1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
|
@echo off
rem This script starts textadept in an adapted mode, stripped from all the stuff we don't need,
rem geared at the file formats that context deals with. The reason for this is that first of
rem all we come from scite, but also because the average user doesn't need that much and can
rem get confused by all kind of options that are irrelevant for editing text files.
rem This startup script assumes that the files can be found relative to this script. It's kind
rem of tricky because textadept, while being quite configurable, is not really made for such a
rem real bare startup situation but after some trial and error, so far it works out ok. There
rem are still some issues due to assumptions in the original code. In the meantime processing
rem a file from within the editing sessions works ok which is a huge improvement over earlier
rem versions of textadept (it was actually a show stopper) so now textadept can be used as a
rem drop in for scite. We're getting there!
rem Although I like the idea of textadept, it is no longer a simple Lua binding to scintilla
rem and the claim that it is small is no longer true. The number of Lua lines doesn't really
rem say much if there are many third party dll dependencies (at least I see many files in the
rem zip and most of them probably relate to parts of the graphical interface and therefore most
rem is probably not used at all. The more dependencies there are, the less interesting it is to
rem officially support it as one of the reference editors for context, given that tex and friends
rem aim at long term stability. It's huge and unless I'm mistaken there is no minimal lightweight
rem variant for building a stripped down variant (in editing with mono spaced fonts we don't need
rem all that stuff). A small static stripped binary would be really nice to have (and I'd
rem probably default to using textadept then). I might at some point decide to strip more and just
rem provide what we only need (which is less than is there now). We'll see how it evolves.
rem In the meantime support for scintillua has been dropped which makes scite vulnerable as there
rem is no default scite (yet) with lpeg built in. Anyway, it means that we will not provide an
rem installer for scite or textadept which does the reference highlighting we've been using for
rem decades. It is up to the user: use lightweight scite or a more dependent but also more
rem configurable texadept. It would be really nice to have multiple options for editing (read: if
rem scite would have scintillua on board.) The same is true for notepad++. Each of them has its
rem advantage (and each is used by context users).
rem Unless the textadept api changes fundamentally (as happened a couple of times before) this
rem should work:
start textadept -u %~dp0 %*
rem I still need to port some of the extra functionality that we have in scite to textadept, which
rem will happen in due time. We use our own lexers because they are more efficient and have some
rem extra options (they were also much faster at that time and could handle very large files; they
rem also build on already existing code in context verbatim mode). By the way, editing char-def.lua
rem in textadept is actually now faster than in scite (using the same lpeg lexers), which is nice.
rem There is no language strip functionality yet as there is no strip (bottom area) as in scite.
rem The macros.lua file has some hard coded assumptions wrt menu items and the event crashes with a
rem error message that we can't get rid of. I need to figure out a way to close that buffer but
rem somehow the first buffer is closed anyway which is kind of weird. One way out is to just
rem comment:
rem
rem -- textadept.menu.menubar[_L['_Tools']][_L['Select Co_mmand']][2],
rem
rem Maybe I should just copy all the files and remove code we don't need but ... let's delay that
rem as it might get fixed. I'm in no hurry.
|