these files were moved from their old home in extensions/cookie. the CVS
records cannot be simply moved because attic files exist in the new location,
and so i will unfortunately take cvs blame for all. that's unfair to dwitte
who has made so many wonderful changes to the cookies backend code! ;-)
so, to find out the real cvs blame for these files, please look at the
following cvs remove'd files:
mozilla/extensions/cookie/nsCookieService.cpp
mozilla/extensions/cookie/nsCookieService.h
mozilla/extensions/cookie/nsCookie.cpp
mozilla/extensions/cookie/nsCookie.h
mozilla/extensions/cookie/nsICookie2.idl
mozilla/extensions/cookie/nsICookieConsent.idl
mozilla/extensions/cookie/nsICookie.idl
mozilla/extensions/cookie/nsICookieManager2.idl
mozilla/extensions/cookie/nsICookieManager.idl
mozilla/extensions/cookie/nsICookiePermission.idl
git-svn-id: svn://10.0.0.236/trunk@147859 18797224-902f-48f8-a5cc-f745e15eee43
fixing a crasher regression from yesterday's landing of bug 209475.
patch by mvl; r=dwitte, sr=bz, a=dbaron.
git-svn-id: svn://10.0.0.236/trunk@144690 18797224-902f-48f8-a5cc-f745e15eee43
Changes nsIPermissionManager.idl to accept type strings rather than integers; this allows consumers to register unique types more easily, without fear of conflicting with an already-existing one.
Also fixes some bounds-checking fu (sr=bz on irc for those additional portions).
patch by mvl; r=dwitte, sr=bz.
git-svn-id: svn://10.0.0.236/trunk@144606 18797224-902f-48f8-a5cc-f745e15eee43
This changes the nsCookieService singleton fu to AddRef only once, rather than twice (and removes the corresponding Release in the module dtor). The double-addref was making the cookie module stick around until XPCOM starts shutting down modules, which is too late to release things we own (observer service, prefbranch, etc). So we were violating XPCOM shutdown rules without knowing it.
So, the tracemalloc leak stats on Brad should halve. :)
b=209571, r=alecf, sr=darin.
git-svn-id: svn://10.0.0.236/trunk@144084 18797224-902f-48f8-a5cc-f745e15eee43
UTF-8 in all XUL files to make them to be genuinely encoded in UTF-8. All the
changes are in comments so that this shouldn't affect anything at run-time.
(r=hwaara)
git-svn-id: svn://10.0.0.236/trunk@144053 18797224-902f-48f8-a5cc-f745e15eee43
- moves core cookie code into nsCookieService.{h,cpp}, and kills nsCookies & nsCookieManager
- makes nsCookieService a singleton object that implements nsICookieManager also
- improves mem efficiency of the nsCookie object, and removes unneeded conversions
- adds an nsICookie2 interface to extend the sucky nsICookie.
- fixes a few (unrelated) trivial things while I'm in there (use ->ASCII instead of ->UTF8, and remove some erroneous stuff in nsPermissionManager.h)
cvs removal of nsCookies.{h,cpp} and nsCookieManager.{h,cpp} will follow; to find old blame info and logs for those files, look in attic.
burn the witch!
b=200632, r=alecf, sr=darin.
git-svn-id: svn://10.0.0.236/trunk@143712 18797224-902f-48f8-a5cc-f745e15eee43
Content interface out-parameters should use ptr-to-ptrs instead of ref-to-ptrs and some small deCOMtamination.
r+sr=jst@netscape.com
git-svn-id: svn://10.0.0.236/trunk@143683 18797224-902f-48f8-a5cc-f745e15eee43
makes the cookie parser tolerant of spaces in the token name strings, which some sites rely on.
r=mvl, sr=alecf, a=asa.
git-svn-id: svn://10.0.0.236/trunk@142759 18797224-902f-48f8-a5cc-f745e15eee43
makes the "do you want to accept this cookie?" dialog more user-friendly, by not touching the permission list if you dismiss the dialog without clicking a button.
patch by mvl, r=neil, sr=jag, a=asa.
git-svn-id: svn://10.0.0.236/trunk@142735 18797224-902f-48f8-a5cc-f745e15eee43
polishes cookie expiry behavior and cleans up a little expiry code.
r=mvl, sr=darin, a=sspitzer.
git-svn-id: svn://10.0.0.236/trunk@142065 18797224-902f-48f8-a5cc-f745e15eee43
