Merge remote-tracking branch 'upstream/master' into rotary_skip_ytdlp

2025-04-20 23:56:34 -04:00 · 2024-05-13 18:34:50 -07:00 · 2024-05-13 18:34:50 -07:00 · c66fec4c6d
commit c66fec4c6d
parent ff02480a13 d51d09c640
2 changed files with 10 additions and 1 deletions
--- a/brozzler/behaviors.yaml
+++ b/brozzler/behaviors.yaml
@ -49,6 +49,15 @@
      - selector: a.coreSpriteRightPaginationArrow
        repeatSameElement: true
        rmSelector: '.RnEpo'
+-
+  url_regex: '^https?://www.opengov\.nsw\.gov\.au/publications.*$'
+  behavior_js_template: umbraBehavior.js.j2
+  request_idle_timeout_sec: 10
+  default_parameters:
+    interval: 1000
+    actions:
+      - selector: span.DV-trigger.DV-next
+        repeatSameElement: true
 -
  url_regex: '^https?://americaspresidents\.si\.edu/gallery.*$'
  behavior_js_template: umbraBehavior.js.j2
--- a/setup.py
+++ b/setup.py
@ -34,7 +34,7 @@ def find_package_data(package):

 setuptools.setup(
    name="brozzler",
-    version="1.5.47",
+    version="1.5.48",
    description="Distributed web crawling with browsers",
    url="https://github.com/internetarchive/brozzler",
    author="Noah Levitt",