summaryrefslogtreecommitdiffstats
path: root/chrome/browser
Commit message (Collapse)AuthorAgeFilesLines
* Fix interactive test crashes.beng@google.com2008-11-011-0/+1
| | | | | | | | TBR=glen Review URL: http://codereview.chromium.org/9002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4362 0039d316-1c4b-4281-b951-d872f2087c98
* Second attempt at fixing dist crash... more accurately control the lifetime ↵ben@chromium.org2008-11-012-3/+3
| | | | | | | | | | of the dragged tab container. http://crbug.com/3939 Review URL: http://codereview.chromium.org/8800 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4352 0039d316-1c4b-4281-b951-d872f2087c98
* On first run process the master prefs json, only one entry matters to us.cpu@google.com2008-10-316-22/+93
| | | | | | | | | | | - skip or not first run dialgos - on skip import silently search provider only BUG=1394863 Review URL: http://codereview.chromium.org/8758 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4346 0039d316-1c4b-4281-b951-d872f2087c98
* This fixes the VerifyHistoryLength ui test flakiness. The test initiates ↵ananta@chromium.org2008-10-311-2/+0
| | | | | | | | | | | | | | page navigations in timers and reads the history length to validate it. When we receive a request to initiate a navigation in the browser, we send out the request to the renderer and then immediately read the history length and send it out in an update history length request. This causes the test to fail at times as it reads a stale history length. When we receive the DidAddHistoryItem notification in the renderer, we should not update the history length for the start page navigation. This results in the test failure at times as it reads an incorrect history length. R=jam Review URL: http://codereview.chromium.org/8898 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4344 0039d316-1c4b-4281-b951-d872f2087c98
* Takes steps to make our PlatformScreen implementation more portable.darin@chromium.org2008-10-312-52/+56
| | | | | | | | | | | | | | | | | | Introduces ChromiumBridge as a means for our WebCore port to depend on the embedder indirectly. This will be extended to support the rest of our port. WebWidgetImpl and ChromeClientImpl both needed to have their platformWindow getter implemented. This fixes a regression related to the most recent merge. Removes the orphaned Language.cpp (see the real one in the platform/chromium directory. Changed webkit_glue::GetMonitorInfoForWindow to webkit_glue::GetScreenInfo. This resulted in a varied amount of plumbing changes. It also pushes the platform specific bits up into the browser where they belong. ScreenInfo is a struct that is part of the webkit/glue API. R=dglazkov,eseidel Review URL: http://codereview.chromium.org/8761 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4335 0039d316-1c4b-4281-b951-d872f2087c98
* Some code in MessageBoxView was focusing the first focusable element in the ↵jcampan@chromium.org2008-10-311-2/+7
| | | | | | | | | | | | | view, overidding the focus set from the DialogDelegate. We now only rely on the dialog delegate. Also changed dialog delegate so the default button is also the focused button. BUG=98 TEST=Open a javascript alert then confirm dialog. A button should be focused. Open all dialogs in Chrome. A button should be focused by default. Review URL: http://codereview.chromium.org/8786 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4334 0039d316-1c4b-4281-b951-d872f2087c98
* Adds support for keyboard-triggered (through VK_APPS and SHIFT+VK_F10) ↵klink@chromium.org2008-10-312-2/+12
| | | | | | | | right-click menu, specifically on the toolbar's back/forward buttons. Review URL: http://codereview.chromium.org/8942 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4326 0039d316-1c4b-4281-b951-d872f2087c98
* Fix a bug where languages that have a long string for thepaulg@google.com2008-10-314-79/+120
| | | | | | | | | | | | | | | | dangerous download warning cause an overlap with the label for that download's URL. We now use a locale-specific pixel value which changes the height of each download's view on the download page so that this overlap no longer happens. BUG=3824 (http://code.google.com/p/chromium/issues/detail?id=3824) Review URL: http://codereview.chromium.org/8933 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4323 0039d316-1c4b-4281-b951-d872f2087c98
* Fix regression of b/issue?id=1280317. We need to do a case-insensitivempcomplete@google.com2008-10-311-4/+9
| | | | | | | | | compare of filenames to test if a plugin is Gears. BUG=1280317 Review URL: http://codereview.chromium.org/8784 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4322 0039d316-1c4b-4281-b951-d872f2087c98
* 2nd part of the reorganize options under 'UnderTheHood' (bug 3655):finnur@google.com2008-10-311-37/+34
| | | | | | | | | | The link 'learn more' now points to a yet-to-be-created article in the help center. I moved it also, so that it appears to the right of the label (instead of below the label). And finally, I added missing definitions for the Bookmark Manager, which I noticed while I was changing this. Review URL: http://codereview.chromium.org/8777 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4317 0039d316-1c4b-4281-b951-d872f2087c98
* Add the comment for why we detach plugin windows when the WebContents isbrettw@google.com2008-10-311-4/+15
| | | | | | | destroyed. This is from the commit log of the patch checking this code in in the old repo. git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4314 0039d316-1c4b-4281-b951-d872f2087c98
* Background:nsylvain@chromium.org2008-10-311-1/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | 1. We can tell chrome to "always open this type of file" when downloading. This is disabled for executable for security reasons. 2. When downloading an executable files we download a .download instead. -=-= The problem is that the "always open this type of file" is enabled for executables during the download because we check the extension of the file to know if it's an executable or not, and during download we see .download, and we don't consider it an executable. The fix is to look at the "original_name" when it's present to get the extension of the real file, not the .download. BUG:3814 Review URL: http://codereview.chromium.org/8918 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4306 0039d316-1c4b-4281-b951-d872f2087c98
* A fix for Issue 3156 in chromium: "OmniBox: NavSuggest doesn't work fine ↵hbono@chromium.org2008-10-311-0/+20
| | | | | | | | when you select NavSuggest suggested item before finalizing IME."To investigate this issue today, a rich-edit control of Windows XP does not finish an ongoing composition before changing its text with a SetWindowText() call. (Thanks pkasting for colleting my corrupted comments.)To solve this issue, this change manually finishes an ongoing composition before calling the function. Review URL: http://codereview.chromium.org/8669 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4300 0039d316-1c4b-4281-b951-d872f2087c98
* Attempt to fix dist crashes or at least make them crash with more ↵ben@chromium.org2008-10-311-2/+5
| | | | | | | | understandable stack traces. Destroy the dragged representation immediately, not after returning to the message loop. Review URL: http://codereview.chromium.org/8926 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4295 0039d316-1c4b-4281-b951-d872f2087c98
* Fix RTL issues in the About box (bug 3756).finnur@google.com2008-10-313-94/+94
| | | | | | | | I used the BiDiLineIterator, which I moved out of the AutocompletePopup and into the l10n_util file. I also added a wrapper function around ubidi_getLogicalRun. Review URL: http://codereview.chromium.org/8727 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4294 0039d316-1c4b-4281-b951-d872f2087c98
* Adding preferences and clear browsing data for form autofill.jcampan@chromium.org2008-10-318-7/+88
| | | | | | | | BUG=None TEST=Open the 'Clear browsing data' dialog, there should be a check box for from data. Open the options dialogue, 2nd tab, there should be an option for autofill. Review URL: http://codereview.chromium.org/8740 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4292 0039d316-1c4b-4281-b951-d872f2087c98
* Display "No suggestions found" for misspelled words with no suggestions.sidchat@google.com2008-10-311-0/+5
| | | | | | | Issue=1677 Review URL: http://codereview.chromium.org/8408 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4289 0039d316-1c4b-4281-b951-d872f2087c98
* Remove an hack to enable the parent window. It seems to cause crashes in the ↵maruel@chromium.org2008-10-311-2/+0
| | | | | | | | field. I wonder why I had kept it active. Review URL: http://codereview.chromium.org/8963 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4285 0039d316-1c4b-4281-b951-d872f2087c98
* Adds the ability for ObserverList to not notify observers added duringsky@google.com2008-10-311-0/+1
| | | | | | | | | | | | | notification. I need this for bookmarks. If a new observer is added while the bookmark model is in the process of sending out notification the newly added observer gets confused. BUG=674 TEST=none Review URL: http://codereview.chromium.org/8919 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4267 0039d316-1c4b-4281-b951-d872f2087c98
* This fixes http://code.google.com/p/chromium/issues/detail?id=3769, whichananta@chromium.org2008-10-311-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | is to allow users to download the java plugin via our plugin installer. The plugin installer currently supports URLs which point to executable files. However the java runtime path is timed and changes frequently. As a stop gap fix for this, I added a facility in the plugin installer which indicates if a URL is a display URL. If yes then we navigate to this URL in a new tab, and allow the user to download the plugin from there. The format of the chrome_plugins_file.xml has been changed to have each plugin node specify a list of mime types. This reduces a great deal of duplicated information in the plugins file. Changed the plugin finder URL to http://dl.google.com/chrome/plugins/plugins2.xml to ensure that older clients continue to work. Bug=3769 R=jam Review URL: http://codereview.chromium.org/8729 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4260 0039d316-1c4b-4281-b951-d872f2087c98
* Make it so that you can refresh and see changes in Greasemonkeyaa@chromium.org2008-10-304-33/+55
| | | | | | | | scripts immediately, without having to restart the renderer. Review URL: http://codereview.chromium.org/8908 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4256 0039d316-1c4b-4281-b951-d872f2087c98
* Begin the first small step towards using FilePath everywhere:evanm@google.com2008-10-302-5/+6
| | | | | | | | | | - Add some transition APIs. - Start migrating some code to transition APIs. Review URL: http://codereview.chromium.org/8825 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4254 0039d316-1c4b-4281-b951-d872f2087c98
* Include webkit licence changes, make page prettier with expando licence ↵glen@chromium.org2008-10-301-31/+236
| | | | | | | | sections and homepage links. Review URL: http://codereview.chromium.org/8905 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4251 0039d316-1c4b-4281-b951-d872f2087c98
* Fix for HTTP AUTH prompts allowing save-password infobars to be shown after ↵glen@chromium.org2008-10-303-7/+14
| | | | | | | | | | | a transition away from the appropriate page while that page was loading. The issue was the LoginPrompt set up a provisional password save prior to the start of a page load, and DidNavigate is not called early enough when a user navigates, which they have an opportunity to do in the HTTP auth case (details in bug comments). BUG=2792 Review URL: http://codereview.chromium.org/8647 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4250 0039d316-1c4b-4281-b951-d872f2087c98
* When acting as a drop target, the webview keeps a cache oftc@google.com2008-10-302-7/+33
| | | | | | | | | | | | | the drop data so we don't have to send it over the IPC for each drag move. However, if a page load happens during the drop, the webview changes and we lose the cached data. To handle this, we keep track of which RVH we're dragging over and if it changes, we re-send the drop data. Review URL: http://codereview.chromium.org/8744 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4249 0039d316-1c4b-4281-b951-d872f2087c98
* SCons config for recent bookmark changes.sgk@google.com2008-10-302-1/+5
| | | | | | Review URL: http://codereview.chromium.org/8921 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4228 0039d316-1c4b-4281-b951-d872f2087c98
* Roll deps to pull in new third_party/WebKitojan@google.com2008-10-301-1/+3
| | | | git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4226 0039d316-1c4b-4281-b951-d872f2087c98
* Accumulated SCons changes for Windows chrome build, round II:sgk@google.com2008-10-302-0/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Includes other recent additions, and doesn't fix everything, but gets significant pieces back on track: * New/additional library dependencies in various test executables (too many to be worth listing). * Portability: put library base names in $LIBS instead of explicitly listing .lib files. * Have interactive_ui re-use various object files from test/ui instead of (re-)listing the .cc files. * Add new files: chrome\browser\bookmarks\bookmark_drag_data.cc chrome\browser\bookmarks\bookmark_drag_utils.cc chrome\browser\bookmarks\bookmark_table_model.cc chrome\browser\browser_trial.cc chrome\browser\download\download_request_manager_unittest.cc chrome\browser\greasmonkey_master.cc chrome\browser\spellcheck_unittest.cc chrome\browser\views\find_bar_view.cc chrome\browser\views\find_bar_win.cc chrome\renderer\greasemonkey_slave.cc * Remove old files: chrome\browser\find_in_page_controller.cc chrome\browser\find_in_page_view.cc chrome\renderer\spellcheck_unittest.cc * Remove relative reference to ../../app/google_update_settings$OBJSUFFIX. * Get rid of by-hand /IMPLIB and /MAP flags. * Prefix / suffix portability in the jscre library. * Rename $LIBXML_DIR/scons => $LIBXML_DIR/DerivedSources. * Add $CHROME_DIR/app/resources to CPPPATH. * Add icu38 directories to CPPPATH for building chrome/views files. Review URL: http://codereview.chromium.org/8742 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4215 0039d316-1c4b-4281-b951-d872f2087c98
* Fix Issue 2918 in chromium: Chrome dialog displayed in large size with ↵hbono@chromium.org2008-10-303-6/+5
| | | | | | | | Meiryo font.This issue is caused by a font whose text metric |tmAveCharWidth| is not precise as noted in Microsoft Knowledge Base 145994 (*1) To handle this case, this change uses the ChromeFont::horizontal_dlus_to_pixels() function instead of the ChromeFont::ave_char_width() function.(*1) http://support.microsoft.com/kb/145994BUG=2918 Review URL: http://codereview.chromium.org/8051 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4213 0039d316-1c4b-4281-b951-d872f2087c98
* Makes form resubmission dialog work again. This was a regressionsky@google.com2008-10-308-19/+13
| | | | | | | | | | | introduced during NavigationController refactoring. BUG=2616 TEST=see bug Review URL: http://codereview.chromium.org/8865 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4212 0039d316-1c4b-4281-b951-d872f2087c98
* Revert r4205 to track down buildbot unit_tests failure.sgk@google.com2008-10-302-11/+0
| | | | | | Review URL: http://codereview.chromium.org/8730 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4206 0039d316-1c4b-4281-b951-d872f2087c98
* Accumulated SCons changes for Windows chrome build and othersgk@google.com2008-10-302-0/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | recent additions (doesn't fix everything, but gets significant pieces back on track). * New/additional library dependencies in various test executables (too many to be worth listing). * Portability: put library base names in $LIBS instead of explicitly listing .lib files. * Have interactive_ui re-use various object files from test/ui instead of (re-)listing the .cc files. * Add new files: chrome\browser\bookmarks\bookmark_drag_data.cc chrome\browser\bookmarks\bookmark_drag_utils.cc chrome\browser\bookmarks\bookmark_table_model.cc chrome\browser\browser_trial.cc chrome\browser\download\download_request_manager_unittest.cc chrome\browser\greasmonkey_master.cc chrome\browser\spellcheck_unittest.cc chrome\browser\views\find_bar_view.cc chrome\browser\views\find_bar_win.cc chrome\renderer\greasemonkey_slave.cc * Remove old files: chrome\browser\find_in_page_controller.cc chrome\browser\find_in_page_view.cc chrome\renderer\spellcheck_unittest.cc * Remove relative reference to ../../app/google_update_settings$OBJSUFFIX. * Get rid of by-hand /IMPLIB and /MAP flags. * Prefix / suffix portability in the jscre library. * Rename $LIBXML_DIR/scons => $LIBXML_DIR/DerivedSources. * Add $CHROME_DIR/app/resources to CPPPATH. * Add icu38 directories to CPPPATH for building chrome/views files. Review URL: http://codereview.chromium.org/8725 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4205 0039d316-1c4b-4281-b951-d872f2087c98
* Changes session restore to account for apps and popups. If you havesky@google.com2008-10-3011-81/+182
| | | | | | | | | | | | | | | | | | | | | session restore enabled any time you transition from no tabbed browsers to a tabbed browser (in the same profile) we restore your last session, where the last session was any previously open tabbed browsers. For example, if you start Chrome from an installed app shortcut then create a new tabbed browser (by any means) we'll restore your last session. This gives the illustion that apps are running in their own process. I would love to have test coverage of this, but I'm loathe to write anymore flakey UI tests. Need to revisit post 1.0. BUG=1883 TEST=thoroughly test session restore, especially with application shortcuts and popups. Review URL: http://codereview.chromium.org/8856 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4202 0039d316-1c4b-4281-b951-d872f2087c98
* Make sure titles and bodies of web pages that come into the history system verybrettw@google.com2008-10-305-32/+139
| | | | | | | | | late are indexed. Very slow pages (>20 seconds) would previously not get indexed. BUG=3835 Review URL: http://codereview.chromium.org/8899 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4201 0039d316-1c4b-4281-b951-d872f2087c98
* Remove downloads from the completed dangerous downloads listpaulg@google.com2008-10-301-0/+6
| | | | | | | | | | | | | | | | when they are deleted via "Clear browsing data". There was definite crash a shutdown, and sporadic crashes when idling (but after a dangerous download was complete but not yet confirmed) after a user cleared their browsing data, since the dangerous list kept around a pointer to a download that had been deleted elsewhere. BUG=3498 (http://code.google.com/p/chromium/issues/detail?id=3498) Review URL: http://codereview.chromium.org/8728 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4200 0039d316-1c4b-4281-b951-d872f2087c98
* Add compatible version support since it was only halfway in place, and try ↵pkasting@chromium.org2008-10-308-50/+82
| | | | | | | | | | | | | and make our database versioning code and logging more similar across various consumers. The compatible version support isn't really used yet. It was going to be used for my cookie change until we decided that the old code was too busted to be forward-compatible. It seems worthwhile to put this in but maybe I am wrong. The logging similarity stuff is fairly useful. In a couple consumers in the old code, we DLOGed instead of LOGing, which meant that most people would get nothing in the log at all. I think it's a little weird that in a lot of these consumers, logging is all we do; for example, if you use a too-new cookie DB, you get output in the log, but no actual dialog box while the browser is running -- your cookies just silently don't get saved to disk. Seems bad, but I'm not prepared to try and do major surgery to address that (and add translated strings, etc.). At least now we'll actually get log messages in release builds instead of nothing at all. Because my last-access change touches this code, I'm considering asking that this change be merged back to the branch. Review URL: http://codereview.chromium.org/8712 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4195 0039d316-1c4b-4281-b951-d872f2087c98
* Add more UMA collection to the New Tab Pageglen@chromium.org2008-10-293-31/+137
| | | | | | | | BUG=1450986 Review URL: http://codereview.chromium.org/8686 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4192 0039d316-1c4b-4281-b951-d872f2087c98
* First cut at the bookmark manager. There are still a fair number ofsky@google.com2008-10-2927-849/+2882
| | | | | | | | | | | | | | | | | | | | | | | | | rough edges, but I'm at a good point where I want to land what I have. Here's what is left: . Flicker on show, likely the result of restoring window placement. . tree flickers when dragging splitter. . table/tree need to autoscroll when drop cursor held at bottom of view. . prompts for deleting. . When you move an item the table snaps to the top, this is because I'm sending out model changed. need a better notification. . Operations in menu to add need to change selection. . Remember split location. I would have preferred to split this up into a couple of reviews, but everything is intertwined now. Sorry. BUG=674 TEST=don't test the bookmark manager yet, but make sure bookmark bar still works. Review URL: http://codereview.chromium.org/8197 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4191 0039d316-1c4b-4281-b951-d872f2087c98
* Fix more problems with search providers.pkasting@chromium.org2008-10-291-5/+5
| | | | | | | | | | | onet.pl apparently moved their search service out of beta, and changed the URLs accordingly, since the time I put the original list together. In the other two cases, changing encodings to UTF-8 seemed to fix issues with words in the local languages, and who am I to argue with the results. BUG=3804,3809,3816 Review URL: http://codereview.chromium.org/8719 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4190 0039d316-1c4b-4281-b951-d872f2087c98
* Fixing some windows scons build breaks.bradnelson@chromium.org2008-10-292-9/+2
| | | | | | Review URL: http://codereview.chromium.org/8723 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4186 0039d316-1c4b-4281-b951-d872f2087c98
* Only show the language restart warning once, and only if needed.glen@chromium.org2008-10-292-3/+10
| | | | | | | | BUG=2835 Review URL: http://codereview.chromium.org/8195 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4179 0039d316-1c4b-4281-b951-d872f2087c98
* Fix for Save Page downloads not notifying their observers on completion.glen@chromium.org2008-10-291-0/+1
| | | | | | | | BUG=3417 Review URL: http://codereview.chromium.org/8804 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4177 0039d316-1c4b-4281-b951-d872f2087c98
* The SharedMemory is owned by the GreasemonkeyMaster, not the unit test. My ↵evanm@google.com2008-10-291-5/+5
| | | | | | | | | previous attempt at fixing the purify problem was overzealous; I just needed to fix the uninitialized boolean that my previous "fix" did. This change halfway-reverts to before that state. Review URL: http://codereview.chromium.org/8893 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4172 0039d316-1c4b-4281-b951-d872f2087c98
* Free memory in unit test, so purify doesn't complain.evanm@google.com2008-10-292-5/+6
| | | | | | | Review URL: http://codereview.chromium.org/8405 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4163 0039d316-1c4b-4281-b951-d872f2087c98
* Fix a couple of tab related bugs:ben@chromium.org2008-10-293-0/+13
| | | | | | | | | | | | - moving the window while a tab is animating to its final resting position disembodies the tab - it's possible to make a tab disappear while clicking on another tab while a tab is animating to its final resting position http://crbug.com/2388 http://crbug.com/1142 Review URL: http://codereview.chromium.org/8689 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4155 0039d316-1c4b-4281-b951-d872f2087c98
* Trivial patch: add whitespace to some SQL statements to turn "INSERT INTO ↵pkasting@chromium.org2008-10-295-15/+13
| | | | | | | | | foo(bar,baz)VALUES(?,?)" into "INSERT INTO foo (bar, baz) VALUES (?,?)" (which is the way we have it in most of our code). This was annoying me while I tried to track down cases where we omitted the appropriate fieldnames from our SQL statements. Review URL: http://codereview.chromium.org/8855 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4145 0039d316-1c4b-4281-b951-d872f2087c98
* Add option to disable/enable Spell Check. In addition, remove additional ↵sidchat@google.com2008-10-299-149/+123
| | | | | | | | | | code in resource message filter, which was initilaizing spellchecker in the filter if it was NULL. This part of the code is not required since the resource message filter should not initialize spellchecker; it is up to the profile to give the spellchecker to the resource message filter. Issue = 120 Review URL: http://codereview.chromium.org/7935 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4140 0039d316-1c4b-4281-b951-d872f2087c98
* Fix a minor memory leak.paulg@google.com2008-10-294-8/+6
| | | | | | | | | BUG=1340229 Review URL: http://codereview.chromium.org/8690 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4117 0039d316-1c4b-4281-b951-d872f2087c98
* Fix crbug.com/3684.tim@chromium.org2008-10-282-21/+25
| | | | | | | | | | | | | | | | Change 6053 involved making the PasswordManagerView's table model contain a vector of stack-allocated PasswordRow that owned a heap allocated PasswordForm, but std::vector creates temp PasswordRow's internally (on the stack) that were deleting the PasswordForm in their dtor. This accidental vector-internal delete caused subsequent operations on saved_signons_ to blow up. My fix is to make PasswordRows a vector of heap allocated PasswordRow instead, and have a STLElementDeleter member handle cleanup. Review URL: http://codereview.chromium.org/8651 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4103 0039d316-1c4b-4281-b951-d872f2087c98
* Don't automatically deny subresource requests that have minorwtc@google.com2008-10-282-26/+43
| | | | | | | | | | | | | | | | certificate errors (ERR_CERT_NO_REVOCATION_MECHANISM and ERR_CERT_UNABLE_TO_CHECK_REVOCATION). This requires moving the code that automatically denies subresource requests into only the SSLPolicy subclasses that need it. To avoid duplicating that code in five classes, I added convenience methods for handling overridable certificate errors and fatal errors. R=jcampan,abarth Review URL: http://codereview.chromium.org/8085 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4098 0039d316-1c4b-4281-b951-d872f2087c98
generated by cgit v1.1 at 2025-01-19 12:26:53 +0000