line-numbers mixin
Specifically, when both are turned on, the #lang line now shows
the line number and shifts the "#lang ..." over to the same spot
where it is drawn when the first line is natually visible
Also, rackety
pixel of space in between lines in DrRacket.
This change is based on Matthew's experience having a look
at the font setup on the three platforms.
He writes:
> * Mac OS X: the convention seems to be to add space between lines.
> TextEdit, for example, looks like DrRacket: the maze has spaces.
>
> (I can't find a font that makes the maze look right, actually, even
> if I adjust the line spacing.)
>
> * Windows: the convention seems to be that space is built into the
> font. DrRacket (and SirMail) draw lines more sparsely than Notepad.
>
> Perhaps consistent with the differing conventions, the height of
> "Courier New" at 11-pixel size is 14 on Windows, 13 on Mac OS X.
>
> * Unix: the convention seems to be to add space. DrRacket looks like
> the default Terminal and Text Editor programs on Ubuntu.
>
> The maze nevertheless looks right everywhere, because the glyphs
> extend an extra pixel above the declared bounding box!
For example, autocomplete of `cur-dir' will produce the same list of results as the documentation search, with `current-directory' first, instead of producing no results.
line terminators when:
a) running under windows, and
b) the file on the disk (when DrRacket first opens it)
contains all CRLF line endings.
In all other situations, the file is saved with LF terminators.
Before this commit, DrRacket would always use the platform-specific
convention, regardless of the original content of the file.
closes PR 12242
specifically add an argument to move/copy-to-edit so that it
does copy-only and export a function for trigger a submission
to the input port of a text:ports<%> object programmatically
then, use that to change how it works for the scheme mode (and also another variation for the REPL to
cope with the prompt)
I spent a while trying to make this work at the keymap% level (ie putting different keybindings for "home"
and "c:a" into different keymaps) but this just turned out to be far too confusing and fragile, so went
with this alternative (one keybinding, but that delegates to an overridable method)
closes PR 11446
