Releases: lightpanda-io/browser
v0.2.6
v0.2.5
Initial iframe support
This version starts supporting iframes.
You can dump iframe's content using the fetch command:
lightpanda fetch --dump html --with_frames <URL>
First MCP server iteration
Start integrated stdio MCP server with lightpanda mcp. Some initial tools are available, notably goto and evaluate.
custom CDP command: LP.getMarkdown
First custom CDP command available to dump a page in markdown format.
v0.2.4
First version of multi thread support
Thanks to #1353 you can now have multiple CDP connections to the same Lightpanda process.
Every CDP connection gets its own full browser, but you can still only open 1 context and 1 page at a time per connection.
Everything is cleaned up on disconnect.
https://lightpanda.io/blog/posts/how-we-built-multiclient
Dump web page in markdown format
Thanks to #1560 you can now dump a fetched page into markdown format.
Use the --dump markdown option with the fetch command.
v0.2.3
robots.txt support
This version respect robots.txt if you run it with option --obey_robots.
v0.2.2
Merge pull request #1435 from lightpanda-io/handle_invalid_attribute_…
v0.2.1
Merge pull request #1359 from lightpanda-io/crash_handler Improve crash handling
v0.2.0
v0.1.0
This release is the last version of Lightpanda using Netsurf's LibDOM for DOM implementation and Netsurf's Hubbub.
Thanks to #1181, future versions of Lightpanda use html5ever for HTML parsing and a pure Zig DOM implementation.
This release must be used as reference to compare behaviors with future versions.
nightly
Merge pull request #246 from lightpanda-io/mutation-observer dom: implement MutationObserver