Welcome to mirror list, hosted at ThFree Co, Russian Federation.

import_ply.py « io_mesh_ply - git.blender.org/blender-addons.git - Unnamed repository; edit this file 'description' to name the repository.
summaryrefslogtreecommitdiff
blob: 1ec4c0735682ccbc3bf43f9a4b5ccabcab0989b3 (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
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
# ##### BEGIN GPL LICENSE BLOCK #####
#
#  This program is free software; you can redistribute it and/or
#  modify it under the terms of the GNU General Public License
#  as published by the Free Software Foundation; either version 2
#  of the License, or (at your option) any later version.
#
#  This program is distributed in the hope that it will be useful,
#  but WITHOUT ANY WARRANTY; without even the implied warranty of
#  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
#  GNU General Public License for more details.
#
#  You should have received a copy of the GNU General Public License
#  along with this program; if not, write to the Free Software Foundation,
#  Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
#
# ##### END GPL LICENSE BLOCK #####

# <pep8 compliant>

import re
import struct


class element_spec(object):
    __slots__ = ("name",
                 "count",
                 "properties",
                 )

    def __init__(self, name, count):
        self.name = name
        self.count = count
        self.properties = []

    def load(self, format, stream):
        if format == b'ascii':
            stream = re.split(b'\s+', stream.readline())
        return [x.load(format, stream) for x in self.properties]

    def index(self, name):
        for i, p in enumerate(self.properties):
            if p.name == name:
                return i
        return -1


class property_spec(object):
    __slots__ = ("name",
                 "list_type",
                 "numeric_type",
                 )

    def __init__(self, name, list_type, numeric_type):
        self.name = name
        self.list_type = list_type
        self.numeric_type = numeric_type

    def read_format(self, format, count, num_type, stream):
        if format == b'ascii':
            if num_type == 's':
                ans = []
                for i in range(count):
                    s = stream[i]
                    if len(s) < 2 or s[0] != '"' or s[-1] != '"':
                        print('Invalid string', s)
                        print('Note: ply_import.py does not handle whitespace in strings')
                        return None
                    ans.append(s[1:-1])
                stream[:count] = []
                return ans
            if num_type == 'f' or num_type == 'd':
                mapper = float
            else:
                mapper = int
            ans = [mapper(x) for x in stream[:count]]
            stream[:count] = []
            return ans
        else:
            if num_type == 's':
                ans = []
                for i in range(count):
                    fmt = format + 'i'
                    data = stream.read(struct.calcsize(fmt))
                    length = struct.unpack(fmt, data)[0]
                    fmt = '%s%is' % (format, length)
                    data = stream.read(struct.calcsize(fmt))
                    s = struct.unpack(fmt, data)[0]
                    ans.append(s[:-1])  # strip the NULL
                return ans
            else:
                fmt = '%s%i%s' % (format, count, num_type)
                data = stream.read(struct.calcsize(fmt))
                return struct.unpack(fmt, data)

    def load(self, format, stream):
        if self.list_type is not None:
            count = int(self.read_format(format, 1, self.list_type, stream)[0])
            return self.read_format(format, count, self.numeric_type, stream)
        else:
            return self.read_format(format, 1, self.numeric_type, stream)[0]


class object_spec(object):
    __slots__ = ("specs",
                )
    'A list of element_specs'
    def __init__(self):
        self.specs = []

    def load(self, format, stream):
        return dict([(i.name, [i.load(format, stream) for j in range(i.count)]) for i in self.specs])

        '''
        # Longhand for above LC
        answer = {}
        for i in self.specs:
            answer[i.name] = []
            for j in range(i.count):
                if not j % 100 and meshtools.show_progress:
                    Blender.Window.DrawProgressBar(float(j) / i.count, 'Loading ' + i.name)
                answer[i.name].append(i.load(format, stream))
        return answer
            '''


def read(filepath):
    format = b''
    version = b'1.0'
    format_specs = {b'binary_little_endian': '<',
                    b'binary_big_endian': '>',
                    b'ascii': b'ascii'}
    type_specs = {b'char': 'b',
                  b'uchar': 'B',
                  b'int8': 'b',
                  b'uint8': 'B',
                  b'int16': 'h',
                  b'uint16': 'H',
                  b'ushort': 'H',
                  b'int': 'i',
                  b'int32': 'i',
                  b'uint': 'I',
                  b'uint32': 'I',
                  b'float': 'f',
                  b'float32': 'f',
                  b'float64': 'd',
                  b'double': 'd',
                  b'string': 's'}
    obj_spec = object_spec()

    file = open(filepath, 'rb')  # Only for parsing the header, not binary data
    signature = file.readline()

    if not signature.startswith(b'ply'):
        print('Signature line was invalid')
        return None

    while 1:
        tokens = re.split(br'[ \r\n]+', file.readline())

        if len(tokens) == 0:
            continue
        if tokens[0] == b'end_header':
            break
        elif tokens[0] == b'comment' or tokens[0] == b'obj_info':
            continue
        elif tokens[0] == b'format':
            if len(tokens) < 3:
                print('Invalid format line')
                return None
            if tokens[1] not in format_specs:  # .keys(): # keys is implicit
                print('Unknown format', tokens[1])
                return None
            if tokens[2] != version:
                print('Unknown version', tokens[2])
                return None
            format = tokens[1]
        elif tokens[0] == b'element':
            if len(tokens) < 3:
                print(b'Invalid element line')
                return None
            obj_spec.specs.append(element_spec(tokens[1], int(tokens[2])))
        elif tokens[0] == b'property':
            if not len(obj_spec.specs):
                print('Property without element')
                return None
            if tokens[1] == b'list':
                obj_spec.specs[-1].properties.append(property_spec(tokens[4], type_specs[tokens[2]], type_specs[tokens[3]]))
            else:
                obj_spec.specs[-1].properties.append(property_spec(tokens[2], None, type_specs[tokens[1]]))

    if format != b'ascii':
        file.close()  # was ascii, now binary
        file = open(filepath, 'rb')

        # skip the header...
        while not file.readline().startswith(b'end_header'):
            pass

    obj = obj_spec.load(format_specs[format], file)

    file.close()

    return obj_spec, obj


import bpy


def load_ply(filepath):
    import time
    from bpy_extras.io_utils import unpack_list, unpack_face_list
    from bpy_extras.image_utils import load_image

    t = time.time()
    obj_spec, obj = read(filepath)
    if obj is None:
        print('Invalid file')
        return

    uvindices = colindices = None
    colmultiply = None

    # noindices = None # Ignore normals

    for el in obj_spec.specs:
        if el.name == b'vertex':
            vindices = vindices_x, vindices_y, vindices_z = el.index(b'x'), el.index(b'y'), el.index(b'z')
            # noindices = (el.index('nx'), el.index('ny'), el.index('nz'))
            # if -1 in noindices: noindices = None
            uvindices = (el.index(b's'), el.index(b't'))
            if -1 in uvindices:
                uvindices = None
            colindices = el.index(b'red'), el.index(b'green'), el.index(b'blue')
            if -1 in colindices:
                colindices = None
            else: # if not a float assume uchar
                colmultiply = [1.0 if el.properties[i].numeric_type in ('f', 'd') else (1.0 / 256.0) for i in colindices]

        elif el.name == b'face':
            findex = el.index(b'vertex_indices')

    mesh_faces = []
    mesh_uvs = []
    mesh_colors = []

    def add_face(vertices, indices, uvindices, colindices):
        mesh_faces.append(indices)
        if uvindices:
            mesh_uvs.append([(vertices[index][uvindices[0]], 1.0 - vertices[index][uvindices[1]]) for index in indices])
        if colindices:
            mesh_colors.append([(vertices[index][colindices[0]] * colmultiply[0],
                                 vertices[index][colindices[1]] * colmultiply[1],
                                 vertices[index][colindices[2]] * colmultiply[2],
                                 ) for index in indices])

    if uvindices or colindices:
        # If we have Cols or UVs then we need to check the face order.
        add_face_simple = add_face

        # EVIL EEKADOODLE - face order annoyance.
        def add_face(vertices, indices, uvindices, colindices):
            if len(indices) == 4:
                if indices[2] == 0 or indices[3] == 0:
                    indices = indices[2], indices[3], indices[0], indices[1]
            elif len(indices) == 3:
                if indices[2] == 0:
                    indices = indices[1], indices[2], indices[0]

            add_face_simple(vertices, indices, uvindices, colindices)

    verts = obj[b'vertex']

    if b'face' in obj:
        for f in obj[b'face']:
            ind = f[findex]
            len_ind = len(ind)
            if len_ind <= 4:
                add_face(verts, ind, uvindices, colindices)
            else:
                # Fan fill the face
                for j in range(len_ind - 2):
                    add_face(verts, (ind[0], ind[j + 1], ind[j + 2]), uvindices, colindices)

    ply_name = bpy.path.display_name_from_filepath(filepath)

    mesh = bpy.data.meshes.new(name=ply_name)

    mesh.vertices.add(len(obj[b'vertex']))

    mesh.vertices.foreach_set("co", [a for v in obj[b'vertex'] for a in (v[vindices_x], v[vindices_y], v[vindices_z])])

    if mesh_faces:
        mesh.faces.add(len(mesh_faces))
        mesh.faces.foreach_set("vertices_raw", unpack_face_list(mesh_faces))

        if uvindices or colindices:
            if uvindices:
                uvlay = mesh.uv_textures.new()
            if colindices:
                vcol_lay = mesh.vertex_colors.new()

            if uvindices:
                for i, f in enumerate(uvlay.data):
                    ply_uv = mesh_uvs[i]
                    for j, uv in enumerate(f.uv):
                        uv[0], uv[1] = ply_uv[j]

            if colindices:
                for i, f in enumerate(vcol_lay.data):
                    # XXX, colors dont come in right, needs further investigation.
                    ply_col = mesh_colors[i]
                    if len(ply_col) == 4:
                        f_col = f.color1, f.color2, f.color3, f.color4
                    else:
                        f_col = f.color1, f.color2, f.color3

                    for j, col in enumerate(f_col):
                        col.r, col.g, col.b = ply_col[j]

    mesh.validate()
    mesh.update()

    scn = bpy.context.scene
    #scn.objects.selected = [] # XXX25

    obj = bpy.data.objects.new(ply_name, mesh)
    scn.objects.link(obj)
    scn.objects.active = obj
    obj.select = True

    print('\nSuccessfully imported %r in %.3f sec' % (filepath, time.time() - t))


def load(operator, context, filepath=""):
    load_ply(filepath)
    return {'FINISHED'}