summaryrefslogtreecommitdiff
path: root/cad/src/exprs/scratch/dna_ribbon_view_scraps.py
blob: e61881eb9621de482374c6e7179ecbbfaa08c198 (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
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
# Copyright 2007 Nanorex, Inc.  See LICENSE file for details.
"""
dna_ribbon_view_scraps.py

$Id$

outtakes from dna_ribbon_view.py which are importable now [070201] ###UNTESTED
and might be worth finishing and testing, or salvaging useful code/ideas from.

Any old code from demo_dna-outtakes.py that was still in dna_ribbon_view.py
is now only here.
"""

# ==

from basic import *
from basic import _self

from Overlay import Overlay

from TextRect import TextRect

from OpenGL.GL import * #e move what needs this into draw_utils

from dna_ribbon_view import * ###e reloadable -- needed at least for IorE, probably more
from utilities import debug_flags
# ==

Corkscrew = Stub
class Ribbon_oldcode_for_edges(Corkscrew): # generates a sequence of rects (quads) from two corkscrews, then draws a ribbon using them
    def draw(self, **mods):
        if 1:
            # draw the ribbon-edges; looks slightly better this way in some ways, worse in other ways --
            # basically, looks good on egdes that face you, bad on edges that face away (if the ribbons had actual thickness)
            # (guess: some kluge with lighting and normals could fix this)
            Corkscrew.draw(self, color = interior_color)
            if 0:
                glTranslate(offset, 0,0)
                Corkscrew.draw(self, color = interior_color) ### maybe we should make the two edges look different, since they are (major vs minor groove)
                glTranslate(-offset, 0,0)

# ==

##class Ribbon2(Ribbon): # draws, using the same params needed for a corkscrew, a pair of ribbons that look like a DNA double helix
##    # radius, axis, turn, n
##    pass

class Rotate(IorE):#e refile with Translate -- in fact, reexpress it as interposing on draw and lbox methods, ignoring the rest...
        # (as discussed extensively elsewhere, not sure if in code or notesfile or paper, 1-3 days before 070131)
    # needs to transform other things too, like lbox -- same issue as Translate
    thing = Arg(Widget)
    angle = Arg(float)
    axis = ArgOrOption(Vector, DZ) ###e or LineSegment, then do translate too
    ###e should normalize axis and check for 0,0,0
    def draw(self):
        glRotatef(angle, axis[0], axis[1], axis[2]) # angle is in degrees, I guess
        self.drawkid( self.thing) ## self.thing.draw(self)
        glRotatef(-angle, axis[0], axis[1], axis[2]) # might be optional, I forget the semantics of things like Overlay ###k
    pass

call_lambda_Expr = Stub
lambda_Expr = Stub
ShareInstances = Stub

class Ribbon2_try1(Macro): ###e perhaps needs some axis -> axisvector
    """Ribbon2(thing1, thing2) draws a thing1 instance in red and a thing2 instance in blue.
    If thing2 is not supplied, a rotated thing1 instance is used for it, and also drawn in blue.
    """
    angle = 150.0
    arg1 = ArgExpr(Widget) # but it better have .axis if we use it to make arg2! BUT if it's an expr, how can it have that yet?
        #### are some attrs of exprs legal if they don't require instanceness to be computed, depending on the type?
        #### (typical for geometric stuff) or only if this kind of expr is "born an instance"? But if it's highlightable, it's not...
    # but it might have an axis anyway... but do we really mean arg2's axis? no... maybe a special instance of arg1 used in this expr alone? yes.
    _arg1_for_arg2 = Instance(arg1) # only used if arg2's dflt expr is needed; has to be an instance so we can ask for its .axis
        #e maybe this requirement can be relaxed since.axis does not depend on self?? not sure -- anyway it might... since it
        # might depend on local coords -- no, it's defined to be rel to local coords, so it doesn't.... ###k
    arg2 = ArgExpr(Widget, Rotate( _arg1_for_arg2, angle, _arg1_for_arg2.axis)) # in old code we passed a single Ribbon, used it twice
    delegate = Overlay(arg1(color = red), arg2(color = blue)) ###PROBLEM: too late to customize arg2 if it's an Instance!
        ## (unless this sets some state and is thereby possible on an instance... seems fishy even so.)
        # is the real intent of _arg1_for_arg2 to be an instance? what if arg2, using it, was made multiple times?
        # so no, that's the bug -- _arg1_for_arg2 is an expr which when made in those two places in arg2 dflt should be shared --
        # but it's not already an instance. It's like an ArgExpr with all ipath components added locally in this class already fixed
        # rather than coming from whatever gets built up -- as if we wrapped it when it came in and we did this,
        # then were able to strip off whatever ipath it got -- or equivly, we wrap it with "ignore outer ipath, just use this one".
        # As if I said _arg1_for_arg2 = ShareInstances(arg1) -- it gets an ipath but then ignores furtherly added ones. Hmm...
        # probably easy to implem if I decide it's right!
        #
        # [then I did Ribbon2_try2, then I extended this to say:]
        ### BUT IT'S WRONG, since if we use arg2 (expr) in several places here, how do we let each of *them* use different instances
        # of _arg1_for_arg2? We'd somehow have to say in arg2 dflt that we use the same instance of _arg1_for_arg2 for each inst of it --
        # but only for that, not for other instances of it. As if it had a quantifier for "what inst of arg1 to use in here".
        # As our lambda_Expr would make possible... ###DIGR can we let a call_Expr imply a lambda_Expr if arg1 is a lambda??
    if 'do this instead':
        dflt_arg2 = call_lambda_Expr( lambda myarg1: Rotate( myarg1, angle, myarg1.axis), arg1 )
        # or maybe a lambda_Expr is callable and doing so makes a call expr -- as opposed to being customizable (until called):
        dflt_arg2 = 0 and lambda_Expr( lambda arg1: Rotate( arg1, angle, arg1.axis))( arg1 ) # lambda is really called on an Instance
            # without '0 and', due to Stub bug, we get AssertionError: not permitted to re-supply already supplied positional args, in <Widget2D#17093(a)>

        arg2 = ArgExpr(Widget, dflt_arg2)
        _customized_arg2 = arg2(color=blue) ### PROBLEM: it might *still* be too late to customize,
        # since color=whatever won't burrow into things like localipathmod, let alone Rotate! #########
        # this current expr could only work if customizations on certain exprs would get passed into the interior
        # to be used by, or before, instantiations called for inside them. Hmm, same issue for If_expr(cond, expr1, expr2)(color=blue)
        # assuming expr[12] accept color customization. ####### if these custs became an OpExpr, eval could eval its arg1... might work.
        #
        # later, 070127: passthru customization seems unclear & hard but wanted:
        # maybe the feeling that customization ought to burrow into Rotate or localipathmod is related to them delegating to an arg --
        # since attr accesses delegate, so should attr definitions (but only for the attrs that actually delegate). That does make sense.
        # It's not easy to implement -- we have no list of attrs that don't delegate, we just try getattr and see if we get into __getattr__.
        # So to customize an attr -- wait, would it work to customize it at the outer level? No, because the draw call occurs inside.
        # It's as if we wanted to constrain X and Rotate(X) to have the same def and value of attr (if it delegates), e.g. color.
        # But the fact that X.draw uses X.color is not even dependent on whether Rotate delegates color! Ignoring that last point,
        # can customize mean "find the internal users of the def of color, and make them use this different def"? These internal users
        # are kid instances. They are self-contained in ability to use color. So parent would have to modify them -- presumably when it makes them.
        # So parent(color = cust) would have to instantiate, not its normal kid delegate, but kid(color = cust), I guess. I don't see an
        # easy implem, and I'm also suspicious it's the right idea.

        # So maybe the only soln is to do the color cust first, as the docstring had to say to describe the intent, anyway.
    pass

class Ribbon2_try2(Macro):###e perhaps needs some axis -> axisvector
    ###IMPLEM ShareInstances  - or not, if that lambda_Expr in _try1 makes it unneeded... it might be wrong anyway
    # if it provides no way to limit the sharing within the class that says to do it, or if doing that is too cumbersome or unclear.
    """Ribbon2(thing1, thing2) draws a thing1 instance in red and a thing2 instance in blue.
    If thing2 is not supplied, a rotated thing1 instance is used for it, and also drawn in blue.
    """
    angle = 150.0
    arg1 = ArgExpr(Widget)
    _drawn_arg1 = Instance(arg1(color=red)) ###e digr: could I have said Instance(arg1, color=red)? I bet people will try that...
    _arg1_for_arg2 = ShareInstances(arg1(color=blue)) # still an expr; in fact, this ASSUMES arg1 is passed in as an expr, not as an instance! Hmm#####
    ## arg2 = ArgExpr(Widget, Rotate( _arg1_for_arg2, angle, _arg1_for_arg2.axis))
    arg2 = ArgExpr(Widget, Rotate( _arg1_for_arg2, angle, getattr_Expr( _arg1_for_arg2, 'axis') ))
    _drawn_arg2 = Instance(arg2(color=blue)) ####KLUGE: add color here in case arg2 was supplied, but in dflt expr in case we used that
    delegate = Overlay(_drawn_arg1, _drawn_arg2)

class Ribbon2_try3(Macro): #070129 ###e perhaps needs some axis -> axisvector
    """Ribbon2(thing1, thing2) draws a thing1 instance in red and a thing2 instance in blue, assuming they are color-customizable.
    If thing2 is not supplied, a rotated thing1 instance (different drawable instance, same model object instance)
    is used for it (also drawn in blue).
    """
    angle = 150.0
    arg1 = Arg(ModelObject3D) ###k type implies that it has 3d shape etc, determines how much gets instantiated here (axis, not color??)
    arg2 = Arg(ModelObject3D, Rotate(arg1, angle, arg1.axis) ) # doesn't yet work... see below
        # Rotate works on a ModelObject3D instance which is like a Drawable expr
        # note: Rotate's default axis is the Z axis, not arg1.axis even if that exists
    delegate = Overlay(arg1(color = red), arg2(color = blue)) # doesn't yet work... see below
        # For all this to work requires two proposed semantic changes:
        # - Rotate delegation more like customization -- instance of Rotate produces no separate instance of its delegate
        # - partial instantiation by Arg, only of attrs that are part of its declared type (mentioned above)
        # (these changes make it likely that the 'delegate' attr will be renamed, too)
    pass


# very old cmt:
# Ribbon2 has: radius, axis (some sort of length - of one bp?), turn (angle??), n, color1, color2, and full position/orientation
# and it will have display modes, incl cyl with helix/base/groove texture, and flexible coloring;
# and ability to show the units in it, namely strands, or basepairs, or bases (selected/opd by type)
# and for us to add cmd bindings like "make neighbor strand" (if room) or "make crossover here" (maybe only on a base?)

# but as a first step, we can select it as a unit, copy/paste, deposit, move, etc.
# In this, it is sort of like an atom or set of co-selected atoms... some relation to chunk or jig too.
# I'd get best results by letting it be its own thing... but fastest, by borrowing one of those...


class obs:
    path = StateArg(Cylinder_HelicalPath)
    path.someattr = Option(Something, 'dflt') # set path.someattr to Option(...) -- ExprsMeta scanner would need to see this --
        # but what attrname would it use in the Option?? maybe the _try1 version is better since it says the name,
        # or maybe you can give it in Option as string arg1.

class todo:
    # might need other state, like some colors

    # and links to things made from this guide shape -- or a superclass or whatever that says we are a guide shape
    # (and all of them can have links like that)

    # and ops to make attached things like crossovers, to give posns for potential ones
    # and display styles for self and those things...
    # for now just have default drawing code, using the Ribbon classes above.
    pass

# ==

import Command_scratch_1

# ==

if 0:
    # old example usage
    Ribbon2(1, 0.2, 1/10.5, 50, blue, color2 = green), # this color2 arg stuff is a kluge
    Highlightable( Ribbon2(1, 0.2, 1/10.5, 50, yellow, color2 = red), sbar_text = "bottom ribbon2" ),

# end