50
50
|
"domhandler": "^6.0.0",
|
|
@@ -63,5 +63,5 @@
|
|
|
63
63
|
}
|
|
64
64
|
}
|
|
65
65
|
},
|
|
66
|
-
"gitHead": "
|
|
66
|
+
"gitHead": "669f726126327dee3b1d6b950663fd4fceace651"
|
|
67
67
|
}
|
|
File without changes
|
|
File without changes
|