Extractors: Troubleshooting

Common issues and their solutions for extractors

Henrik Hofmeister avatar
Written by Henrik Hofmeister
Updated over a week ago

This document is your critical resource for troubleshooting unexpected results or execution failures. While our support team is happy to answer questions or offer guidance, check here first for solutions to common issues.

General Tips

The following tips are good first things to try out if a website doesn't load in the Extractor editor or doesn't function like in a normal browser:

  • On the "Settings" tab, switch the browser engine from "webkit" to "chrome". Currently, the "webkit" engine is faster but the "chrome" engine supports more sites.

  • On the "Settings" tab, enable "single page" mode. Read more about single page applications.

  • On the settings tab, change the "proxies" field to force the robot to load the website from another location. Read more about proxies.

  • Disable all network filters - or as a minimum the ones that could block requests for required resources.

  • Add network filters for things like live chat widgets, video, tracking, analytics, etc. that can make the editor unresponsive.

  • Set a fixed user agent. See e.g. this list of user agent strings.

For other specific specific problems, see the sections below. A lot of them can be traced to incorrect CSS selectors.

Robot fails after “Querying dom” message

The execution log shows an entry for Querying dom: and immediately after an error message such as error_screen|<binary> appears.

Thu May 05 20:07:35 UTC 2016|log|Querying dom: div.review-container:nth-child(8) > div.review-user > b
Thu May 05 20:07:35 UTC 2016|error_screen|<binary>
Thu May 05 20:07:36 UTC 2016|error_report|<base64>
Thu May 05 20:07:36 UTC 2016|error|Extraction element not found
Thu May 05 20:07:36 UTC 2016|done|false

These problems always happen when the scope for a child element in a loop is assigned improperly. Ensure that the selector in the child node takes the parent node’s selector into account.

This error usually happens when selecting an element for extraction inside a loop using the Click to select new element button located in the step's Edit panel. Be careful when using this button; it will resolve the path to the absolute path (using the document context) and not the relative path (using the loop element context). Most of the time you must delve into the CSS selector to make sure that you’re using the right context.

Solving page iteration problems

When creating a new Iterate Pages step, the extractor editor will determine the unique CSS selector representing a page element. However, a common pattern found in paginated Web interface design is that once you navigate to Page 2, a Previous button appears. This new page element may cause the Iterate Pages step’s CSS path to become ambiguous or incorrect upon loading Page 2. It may even point to the Previous button instead of the Next button as intended. In practice, this means that the robot will first navigate to Page 2, but will then navigate back to Page 1. To solve this problem, navigate to Page 2 in the extractor editor and verify that the CSS selector specified for the Iterate Pages step still points to the Next button. If it doesn't, modify the selector accordingly. This is often as simple as adding :last to the end of the path.

A Loop Through Elements step works only on the first page

The extractor editor tries its best to determine the correct common path for multiple elements, but sometimes it fails to select the optimal choice. If you experience issues where a loop is not working on certain pages - or all pages except the first - investigate the CSS selector of the Loop Through Elements step and ensure it’’s configured with the correct CSS selector for the list of elements you're looping through.

An Element not found error occurs, but I can see the element in the extractor editor

The extractor editor tries to identify the unique CSS selector for an element, but it can fail. If certain elements aren’t found - either in the editor or during your executions - double-check the CSS selectors specified for the relevant steps.

No value found for input field

No input values were configured for the run. Visit the run configuration screen, click the Inputs tab, and add input values to the run configuration.

No elements found in loop

The CSS selector specified for the relevant Loop through elements step is incorrect. In the extractor editor, edit the step to enter the correct CSS selector in the Element Path field.

No output available

The robot failed to extract results, so the Save Current Output step had no data to save.

  • In the extractor editor, verify the CSS selectors for each suspect extraction step.

  • If setting extraction steps’ error handling to *Ignore & Continue* throughout an extraction sequence, ensure this setting is applied to the *Save Current Output* step, as well

Extraction element not found

The robot failed to find the page element to extract data from.

  • In the extractor editor, verify the CSS selector specified for this extraction step.

  • Review the Web page at which the error occurred to ensure the CSS selector pointing to the desired data doesn’t differ from the page upon which you based the robot design.

Failed to query using CSS selector

The CSS selector syntax is incorrect.

  • In the extractor editor, verify the syntax used in the CSS selector specified for the relevant step.

Did this answer your question?