@crawlee/playwright 4.0.0-beta.184 → 4.0.0-beta.185

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.
@@ -185,12 +185,8 @@ export class AdaptivePlaywrightCrawler extends BasicCrawler {
185
185
  });
186
186
  this.#staticCrawler = staticCrawler;
187
187
  this.#browserCrawler = browserCrawler;
188
- this.#staticContextPipeline = staticCrawler.contextPipeline.compose({
189
- action: this.adaptCheerioContext.bind(this),
190
- });
191
- this.#browserContextPipeline = browserCrawler.contextPipeline.compose({
192
- action: this.adaptPlaywrightContext.bind(this),
193
- });
188
+ this.#staticContextPipeline = staticCrawler.contextPipeline.compose(this.adaptCheerioContext.bind(this));
189
+ this.#browserContextPipeline = browserCrawler.contextPipeline.compose(this.adaptPlaywrightContext.bind(this));
194
190
  }
195
191
  async init() {
196
192
  // A crawler can be run again after a teardown.
@@ -202,34 +198,32 @@ export class AdaptivePlaywrightCrawler extends BasicCrawler {
202
198
  }
203
199
  buildContextPipeline() {
204
200
  const errorMessage = (prop) => `The \`${prop}\` property is not available on the outer context pipeline of AdaptivePlaywrightCrawler - it is provided by the inner (static/browser) pipelines`;
205
- return super.buildContextPipeline().compose({
206
- action: async ({ request }) => ({
207
- get request() {
208
- return request;
209
- },
210
- get response() {
211
- throw new Error(errorMessage('response'));
212
- },
213
- get page() {
214
- throw new Error(errorMessage('page'));
215
- },
216
- get querySelector() {
217
- throw new Error(errorMessage('querySelector'));
218
- },
219
- get querySelectorAll() {
220
- throw new Error(errorMessage('querySelectorAll'));
221
- },
222
- get waitForSelector() {
223
- throw new Error(errorMessage('waitForSelector'));
224
- },
225
- get parseWithCheerio() {
226
- throw new Error(errorMessage('parseWithCheerio'));
227
- },
228
- get enqueueLinks() {
229
- throw new Error(errorMessage('enqueueLinks'));
230
- },
231
- }),
232
- });
201
+ return super.buildContextPipeline().compose(async ({ request }) => ({
202
+ get request() {
203
+ return request;
204
+ },
205
+ get response() {
206
+ throw new Error(errorMessage('response'));
207
+ },
208
+ get page() {
209
+ throw new Error(errorMessage('page'));
210
+ },
211
+ get querySelector() {
212
+ throw new Error(errorMessage('querySelector'));
213
+ },
214
+ get querySelectorAll() {
215
+ throw new Error(errorMessage('querySelectorAll'));
216
+ },
217
+ get waitForSelector() {
218
+ throw new Error(errorMessage('waitForSelector'));
219
+ },
220
+ get parseWithCheerio() {
221
+ throw new Error(errorMessage('parseWithCheerio'));
222
+ },
223
+ get enqueueLinks() {
224
+ throw new Error(errorMessage('enqueueLinks'));
225
+ },
226
+ }));
233
227
  }
234
228
  async adaptCheerioContext(cheerioContext) {
235
229
  return {
@@ -306,12 +300,16 @@ export class AdaptivePlaywrightCrawler extends BasicCrawler {
306
300
  Object.defineProperty(subCrawlerContext, key, { ...descriptor, configurable: false });
307
301
  }
308
302
  try {
303
+ // Any failure - middleware or handler - ends up in the `ok: false` branch below.
304
+ const rethrow = (error) => {
305
+ throw error;
306
+ };
309
307
  const callAdaptiveRequestHandler = async () => {
310
308
  if (renderingType === 'static') {
311
- await this.#staticContextPipeline.call(subCrawlerContext, this.requestHandler.bind(this));
309
+ await this.#staticContextPipeline.call(subCrawlerContext, this.requestHandler.bind(this), rethrow);
312
310
  }
313
311
  else if (renderingType === 'clientOnly') {
314
- await this.#browserContextPipeline.call(subCrawlerContext, this.requestHandler.bind(this));
312
+ await this.#browserContextPipeline.call(subCrawlerContext, this.requestHandler.bind(this), rethrow);
315
313
  }
316
314
  };
317
315
  // this crawler overrides `runRequestHandler` and times each rendering-type run itself, so it has
@@ -107,7 +107,7 @@ export class PlaywrightCrawler extends BrowserCrawler {
107
107
  });
108
108
  }
109
109
  buildContextPipeline() {
110
- return super.buildContextPipeline().compose({ action: this.enhanceContext.bind(this) });
110
+ return super.buildContextPipeline().compose(this.enhanceContext.bind(this));
111
111
  }
112
112
  async navigationHandler(crawlingContext, gotoOptions) {
113
113
  return gotoExtended(crawlingContext.page, crawlingContext.request, gotoOptions);
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@crawlee/playwright",
3
- "version": "4.0.0-beta.184",
3
+ "version": "4.0.0-beta.185",
4
4
  "description": "The scalable web crawling and scraping library for JavaScript/Node.js. Enables development of data extraction and web automation jobs (not only) with headless Chrome and Puppeteer.",
5
5
  "engines": {
6
6
  "node": ">=22.0.0"
@@ -49,13 +49,13 @@
49
49
  "dependencies": {
50
50
  "@apify/datastructures": "^3.0.1",
51
51
  "@apify/timeout": "^1.0.1",
52
- "@crawlee/basic": "4.0.0-beta.184",
53
- "@crawlee/browser": "4.0.0-beta.184",
54
- "@crawlee/browser-pool": "4.0.0-beta.184",
55
- "@crawlee/cheerio": "4.0.0-beta.184",
56
- "@crawlee/core": "4.0.0-beta.184",
57
- "@crawlee/types": "4.0.0-beta.184",
58
- "@crawlee/utils": "4.0.0-beta.184",
52
+ "@crawlee/basic": "4.0.0-beta.185",
53
+ "@crawlee/browser": "4.0.0-beta.185",
54
+ "@crawlee/browser-pool": "4.0.0-beta.185",
55
+ "@crawlee/cheerio": "4.0.0-beta.185",
56
+ "@crawlee/core": "4.0.0-beta.185",
57
+ "@crawlee/types": "4.0.0-beta.185",
58
+ "@crawlee/utils": "4.0.0-beta.185",
59
59
  "cheerio": "^1.0.0",
60
60
  "jquery": "^3.7.1",
61
61
  "ml-logistic-regression": "^2.0.0",
@@ -80,5 +80,5 @@
80
80
  }
81
81
  }
82
82
  },
83
- "gitHead": "19ebeeb9bb90e38496cbe90565c8ebb02e102193"
83
+ "gitHead": "642c670956a5a60384d5ed5e67ee809927cc419b"
84
84
  }