added support for almost all missing DXF features - incl. 3D models

added a FontCache object for incremental font loading (useful for DXF files with text elements)


git-svn-id: https://pycam.svn.sourceforge.net/svnroot/pycam/trunk@1017 bbaffbd6-741e-11dd-a85d-61de82d9cad9

Changelog

@@ -11,7 +11,9 @@ Version 0.5 - UNRELEASED
  * added automatic repair of inconsistent polygon winding (inside/outside detection)
  * added toolpath cropping 
  * added toolpath grid pattern: clone a single toolpath in rows and columns
- * added support for DXF features "LWPOLYLINE" and "ARC"
+ * added support for more DXF features:
+  * 2D: "LWPOLYLINE", "ARC", "TEXT", "MTEXT", "CIRCLE", "POLYLINE"
+  * 3D: "3DFACE"
  * added a configuration setting for automatically loading a custom task settings file on startup
  * added a simple "undo" feature for reversing model manipulations
  * GCode features:

src/pycam/Geometry/Matrix.py

@@ -39,6 +39,9 @@ TRANSFORMATIONS = {
     "x_swap_y": ((0, 1, 0, 0), (1, 0, 0, 0), (0, 0, 1, 0)),
     "x_swap_z": ((0, 0, 1, 0), (0, 1, 0, 0), (1, 0, 0, 0)),
     "y_swap_z": ((1, 0, 0, 0), (0, 0, 1, 0), (0, 1, 0, 0)),
+    "xy_mirror": ((1, 0, 0, 0), (0, 1, 0, 0), (0, 0, -1, 0)),
+    "xz_mirror": ((1, 0, 0, 0), (0, -1, 0, 0), (0, 0, 1, 0)),
+    "yz_mirror": ((-1, 0, 0, 0), (0, 1, 0, 0), (0, 0, 1, 0)),
 }
 
 

src/pycam/Geometry/Model.py

@@ -481,7 +481,7 @@ class ContourModel(BaseModel):
             if progress_callback and progress_callback():
                 self.reset_cache()
                 return
-        log.info("The winding of %d polygons was fixed." % change_counter)
+        log.info("The winding of %d polygon(s) was fixed." % change_counter)
         self.reset_cache()
 
     def reverse_directions(self, callback=None):

src/pycam/Geometry/Point.py

@@ -58,8 +58,10 @@ class Point(object):
         Otherwise the result is based on the individual x/y/z comparisons.
         """
         if self.__class__ == other.__class__:
-            if (_is_near(self.x, other.x)) and (_is_near(self.y, other.y)) \
-                    and (_is_near(self.z, other.z)):
+            if (self.id == other.id) or \
+                    ((_is_near(self.x, other.x)) and \
+                        (_is_near(self.y, other.y)) and \
+                        (_is_near(self.z, other.z))):
                 return 0
             elif not _is_near(self.x, other.x):
                 return cmp(self.x, other.x)
@@ -71,12 +73,19 @@ class Point(object):
             return cmp(str(self), str(other))
 
     def transform_by_matrix(self, matrix, transformed_list=None, callback=None):
+        # accept 3x4 matrices as well as 3x3 matrices
+        offsets = []
+        for column in matrix:
+            if len(column) < 4:
+                offsets.append(0)
+            else:
+                offsets.append(column[3])
         x = self.x * matrix[0][0] + self.y * matrix[0][1] \
-                + self.z * matrix[0][2] + matrix[0][3]
+                + self.z * matrix[0][2] + offsets[0]
         y = self.x * matrix[1][0] + self.y * matrix[1][1] \
-                + self.z * matrix[1][2] + matrix[1][3]
+                + self.z * matrix[1][2] + offsets[1]
         z = self.x * matrix[2][0] + self.y * matrix[2][1] \
-                + self.z * matrix[2][2] + matrix[2][3]
+                + self.z * matrix[2][2] + offsets[2]
         self.x = x
         self.y = y
         self.z = z

src/pycam/Gui/Project.py

@@ -42,6 +42,7 @@ from pycam.Geometry.Letters import TEXT_ALIGN_LEFT, TEXT_ALIGN_CENTER, \
 import pycam.Geometry.Model
 from pycam.Utils import ProgressCounter, check_uri_exists
 from pycam.Toolpath import Bounds
+import pycam.Utils.FontCache
 from pycam import VERSION
 import pycam.Physics.ode_physics
 # this requires ODE - we import it later, if necessary
@@ -100,6 +101,11 @@ FILTER_MODEL = (("All supported model filetypes",
 FILTER_CONFIG = (("Config files", "*.conf"),)
 FILTER_EMC_TOOL = (("EMC tool files", "*.tbl"),)
 
+CLIPBOARD_TARGETS = {
+        "svg": ("image/x-inkscape-svg", "image/svg+xml"),
+        "filename_drag": ("text/uri-list", "text-plain"),
+}
+
 PREFERENCES_DEFAULTS = {
         "enable_ode": False,
         "boundary_mode": -1,
@@ -237,37 +243,6 @@ def get_font_dir():
                     + "No fonts will be available.") % FONT_DIR_FALLBACK)
             return None
 
-def get_font_files():
-    font_dir = get_font_dir()
-    if not font_dir:
-        return []
-    log.info("Loading font files from '%s'." % font_dir)
-    result = []
-    files = os.listdir(font_dir)
-    for fname in files:
-        filename = os.path.join(font_dir, fname)
-        if filename.lower().endswith(".cxf") and os.path.isfile(filename):
-            result.append(filename)
-    result.sort()
-    return result
-
-def load_fonts(callback=None):
-    fonts = {}
-    all_font_files = get_font_files()
-    if not callback is None:
-        progress_counter = ProgressCounter(len(all_font_files), callback)
-    else:
-        progress_counter = None
-    for font_file in all_font_files:
-        charset = pycam.Importers.CXFImporter.import_font(font_file,
-                callback=progress_counter.update)
-        if (not progress_counter is None) and progress_counter.increment():
-            break
-        if not charset is None:
-            for name in charset.get_names():
-                fonts[name] = charset
-    return fonts
-
 
 class ProjectGui:
 
@@ -295,6 +270,9 @@ class ProjectGui:
         self._progress_cancel_requested = False
         self._last_gtk_events_time = None
         self._undo_states = []
+        self.clipboard = gtk.clipboard_get()
+        self._fonts_cache = pycam.Utils.FontCache.FontCache(get_font_dir(),
+                callback=self.update_progress_bar)
         self.gui = gtk.Builder()
         gtk_build_file = get_data_file_location(GTKBUILD_FILE)
         if gtk_build_file is None:
@@ -451,7 +429,6 @@ class ProjectGui:
                     self.update_font_dialog_preview)
         self._font_dialog_window_visible = False
         self._font_dialog_window_position = None
-        self._fonts = None
         # set defaults
         self.model = None
         self.toolpath = pycam.Toolpath.ToolpathList()
@@ -1945,14 +1922,13 @@ class ProjectGui:
         if state is None:
             state = not self._font_dialog_window_visible
         if state:
-            if self._fonts is None:
+            if not self._fonts_cache.is_loading_complete():
                 self.update_progress_bar("Initializing fonts")
-                self._fonts = load_fonts(callback=self.update_progress_bar)
                 # create it manually to ease access
                 font_selector = gtk.combo_box_new_text()
                 self.gui.get_object("FontSelectionBox").pack_start(
                         font_selector, expand=False, fill=False)
-                sorted_keys = self._fonts.keys()
+                sorted_keys = list(self._fonts_cache.get_font_names())
                 sorted_keys.sort(key=lambda x: x.upper())
                 for name in sorted_keys:
                     font_selector.append_text(name)
@@ -1964,7 +1940,7 @@ class ProjectGui:
                         self.update_font_dialog_preview)
                 font_selector.show()
                 self.font_selector = font_selector
-            if self._fonts:
+            if len(self._fonts_cache) > 0:
                 # show the dialog only if fonts are available
                 if self._font_dialog_window_position:
                     self.font_dialog_window.move(
@@ -2002,8 +1978,9 @@ class ProjectGui:
                     align = value
                     input_field.set_justification(justification)
             font_name = self.font_selector.get_active_text()
-            return self._fonts[font_name].render(text, skew=skew,
-                    line_spacing=line_space, pitch=pitch, align=align)
+            charset = self._fonts_cache.get_font(font_name)
+            return charset.render(text, skew=skew, line_spacing=line_space,
+                    pitch=pitch, align=align)
         else:
             # empty text
             return None
@@ -2024,23 +2001,27 @@ class ProjectGui:
             text_buffer = StringIO.StringIO()
             text_model.export(comment=self.get_meta_data(),
                     unit=self.settings.get("unit")).write(text_buffer)
-            clipboard_target = "image/svg+xml"
-            clipboard = gtk.clipboard_get()
-            targets = [(clipboard_target, gtk.TARGET_OTHER_WIDGET, 0)]
-            def get_func(clipboard, selectiondata, info, text):
-                text.seek(0)
-                selectiondata.set("STRING", 8, text.read())
-            result = clipboard.set_with_data(targets, get_func,
-                    lambda *args: None, text_buffer)
-            clipboard.store()
+            text_buffer.seek(0)
+            text = text_buffer.read()
+            self._copy_text_to_clipboard(text, CLIPBOARD_TARGETS["svg"])
+
+    def _copy_text_to_clipboard(self, text, targets):
+        targets = [(key, gtk.TARGET_OTHER_WIDGET, index)
+                for index, key in enumerate(targets)]
+        def get_func(clipboard, selectiondata, info, text):
+            # Inkscape for Windows strictly requires the BITMAP type
+            selectiondata.set(gtk.gdk.SELECTION_TYPE_BITMAP, 8, text.read())
+        result = self.clipboard.set_with_data(targets, get_func,
+                lambda *args: None, text)
+        self.clipboard.store()
 
     @gui_activity_guard
     def update_font_dialog_preview(self, widget=None, event=None):
-        if not self._fonts:
-            # not initialized or empty
+        if len(self._fonts_cache) == 0:
+            # empty
             return
         font_name = self.font_selector.get_active_text()
-        font = self._fonts[font_name]
+        font = self._fonts_cache.get_font(font_name)
         self.gui.get_object("FontAuthorText").set_label(
                 "\n".join(font.get_authors()))
         preview_widget = self.gui.get_object("FontDialogPreview")
@@ -2158,8 +2139,7 @@ class ProjectGui:
                 columns.append(model.get_value(it, column))
             content.append(" ".join(columns))
         self.log_model.foreach(copy_row, content)
-        clipboard = gtk.Clipboard()
-        clipboard.set_text(os.linesep.join(content))
+        self.clipboard.set_text(os.linesep.join(content))
         self.gui.get_object("StatusBarWarning").hide()
 
     @gui_activity_guard
@@ -2932,8 +2912,8 @@ class ProjectGui:
     def configure_drag_and_drop(self, obj):
         obj.connect("drag-data-received", self.handle_data_drop)
         flags = gtk.DEST_DEFAULT_ALL
-        targets = [("text/uri-list", gtk.TARGET_OTHER_APP, 0),
-                ("text/plain", gtk.TARGET_OTHER_APP, 1)]
+        targets = [(key, gtk.TARGET_OTHER_APP, index)
+                for index, key in enumerate(CLIPBOARD_TARGETS["filename_drag"])]
         actions = gtk.gdk.ACTION_COPY | gtk.gdk.ACTION_LINK | \
                 gtk.gdk.ACTION_DEFAULT | gtk.gdk.ACTION_PRIVATE | \
                 gtk.gdk.ACTION_ASK
@@ -2998,6 +2978,7 @@ class ProjectGui:
                 if self.load_model(importer(filename,
                         program_locations=program_locations,
                         unit=self.settings.get("unit"),
+                        fonts_cache=self._fonts_cache,
                         callback=self.update_progress_bar)):
                     self.set_model_filename(filename)
                     self.add_to_recent_file_list(filename)
@@ -3565,7 +3546,7 @@ class ProjectGui:
             self.menubar.set_sensitive(False)
             self.task_pane.set_sensitive(False)
             self._progress_start_time = time.time()
-            self.update_progress_bar("", 0)
+            self.update_progress_bar(text="", percent=0)
             self.progress_cancel_button.set_sensitive(True)
             # enable "pulse" mode for a start (in case of unknown ETA)
             self.progress_bar.pulse()
@@ -3585,7 +3566,7 @@ class ProjectGui:
         if not percent is None:
             percent = min(max(percent, 0.0), 100.0)
             self.progress_bar.set_fraction(percent/100.0)
-        if (percent is None) and (self.progress_bar.get_fraction() == 0):
+        if (not percent) and (self.progress_bar.get_fraction() == 0):
             # use "pulse" mode until we reach 1% of the work to be done
             self.progress_bar.pulse()
         # update the GUI

src/pycam/Importers/DXFImporter.py

@@ -20,16 +20,32 @@ You should have received a copy of the GNU General Public License
 along with PyCAM.  If not, see <http://www.gnu.org/licenses/>.
 """
 
+from pycam.Geometry.Triangle import Triangle
 from pycam.Geometry.Point import Point
 from pycam.Geometry.Line import Line
 import pycam.Geometry.Model
+import pycam.Geometry.Matrix
 import pycam.Geometry
 import pycam.Utils.log
 from pycam.Utils import open_url
+import math
+import re
+import os
 
 log = pycam.Utils.log.get_logger()
 
 
+def _unescape_control_characters(text):
+    # see http://www.kxcad.net/autodesk/autocad/AutoCAD_2008_Command_Reference/d0e73428.htm
+    # and QCad: qcadlib/src/filters/rs_filterdxf.cpp
+    for src, dest in (("%%d", u"\u00B0"), ("%%p", u"\u00B1"),
+            ("%%c", u"\u2205"), (r"\P", os.linesep), (r"\~", " ")):
+        text = text.replace(src, dest)
+    # convert "\U+xxxx" to unicode characters
+    return re.sub(r"\\U\+([0-9a-fA-F]{4})",
+            lambda hex_in: unichr(int(hex_in.groups()[0], 16)), text)
+
+
 class DXFParser(object):
 
     # see http://www.autodesk.com/techpubs/autocad/acad2000/dxf/group_code_value_types_dxf_01.htm
@@ -37,30 +53,63 @@ class DXFParser(object):
 
     KEYS = {
         "MARKER": 0,
-        "START_X": 10,
-        "START_Y": 20,
-        "START_Z": 30,
-        "END_X": 11,
-        "END_Y": 21,
-        "END_Z": 31,
+        "DEFAULT": 1,
+        "TEXT_MORE": 3,
+        "TEXT_FONT": 7,
+        "P1_X": 10,
+        "P1_Y": 20,
+        "P1_Z": 30,
+        "P2_X": 11,
+        "P2_Y": 21,
+        "P2_Z": 31,
+        "P3_X": 12,
+        "P3_Y": 22,
+        "P3_Z": 32,
+        "P4_X": 13,
+        "P4_Y": 23,
+        "P4_Z": 33,
         "RADIUS": 40,
+        "TEXT_HEIGHT": 40,
+        "TEXT_WIDTH_FINAL": 41,
         "ANGLE_START": 50,
+        "TEXT_ROTATION": 50,
         "ANGLE_END": 51,
+        "TEXT_SKEW_ANGLE": 51,
         "COLOR": 62,
+        "TEXT_MIRROR_FLAGS": 71,
+        "MTEXT_ALIGNMENT": 71,
+        "TEXT_ALIGN_HORIZONTAL": 72,
+        "TEXT_ALIGN_VERTICAL": 73,
     }
 
-    def __init__(self, inputstream, color_as_height=False, callback=None):
+    IGNORE_KEYS = ("DICTIONARY", "VPORT", "LTYPE", "STYLE", "APPID", "DIMSTYLE",
+            "BLOCK_RECORD", "BLOCK", "ENDBLK", "ACDBDICTIONARYWDFLT", "POINT",
+            "ACDBPLACEHOLDER", "LAYOUT", "MLINESTYLE", "DICTIONARYVAR", "CLASS",
+            "HATCH")
+
+    def __init__(self, inputstream, color_as_height=False, fonts_cache=None,
+            callback=None):
         self.inputstream = inputstream
         self.line_number = 0
         self.lines = []
+        self.triangles = []
         self._input_stack = []
         self._color_as_height = color_as_height
-        self.callback = callback
+        if callback:
+            # no "percent" updates - just pulse ...
+            callback_wrapper = lambda text="", percent=None: callback()
+            self.callback = callback_wrapper
+        else:
+            self.callback = None
+        self._fonts_cache = fonts_cache
+        self._open_sequence = None
+        self._open_sequence_items = []
+        # run the parser
         self.parse_content()
         self.optimize_line_order()
 
     def get_model(self):
-        return {"lines": self.lines}
+        return {"lines": self.lines, "triangles": self.triangles}
 
     def optimize_line_order(self):
         groups = []
@@ -132,9 +181,10 @@ class DXFParser(object):
             log.warn("DXFImporter: Invalid key in line " \
                     + "%d (int expected): %s" % (self.line_number, line1))
             return None, None
-        if line1 in [self.KEYS[key] for key in ("START_X", "START_Y", "START_Z",
-                "END_X", "END_Y", "END_Z", "RADIUS", "ANGLE_START",
-                "ANGLE_END")]:
+        if line1 in [self.KEYS[key] for key in ("P1_X", "P1_Y", "P1_Z",
+                "P2_X", "P2_Y", "P2_Z", "RADIUS", "ANGLE_START", "ANGLE_END",
+                "TEXT_HEIGHT", "TEXT_WIDTH_FINAL", "TEXT_ROTATION",
+                "TEXT_SKEW_ANGLE")]:
             try:
                 line2 = float(line2)
             except ValueError:
@@ -142,7 +192,9 @@ class DXFParser(object):
                         + "%d (float expected): %s" % (self.line_number, line2))
                 line1 = None
                 line2 = None
-        elif line1 in (self.KEYS["COLOR"],):
+        elif line1 in [self.KEYS[key] for key in ("COLOR", "TEXT_MIRROR_FLAGS",
+                "TEXT_ALIGN_HORIZONTAL", "TEXT_ALIGN_VERTICAL",
+                "MTEXT_ALIGNMENT")]:
             try:
                 line2 = int(line2)
             except ValueError:
@@ -150,6 +202,8 @@ class DXFParser(object):
                         + "%d (int expected): %s" % (self.line_number, line2))
                 line1 = None
                 line2 = None
+        elif line1 in [self.KEYS[key] for key in ("DEFAULT", "TEXT_MORE")]:
+            line2 = _unescape_control_characters(line2)
         else:
             line2 = line2.upper()
         self.line_number += 2
@@ -168,16 +222,93 @@ class DXFParser(object):
                 elif value == "LINE":
                     self.parse_line()
                 elif value == "LWPOLYLINE":
-                    self.parse_polyline()
+                    self.parse_lwpolyline()
+                elif value == "POLYLINE":
+                    self.parse_polyline(True)
+                elif value == "VERTEX":
+                    self.parse_vertex()
+                elif value == "SEQEND":
+                    self.close_sequence()
                 elif value == "ARC":
                     self.parse_arc()
+                elif value == "CIRCLE":
+                    self.parse_arc(circle=True)
+                elif value == "TEXT":
+                    self.parse_text()
+                elif value == "MTEXT":
+                    self.parse_mtext()
+                elif value == "3DFACE":
+                    self.parse_3dface()
+                elif value in self.IGNORE_KEYS:
+                    log.debug("DXFImporter: Ignored a blacklisted element " \
+                            + "in line %d: %s" % (self.line_number, value))
                 else:
                     # not supported
-                    log.warn("DXFImporter: Ignored unsupported element in " \
-                            + "line %d: %s" % (self.line_number, value))
+                    log.warn("DXFImporter: Ignored unsupported element " \
+                            + "in line %d: %s" % (self.line_number, value))
             key, value = self._read_key_value()
 
-    def parse_polyline(self):
+    def close_sequence(self):
+        start_line = self.line_number
+        if self._open_sequence == "POLYLINE":
+            self.parse_polyline(False)
+        else:
+            log.warn("DXFImporter: unexpected SEQEND found at line %d" + \
+                    start_line)
+
+    def parse_vertex(self):
+        start_line = self.line_number
+        point = [None, None, 0]
+        color = None
+        key, value = self._read_key_value()
+        while (not key is None) and (key != self.KEYS["MARKER"]):
+            if key == self.KEYS["P1_X"]:
+                point[0] = value
+            elif key == self.KEYS["P1_Y"]:
+                point[1] = value
+            elif key == self.KEYS["P1_Z"]:
+                point[2] = value
+            elif key == self.KEYS["COLOR"]:
+                color = value
+            else:
+                pass
+            key, value = self._read_key_value()
+        end_line = self.line_number
+        if not key is None:
+            self._push_on_stack(key, value)
+        if self._color_as_height and (not color is None):
+            # use the color code as the z coordinate
+            point[2] = float(color) / 255
+        if None in point:
+            log.warn("DXFImporter: Missing attribute of VERTEX item" + \
+                    "between line %d and %d" % (start_line, end_line))
+        else:
+            self._open_sequence_items.append(
+                    Point(point[0], point[1], point[2]))
+
+    def parse_polyline(self, init):
+        start_line = self.line_number
+        if init:
+            self._open_sequence = "POLYLINE"
+            self._open_sequence_items = []
+            key, value = self._read_key_value()
+            while (not key is None) and (key != self.KEYS["MARKER"]):
+                key, value = self._read_key_value()
+            if not key is None:
+                self._push_on_stack(key, value)
+        else:
+            # closing
+            points = self._open_sequence_items
+            for index in range(len(points) - 1):
+                point = points[index]
+                next_point = points[index + 1]
+                if point != next_point:
+                    self.lines.append(Line(point, next_point))
+            self._open_sequence_items = []
+            self._open_sequence = None
+
+    def parse_lwpolyline(self):
+        start_line = self.line_number
         points = []
         def add_point(p_array):
             # fill all "None" values with zero
@@ -189,15 +320,14 @@ class DXFParser(object):
                                 (self.line_number, p_array))
                     p_array[index] = 0
             points.append(Point(p_array[0], p_array[1], p_array[2]))
-        start_line = self.line_number
         current_point = [None, None, None]
         key, value = self._read_key_value()
         while (not key is None) and (key != self.KEYS["MARKER"]):
-            if key == self.KEYS["START_X"]:
+            if key == self.KEYS["P1_X"]:
                 axis = 0
-            elif key == self.KEYS["START_Y"]:
+            elif key == self.KEYS["P1_Y"]:
                 axis = 1
-            elif not self._color_as_height and (key == self.KEYS["START_Z"]):
+            elif not self._color_as_height and (key == self.KEYS["P1_Z"]):
                 axis = 2
             elif self._color_as_height and (key == self.KEYS["COLOR"]):
                 # interpret the color as the height
@@ -238,6 +368,352 @@ class DXFParser(object):
                             + "(between input line %d and %d): %s" \
                             % (start_line, end_line, point))
 
+    def parse_mtext(self):
+        start_line = self.line_number
+        # the z-level defaults to zero (for 2D models)
+        ref_point = [None, None, 0]
+        direction_vector = [None, None, None]
+        color = None
+        text_groups_start = []
+        text_end = []
+        text_height = None
+        rotation = 0
+        width_final = None
+        font_name = "normal"
+        alignment = 0
+        key, value = self._read_key_value()
+        while (not key is None) and (key != self.KEYS["MARKER"]):
+            if key == self.KEYS["DEFAULT"]:
+                text_end = value
+            elif key == self.KEYS["TEXT_MORE"]:
+                text_groups_start.append(value)
+            elif key == self.KEYS["P1_X"]:
+                ref_point[0] = value
+            elif key == self.KEYS["P1_Y"]:
+                ref_point[1] = value
+            elif key == self.KEYS["P1_Z"]:
+                ref_point[2] = value
+            elif key == self.KEYS["P2_X"]:
+                direction_vector[0] = value
+                # according to DXF spec: the last one wins
+                rotation = None
+            elif key == self.KEYS["P2_Y"]:
+                direction_vector[1] = value
+                # according to DXF spec: the last one wins
+                rotation = None
+            elif key == self.KEYS["P2_Z"]:
+                direction_vector[2] = value
+                # according to DXF spec: the last one wins
+                rotation = None
+            elif key == self.KEYS["COLOR"]:
+                color = value
+            elif key == self.KEYS["TEXT_HEIGHT"]:
+                text_height = value
+            elif key == self.KEYS["TEXT_ROTATION"]:
+                rotation = value
+                # according to DXF spec: the last one wins
+                direction_vector = [None, None, None]
+            elif key == self.KEYS["TEXT_FONT"]:
+                font_name = value
+            elif key == self.KEYS["MTEXT_ALIGNMENT"]:
+                alignment = value
+            elif key == self.KEYS["TEXT_WIDTH_FINAL"]:
+                width_final = value
+            else:
+                pass
+            key, value = self._read_key_value()
+        end_line = self.line_number
+        # The last lines were not used - they are just the marker for the next
+        # item.
+        text = "".join(text_groups_start) + text_end
+        if not key is None:
+            self._push_on_stack(key, value)
+        if None in ref_point:
+            log.warn("DXFImporter: Incomplete MTEXT definition between line " \
+                    + "%d and %d: missing location point" % \
+                    (start_line, end_line))
+        elif not text:
+            log.warn("DXFImporter: Incomplete MTEXT definition between line " \
+                    + "%d and %d: missing text" % (start_line, end_line))
+        elif not text_height:
+            log.warn("DXFImporter: Incomplete MTEXT definition between line " \
+                    + "%d and %d: missing height" % (start_line, end_line))
+        else:
+            if self._color_as_height and (not color is None):
+                # use the color code as the z coordinate
+                ref_point[2] = float(color) / 255
+            if self._fonts_cache:
+                font = self._fonts_cache.get_font(font_name)
+            else:
+                font = None
+            if not font:
+                log.warn("DXFImporter: No fonts are available - skipping " + \
+                        "MTEXT item between line %d and %d" % \
+                        (start_line, end_line))
+                return
+            model = font.render(text)
+            if (model.minx is None) or (model.miny is None) or \
+                    (model.minz is None) or (model.minx == model.maxx) or \
+                    (model.miny == model.maxy):
+                log.warn("DXFImporter: Empty rendered MTEXT item between " + \
+                        "line %d and %d" % (start_line, end_line))
+                return
+            model.scale(text_height / (model.maxy - model.miny),
+                    callback=self.callback)
+            # this setting seems to refer to a box - not the text width - ignore
+            if False and width_final:
+                scale_x = width_final / (model.maxx - model.minx)
+                model.scale(scale_x, callback=self.callback)
+            if rotation:
+                matrix = pycam.Geometry.Matrix.get_rotation_matrix_axis_angle(
+                        (0, 0, 1), rotation)
+            elif not None in direction_vector:
+                # Due to the parsing code above only "rotation" or
+                # "direction_vector" is set at the same time.
+                matrix = pycam.Geometry.Matrix.get_rotation_matrix_from_to(
+                        (1, 0, 0), direction_vector)
+            else:
+                matrix = None
+            if matrix:
+                model.transform_by_matrix(matrix, callback=self.callback)
+            # horizontal alignment
+            if alignment % 3 == 1:
+                offset_horiz = 0
+            elif alignment % 3 == 2:
+                offset_horiz = -(model.maxx - model.minx) / 2
+            else:
+                offset_horiz = -(model.maxx - model.minx)
+            # vertical alignment
+            if alignment <= 3:
+                offset_vert = -(model.maxy - model.miny)
+            elif alignment <= 6:
+                offset_vert = -(model.maxy - model.miny) / 2
+            else:
+                offset_vert = 0
+            # shift the text to its final destination
+            shift_x = ref_point[0] - model.minx + offset_horiz
+            shift_y = ref_point[1] - model.miny + offset_vert
+            shift_z = ref_point[2] - model.minz
+            model.shift(shift_x, shift_y, shift_z, callback=self.callback)
+            for polygon in model.get_polygons():
+                for line in polygon.get_lines():
+                    self.lines.append(line)
+
+    def parse_text(self):
+        start_line = self.line_number
+        # the z-level defaults to zero (for 2D models)
+        ref_point = [None, None, 0]
+        ref_point2 = [None, None, 0]
+        color = None
+        text = None
+        text_height = None
+        rotation = 0
+        width_final = None
+        skew_angle = 0
+        font_name = "normal"
+        mirror_flags = 0
+        align_horiz = 0
+        align_vert = 0
+        key, value = self._read_key_value()
+        while (not key is None) and (key != self.KEYS["MARKER"]):
+            if key == self.KEYS["DEFAULT"]:
+                text = value
+            elif key == self.KEYS["P1_X"]:
+                ref_point[0] = value
+            elif key == self.KEYS["P1_Y"]:
+                ref_point[1] = value
+            elif key == self.KEYS["P1_Z"]:
+                ref_point[2] = value
+            elif key == self.KEYS["P2_X"]:
+                ref_point2[0] = value
+            elif key == self.KEYS["P2_Y"]:
+                ref_point2[1] = value
+            elif key == self.KEYS["P2_Z"]:
+                ref_point2[2] = value
+            elif key == self.KEYS["COLOR"]:
+                color = value
+            elif key == self.KEYS["TEXT_HEIGHT"]:
+                text_height = value
+            elif key == self.KEYS["TEXT_ROTATION"]:
+                rotation = value
+            elif key == self.KEYS["TEXT_SKEW_ANGLE"]:
+                skew_angle = value
+            elif key == self.KEYS["TEXT_FONT"]:
+                font_name = value
+            elif key == self.KEYS["TEXT_MIRROR_FLAGS"]:
+                mirror_flags = value
+            elif key == self.KEYS["TEXT_ALIGN_HORIZONTAL"]:
+                align_horiz = value
+            elif key == self.KEYS["TEXT_ALIGN_VERTICAL"]:
+                align_vert = value
+            elif key == self.KEYS["TEXT_WIDTH_FINAL"]:
+                width_final = value
+            else:
+                pass
+            key, value = self._read_key_value()
+        end_line = self.line_number
+        # The last lines were not used - they are just the marker for the next
+        # item.
+        if not key is None:
+            self._push_on_stack(key, value)
+        if (None in ref_point2) and (ref_point != ref_point2):
+            # just a warning - continue as usual
+            log.warn("DXFImporter: Second alignment point is not " + \
+                    "implemented for TEXT items - the text specified " + \
+                    "between line %d and %d may be slightly misplaced" % \
+                    (start_line, end_line))
+        if None in ref_point:
+            log.warn("DXFImporter: Incomplete TEXT definition between line " \
+                    + "%d and %d: missing location point" % \
+                    (start_line, end_line))
+        elif not text:
+            log.warn("DXFImporter: Incomplete TEXT definition between line " \
+                    + "%d and %d: missing text" % (start_line, end_line))
+        elif not text_height:
+            log.warn("DXFImporter: Incomplete TEXT definition between line " \
+                    + "%d and %d: missing height" % (start_line, end_line))
+        else:
+            if self._color_as_height and (not color is None):
+                # use the color code as the z coordinate
+                ref_point[2] = float(color) / 255
+            if self._fonts_cache:
+                font = self._fonts_cache.get_font(font_name)
+            else:
+                font = None
+            if not font:
+                log.warn("DXFImporter: No fonts are available - skipping " + \
+                        "TEXT item between line %d and %d" % \
+                        (start_line, end_line))
+                return
+            if skew_angle:
+                # calculate the "skew" factor
+                if (skew_angle <= -90) or (skew_angle >= 90):
+                    log.warn("DXFImporter: Invalid skew angle for TEXT " + \
+                            "between line %d and %d" % (start_line, end_line))
+                    skew = 0
+                else:
+                    skew = math.tan(skew_angle)
+            else:
+                skew = 0
+            model = font.render(text, skew=skew)
+            if (model.minx is None) or (model.miny is None) or \
+                    (model.minz is None) or (model.minx == model.maxx) or \
+                    (model.miny == model.maxy):
+                log.warn("DXFImporter: Empty rendered TEXT item between " + \
+                        "line %d and %d" % (start_line, end_line))
+                return
+            model.scale(text_height / (model.maxy - model.miny),
+                    callback=self.callback)
+            if mirror_flags & 2:
+                # x mirror left/right
+                model.transform_by_template("yz_mirror", callback=self.callback)
+            if mirror_flags & 4:
+                # y mirror upside/down
+                model.transform_by_template("xz_mirror", callback=self.callback)
+            # this setting seems to refer to a box - not the text width - ignore
+            if False and width_final:
+                scale_x = width_final / (model.maxx - model.minx)
+                model.scale(scale_x, callback=self.callback)
+            if rotation:
+                matrix = pycam.Geometry.Matrix.get_rotation_matrix_axis_angle(
+                        (0, 0, 1), rotation)
+                model.transform_by_matrix(matrix, callback=self.callback)
+            # horizontal alignment
+            if align_horiz == 0:
+                offset_horiz = 0
+            elif align_horiz == 1:
+                offset_horiz = - (model.maxx - model.minx) / 2
+            elif align_horiz == 2:
+                offset_horiz = - (model.maxx - model.minx)
+            else:
+                log.warn("DXFImporter: Horizontal TEXT justifications " + \
+                        "(3..5) are not supported - ignoring (between line " + \
+                        "%d and %d)" % (start_line, end_line))
+                offset_horiz = 0
+            # vertical alignment
+            if align_vert in (0, 1):
+                # we don't distinguish between "bottom" and "base"
+                offset_vert = 0
+            elif align_vert == 2:
+                offset_vert = - (model.maxy - model.miny) / 2
+            elif align_vert == 3:
+                offset_vert = - (model.maxy - model.miny)
+            else:
+                log.warn("DXFImporter: Invalid vertical TEXT justification " + \
+                        " between line %d and %d" % (start_line, end_line))
+                offset_vert = 0
+            # shift the text to its final destination
+            shift_x = ref_point[0] - model.minx + offset_horiz
+            shift_y = ref_point[1] - model.miny + offset_vert
+            shift_z = ref_point[2] - model.minz
+            model.shift(shift_x, shift_y, shift_z, callback=self.callback)
+            for polygon in model.get_polygons():
+                for line in polygon.get_lines():
+                    self.lines.append(line)
+
+    def parse_3dface(self):
+        start_line = self.line_number
+        # the z-level defaults to zero (for 2D models)
+        p1 = [None, None, 0]
+        p2 = [None, None, 0]
+        p3 = [None, None, 0]
+        p4 = [None, None, 0]
+        color = None
+        key, value = self._read_key_value()
+        while (not key is None) and (key != self.KEYS["MARKER"]):
+            if key == self.KEYS["P1_X"]:
+                p1[0] = value
+            elif key == self.KEYS["P1_Y"]:
+                p1[1] = value
+            elif key == self.KEYS["P1_Z"]:
+                p1[2] = value
+            elif key == self.KEYS["P2_X"]:
+                p2[0] = value
+            elif key == self.KEYS["P2_Y"]:
+                p2[1] = value
+            elif key == self.KEYS["P2_Z"]:
+                p2[2] = value
+            elif key == self.KEYS["P3_X"]:
+                p3[0] = value
+            elif key == self.KEYS["P3_Y"]:
+                p3[1] = value
+            elif key == self.KEYS["P3_Z"]:
+                p3[2] = value
+            elif key == self.KEYS["P4_X"]:
+                p4[0] = value
+            elif key == self.KEYS["P4_Y"]:
+                p4[1] = value
+            elif key == self.KEYS["P4_Z"]:
+                p4[2] = value
+            else:
+                pass
+            key, value = self._read_key_value()
+        end_line = self.line_number
+        # The last lines were not used - they are just the marker for the next
+        # item.
+        if not key is None:
+            self._push_on_stack(key, value)
+        if (None in p1) or (None in p2) or (None in p3):
+            log.warn("DXFImporter: Incomplete 3DFACE definition between line " \
+                    + "%d and %d" % (start_line, end_line))
+        else:
+            # no color height adjustment for 3DFACE
+            point1 = Point(p1[0], p1[1], p1[2])
+            point2 = Point(p2[0], p2[1], p2[2])
+            point3 = Point(p3[0], p3[1], p3[2])
+            triangles = []
+            triangles.append((point1, point2, point3))
+            if not None in p4:
+                point4 = Point(p4[0], p4[1], p4[2])
+                triangles.append((point3, point4, point1))
+            for t in triangles:
+                if (t[0] != t[1]) and (t[0] != t[2]) and (t[1] != t[2]):
+                    self.triangles.append(Triangle(t[0], t[1], t[2]))
+                else:
+                    log.warn("DXFImporter: Ignoring zero-sized 3DFACE " + \
+                            "(between input line %d and %d): %s" % \
+                            (start_line, end_line, t))
+
     def parse_line(self):
         start_line = self.line_number
         # the z-level defaults to zero (for 2D models)
@@ -246,17 +722,17 @@ class DXFParser(object):
         color = None
         key, value = self._read_key_value()
         while (not key is None) and (key != self.KEYS["MARKER"]):
-            if key == self.KEYS["START_X"]:
+            if key == self.KEYS["P1_X"]:
                 p1[0] = value
-            elif key == self.KEYS["START_Y"]:
+            elif key == self.KEYS["P1_Y"]:
                 p1[1] = value
-            elif key == self.KEYS["START_Z"]:
+            elif key == self.KEYS["P1_Z"]:
                 p1[2] = value
-            elif key == self.KEYS["END_X"]:
+            elif key == self.KEYS["P2_X"]:
                 p2[0] = value
-            elif key == self.KEYS["END_Y"]:
+            elif key == self.KEYS["P2_Y"]:
                 p2[1] = value
-            elif key == self.KEYS["END_Z"]:
+            elif key == self.KEYS["P2_Z"]:
                 p2[2] = value
             elif key == self.KEYS["COLOR"]:
                 color = value
@@ -283,21 +759,26 @@ class DXFParser(object):
                 log.warn("DXFImporter: Ignoring zero-length LINE (between " \
                         + "input line %d and %d): %s" % (start_line, end_line,
                         line))
-    def parse_arc(self):
+
+    def parse_arc(self, circle=False):
         start_line = self.line_number
         # the z-level defaults to zero (for 2D models)
         center = [None, None, 0]
         color = None
         radius = None
-        angle_start = None
-        angle_end = None
+        if circle:
+            angle_start = 0
+            angle_end = 360
+        else:
+            angle_start = None
+            angle_end = None
         key, value = self._read_key_value()
         while (not key is None) and (key != self.KEYS["MARKER"]):
-            if key == self.KEYS["START_X"]:
+            if key == self.KEYS["P1_X"]:
                 center[0] = value
-            elif key == self.KEYS["START_Y"]:
+            elif key == self.KEYS["P1_Y"]:
                 center[1] = value
-            elif key == self.KEYS["START_Z"]:
+            elif key == self.KEYS["P1_Z"]:
                 center[2] = value
             elif key == self.KEYS["RADIUS"]:
                 radius = value
@@ -349,8 +830,8 @@ class DXFParser(object):
             return None
 
 
-def import_model(filename, program_locations=None, unit=None,
-        color_as_height=False, callback=None):
+def import_model(filename, color_as_height=False, fonts_cache=None,
+        callback=None, **kwargs):
     try:
         infile = open_url(filename)
     except IOError, err_msg:
@@ -359,15 +840,31 @@ def import_model(filename, program_locations=None, unit=None,
         return None
 
     result = DXFParser(infile, color_as_height=color_as_height,
-            callback=callback)
+            fonts_cache=fonts_cache, callback=callback)
 
-    lines = result.get_model()["lines"]
+    model_data = result.get_model()
+    lines = model_data["lines"]
+    triangles = model_data["triangles"]
 
     if callback and callback():
         log.warn("DXFImporter: load model operation was cancelled")
         return None
 
-    if lines:
+    # 3D models are preferred over 2D models
+    if triangles:
+        if lines:
+            log.warn("DXFImporter: Ignoring 2D elements in DXF file: " + \
+                    "%d lines" % len(lines))
+        model = pycam.Geometry.Model.Model()
+        for index, triangle in enumerate(triangles):
+            model.append(triangle)
+            # keep the GUI smooth
+            if callback and (index % 50 == 0):
+                callback()
+        log.info("DXFImporter: Imported DXF model (3D): %d triangles" % \
+                len(model.triangles()))
+        return model
+    elif lines:
         model = pycam.Geometry.Model.ContourModel()
         for index, line in enumerate(lines):
             model.append(line)
@@ -388,8 +885,9 @@ def import_model(filename, program_locations=None, unit=None,
             if callback:
                 callback(text="Shifting 2D model down to to z=0")
             model.shift(0, 0, -model.minz, callback=callback)
-        log.info("DXFImporter: Imported DXF model: %d lines / %d polygons" \
-                % (len(lines), len(model.get_polygons())))
+        log.info("DXFImporter: Imported DXF model (2D): " + \
+                "%d lines / %d polygons" % \
+                (len(lines), len(model.get_polygons())))
         return model
     else:
         link = "http://sf.net/apps/mediawiki/pycam/?title=SupportedFormats"

src/pycam/Importers/PSImporter.py

@@ -29,7 +29,8 @@ import os
 log = pycam.Utils.log.get_logger()
 
 
-def import_model(filename, program_locations=None, unit="mm", callback=None):
+def import_model(filename, program_locations=None, unit="mm", callback=None,
+        **kwargs):
     if not check_uri_exists(filename):
         log.error("PSImporter: file (%s) does not exist" % filename)
         return None

src/pycam/Importers/STLImporter.py

@@ -52,8 +52,7 @@ def UniqueVertex(x, y, z):
         vertices += 1
         return Point(x, y, z)
 
-def ImportModel(filename, use_kdtree=True, program_locations=None, unit=None,
-        callback=None):
+def ImportModel(filename, use_kdtree=True, callback=None, **kwargs):
     global vertices, edges, kdtree
     vertices = 0
     edges = 0

src/pycam/Importers/SVGImporter.py

@@ -81,7 +81,8 @@ def convert_eps2dxf(eps_filename, dxf_filename, location=None, unit="mm"):
                 process.stderr.read()))
         return False
 
-def import_model(filename, program_locations=None, unit="mm", callback=None):
+def import_model(filename, program_locations=None, unit="mm", callback=None,
+        **kwargs):
     if not check_uri_exists(filename):
         log.error("SVGImporter: file (%s) does not exist" % filename)
         return None

src/pycam/Utils/FontCache.py

+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+"""
+$Id$
+
+Copyright 2010 Lars Kruse <devel@sumpfralle.de>
+
+This file is part of PyCAM.
+
+PyCAM 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 3 of the License, or
+(at your option) any later version.
+
+PyCAM 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 PyCAM.  If not, see <http://www.gnu.org/licenses/>.
+"""
+
+import pycam.Utils.log
+import os
+
+
+DEFAULT_NAMES = ("normal", "default", "standard")
+
+
+log = pycam.Utils.log.get_logger()
+
+
+class FontCache(object):
+    """ The FontCache gradually loads fonts. This is more efficient than an
+    immeadiate initialization of all fonts for the DXF importer.
+    Use "get_font" for loading (incrementally) fonts until the requested font
+    name was found.
+    The functions "get_names" and "len()" trigger an immediate initialization
+    of all available fonts.
+    """
+
+    def __init__(self, font_dir=None, callback=None):
+        self.font_dir = font_dir
+        self.fonts = {}
+        self.callback = callback
+        self._unused_font_files = list(self._get_font_files())
+
+    def is_loading_complete(self):
+        return len(self._unused_font_files) == 0
+
+    def _get_font_files(self):
+        if self.font_dir is None:
+            return []
+        log.info("Font directory: %s" % self.font_dir)
+        result = []
+        files = os.listdir(self.font_dir)
+        for fname in files:
+            filename = os.path.join(self.font_dir, fname)
+            if filename.lower().endswith(".cxf") and os.path.isfile(filename):
+                result.append(filename)
+        result.sort()
+        return result
+
+    def __len__(self):
+        self._load_all_files()
+        return len(self.fonts)
+
+    def _get_font_without_loading(self, name):
+        for font_name in self.fonts:
+            if font_name.lower() == name.lower():
+                return self.fonts[font_name]
+        else:
+            return None
+
+    def get_font_names(self):
+        self._load_all_files()
+        return self.fonts.keys()
+
+    def get_font(self, name):
+        font = self._get_font_without_loading(name)
+        while not font and not self.is_loading_complete():
+            self._load_next_file()
+            font = self._get_font_without_loading(name)
+        if font:
+            return font
+        else:
+            # no font with that name is available
+            for other_name in DEFAULT_NAMES:
+                font = self._get_font_without_loading(other_name)
+                if font:
+                    return font
+            else:
+                if self.fonts:
+                    # return the first (random) font in the dictionary
+                    return self.fonts.values()[0]
+
+    def _load_all_files(self):
+        while not self.is_loading_complete():
+            self._load_next_file()
+
+    def _load_next_file(self):
+        if self.is_loading_complete():
+            return
+        filename = self._unused_font_files.pop(0)
+        charset = pycam.Importers.CXFImporter.import_font(filename,
+                callback=self.callback)
+        if not charset is None:
+            for name in charset.get_names():
+                self.fonts[name] = charset
+

src/pycam/Utils/log.py

@@ -112,7 +112,7 @@ class RepetitionsFilter(logging.Filter):
             if self._suppressed_messages_counter > 0:
                 # inject a message regarding the previously suppressed messages
                 self._last_record.msg =  \
-                        "*** %d similar messages were suppressed ***" % \
+                        "*** skipped %d similar message(s) ***" % \
                         self._suppressed_messages_counter
                 self._handler.emit(self._last_record)
             self._last_record = record