A content filter for websites, emulating dansguardian’s score-based filter in JavaScript.
Tested (Linux): 
- reported by @NovaLibertas in #15
- include example file
meta/@wsf.json. Place this in/usr/lib/mozilla/managed-storageto set fixed values.
if it was triggered multiple times, the blockpage was shown
by scanning the blockpage, which contained all the words, triggering on that, so that just the cached block page was shown.
searching of iframes, e.g. in Google Translate
to show the score if an ampersand, &, was in the blocked phrases
to persist the blockpage, credits to hllwdfr@github
speed: Was perceived as too slow to block, due to when the script loaded, and when it showed the block page.
- to load the blockpage in all cases
- to save/not save and load undefined/empty values
possibility to set another block page target (like about:blank)
- This needs to be a valid URL. You have been warned (™ ☺).
block reasons were not shown
refactoring: settings object was in main code file
- break on FF Android due to missing application ID
- block cache + test
- set the block page multiple times
- test for preset file loading
- less refactored
- tests
- less refactored
- logo needs to be IMG
- capitalization like JSDoc
- popup.html seems to be blocked on Windows 10.
- Had layout in HTML.
- HTML was invalid
Popup lacked link to options.
managed storage for chrome
settings seem to not have worked on windows
Had settings object hardcoded in main background script, which was not needed if in storage. (Request by Mozilla review)
WebExt lacked access to words for scoring.
Limit option
back button overrode block page
Whitelist option
zip/xpi lacked icon file
name (from Score-based web filter) to jsGuardian
webextension (only chromium)
safe search (use separate safe search addon for this)
Lacked toggle to turn safe search on/off
lacked (moderate) safe search on youtube
lacked safe search (google, yahoo, bing, ddg)
previous versions lacked whitelist
multiprocess as permission … ;-)
multiprocess spelling had .gitignore
?: multi-process was not enabled default was at 60, which is for small children, now it’s at 160
Bug: counted different cases differently
Doc: did not copy dg appropriate values for age groups
did not find multi-line matches
old mods were not destroyed on re-setting limit.
quotes surrounding single matches
Changing score did not affect anything.
Sites like google with many <script> tags made loading very slow.
Was not installable on recent android firefox (45).
Previous version used innerHTML.
Had no message which parts triggered filter if blocked.