Commit Graph

62 Commits (99225e1688244de41ee76095367234560497617b)

Author SHA1 Message Date
gorhill 12c5acd287 this fixes #325 9 years ago
gorhill ff36607354 even better, just skip completely nsIContentPolicy for FF35+ 9 years ago
gorhill 51d18b4c47 shouldLoad needed only for FF34 and less (no need for sync messaging with e10s) 9 years ago
gorhill 1f4ba9dda9 this extends compatibility to FF-24-34
imported code from uBlock, including legacy toolbar support in 34485a04658ae7d8357dabe5e42b18e3e066a4b9,
which is itself directly based on work from @AlexVallat:
- 620a7d78d1
- d9b23d4f10
- 5ac4ec21e9
9 years ago
gorhill e088e9cf70 this should fix #284 9 years ago
gorhill bbfef4f6cf this fixes #234 + other fixes imported from uBlock 9 years ago
gorhill 714c02232a script injection can throw when not a real HTML document 9 years ago
gorhill 4ab8f52b44 this fixes #189: reverted changes from 6123335c51 9 years ago
gorhill 57de4fdf14 code review 9 years ago
gorhill 1d2954c39c cleaning up last committed code 9 years ago
gorhill 2cd6961ed4 firefox: draft work for fixing #165 9 years ago
gorhill 1f4ff9c809 more refactoring, code review 9 years ago
gorhill 4fffc97954 untangling fennec code 9 years ago
gorhill beb70b8343 code review 9 years ago
gorhill a26c893339 this fixes #43 9 years ago
gorhill 73dc3ba2b0 removed spurious lines of code 9 years ago
gorhill 231afb565a translation work from https://crowdin.com/project/ublock 9 years ago
AlexVallat 75ac4b979e Changes following review 9 years ago
gorhill 22bad6201e translation work from https://crowdin.com/project/ublock 9 years ago
AlexVallat 822d78be1b More restrictive monitoring.
Probably doesn't make a difference, but at least theoretically more efficient.
9 years ago
AlexVallat c5546e322f Use a nsIWebProgressListener instead of a tabsProgressListener for location change monitoring 9 years ago
Deathamns fd3e24713f Firefox: load content-scripts on extension start 9 years ago
Deathamns d3223766c8 Firefox: abp:subscribe 9 years ago
gorhill 7243aedf17 this fixes #224 for Chromium 9 years ago
Deathamns b1acfb08e6 Firefox/Fennec fixes 9 years ago
Deathamns 7b5ea26f28 Firefox/Fennec code cleanup 9 years ago
Deathamns df2ffce8aa Firefox: match Chrome's behavior with frame ids 9 years ago
gorhill 1b66ce8b40 #795: code review as per @Deathamns feedback 9 years ago
gorhill 0f629c9df8 this fixes #795 9 years ago
Deathamns a6dfe3f416 Firefox: remove unused arguments 9 years ago
Deathamns 1e6f29e64e Firefox: don't close popups after user interaction 9 years ago
Deathamns 052f296181 Doesn't work 9 years ago
Deathamns eecc860be7 Firefox: use correct data for parentFrameId 9 years ago
Deathamns 65cecfe263 Firefox: each frame should have a unique ID 9 years ago
Deathamns 45cd3305d1 Firefox: remove evalInSandbox
Probably we won't eval plain text code, which makes it unnecessary, also
this way the validator on AMO won't complain about it.
9 years ago
Deathamns cd52352995 Update frameModule.js 9 years ago
Deathamns 92838e16d3 Firefox: behind-the-scene requests / non-http tabs 9 years ago
Deathamns ae66f96570 Firefox: minimal effort to support older versions 9 years ago
Deathamns 6250109a78 Firefox: fix tab closing issue (pop-up blocking) 9 years ago
gorhill f482164601 removed unused code for the time being 9 years ago
Deathamns c46d25874d URL handling changes 9 years ago
Deathamns 46883756e5 Miscellaneous changes 9 years ago
Deathamns f51bb42fa4 Firefox: fixed content scripts? 9 years ago
Deathamns de1f643510 Firefox: improvements for content scripts 9 years ago
Deathamns f570bc7f48 Firefox: simplify vAPI.app.restart 9 years ago
Deathamns 6571a0700f Firefox: workaround for the e10s crash 9 years ago
Deathamns 9f4e9d48da Firefox: blocking improvements / other fixes
- Implement pop-up blocking
- Support blocking redirected requests
- Fix Local mirroring and inline-script blocking
- Block content on data: and about:blank pages
9 years ago
Deathamns 40b191e9f4 Code styling 9 years ago
Deathamns de2f0bb292 Firefox: ownsWeak=true for observers, +minor mods 9 years ago
Deathamns 1b71f8bba3 Firefox: new method for request handling
Now both nsIContentPolicy and on-http-* observers are used for net request
monitoring.

Reasons:
 - In many cases, nsIContentPolicy.shouldLoad is invoked twice for the same
   resource, because of the speculative parsing.
 - nsIContentPolicy.shouldLoad don't have information about the channel,
   so it can't redirect the request, nor change its headers, however
   on-http-opening-request can.

Also, local mirroring and inline-script blocking has been implemented.
9 years ago