summaryrefslogtreecommitdiff
path: root/dviware/dvi2bitmap/TODO
blob: ffe4650ef06165ea815c56315daf2dcd06afa26a (plain)
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

Implement PostscriptImage, either using PipeStream, or using libgs
    [potentially messy, but not terribly hard]

Handle multiple DVI files on the command line.  Similarly, have it able
    to read more than one DVI file from the same source: that would
    mean that it would be possible to have dvi2bitmap sitting reading
    from a FIFO, while a succession of TeX processes write to the
    FIFO. [a bit harder than it might seem, as a variety of things
    might have to be re-initialised, and the code isn't currently
    structured with this in mind]

Have dvi2bitmap able to write one of its bitmaps to the stdout.  This
    would make it possible to regard dvi2bitmap as a filter [this
    should be pretty easy, and only depends on the output functions
    (essentially libpng) accepting file handles rather than file
    names, which libpng, at least, does]

Support PBM graphics; support PICT graphics on the Mac. [PICT would be
    a lot harder than PBM, because there's a lot more Mac apparatus
    has to be learned about and built into the build procedures.  This
    is a rather low priority, I think, and as far as PICT goes, more
    trouble than it may be worth]

Perhaps have a configuration file, rather than all those options

Improve antialiasing, if there is in fact a more sophisticated
    algorithm available than the simple scale-it-up-then-
    scale-it-down used at present.  At least add a --antialiasing
    option to abbreviate --scale + --magnification [the latter at
    least is quite trivial]

Improve BitmapImage documentation.  Improve?  Write it!

Create a process-DVI-file main loop as a function, allowing hooks to
    be registered for each of the possible events.

----------------------------------------------------------------------

Refactor runCommandPipe as PipeStream, a subclass of InputByteStream
    Done, 3 August 2003

Reimplement kpathsea as a singleton object
    Done, 3 August 2003

Have InputByteStream mmap its target file, if possible [very easy,
    really just a matter of the right bookkeeping]
    Done, 22 August 2003 (pretty easy in fact, barring a couple of gotchas)

Implement reference points in output bitmaps, reported by --output=bitmap
    Done 30 September 2003