mirror of
https://github.com/webrecorder/browsertrix-crawler.git
synced 2025-10-19 06:23:16 +00:00
Compare commits
2 commits
cc2d890916
...
2270964996
Author | SHA1 | Date | |
---|---|---|---|
![]() |
2270964996 | ||
![]() |
fd49041f63 |
3 changed files with 6 additions and 13 deletions
|
@ -129,8 +129,6 @@ export class Crawler {
|
|||
limitHit = false;
|
||||
pageLimit: number;
|
||||
|
||||
dupeSeedsFound = false;
|
||||
|
||||
saveStateFiles: string[] = [];
|
||||
lastSaveTime: number;
|
||||
|
||||
|
@ -2487,10 +2485,6 @@ self.__bx_behaviors.selectMainBehavior();
|
|||
return false;
|
||||
|
||||
case QueueState.DUPE_URL:
|
||||
if (!this.dupeSeedsFound && depth === 0) {
|
||||
logger.error("Duplicate seed URLs found and skipped");
|
||||
this.dupeSeedsFound = true;
|
||||
}
|
||||
logger.debug(
|
||||
"Page URL not queued, already seen",
|
||||
{ url, ...logDetails },
|
||||
|
|
|
@ -368,7 +368,7 @@ class Flow {
|
|||
case StepType.DoubleClick:
|
||||
await locator(step)
|
||||
.setTimeout(timeout * 1000)
|
||||
//.on('action', () => startWaitingForEvents())
|
||||
.setEnsureElementIsInTheViewport(true)
|
||||
.click({
|
||||
count: 2,
|
||||
button: step.button && mouseButtonMap.get(step.button),
|
||||
|
@ -392,7 +392,7 @@ class Flow {
|
|||
|
||||
await locator(step)
|
||||
.setTimeout(timeout * 1000)
|
||||
//.on('action', () => startWaitingForEvents())
|
||||
.setEnsureElementIsInTheViewport(true)
|
||||
.click({
|
||||
delay: step.duration,
|
||||
button: step.button && mouseButtonMap.get(step.button),
|
||||
|
@ -410,7 +410,7 @@ class Flow {
|
|||
case StepType.Hover:
|
||||
await locator(step)
|
||||
.setTimeout(timeout * 1000)
|
||||
//.on('action', () => startWaitingForEvents())
|
||||
.setEnsureElementIsInTheViewport(true)
|
||||
.hover();
|
||||
break;
|
||||
|
||||
|
@ -426,15 +426,14 @@ class Flow {
|
|||
|
||||
case StepType.Change:
|
||||
await locator(step)
|
||||
//.on('action', () => startWaitingForEvents())
|
||||
.setTimeout(timeout * 1000)
|
||||
.setEnsureElementIsInTheViewport(true)
|
||||
.fill(step.value);
|
||||
break;
|
||||
|
||||
case StepType.Scroll: {
|
||||
if ("selectors" in step) {
|
||||
await locator(step)
|
||||
//.on('action', () => startWaitingForEvents())
|
||||
.setTimeout(timeout * 1000)
|
||||
.scroll({
|
||||
scrollLeft: step.x || 0,
|
||||
|
|
|
@ -43,8 +43,8 @@ test("test custom selector crawls JS files as pages", async () => {
|
|||
]);
|
||||
|
||||
const expectedExtraPages = new Set([
|
||||
"https://www.iana.org/_js/jquery.js",
|
||||
"https://www.iana.org/_js/iana.js",
|
||||
"https://www.iana.org/static/_js/jquery.js",
|
||||
"https://www.iana.org/static/_js/iana.js",
|
||||
]);
|
||||
|
||||
expect(pages).toEqual(expectedPages);
|
||||
|
|
Loading…
Add table
Add a link
Reference in a new issue