Addresses #597, Don't iterate Array with for...in
Also fixes child items displaying twice in saved search reports and adds proper collation support in reports
- Added asTempTable parameter to Zotero.Search.search() to save results to a temporary table -- avoids large inserts with post-search filters
- Added proper collation sort support for tags
Also, child items in search results now always display in context, with the non-matching rows in gray. May later implement Elena's idea to show an ellipsis to toggle non-matching items.
New method Search.setScope(searchObj) to pass one search as the scope for another -- used to make quicksearch and tag selector properly search within the current view.
Note: These changes need testing -- I may not have gotten all the various conditions (view mode, quicksearch, tag selector, advanced search, drag-and-drop) exactly right.
- Items added with Zotero pane closed did not display in the library until clicking away and back
- Height of Zotero pane and tag selector are now reduced automatically when unhiding if larger than the available space
- New Actions menu option, "Report Errors...", to send chrome errors from the Error Console to the Zotero server via the error reporting wizard -- no more Advanced Instructions from the Reporting Bugs page
- Disabled unresponsive script indicator during schema upgrades, since that seems to be what caused problems during the Beta 4 upgrade
- Added repair steps for interrupted Beta 4 upgrades
- Speed up b4.r2 itemNotes repair step, which may have been what was causing the unresponsive script warning on large libraries
- Moved some Zotero.UnresponsiveScriptIndicator.enable() calls into finally{} blocks to make sure they get triggered
- New method Zotero.getSystemInfo() to get some info on the environment (platform, version, app, app version, locale) as a string for inclusion in bug reports
- Added public property skipBackup to Zotero.DB so a startup error can cancel the shutdown backup
Closes#321, Notification window to show when rebuilding cache
- Properly handle missing custom data directory on Windows
- Disable unresponsive script indicator during schema update
- Back up the database before userdata schema upgrades (and delete all but last pre-upgrade backup on successful upgrade)
- Delete tmp file in backupDatabase if corrupt
- Remove backup file test on startup -- seems to be working
Closes#560, Method for inserting bibliography/export format into text box on current page
Dragging items to external text apps is also possible now (the one oddity being that the format has to be set at the beginning of the drag session, so if the window's current page matches a site-specific setting, the dragged references will use that format even the if the text is dropped into an external app or another window)
- Fix PDF indexing for importFromDocument() files
(Reverted to using saveURI() rather than WebPageDump for non-HTML files in Attachments.importFromDocument(), since custom filename logic was being lost and it was just calling saveURI() itself anyway)
I just set a min-height to not collapse the top part when it's empty. Could add the parent item type in brackets, but I don't know that it's necessary.
