Implements uBO style polyfills for requests redirects #29
Conversation
…direct` option Closes #27
|
@snyderp if you review Also added a test that checks parsing extracts all 110 resources for the snapshot of |
src/blocker.rs
Outdated
| } | ||
| Some(data_url.trim().to_owned()) | ||
| } else { | ||
| // TOOD: handle error - throw? |
There was a problem hiding this comment.
Could this be logged somehow, even if its just mirroring the "didn't understand filter: X" stuff the current lib does? Would be a nice, noisy reminder if there is some new filter format we don't support, something like that
|
Also, is this exposed through the node API? |
Updates README to demonstrate redirect handling, includes logging to stderr for some of the worse failure cases
AndriusA
force-pushed
the
feature-redirect-injection
branch
from
6ae14dd
to
d6530eb
Compare
|
@snyderp should we merge and close? It is backwards-compatible, so don't see any harm in that, FFI and browser can be wired up in due course |
|
Yep, looks good to me, though it looks like there are some conflicts now. But I think the change is terrific! |
Make sure deserialization is backwards-comaptible by not deserializing redirect polyfill resources
|
Merged master to resolve conflicts Also to ensure deserialization is backwards compatible have to ignore deserializing of redirect resources for now - included the required changes, tagging @bbondy for review |
|
Switching to MessagePack for serialization to have a cleaner way of adding new fields to blocker structures over time - this works as intended with |
AndriusA
force-pushed
the
feature-redirect-injection
branch
from
63c2eb4
to
7d64942
Compare
returns request content to use for requests matched on rules with
redirectoptionCloses #27