Commands can prepare the document for saving in this callback, e.g. strip trailing whitespace (issue #35), ensure the line always has a newline character as the last byte in the document (issue #76), or similar.
Presently having a “will-save” command fail does not abort saving (but likely should). Also, there is no check to see if the command called requires the document to be saved before running, so that likely leads to an infinite loop.
The text view needs to do a little work before saving a document (serialize folded text) and with the window as part of the user data, we can now skip that for text views in windows for which documents are not going to be saved.
This is not an ideal solution; long-term saving will be moved up to the DocumentWindowController which has the full view of all documents and text views.
Delays may still be experienced when a file is opened or its “modified” state is changed but general edits should no longer be affected by how much is showing in the file browser.
This addresses issue #463.
We now set the isExpanded property during initialize and use a property to reflect the current user preference for wether or not to collapse when there is just one tab.
The methods going to “begin of indented line” will go to the first non-whitespace character on the line, unless the caret is already there or to the left of this character, in which case it will go to the actual beginning of the line.
The “end of indented line” methods work similarly.
If you want [⇪]⌘⇠/⇢ and ⌘⌫/⌦ to use this behavior, you can add the following to your key bindings file:
"@\UF702" = "moveToBeginningOfIndentedLine:";
"$@\UF702" = "moveToBeginningOfIndentedLineAndModifySelection:";
"@\UF703" = "moveToEndOfIndentedLine:";
"$@\UF703" = "moveToEndOfIndentedLineAndModifySelection:";
"@\U007F" = "deleteToBeginningOfIndentedLine:";
"@\UF728" = "deleteToEndOfIndentedLine:";
Initially the environment was clean but it was found that some variables should be inherited, e.g. SSH_AUTH_SOCK.
Therefor a blacklist was introduced, but it seems futile trying to keep it updated with all the variables that may affect the behavior of commands executed from within TextMate, therefor we now use a whitelist.
Probably the whitelist should be user configurable as launchd uses environment variables to communicate dynamically allocated sockets (like SSH_AUTH_SOCK, but there are also some Apple-prefixed variables pointing to named sockets).
Issue #238.
Since the std::map hosting the basic environment was using a (function local) static initializer, it would be destroyed during exit — though there might be threads using the basic environment, and joining these might happen after destroying the basic environment.
This should explain the crash in network::launch_tbz.
For example moving the selection often require padding to be inserted, previously this padding would become part of the selection.
This fixes issue #72.
Since it moves one column left/right and east asian scripts have their glyphs count as two columns, it would effectively move half a column, which in practice would be done by inserting a space.
Fixes issue #469.
Command-clicking an existing caret normally toggles, unless there only is a single caret, in which case it adds a caret. This would thus allow you to add a new caret sharing the position of the existing, and command-clicking again would then remove both the two existing carets, leading to an assertion failure (as there were no carets).
Fixes issue #466.
There are several crashes reported from firstRectForCharacterRange: which indicate that the system may ask for indexes “out of range”, hence why we now effectively just cap such request.
