| Commit message (Collapse) | Author | Age |
| ... | |
| | | |\ \ \ \
| | | | | | |
| | | | | | | |
Fix host-only cookie duplication & leak
|
| | | | |/ / /
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Currently we create duplicate cookies (!) for host-only cookies, one
secure, one insecure. This is because we're always setting the domain
parameter, regardless of the hostOnly value.
See: http://developer.chrome.com/extensions/cookies.html#method-set
Closes issue #159
Signed-off-by: Nick Semenkovich <semenko@alum.mit.edu>
|
| | | |\ \ \ \
| | | |_|/ /
| | |/| | | |
Run Chrome incognito in split mode
|
| | | | |/ /
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Since we now have a LRU cache, be very paranoid about the
incognito <-> normal browsing boundary. This runs the extension
in a separate process in incognito mode (no shared cache, etc.)
Signed-off-by: Nick Semenkovich <semenko@alum.mit.edu>
|
| | | |/ /
| | | |
| | | |
| | | | |
Updated: about.tor_lang_code to "en", since there are no "sv" verision of Tor webpage.
|
| | | | | |
|
| | | |/ |
|
| | | | |
|
| | | |\
| | | |
| | | | |
Add HTTPS Switch Planner mode to devtools
|
| | | | | |
|
| | | | | |
|
| | | | |
| | | |
| | | |
| | | | |
Planner mode
|
| | | | |\
| | | | |
| | | | |
| | | | | |
switchplanner
|
| | | | | | |
|
| | | | | | |
|
| | | | | | |
|
| | | | |\ \
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
switchplanner
Conflicts:
chromium/background.js
|
| | | | |\ \ \
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
switchplanner
Conflicts:
chromium/background.js
|
| | | | | | | | |
|
| | | | | | | | |
|
| | | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
semi-colon, remove duplicate TODO
|
| | | | | | | | |
|
| | |/ / / / /
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
and automatically loaded when the browser / extension starts up.
TODO:
This change does not allow users to edit/delete existing rules or to submit them to a central location.
This change doesn't support exceptions or cookie rules yet (but no reason they couldn't be easily added).
I also introduced a structure for a JS object that can be used to represent rules.
|
| | | | | | | |
|
| | | | | | | |
|
| | | | | | | |
|
| | | | | | | |
|
| | | | | | | |
|
| | | | | | | |
|
| | | | | | | |
|
| |/ / / / / |
|
| | | | | | |
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
I think this was matching anyway, but let's be paranoid and keep it here for now.
Signed-off-by: Nick Semenkovich <semenko@alum.mit.edu>
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
For some reason, we were repeatedly testing that dozens of times...
Signed-off-by: Nick Semenkovich <semenko@alum.mit.edu>
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Stop building unused test rules for the TLD itself (e.g. www.example.*)
Also remove an intermediate varible assignment.
Signed-off-by: Nick Semenkovich <semenko@alum.mit.edu>
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Useful as this is a hot path that blocks headers.
Signed-off-by: Nick Semenkovich <semenko@alum.mit.edu>
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
After some testing, it's super hard to hit 2,000 entries.
We'll play it safer with lower-ram systems and limit this to 1,000 for now.
Signed-off-by: Nick Semenkovich <semenko@alum.mit.edu>
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
The exclusion is much more likely.
Signed-off-by: Nick Semenkovich <semenko@alum.mit.edu>
|
| | | | | |
| | | | |
| | | | |
| | | | | |
Signed-off-by: Nick Semenkovich <semenko@alum.mit.edu>
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
A "global_ruleset" would've been a totally wildcard rule, "*".
There are zero of these in the current code, and if there were, I imagine it'd be an error.
Signed-off-by: Nick Semenkovich <semenko@alum.mit.edu>
|
| | |_|_|/
|/| | |
| | | |
| | | | |
Signed-off-by: Nick Semenkovich <semenko@alum.mit.edu>
|
| | |_|/
|/| | |
|
| | | | |
|
| |\ \ \
| | | |
| | | | |
Get outta the dom
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Explain the reason for this function.
Signed-off-by: Nick Semenkovich <semenko@alum.mit.edu>
|
| | | |/
| |/|
| | |
| | |
| | |
| | |
| | |
| | | |
(Inspired by @jsha's commit 900bedc)
Remove DOM manipulation and use URI() instead.
Signed-off-by: Nick Semenkovich <semenko@alum.mit.edu>
|
| | | | |
|
| | | | |
|
| |\ \ \
| |/ /
| | |
| | |
| | | |
Conflicts:
chromium/background.js
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This mimics the behavior of the Firefox addon.
Usually someone is trying to test whether disabling
a rule fixes some bug on the site, in which case
they would want to reload it anyway. We should
just do it automatically.
|