neurolucida ascii parser extracts more fields

color, closed, guid, resolution

slider/external/neurolucida.py

@@ -41,17 +41,91 @@ import numpy as np
 # WANTED_SECTIONS = {
 #     'Asterix': 0,
 # }
+# UNWANTED_SECTION_NAMES = [
+#     'Color', 'Closed','Stereology','StereologyPropertyExtension','MBFObjectType','FillDensity', 'GUID', 'ImageCoords', 'MBFObjectType',
+#     'Marker', 'Name', 'Resolution', 'Set', 'Description',
+#     'Cross', 'Dot', 'DoubleCircle', 'FilledCircle', 'FilledDownTriangle',
+#     'FilledSquare', 'FilledStar', 'FilledUpTriangle', 'FilledUpTriangle', 'Flower',
+#     'Flower2', 'OpenCircle', 'OpenDiamond', 'OpenDownTriangle', 'OpenSquare', 'OpenStar',
+#     'OpenUpTriangle', 'Plus', 'ShadedStar', 'Splat', 'TriStar', 'Sections', 'SSM', 'SSM2',
+# ]
 UNWANTED_SECTION_NAMES = [
-    'Color', 'Closed','Stereology','StereologyPropertyExtension','MBFObjectType','FillDensity', 'GUID', 'ImageCoords', 'MBFObjectType',
-    'Marker', 'Name', 'Resolution', 'Set', 'Description',
+    'Stereology','StereologyPropertyExtension','MBFObjectType','FillDensity', 'ImageCoords', 'MBFObjectType',
+    'Marker', 'Name', 'Set', 'Description',
     'Cross', 'Dot', 'DoubleCircle', 'FilledCircle', 'FilledDownTriangle',
     'FilledSquare', 'FilledStar', 'FilledUpTriangle', 'FilledUpTriangle', 'Flower',
     'Flower2', 'OpenCircle', 'OpenDiamond', 'OpenDownTriangle', 'OpenSquare', 'OpenStar',
-    'OpenUpTriangle', 'Plus', 'ShadedStar', 'Splat', 'TriStar', 'Sections', 'SSM', 'SSM2',
+    'OpenUpTriangle', 'Plus', 'ShadedStar', 'Splat', 'TriStar', 'Sections', 'SSM', 'SSM2', 'Site'
 ]
 UNWANTED_SECTIONS = {name: True for name in UNWANTED_SECTION_NAMES}
 
 
+def is_number(s):
+    '''Test if string is a number'''
+    try:
+        float(s)
+        return True
+    except ValueError:
+        return False
+
+def is_point(p):
+    '''Test if section looks like a point section'''
+    result = True
+    if len(p) != 5: result = False
+    if not all([is_number(n) for n in p[:-1]]): result = False
+    return result
+
+def parse_point(p):
+    '''Parse a point section'''
+    return [float(p0) for p0 in p[:-1]] + [p[-1]] 
+
+
+def parse_contour(c0):
+    '''Convert contour section into a dictionary'''
+    
+    c0_dict = {
+        'name': c0[0].replace('"',''),
+    }
+
+    points = []
+
+    for p in c0[1:]:
+
+        if p[0] == 'Closed':
+            c0_dict['closed'] = True
+        elif p[0] == 'Color':
+            c0_dict['color'] = p[1]
+        elif p[0] == 'GUID':
+            c0_dict['guid'] = p[1].replace('"','')
+        elif p[0] == 'Resolution':
+            c0_dict['resolution'] = float(p[1])
+        elif is_point(p):
+            points.append(parse_point(p)[:4])
+        else:
+            raise ValueError(f'Unknown property: ({p})')
+
+
+    c0_dict['points'] = np.array(points, dtype=np.float64)
+
+    return c0_dict
+
+def parse_cell(cell):
+    '''Convert cell (asterisk) section into a dictionary'''
+
+    cell_dict = {}
+
+    for prop in cell[1:]:
+        if prop[0] == 'Color':
+            cell_dict['color'] = prop[1]
+        elif is_point(prop):
+            cell_dict['point'] = np.array(parse_point(prop)[:-1])
+        else:
+            raise ValueError(f'Unknown property: ({prop})')
+
+
+    return cell_dict
+
+
 def _match_section(section, match):
     '''checks whether the `type` of section is in the `match` dictionary
 
@@ -178,15 +252,16 @@ def _flatten_subsection(subsection, _type, offset, parent):
                     yield _row
 
 
-def _to_data(sections):
+def to_data(sections):
+    # convert sections to dicts
     cells    = []
     contour = []
     
     for section in sections:
         if section[0] == 'Asterisk':  # Cell
-            cells.append( np.asarray(section[2][:4],dtype=np.float64) )           
+            cells.append(parse_cell(section))           
         else:
-            contour.append((section[0].replace('"',''), np.asarray([s[:4] for s in section[1:]],dtype=np.float64)))
+            contour.append(parse_contour(section))
 
     cells = np.asarray(cells)
 
@@ -199,5 +274,7 @@ def read(file):
 
     with open(file, encoding='utf-8', errors='replace') as fd:
         sections = _parse_sections(fd)
-    contour, cells = _to_data(sections)
-    return contour, cells
\ No newline at end of file
+    contour, cells = to_data(sections)
+    return contour, cells
+
+