unprint 0.16.4-beta → 0.17.0

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.
Files changed (3) hide show
  1. package/README.md +11 -3
  2. package/package.json +1 -1
  3. package/src/app.js +6 -4
package/README.md CHANGED
@@ -205,15 +205,23 @@ Extracts the CSS `url()` background from a style attribute. Alias for `query.sty
205
205
  * `unprint.request(url, body, [options], [method])`
206
206
 
207
207
  Options
208
- * `select`: Pre-query and initialize a specific element on the page
209
- * `selectAll`: Pre-query and initialize multiple specific element on the page
208
+ * `select`: Pre-query and initialize a specific element on the page.
209
+ * `selectAll`: Pre-query and initialize multiple specific element on the page.
210
210
 
211
211
  Use Playwright with Chromium (experimental)
212
212
  * `unprint.browserRequest(url, [options])`
213
213
  * `unprint.closeAllBrowsers()`
214
214
 
215
215
  Additional options
216
- * `browser`: Options object passed to Playwright
216
+ * `scope`: Browser instance to (re)use, default `main`.
217
+ * `browser`: Options object passed to Playwright's `launch`, requires new scope.
218
+ * `browser.headless`: Headless mode, set to `false` to launch visible browser, default `true`.
219
+ * `context`: Options object passed to Playwright's `newContext`, requires new scope.
220
+ * `page`: Options object passed to Playwright's `goto`.
221
+
222
+ This requires you to install the Chromium executable:
223
+ * `sudo npx patchright install-deps`
224
+ * `npx patchright install`
217
225
 
218
226
  Returns
219
227
  ```javascript
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "unprint",
3
- "version": "0.16.4-beta",
3
+ "version": "0.17.0",
4
4
  "description": "Simplify common web scraping tasks while staying in control of the data.",
5
5
  "main": "src/app.js",
6
6
  "scripts": {},
package/src/app.js CHANGED
@@ -1046,17 +1046,19 @@ function setProxy(instance, options, url) {
1046
1046
  const clients = new Map();
1047
1047
 
1048
1048
  /* eslint-enable no-param-reassign */
1049
- async function getBrowserInstance(scope) {
1049
+ async function getBrowserInstance(scope, options) {
1050
1050
  if (clients.has(scope)) {
1051
1051
  return clients.get(scope);
1052
1052
  }
1053
1053
 
1054
1054
  const browser = await chromium.launch({
1055
- headless: false,
1055
+ headless: true,
1056
+ ...options.browser,
1056
1057
  });
1057
1058
 
1058
1059
  const context = await browser.newContext({
1059
1060
  userAgent: 'unprint',
1061
+ ...options.context,
1060
1062
  });
1061
1063
 
1062
1064
  const client = { context, browser };
@@ -1126,11 +1128,11 @@ async function browserRequest(url, customOptions = {}) {
1126
1128
  };
1127
1129
 
1128
1130
  return limiter.schedule(async () => {
1129
- const { context, browser } = await getBrowserInstance(options.scope);
1131
+ const { context, browser } = await getBrowserInstance(options.scope, options);
1130
1132
  const page = await context.newPage();
1131
1133
 
1132
1134
  const res = await page.goto(url, {
1133
- ...options.browser,
1135
+ ...options.page,
1134
1136
  });
1135
1137
 
1136
1138
  const status = res.status();