def populate_root_frame(self, frame, pres_style, draw_style):
frame_style = self.style_src.find("office:automatic-styles")\
.find({"style:style"}, {"style:name": pres_style})
parent_frame_style = self.pres.styles.find("office:styles")\
.find({"style:style"}, {"style:name": frame_style["style:parent-style-name"]})
parent_text_props = parent_frame_style.find({"style:text-properties"}, recursive=False)
parent_para_props = parent_frame_style.find({"style:paragraph-properties"}, recursive=False)
frame_text_style = self.style_src.find("office:automatic-styles")\
.find({"style:style"}, {"style:name": draw_style})
frame_text_props = frame_text_style.find({"style:text-properties"}, recursive=False)
frame_para_props = frame_text_style.find({"style:paragraph-properties"}, recursive=False)
for prop in TEXT_PROPS:
if frame_text_props and prop in frame_text_props.attrs:
frame[prop] = frame_text_props[prop]
elif parent_text_props and prop in parent_text_props.attrs:
frame[prop] = parent_text_props[prop]
for prop in PARA_PROPS:
if frame_para_props and prop in frame_para_props.attrs:
frame[prop] = frame_para_props[prop]
elif parent_para_props and prop in parent_para_props.attrs:
frame[prop] = parent_para_props[prop]
for prop in PARA_PROPS_U:
if frame_para_props and prop in frame_para_props.attrs:
frame[prop] = units_to_float(frame_para_props[prop])
elif parent_para_props and prop in parent_para_props.attrs:
frame[prop] = units_to_float(parent_para_props[prop])
def parse_page_animations(self, timing_root, json_data, page_json_data):
main_seq = timing_root.find("anim:seq")
click_anims = main_seq.findChildren({"anim:par"}, recursive=False)
# Keep track of initially hidden and visible elements that have associated animations
init_visible, init_hidden = [], []
for click_anim in click_anims:
timed_anims = click_anim.findChildren({"anim:par"}, recursive=False)
first_timed_anim = None
begin_at = "indefinite" # First node is initiated with a click
anim_json_data = {}
anim_order = {}
for timed_anim in timed_anims:
parallel_anims = timed_anim.findChildren({"anim:par"}, recursive=False)
begin_delay = units_to_float(timed_anim["smil:begin"])
for anim_data in parallel_anims:
# Find animation target
anim_subnode = anim_data.findChild()
anim_target = "obj_" + anim_subnode["smil:targetelement"]
anim_preset = anim_data["presentation:preset-id"]
anim_delay = begin_delay + units_to_float(anim_data["smil:begin"])
if anim_target in self.xml_ids:
if first_timed_anim:
begin_at = first_timed_anim + ".begin+" + str(anim_delay) + "s"
anim_id = self.animator.add_animation(\
self, self.xml_ids[anim_target], anim_data, begin_at)
if anim_id:
if not first_timed_anim:
first_timed_anim = anim_id
anim_json_data["id"] = anim_id
if anim_preset.find("ooo-entrance") == -1\
and anim_target not in init_hidden:
init_visible.append(anim_target)
elif anim_target not in init_visible and anim_target not in init_hidden:
init_hidden.append(anim_target)
# Track order of sub-animations
if anim_delay in anim_order:
anim_order[anim_delay].append(anim_id)
else:
anim_order[anim_delay] = [anim_id]
anim_json_data["anim_order"] = []
for time_index in sorted(anim_order.keys()):
anim_json_data["anim_order"] += anim_order[time_index]
if "id" in anim_json_data:
# Condition prevents "empty" animations begin added (those that refer to
# un-implemented types of animation)
page_json_data["animations"].append(anim_json_data)
page_json_data["init_hidden"] = init_hidden
page_json_data["init_visible"] = init_visible
def fill_hatch(cls, dwg, elt, pres, attrs, e_width, e_height, style_tag):
hatch_node = pres.styles.find("office:styles")\
.find({"draw:hatch"}, {"draw:name": attrs["draw:fill-hatch-name"]})
hatch_angle = int(hatch_node["draw:rotation"]) / 10
hatch_dist = units_to_float(str(hatch_node["draw:distance"]))
# Print background color first
pattern = dwg.pattern(insert=(0, 0), size=(e_width, e_height), \
patternUnits="userSpaceOnUse", patternContentUnits="userSpaceOnUse")
if style_tag.find(
"style:graphic-properties")["draw:fill-hatch-solid"] == "true":
pattern.add(dwg.rect((0, 0), (e_width, e_height),\
fill=attrs["draw:fill-color"]))
else:
pattern.add(dwg.rect((0, 0), (e_width, e_height), fill='#ffffff'))
# Single line hatch
FillFactory.draw_hatching(dwg, pattern, hatch_angle, hatch_dist,\
hatch_node["draw:color"], 1/DPCM, e_width, e_height)
# Double line hatch
if hatch_node["draw:style"] in ["double", "triple"]:
FillFactory.draw_hatching(dwg, pattern, hatch_angle + 90, hatch_dist,\
hatch_node["draw:color"], 1/DPCM, e_width, e_height)
# Triple line hatch
if hatch_node["draw:style"] == "triple":
FillFactory.draw_hatching(dwg, pattern, hatch_angle + 135, hatch_dist,\
hatch_node["draw:color"], 1/DPCM, e_width, e_height)
dwg.defs.add(pattern)
elt.fill(pattern.get_paint_server())
def parse_line(self, item, layer_g, style_src):
line_x1 = units_to_float(item.attrs["svg:x1"])
line_y1 = units_to_float(item.attrs["svg:y1"])
line_x2 = units_to_float(item.attrs["svg:x2"])
line_y2 = units_to_float(item.attrs["svg:y2"])
line = self.dwg.line(start=(line_x1, line_y1), end=(line_x2, line_y2))
StrokeFactory.stroke(self, self.dwg, item, line, 1, style_src)
if "xml:id" in item.attrs:
self.xml_ids["obj_"+item["xml:id"]] = {
"item": line,
"x": min(line_x1, line_x2),
"y": min(line_y1, line_y2),
"width": abs(line_x2 - line_x1),
"height": abs(line_y2 - line_y1)
}
line.__setitem__("id", "obj_"+item["xml:id"])
layer_g.add(line)
def parse_polygon(self, item, layer_g, style_src):
polygon_w = units_to_float(item.attrs["svg:width"])
polygon_h = units_to_float(item.attrs["svg:height"])
polygon_vb = [int(x) for x in item.attrs["svg:viewbox"].split()]
# TODO: Check assumption - this is always scaled equally on both axes
polygon_pts = item.attrs["draw:points"].replace(',', ' ').split()
polygon_d = "M " + ' '.join(polygon_pts[0:2]) + " L " + \
' '.join(polygon_pts[2:]) + ' Z'
# TODO: Add in fill
polygon = self.dwg.path(d=polygon_d, fill='none')
if "svg:x" in item.attrs and "svg:y" in item.attrs:
polygon.translate(units_to_float(item.attrs["svg:x"]),\
units_to_float(item.attrs["svg:y"]))
elif "draw:transform" in item.attrs:
ShapeParser.transform_shape(item, polygon)
polygon_scale = polygon_w / (polygon_vb[2]-polygon_vb[0])
polygon.scale(polygon_scale)
style_tag = style_src.find("office:automatic-styles")\
.find({"style:style"}, {"style:name": item["draw:style-name"]})
attrs = style_tag.find("style:graphic-properties").attrs
FillFactory.fill(self.dwg, polygon, self, attrs, \
polygon_w/polygon_scale, polygon_h/polygon_scale, style_tag)
StrokeFactory.stroke(self, self.dwg, item, polygon, 1/polygon_scale, style_src)
if "xml:id" in item.attrs:
self.xml_ids["obj_"+item["xml:id"]] = {
"item": polygon,
"x": units_to_float(item.attrs["svg:x"]),
"y": units_to_float(item.attrs["svg:y"]),
"width": polygon_w,
"height": polygon_h
}
polygon.__setitem__("id", "obj_"+item["xml:id"])
layer_g.add(polygon)
def parse_path(self, item, layer_g, style_src):
path_w = units_to_float(item.attrs["svg:width"])
path_vb = [int(x) for x in item.attrs["svg:viewbox"].split()]
# TODO: Check assumption - this is always scaled equally on both axes
path = self.dwg.path(d=item.attrs["svg:d"], fill='none')
if "svg:x" in item.attrs and "svg:y" in item.attrs:
path.translate(units_to_float(item.attrs["svg:x"]),\
units_to_float(item.attrs["svg:y"]))
elif "draw:transform" in item.attrs:
ShapeParser.transform_shape(item, path)
path_scale = path_w / (path_vb[2]-path_vb[0])
path.scale(path_scale)
# TODO: See how dashed paths and end markers for paths work
StrokeFactory.stroke(self, self.dwg, item, path, 1/path_scale, style_src)
if "xml:id" in item.attrs:
# TODO: Configure this to work with draw:transforms
self.xml_ids["obj_"+item["xml:id"]] = {
"item": path,
"x": units_to_float(item.attrs["svg:x"]),
"y": units_to_float(item.attrs["svg:y"]),
"width": path_w,
"height": units_to_float(item.attrs["svg:width"])
}
path.__setitem__("id", "obj_"+item["xml:id"])
layer_g.add(path)
def __init__(self, url, data_store):
self.url = url
self.data_store = data_store
pres_archive = zipfile.ZipFile(url, 'r')
self.styles = BeautifulSoup(pres_archive.read('styles.xml'), \
"lxml", from_encoding='UTF-8')
self.content = BeautifulSoup(pres_archive.read('content.xml'), \
"lxml", from_encoding='UTF-8')
self.font_mgr = font_manager.FontManager()
self.animator = AnimationFactory()
self.xml_ids = {}
# Create SVG drawing of correct size
drawing_size = self.get_document_size()
self.d_width = units_to_float(str(drawing_size[0]))
self.d_height = units_to_float(str(drawing_size[1]))
self.view_box = '0 0 ' + str(self.d_width) + ' ' + str(self.d_height)
self.dwg = svgwrite.Drawing(size=drawing_size, viewBox=(self.view_box))
# Setup iterative variables
self.clip_id = 0
self.sub_g = 0
def parse_polyline(self, item, layer_g, style_src):
polyline_w = units_to_float(item.attrs["svg:width"])
polyline_vb = [int(x) for x in item.attrs["svg:viewbox"].split()]
# TODO: Check assumption - this is always scaled equally on both axes
polyline_pts = item.attrs["draw:points"].replace(',', ' ').split()
polyline_d = "M " + ' '.join(polyline_pts[0:2]) + " L " + ' '.join(polyline_pts[2:])
polyline = self.dwg.path(d=polyline_d, fill='none')
if "draw:transform" in item.attrs:
ShapeParser.transform_shape(item, polyline)
polyline_scale = polyline_w / (polyline_vb[2]-polyline_vb[0])
polyline.scale(polyline_scale)
# TODO: See how dashed polylines and end markers for polylines work
StrokeFactory.stroke(self, self.dwg, item, polyline, 1/polyline_scale, style_src)
if "xml:id" in item.attrs:
self.xml_ids["obj_"+item["xml:id"]] = {
"item": polyline,
"x": units_to_float(item.attrs["svg:x"]),
"y": units_to_float(item.attrs["svg:y"]),
"width": polyline_w,
"height": units_to_float(item.attrs["svg:height"])
}
polyline.__setitem__("id", "obj_"+item["xml:id"])
layer_g.add(polyline)
def populate_stack_frame(self, frame, style_name):
style_tag = self.style_src.find("office:automatic-styles")\
.find({"style:style"}, {"style:name": style_name})
style_text_props = style_tag.find({"style:text-properties"}, recursive=False)
if style_text_props:
for prop in TEXT_PROPS:
if prop in style_text_props.attrs:
frame[prop] = style_text_props[prop]
style_para_props = style_tag.find({"style:paragraph-properties"}, recursive=False)
if style_para_props:
for prop in PARA_PROPS:
if prop in style_para_props.attrs:
frame[prop] = style_para_props[prop]
for prop in PARA_PROPS_U:
if prop in style_para_props.attrs:
frame[prop] = units_to_float(style_para_props[prop])
def fill_bitmap(cls, dwg, elt, pres, attrs, e_width, e_height):
image_node = pres.styles.find("office:styles")\
.find({"draw:fill-image"}, {"draw:name": attrs["draw:fill-image-name"]})
# Extract image to data store
pres_archive = zipfile.ZipFile(pres.url, 'r')
pres_archive.extract(image_node["xlink:href"], pres.data_store)
if "draw:fill-image-width" in attrs:
if attrs["draw:fill-image-width"][-1] == "%":
bitmap_width = e_width * (
int(attrs["draw:fill-image-width"][:-1]) / 100)
bitmap_height = e_height * (
int(attrs["draw:fill-image-height"][:-1]) / 100)
else:
bitmap_width = units_to_float(
str(attrs["draw:fill-image-width"]))
bitmap_height = units_to_float(
str(attrs["draw:fill-image-height"]))
if bitmap_width > 0 and bitmap_height > 0:
pattern_size = (bitmap_width, bitmap_height)
else:
with Image.open(pres.data_store +
image_node["xlink:href"]) as img:
im_width, im_height = img.size
pattern_size = (im_width / DPCM, im_height / DPCM)
if attrs["style:repeat"] == "stretch":
pattern = dwg.pattern(insert=(0, 0), size=(e_width, e_height), \
patternUnits="userSpaceOnUse", patternContentUnits="userSpaceOnUse")
pattern.add(dwg.image(pres.data_store + image_node["xlink:href"],\
insert=(0, 0), size=(e_width, e_height), preserveAspectRatio="none"))
dwg.defs.add(pattern)
elt.fill(pattern.get_paint_server())
elif attrs["style:repeat"] == "no-repeat":
if attrs["draw:fill-image-ref-point"] in [
"top-left", "left", "bottom-left"
]:
image_x = 0
elif attrs["draw:fill-image-ref-point"] in [
"top", "center", "bottom"
]:
image_x = (e_width - bitmap_width) / 2
else:
image_x = e_width - bitmap_width
if attrs["draw:fill-image-ref-point"] in [
"top-left", "top", "top-right"
]:
image_y = 0
elif attrs["draw:fill-image-ref-point"] in [
"left", "center", "right"
]:
image_y = (e_height - bitmap_height) / 2
else:
image_y = e_height - bitmap_height
pattern = dwg.pattern(insert=(0, 0), size=(e_width, e_height), \
patternUnits="userSpaceOnUse", patternContentUnits="userSpaceOnUse")
pattern.add(dwg.image(pres.data_store + image_node["xlink:href"],\
insert=(image_x, image_y), size=(bitmap_width, bitmap_height), \
preserveAspectRatio="none"))
dwg.defs.add(pattern)
elt.fill(pattern.get_paint_server())
else: # Tiled background
# Adjust pattern start point based on reference point
if attrs["draw:fill-image-ref-point"] in [
"top", "center", "bottom"
]:
x_base = (((e_width - pattern_size[0]) / 2) %
pattern_size[0]) / pattern_size[0]
col_count_parity = (((e_width - pattern_size[0]) / 2) //
pattern_size[0] % 2)
if col_count_parity == 0:
tile_col_offset = [1, 0, 1]
else:
tile_col_offset = [0, 1, 0]
elif attrs["draw:fill-image-ref-point"] in [
"top-right", "right", "bottom-right"
]:
x_base = (e_width % pattern_size[0]) / pattern_size[0]
col_count_parity = (e_width // pattern_size[0]) % 2
if col_count_parity == 0:
tile_col_offset = [0, 1, 0]
else:
tile_col_offset = [1, 0, 1]
else: # top-left, center-left, bottom-left
x_base = 0.0
tile_col_offset = [
1, 0, 1
] # First full column (index 1) is not offset
if attrs["draw:fill-image-ref-point"] in [
"left", "center", "right"
]:
y_base = (((e_height - pattern_size[1]) / 2) %
pattern_size[1]) / pattern_size[1]
row_count_parity = (((e_height - pattern_size[1]) / 2) //
pattern_size[1] % 2)
if row_count_parity == 0:
tile_row_offset = [1, 0, 1]
else:
tile_row_offset = [0, 1, 0]
elif attrs["draw:fill-image-ref-point"] in [
"bottom-left", "bottom", "bottom-right"
]:
y_base = (e_height % pattern_size[1]) / pattern_size[1]
row_count_parity = (e_height // pattern_size[1]) % 2
if row_count_parity == 0:
tile_row_offset = [0, 1, 0]
else:
tile_row_offset = [1, 0, 1]
else:
y_base = 0.0
tile_row_offset = [1, 0,
1] # First full row (index 1) is not offset
# Adjust pattern offset
x_offset = x_base + (
(int(attrs["draw:fill-image-ref-point-x"][:-1]) % 100)) / 100
if x_offset >= 1:
x_offset -= 1
if tile_col_offset == [0, 1, 0]:
tile_col_offset = [1, 0, 1]
else:
tile_col_offset = [0, 1, 0]
y_offset = y_base + (
(int(attrs["draw:fill-image-ref-point-y"][:-1]) % 100)) / 100
if y_offset >= 1:
y_offset -= 1
if tile_row_offset == [0, 1, 0]:
tile_row_offset = [1, 0, 1]
else:
tile_row_offset = [0, 1, 0]
# Determine if image has row/col offset or not
offset_type = attrs["draw:tile-repeat-offset"].split(" ")
if offset_type[0] == "0%":
pattern = dwg.pattern(insert=(0, 0), size=pattern_size, \
patternUnits="userSpaceOnUse", patternContentUnits="userSpaceOnUse")
pattern.add(dwg.image(pres.data_store + image_node["xlink:href"],\
insert=(x_offset*pattern_size[0], y_offset*pattern_size[1]),\
size=pattern_size, preserveAspectRatio="none"))
pattern.add(dwg.image(pres.data_store + image_node["xlink:href"],\
insert=((x_offset-1)*pattern_size[0], y_offset*pattern_size[1]),\
size=pattern_size, preserveAspectRatio="none"))
pattern.add(dwg.image(pres.data_store + image_node["xlink:href"],\
insert=(x_offset*pattern_size[0], (y_offset-1)*pattern_size[1]),\
size=pattern_size, preserveAspectRatio="none"))
pattern.add(dwg.image(pres.data_store + image_node["xlink:href"],\
insert=((x_offset-1)*pattern_size[0], (y_offset-1)*pattern_size[1]),\
size=pattern_size, preserveAspectRatio="none"))
else:
tiled_offset = (int(offset_type[0][:-1]) % 100) / 100
if offset_type[1] == "horizontal":
# Horizontal tiling - make fill 1 wide x 2 high
pattern = dwg.pattern(insert=(0, 0), size=(pattern_size[0], 2*pattern_size[1]),\
patternUnits="userSpaceOnUse", patternContentUnits="userSpaceOnUse")
if tiled_offset + x_offset >= 1:
tiled_offset -= 1
# Top row
pattern.add(dwg.image(pres.data_store + image_node["xlink:href"],\
insert=((x_offset + tiled_offset*tile_row_offset[0]) * pattern_size[0],\
(y_offset-1)*pattern_size[1]),\
size=pattern_size, preserveAspectRatio="none"))
pattern.add(dwg.image(pres.data_store + image_node["xlink:href"],\
insert=((x_offset-1 + tiled_offset*tile_row_offset[0]) * pattern_size[0],\
(y_offset-1)*pattern_size[1]),\
size=pattern_size, preserveAspectRatio="none"))
# Centre row
pattern.add(dwg.image(pres.data_store + image_node["xlink:href"],\
insert=((x_offset + tiled_offset*tile_row_offset[1]) * pattern_size[0],\
y_offset*pattern_size[1]),\
size=pattern_size, preserveAspectRatio="none"))
pattern.add(dwg.image(pres.data_store + image_node["xlink:href"],\
insert=((x_offset-1 + tiled_offset*tile_row_offset[1]) * pattern_size[0],\
y_offset*pattern_size[1]),\
size=pattern_size, preserveAspectRatio="none"))
# Bottom row
pattern.add(dwg.image(pres.data_store + image_node["xlink:href"],\
insert=((x_offset + tiled_offset*tile_row_offset[2]) * pattern_size[0],\
(y_offset+1)*pattern_size[1]),\
size=pattern_size, preserveAspectRatio="none"))
pattern.add(dwg.image(pres.data_store + image_node["xlink:href"],\
insert=((x_offset-1 + tiled_offset*tile_row_offset[2]) * pattern_size[0],\
(y_offset+1)*pattern_size[1]),\
size=pattern_size, preserveAspectRatio="none"))
else:
# Vertical tiling - make fill 2 wide x 1 high
pattern = dwg.pattern(insert=(0, 0), size=(2*pattern_size[0], pattern_size[1]),\
patternUnits="userSpaceOnUse", patternContentUnits="userSpaceOnUse")
if tiled_offset + y_offset >= 1:
tiled_offset -= 1
# Left column
pattern.add(dwg.image(pres.data_store + image_node["xlink:href"],\
insert=((x_offset-1) * pattern_size[0],\
(y_offset-1 + tiled_offset*tile_col_offset[0]) * pattern_size[1]),\
size=pattern_size, preserveAspectRatio="none"))
pattern.add(dwg.image(pres.data_store + image_node["xlink:href"],\
insert=((x_offset-1) * pattern_size[0],\
(y_offset + tiled_offset*tile_col_offset[0])*pattern_size[1]),\
size=pattern_size, preserveAspectRatio="none"))
# Centre column
pattern.add(dwg.image(pres.data_store + image_node["xlink:href"],\
insert=(x_offset * pattern_size[0],\
(y_offset-1 + tiled_offset*tile_col_offset[1]) * pattern_size[1]),\
size=pattern_size, preserveAspectRatio="none"))
pattern.add(dwg.image(pres.data_store + image_node["xlink:href"],\
insert=(x_offset * pattern_size[0],\
(y_offset + tiled_offset*tile_col_offset[1]) * pattern_size[1]),\
size=pattern_size, preserveAspectRatio="none"))
# Right column
pattern.add(dwg.image(pres.data_store + image_node["xlink:href"],\
insert=((x_offset+1) * pattern_size[0],\
(y_offset-1 + tiled_offset*tile_col_offset[2]) * pattern_size[1]),\
size=pattern_size, preserveAspectRatio="none"))
pattern.add(dwg.image(pres.data_store + image_node["xlink:href"],\
insert=((x_offset+1) * pattern_size[0],\
(y_offset + tiled_offset*tile_col_offset[2]) * pattern_size[1]),\
size=pattern_size, preserveAspectRatio="none"))
dwg.defs.add(pattern)
elt.fill(pattern.get_paint_server())
def parse_frame(self, item, layer_g, style_src):
# TODO: Work out how to store xml:ids for images with and without borders and text areas
# with and without borders...
print("parse frame")
frame_attrs = item.attrs
if "presentation:placeholder" in frame_attrs \
and frame_attrs["presentation:placeholder"] == "true":
print("Frame is a placeholder - skip!")
return
if "draw:transform" in frame_attrs:
frame_x, frame_y = 0, 0
else:
frame_x = units_to_float(frame_attrs["svg:x"])
frame_y = units_to_float(frame_attrs["svg:y"])
frame_w = units_to_float(frame_attrs["svg:width"])
frame_h = units_to_float(frame_attrs["svg:height"])
# TODO: draw:frame might contain something other than an image...
clip_area = None
if "draw:style-name" in frame_attrs:
style_tag = style_src.find("office:automatic-styles")\
.find({"style:style"}, {"style:name": frame_attrs["draw:style-name"]})\
.find("style:graphic-properties")
if style_tag and "fo:clip" in style_tag.attrs:
clip_area = style_tag["fo:clip"]
if item.find("draw:image"):
print("add image")
image_href = item.find("draw:image").attrs["xlink:href"]
# Extract image to data store
pres_archive = zipfile.ZipFile(self.url, 'r')
pres_archive.extract(image_href, self.data_store)
if clip_area and clip_area[0:4] == "rect":
clip = [units_to_float(x) for x in clip_area[5:-1].split(", ")]
clip_path = self.dwg.defs.add(\
self.dwg.clipPath(id="clip" + str(self.clip_id)))
clip_path.add(self.dwg.rect(
insert=(frame_x, frame_y), size=(frame_w, frame_h)))
img_px = Image.open(self.data_store + image_href).size
img_w = frame_w * img_px[0] / (img_px[0] - DPCM * (clip[1] + clip[3]))
img_h = frame_h * img_px[1] / (img_px[1] - DPCM * (clip[0] + clip[2]))
img_x = frame_x - \
(frame_w * DPCM * clip[3] / (img_px[0] - DPCM * (clip[1] + clip[3])))
img_y = frame_y - \
(frame_h * DPCM * clip[0] / (img_px[1] - DPCM * (clip[0] + clip[2])))
clip_img = self.dwg.image(self.data_store + image_href,\
insert=(img_x, img_y), size=(img_w, img_h),\
preserveAspectRatio="none",\
clip_path="url(#clip" + str(self.clip_id) + ")")
layer_g.add(clip_img)
self.clip_id += 1
else:
clip_img = self.dwg.image(self.data_store + image_href,\
insert=(frame_x, frame_y), size=(frame_w, frame_h),\
preserveAspectRatio="none")
layer_g.add(clip_img)
frame = self.dwg.rect(insert=(frame_x, frame_y), \
size=(frame_w, frame_h), fill='none')
StrokeFactory.stroke(self, self.dwg, item, frame, 1, style_src)
if "draw:transform" in frame_attrs:
ShapeParser.transform_shape(item, clip_img)
ShapeParser.transform_shape(item, frame)
layer_g.add(frame)
elif item.find("draw:text-box"):
tb_rect = self.dwg.rect(insert=(frame_x, frame_y), size=(frame_w, frame_h))
vert_align = "middle"
if "presentation:style-name" in frame_attrs:
tb_style_name = frame_attrs["presentation:style-name"]
elif "draw:style-name" in frame_attrs:
tb_style_name = frame_attrs["draw:style-name"]
else:
tb_style_name = None
if tb_style_name:
frame_style = style_src.find("office:automatic-styles")\
.find({"style:style"}, {"style:name": tb_style_name})
frame_graphics = frame_style.find({"style:graphic-properties"})
FillFactory.fill(self.dwg, tb_rect, self, frame_attrs, frame_w, frame_h, \
frame_style)
layer_g.add(tb_rect)
if "draw:textarea-vertical-align" in frame_graphics.attrs:
vert_align = frame_graphics["draw:textarea-vertical-align"]
tb_parser = TextBoxParser(self.dwg, self, item, self.font_mgr, vert_align, style_src)
tb_parser.visit_textbox(layer_g, "textbox")
def visit_p(self, item_p, is_in_list):
# TODO: Underline, overline - final few variations
p_stack_frame = self.style_stack[len(self.style_stack) - 1].copy() # shallow copy
if "text:style-name" in item_p.attrs:
self.populate_stack_frame(p_stack_frame, item_p["text:style-name"])
self.style_stack.append(p_stack_frame)
if is_in_list:
self.style_stack[-1]["fo:margin-left"] = self.style_stack[-2]["fo:margin-left"]
# Add on before spacing to cur_y
self.cur_y += p_stack_frame["fo:margin-top"]
p_height = 0
first_span = None
line_size, line_indent = 0, 0
queued_spans = []
on_first_line = True
first_descent = 0
highlights = []
decor_lines = []
for item_span in item_p.find_all({"text:span"}, recursive=False):
span_start = 0
span_pos = 0
span_stack_frame = self.style_stack[len(self.style_stack) - 1].copy() # shallow copy
if "text:style-name" in item_span.attrs:
self.populate_stack_frame(span_stack_frame, item_span["text:style-name"])
self.style_stack.append(span_stack_frame)
span_font = self.font_mgr.findfont(FontProperties(\
family=span_stack_frame["style:font-name"],
style=span_stack_frame["fo:font-style"],
weight=span_stack_frame["fo:font-weight"]))
# Text position is either "normal" or a baseline adjustment e.g. "33%" or
# a baseline adjustment and font size adjustment e.g. "-33% 58%"
# Font size only changes in the final of the three cases
if len(span_stack_frame["style:text-position"].split()) == 1:
base_font_size = units_to_float(span_stack_frame["fo:font-size"])
else:
bf_scale = units_to_float(span_stack_frame["style:text-position"].split()[1]) / 100
base_font_size = units_to_float(span_stack_frame["fo:font-size"]) * bf_scale
scaled_font_size = str(base_font_size/DPCM)
i_font = ImageFont.truetype(span_font, math.ceil(base_font_size*96/72))
i_font_height = i_font.font.ascent + i_font.font.descent
# Replace all <text:s></text:s> tags with spaces
for spacer in item_span.find_all({"text:s"}):
spacer.replace_with(" ")
# Replace all <text:tab></text:tab> tags with literal tabs
for tab in item_span.find_all({"text:tab"}):
tab.replace_with("\t")
for line_break in item_span.find_all({"text:line-break"}):
line_break.replace_with("\n")
for fields in item_span.find_all({"presentation:date-time", "presentation:footer",\
"text:page-number"}):
# TODO: Decide whether/how to implement these properly...
fields.replace_with("")
print("++" + str(item_span.contents) + "++")
span_text = ''.join(item_span.contents)
# Cope with line break span
if span_text == "\n":
line_h, line_d, tspan, h_lights, d_lines = self.process_line(queued_spans, \
[line_indent, span_stack_frame["fo:margin-right"]], \
span_stack_frame["fo:line-height"])
queued_spans.clear()
line_size = 0
p_height += line_h
if on_first_line:
on_first_line = False
first_span = tspan
first_descent = line_d
# Cope with empty or whitespace-only spans (tabs and/or spaces)
elif span_text.strip() == "":
if span_text == "":
span_w = 0
else:
span_w = i_font.getsize(span_text)[0]
queued_spans.append({\
"style": self.output_tspan_style(span_stack_frame, scaled_font_size),
"height": i_font_height/DPCM,
"descent": i_font.font.descent/DPCM,
"ascent": i_font.font.ascent/DPCM,
"width": span_w,
"font-size": base_font_size,
"text": span_text,
"stack-frame": span_stack_frame.copy()})
span_text = "" # Ensure that while loop doesn't run
while span_pos != -1:
h_lights, d_lines = [], []
# Recalculate available width to take indents into account
if on_first_line:
line_indent = span_stack_frame["fo:margin-left"] + \
span_stack_frame["fo:text-indent"]
else:
line_indent = span_stack_frame["fo:margin-left"]
line_width = (self.svg_w - line_indent - span_stack_frame["fo:margin-right"])*DPCM
# Find next word break
next_pos = span_text.find(" ", span_pos+1)
if next_pos != -1 and span_text[span_start:next_pos].strip() != "":
size = i_font.getsize(span_text[span_start:next_pos])
if size[0] + line_size > line_width:
# Write out span up to span_pos then start new span on next line
queued_spans.append({\
"style": self.output_tspan_style(span_stack_frame, scaled_font_size),
"height": i_font_height/DPCM,
"descent": i_font.font.descent/DPCM,
"ascent": i_font.font.ascent/DPCM,
"width": i_font.getsize(span_text[span_start:span_pos].rstrip())[0],
"font-size": base_font_size,
"text": span_text[span_start:span_pos].rstrip(),
"stack-frame": span_stack_frame.copy()})
line_h, line_d, tspan, h_lights, d_lines = self.process_line(queued_spans, \
[line_indent, span_stack_frame["fo:margin-right"]], \
span_stack_frame["fo:line-height"])
queued_spans.clear()
span_start = span_pos
line_size = 0
p_height += line_h
if on_first_line:
on_first_line = False
first_span = tspan
first_descent = line_d
elif next_pos == -1 and span_text[span_start:].strip() != "":
# End of span_text has been reached
size = i_font.getsize(span_text[span_start:])
if size[0] + line_size > line_width:
# Write out span up to span_pos then start new span on next line
if span_text[span_start:span_pos] != "":
queued_spans.append({\
"style": self.output_tspan_style(span_stack_frame, \
scaled_font_size),
"height": i_font_height/DPCM,
"descent": i_font.font.descent/DPCM,
"ascent": i_font.font.ascent/DPCM,
"width": i_font.getsize(span_text[span_start:span_pos]\
.rstrip())[0],
"font-size": base_font_size,
"text": span_text[span_start:span_pos].rstrip(),
"stack-frame": span_stack_frame.copy()})
line_h, line_d, tspan, h_lights, d_lines = self.process_line(queued_spans, \
[line_indent, span_stack_frame["fo:margin-right"]], \
span_stack_frame["fo:line-height"])
queued_spans.clear()
# Queue remainder of span
queued_spans.append({\
"style": self.output_tspan_style(span_stack_frame, scaled_font_size),
"height": i_font_height/DPCM,
"descent": i_font.font.descent/DPCM,
"ascent": i_font.font.ascent/DPCM,
"width": i_font.getsize(span_text[span_pos:])[0],
"font-size": base_font_size,
"text": span_text[span_pos:],
"stack-frame": span_stack_frame.copy()})
line_size = i_font.getsize(span_text[span_pos:])[0]
span_start = span_pos
p_height += line_h
if on_first_line:
first_span = tspan
on_first_line = False
first_descent = line_d
else:
# Just queue rest of current span
queued_spans.append({\
"style": self.output_tspan_style(span_stack_frame, scaled_font_size),
"height": i_font_height/DPCM,
"descent": i_font.font.descent/DPCM,
"ascent": i_font.font.ascent/DPCM,
"width": i_font.getsize(span_text[span_start:])[0],
"font-size": base_font_size,
"text": span_text[span_start:],
"stack-frame": span_stack_frame.copy()})
line_size += size[0]
span_pos = next_pos
# Process highlights and decor_lines
for hl in h_lights:
highlights.append(hl)
for dl in d_lines:
decor_lines.append(dl)
# Pop stack frame (span)
self.style_stack.pop()
# Write out any queued spans before going on to next p
h_lights, d_lines = [], []
if queued_spans:
line_h, line_d, tspan, h_lights, d_lines = self.process_line(queued_spans, \
[line_indent, span_stack_frame["fo:margin-right"]], \
span_stack_frame["fo:line-height"])
queued_spans.clear()
p_height += line_h
if on_first_line:
first_span = tspan
first_descent = line_d
# Process highlights and decor_lines
for hl in h_lights:
highlights.append(hl)
for dl in d_lines:
decor_lines.append(dl)
# Add before spacing to first span of paragraph, increase p_height accordingly
if first_span:
first_dy = first_span.__getitem__("dy") + p_stack_frame["fo:margin-top"]
first_span.__setitem__("dy", first_dy)
p_height += p_stack_frame["fo:margin-top"]
# Add after spacing to cur_y
self.cur_y += p_stack_frame["fo:margin-bottom"]
# Pop stack frame (p)
self.style_stack.pop()
return p_height, first_descent, first_span, p_stack_frame["fo:margin-bottom"], \
highlights, decor_lines
def render_shape(cls, dwg, pres, shape, layer, style_src):
# TODO: cope with draw:transform for text boxes
geom = shape.find("draw:enhanced-geometry")
if "svg:viewbox" in geom.attrs:
vb_bounds = [int(x) for x in geom["svg:viewbox"].split()]
else:
vb_bounds = [0, 0, 21600, 21600]
if "svg:x" in shape.attrs:
base_x = units_to_float(str(shape["svg:x"]))
else:
base_x = 0.0
if "svg:y" in shape.attrs:
base_y = units_to_float(str(shape["svg:y"]))
else:
base_y = 0.0
bases = [base_x, base_y]
scale_x = units_to_float(str(shape["svg:width"])) / \
(vb_bounds[2] - vb_bounds[0])
scale_y = units_to_float(str(shape["svg:height"])) / \
(vb_bounds[3] - vb_bounds[1])
scales = [scale_x, scale_y]
if "draw:mirror-horizontal" in geom.attrs and \
geom.attrs["draw:mirror-horizontal"] == "true":
x_adj = vb_bounds[2] - vb_bounds[0]
else:
x_adj = 0
if "draw:mirror-vertical" in geom.attrs and geom.attrs[
"draw:mirror-vertical"] == "true":
y_adj = vb_bounds[3] - vb_bounds[1]
else:
y_adj = 0
adjs = [x_adj, y_adj]
# Step 0 - perform substitutions
modifiers, eq_results = ShapeParser.evaluate_equations(geom, vb_bounds)
# Step 1 - split string into command sections
path_sections = re.findall(r'[a-zA-Z][?\$f0-9 -.]*',
geom["draw:enhanced-path"])
# Step 2 - translate sections from ODP grammar to SVG equivalent
shape_path = dwg.path()
path_start_x, path_start_y = 0, 0
cur_x, cur_y = 0, 0
for section in path_sections:
# Replace any variables in section
section_groups = section.split(" ")
for idx, group in enumerate(section_groups):
if group and group[0] == "$":
section_groups[idx] = str(
eval("modifiers[" + group[1:] + "]"))
elif group and group[0] == "?":
section_groups[idx] = str(
eval("eq_results[" + group[2:] + "]"))
section = ' '.join(section_groups)
# print(section)
# Process section
if section[0] in ["Z"]:
cur_x, cur_y = ShapeParser.closepath\
(section, shape_path, path_start_x, path_start_y)
elif section[0] in ["L", "M"]:
cur_x, cur_y, path_start_x, path_start_y = ShapeParser.draw_linemove\
(section, shape_path, path_start_x, path_start_y, scales, bases, adjs)
elif section[0] in ["C"]:
cur_x, cur_y = ShapeParser.draw_curve(section, shape_path,
scales, bases, adjs)
elif section[0] in ["N"]:
# N = endpath
print("N not supported")
elif section[0] in ["F", "S"]:
# F = nofill, S = nostroke
print("FS not supported")
elif section[0] in ["Q"]:
# TODO: Q
# Q = quadratic-curveto (x1 y1 x y)+ Q (x1 y1 x y)+
print("Q not yet supported")
elif section[0] in ["A", "B", "V", "W"]:
cur_x, cur_y, path_start_x, path_start_y = ShapeParser.draw_arc\
(section, shape_path, path_start_x, path_start_y, scales, bases, adjs)
elif section[0] in ["T", "U"]:
cur_x, cur_y, path_start_x, path_start_y = ShapeParser.draw_ellipse_seg\
(section, shape_path, path_start_x, path_start_y, scales, bases, adjs)
elif section[0] in ["X", "Y"]:
cur_x, cur_y = ShapeParser.draw_quadrant\
(section, shape_path, cur_x, cur_y, scales, bases, adjs)
else:
print("Unrecognised command: " + section)
# Apply stroke and fill
StrokeFactory.stroke(pres, dwg, shape, shape_path, 1, style_src)
style_tag = style_src.find("office:automatic-styles")\
.find({"style:style"}, {"style:name": shape["draw:style-name"]})
attrs = style_tag.find("style:graphic-properties").attrs
FillFactory.fill(dwg, shape_path, pres, attrs, \
units_to_float(str(shape["svg:width"])), \
units_to_float(str(shape["svg:height"])), style_tag)
# Apply transformation to shape if needed
if "draw:transform" in shape.attrs:
ShapeParser.transform_shape(shape, shape_path)
# Store xml:id, if it exists
# TODO: Do we need to group the overlaid text with this...?
if "xml:id" in shape.attrs:
pres.xml_ids["obj_" + shape["xml:id"]] = {
"item": shape_path,
"x": base_x,
"y": base_y,
"width": units_to_float(shape["svg:width"]),
"height": units_to_float(shape["svg:height"])
}
shape_path.__setitem__("id", "obj_" + shape["xml:id"])
# Add custom shape to main drawing
layer.add(shape_path)
# Overlay any text
vert_align = "middle"
if "draw:style-name" in shape.attrs:
tb_style_name = shape["draw:style-name"]
tb_style = style_src.find("office:automatic-styles")\
.find({"style:style"}, {"style:name": tb_style_name})
tb_graphics = tb_style.find({"style:graphic-properties"})
if "draw:textarea-vertical-align" in tb_graphics.attrs:
vert_align = tb_graphics["draw:textarea-vertical-align"]
tb_parser = TextBoxParser(dwg, pres, shape, pres.font_mgr, vert_align,
style_src)
tb_parser.visit_textbox(layer, "shape")
def visit_textbox(self, layer, mode):
if mode == "textbox":
item_tb = self.item.find("draw:text-box")
elif mode == "shape":
item_tb = self.item
self.svg_w = units_to_float(self.item["svg:width"])
self.svg_h = units_to_float(self.item["svg:height"])
if "svg:x" in self.item.attrs:
self.svg_x = units_to_float(self.item["svg:x"])
else:
self.svg_x = 0.0
if "svg:y" in self.item.attrs:
self.svg_y = units_to_float(self.item["svg:y"])
else:
self.svg_y = 0.0
# TODO: Now need to transform based on draw:transform if no svg:x,y
self.svg_w_px = self.svg_w * DPCM
self.cur_y = self.svg_y
self.textbox = self.dwg.text('', insert=(self.svg_x, self.svg_y))
# Populate root layer of style stack, keeping track of font family, size and styles
# through the text box tree. Each stack frame stores the styles on each layer of
# the tree, with missing attributes filled in from parent elements. When going back
# up the tree stack frames are popped.
self.style_stack = []
root_stack_frame = {"style:font-name": "", "fo:font-size": "",\
"fo:font-style": "normal", "fo:font-weight": "normal", \
"fo:color": "#000000", "fo:text-align": "start",\
"fo:margin-left": 0, "fo:margin-right": 0, "fo:text-indent": 0,\
"fo:margin-top": 0, "fo:margin-bottom": 0, "fo:line-height": 100,\
"style:text-outline": "false", "fo:text-shadow": "none",\
"style:text-position": "normal", "fo:background-color": "transparent", \
"style:text-underline-style": "none", \
"style:text-underline-color": "font-color", \
"style:text-underline-type": "single", \
"style:text-underline-width": "auto", \
"style:text-overline-style": "none",\
"style:text-overline-color": "font-color", \
"style:text-overline-type": "single", \
"style:text-overline-width": "auto", \
"style:text-line-through-style": "none", \
"style:text-line-through-color": "font-color", \
"style:text-line-through-type": "single", \
"style:text-line-through-width": "auto", \
"style:font-relief": "none"}
if "presentation:style-name" in self.item.attrs:
self.populate_root_frame(root_stack_frame, \
self.item["presentation:style-name"], self.item["draw:text-style-name"])
elif "draw:style-name" in self.item.attrs:
self.populate_root_frame(root_stack_frame, \
self.item["draw:style-name"], self.item["draw:text-style-name"])
else:
self.populate_stack_frame(root_stack_frame, self.item["draw:text-style-name"])
self.style_stack.append(root_stack_frame)
textbox_h = 0
on_first_item = True
prev_after = 0
self.highlights = []
self.decor_lines = []
first_span = None
first_descent = 0
for p_item in item_tb.find_all({"text:p", "text:list"}, recursive=False):
if p_item.name == "text:p":
item_h, item_d, tspan, after_i, h_lights, d_lines = self.visit_p(p_item, False)
elif p_item.name == "text:list":
item_h, item_d, tspan, after_i, h_lights, d_lines = self.visit_list(p_item, 0, None)
textbox_h += item_h + after_i
if on_first_item:
first_span = tspan
on_first_item = False
first_descent = item_d
else:
if tspan:
item_dy = tspan.__getitem__("dy") + prev_after
tspan.__setitem__("dy", item_dy)
prev_after = after_i
# Process highlights and decor_lines
for hl in h_lights:
self.highlights.append(hl)
for dl in d_lines:
self.decor_lines.append(dl)
# Set textbox vertical align
if self.v_align == "top":
v_adj = 0
elif self.v_align == "middle":
v_adj = (self.svg_h - textbox_h) / 2
else:
v_adj = self.svg_h - textbox_h
v_adj -= first_descent
# TODO: Get this to work properly - use horizontal lines to check calcs
# Doesn't yet appear to work with textboxes that have before-spacing in first paragraph
if first_span:
textbox_dy = first_span.__getitem__("dy") + v_adj
first_span.__setitem__("dy", textbox_dy)
for hl in self.highlights:
hl["baseline"] += v_adj
layer.add(self.dwg.rect(
insert=(hl["x"], hl["baseline"] - hl["height"] + hl["descent"]),
size=(hl["width"], hl["height"]),
fill=hl["color"]))
for dl in self.decor_lines:
dl["y"] += v_adj
stroke_w = dl["font-size"]/(DPCM*12)
if dl["line-width"] == "bold":
stroke_w *= 2
if dl["decor-type"][3:] == "single":
dl_ys = [dl["y"]]
else:
stroke_w *= 2/3
dl_ys = [dl["y"] - stroke_w, dl["y"] + stroke_w]
for y in dl_ys:
d_line = self.dwg.line(
start=(dl["x"], y), end=(dl["x"] + dl["width"], y),
stroke=dl["color"], stroke_width=stroke_w)
if dl["style"] in DASH_ARRAYS:
d_array = DASH_ARRAYS[dl["style"]]
for idx, item in enumerate(d_array):
d_array[idx] = item * dl["font-size"] / DPCM
d_line.dasharray(d_array)
if dl["shadow-color"] != "none":
delta_xy = dl["font-size"] / (12*DPCM)
s_line = self.dwg.line(
start=(dl["x"] + delta_xy, y + delta_xy),
end=(dl["x"] + dl["width"] + delta_xy, y + delta_xy),
stroke=dl["shadow-color"], stroke_width=stroke_w)
if dl["style"] in DASH_ARRAYS:
d_array = DASH_ARRAYS[dl["style"]]
for idx, item in enumerate(d_array):
d_array[idx] = item * dl["font-size"] / DPCM
s_line.dasharray(d_array)
layer.add(s_line)
layer.add(d_line)
layer.add(self.textbox)
def visit_list(self, item_l, level, parent_style):
# TODO: Bullet image
# TODO: Relative indents
# Load in list styles
if "text:style-name" in item_l.attrs:
list_style = item_l["text:style-name"]
else:
list_style = parent_style
l_styles = self.style_src.find("office:automatic-styles")\
.find({"text:list-style"}, {"style:name": list_style}).findChildren(recursive=False)
l_height = 0
first_span = None
first_descent = 0
on_first_item = True
prev_after = 0
highlights = []
decor_lines = []
bullet_font, bullet_color = None, None
list_para_style = l_styles[level].find({"style:list-level-properties"})
if "fo:font-family" in l_styles[level].find("style:text-properties").attrs:
bullet_font = l_styles[level].find("style:text-properties")["fo:font-family"]
if bullet_font == "StarSymbol":
bullet_font = "OpenSymbol" # LibreOffice bug 112948
if "fo:color" in l_styles[level].find("style:text-properties").attrs:
bullet_color = l_styles[level].find("style:text-properties")["fo:color"]
bullet_scale = units_to_float(l_styles[level].find("style:text-properties")\
["fo:font-size"])/100
list_stack_frame = self.style_stack[len(self.style_stack) - 1].copy() # shallow copy
if "text:space-before" in list_para_style.attrs:
space_before = units_to_float(list_para_style["text:space-before"])
else:
space_before = 0
if "text:min-label-width" in list_para_style.attrs:
list_stack_frame["fo:margin-left"] = space_before\
+ units_to_float(list_para_style["text:min-label-width"])
else:
list_stack_frame["fo:margin-left"] = space_before
self.style_stack.append(list_stack_frame)
list_header = item_l.find({"text:list-header"})
if level == 0 and list_header:
for subheader_item in list_header.find_all({"text:p"}, recursive=False):
# TODO: Later add in other children of list-header
if subheader_item.name == "text:p":
item_h, item_d, tspan, after_i, h_lights, d_lines = self.visit_p(subheader_item, True)
l_height += item_h + after_i
if on_first_item:
first_span = tspan
on_first_item = False
first_descent = item_d
else:
if tspan:
item_dy = tspan.__getitem__("dy") + prev_after
tspan.__setitem__("dy", item_dy)
prev_after = after_i
# Process highlights and decor_lines
for hl in h_lights:
highlights.append(hl)
for dl in d_lines:
decor_lines.append(dl)
bullet_count = 0
for list_item in item_l.find_all({"text:list-item"}, recursive=False):
for sublist_item in list_item.find_all({"text:p", "text:list"}, recursive=False):
if sublist_item.name == "text:list":
item_h, item_d, tspan, after_i, h_lights, d_lines = \
self.visit_list(sublist_item, (level+1), list_style)
l_height += item_h + after_i
else:
# Add in bullet point
bullet_count += 1
if "text:bullet-char" in l_styles[level].attrs:
list_bullet = l_styles[level]["text:bullet-char"]
else:
# Create numbered bullet based on format specified
list_bullet = int_to_format(bullet_count, \
l_styles[level]["style:num-format"])
if "style:num-prefix" in l_styles[level].attrs:
list_bullet = l_styles[level]["style:num-prefix"] + list_bullet
if "style:num-suffix" in l_styles[level].attrs:
list_bullet = list_bullet + l_styles[level]["style:num-suffix"]
bullet_span = self.dwg.tspan(list_bullet, x=[self.svg_x+space_before], dy=[0])
self.textbox.add(bullet_span)
item_h, item_d, tspan, after_i, h_lights, d_lines = self.visit_p(sublist_item, True)
l_height += item_h + after_i
if tspan.text.strip() == "":
# Remove empty bullet points by deleting bullet text
bullet_span.text = ""
else:
if on_first_item:
first_span = bullet_span
on_first_item = False
first_descent = item_d
else:
if tspan:
item_dy = tspan.__getitem__("dy") + prev_after
tspan.__setitem__("dy", item_dy)
prev_after = after_i
# Copy attributes to bullet_span from tspan
bullet_span.__setitem__("dy", tspan.__getitem__("dy"))
tspan.__setitem__("dy", 0)
bullet_style = tspan.__getitem__("style")
if bullet_font:
font_start = bullet_style.find("font-family:")
font_end = bullet_style.find(";", font_start + 1)
bullet_style = bullet_style[:font_start+12] + bullet_font + \
bullet_style[font_end:]
size_start = bullet_style.find("font-size:")
size_end = bullet_style.find("pt", size_start + 1)
scaled_size = float(bullet_style[size_start+10:size_end]) * bullet_scale
bullet_style = bullet_style[:size_start+10] + \
str(scaled_size) + bullet_style[size_end:]
if bullet_color:
color_start = bullet_style.find("fill:")
color_end = bullet_style.find(";", color_start + 1)
bullet_style = bullet_style[:color_start+5] + bullet_color + \
bullet_style[color_end:]
bullet_span.__setitem__("style", bullet_style)
# Process highlights and decor_lines
for hl in h_lights:
highlights.append(hl)
for dl in d_lines:
decor_lines.append(dl)
self.style_stack.pop()
# TODO: Get after spacing and bubble up instead of "0"
return l_height, first_descent, first_span, 0, highlights, decor_lines
def stroke(cls, pres, dwg, odp_node, svg_elt, scale_factor, style_src):
# Get stroke parameters from the style tree of odp_node, using the first encountered
# instance of each parameter (i.e. prefer child style over parent style)
style_tag = style_src.find("office:automatic-styles")\
.find({"style:style"}, {"style:name": odp_node["draw:style-name"]})
stroke_params = {}
continue_descent = True
while continue_descent:
style_attrs = style_tag.find({"style:graphic-properties"}).attrs
for x in ["draw:stroke", "draw:stroke-dash", "svg:stroke-color",\
"svg:stroke-width", "svg:stroke-opacity", \
"draw:marker-start", "draw:marker-start-width", "draw:marker-start-center",\
"draw:marker-end", "draw:marker-end-width", "draw:marker-end-center"]:
if x in style_attrs and not x in stroke_params:
stroke_params[x] = style_attrs[x]
if "style:parent-style-name" in style_tag.attrs:
style_tag = pres.styles.find("office:styles")\
.find({"style:style"}, {"style:name": style_tag["style:parent-style-name"]})
else:
continue_descent = False
if stroke_params["draw:stroke"] in ["solid", "dash"]:
stroke_width = units_to_float(
str(stroke_params["svg:stroke-width"]))
if stroke_width == 0.0:
stroke_width = 0.01
stroke_width = stroke_width * scale_factor
svg_elt.stroke(stroke_params["svg:stroke-color"])
svg_elt.stroke(width=stroke_width)
if "svg:stroke-opacity" in stroke_params:
svg_elt.stroke(opacity=int(re.sub(r'[^0-9.]', '', \
str(stroke_params["svg:stroke-opacity"])))/100)
else:
svg_elt.stroke(width=0)
if stroke_params["draw:stroke"] == "dash":
dash_node = pres.styles.find("office:styles")\
.find({"draw:stroke-dash"}, {"draw:name": stroke_params["draw:stroke-dash"]})
dash_array = []
gap_length = dash_node["draw:distance"]
if gap_length[-1] == "%":
gap_length = stroke_width * int(gap_length[:-1]) / 100
else:
gap_length = units_to_float(str(gap_length)) * scale_factor
if "draw:dots1-length" in dash_node.attrs:
dots1_length = dash_node["draw:dots1-length"]
else:
dots1_length = "100%"
if dots1_length[-1] == "%":
dots1_length = stroke_width * int(dots1_length[:-1]) / 100
else:
dots1_length = units_to_float(str(dots1_length)) * scale_factor
for i in range(int(dash_node["draw:dots1"])):
dash_array.append(dots1_length)
dash_array.append(gap_length)
if "draw:dots2" in dash_node.attrs:
if "draw:dots2-length" in dash_node.attrs:
dots2_length = dash_node["draw:dots2-length"]
else:
dots2_length = "100%"
if dots2_length[-1] == "%":
dots2_length = stroke_width * int(dots2_length[:-1]) / 100
else:
dots2_length = units_to_float(
str(dots2_length)) * scale_factor
for j in range(int(dash_node["draw:dots2"])):
dash_array.append(dots2_length)
dash_array.append(gap_length)
# Apply dash array to stroke
svg_elt.dasharray(dash_array)
# Add start and end markers to lines
if odp_node.name in ["draw:line"]:
line_angle = math.degrees(math.atan2(\
units_to_float(odp_node.attrs["svg:y2"]) - \
units_to_float(odp_node.attrs["svg:y1"]),\
units_to_float(odp_node.attrs["svg:x2"]) - \
units_to_float(odp_node.attrs["svg:x1"])))
markers = [None, None, None]
if "draw:marker-start" in stroke_params:
s_marker_style = pres.styles.find("office:styles")\
.find({"draw:marker"}, {"draw:name": stroke_params["draw:marker-start"]})
s_marker_vb = s_marker_style["svg:viewbox"].split()
s_marker_w = units_to_float(
stroke_params["draw:marker-start-width"])
s_marker_vb_w = int(s_marker_vb[2]) - int(s_marker_vb[0])
s_marker_vb_h = int(s_marker_vb[3]) - int(s_marker_vb[1])
s_marker_d = s_marker_style["svg:d"]
s_ref_y = 0.0
if "draw:marker-start-center" in stroke_params:
if stroke_params["draw:marker-start-center"] == 'true':
s_ref_y = 0.5
s_marker = dwg.marker(insert=(0.5 * s_marker_vb_w, s_ref_y * s_marker_vb_h),\
size=(s_marker_w, s_marker_w * s_marker_vb_h / s_marker_vb_w), \
viewBox=(' '.join(s_marker_vb)), markerUnits="userSpaceOnUse",\
fill=stroke_params["svg:stroke-color"], orient=line_angle-90)
s_marker_path = dwg.path(d=s_marker_d)
s_marker.add(s_marker_path)
dwg.defs.add(s_marker)
markers[0] = s_marker
if "draw:marker-end" in stroke_params:
e_marker_style = pres.styles.find("office:styles")\
.find({"draw:marker"}, {"draw:name": stroke_params["draw:marker-end"]})
e_marker_vb = e_marker_style["svg:viewbox"].split()
e_marker_w = units_to_float(
stroke_params["draw:marker-end-width"])
e_marker_vb_w = int(e_marker_vb[2]) - int(e_marker_vb[0])
e_marker_vb_h = int(e_marker_vb[3]) - int(e_marker_vb[1])
e_marker_d = e_marker_style["svg:d"]
e_ref_y = 0.0
if "draw:marker-end-center" in stroke_params:
if stroke_params["draw:marker-end-center"] == 'true':
e_ref_y = 0.5
e_marker = dwg.marker(insert=(0.5 * e_marker_vb_w, e_ref_y * e_marker_vb_h),\
size=(e_marker_w, e_marker_w * e_marker_vb_h / e_marker_vb_w), \
viewBox=(' '.join(e_marker_vb)), markerUnits="userSpaceOnUse",\
fill=stroke_params["svg:stroke-color"], orient=line_angle+90)
e_marker_path = dwg.path(d=e_marker_d)
e_marker.add(e_marker_path)
dwg.defs.add(e_marker)
markers[2] = e_marker
svg_elt.set_markers(tuple(markers))
def process_line(self, queue, margins, line_h):
# Create tspan for each span in queue, including any line spacing
# Return line_height, first_span (ref to first tspan in line)
highlights = []
decor_lines = []
line_height = max(x["height"] for x in queue) * line_h / 100
line_descent = max(x["descent"] for x in queue)
line_ascent = max(x["ascent"] for x in queue)
self.cur_y += line_height
sum_widths = math.fsum(x["width"] for x in queue)
available_width = self.svg_w_px - (margins[0] + margins[1]) * DPCM
t_align = queue[0]["stack-frame"]["fo:text-align"]
if t_align == 'start':
first_x = self.svg_x + margins[0]
elif t_align == 'center':
first_x = (available_width - sum_widths)/(DPCM*2) + self.svg_x + margins[0]
elif t_align == 'justify':
first_x = self.svg_x + margins[0]
else:
first_x = (available_width - sum_widths)/DPCM + self.svg_x + margins[0]
span_xs = [first_x]
# Trim leading and trailing spaces for line
queue[0]["text"] = queue[0]["text"].lstrip()
queue[-1]["text"] = queue[-1]["text"].rstrip()
# Cope with completely blank line by inserting unicode non-breaking space
if len(queue) == 1 and queue[0]["text"] == "":
queue[0]["text"] = "\xa0"
word_lens, space_lens = [0], [0]
else:
# Split spans into words to allow highlights and decor lines to line up
queue_copy = deepcopy(queue)
queue.clear()
total_words = 0
total_spaces = 0
word_lens = []
space_lens = []
prev_type = "space"
for span in queue_copy:
span_parts = re.split(r'([ ]+)', span["text"])
span_font = self.font_mgr.findfont(FontProperties(\
family=span["stack-frame"]["style:font-name"], \
style=span["stack-frame"]["fo:font-style"], \
weight=span["stack-frame"]["fo:font-weight"]))
i_font = ImageFont.truetype(span_font, math.ceil(float(span["font-size"])*96/72))
for part in span_parts:
# Get width of part
if part != "":
size = i_font.getsize(part)
# If word then add to queue and add current value of total width to span_x
if part.strip() != "":
if prev_type == "word":
# Add zero size space
space_lens.append(0)
queue.append({"text": part, "style": span["style"], \
"stack-frame": span["stack-frame"].copy(), "width": size[0], \
"height": span["height"], "font-size": span["font-size"]})
word_lens.append(size[0])
total_words += size[0]
prev_type = "word"
else:
if prev_type == "space":
# Add zero size word
queue.append({"text": "\xa0", "style": span["style"], \
"stack-frame": span["stack-frame"].copy(), "width": 0, \
"height": span["height"], "font-size": span["font-size"]})
space_lens.append(size[0])
total_spaces += size[0]
prev_type = "space"
# Divide available space among spaces
if t_align == "justify" and total_spaces > 0:
space_scale = (available_width - total_words) / total_spaces
else:
space_scale = 1
cur_val = 0
if word_lens:
for idx, space in enumerate(space_lens):
cur_val += (space*space_scale) + word_lens[idx]
span_xs.append(cur_val/DPCM + first_x)
# Write out queued spans
hl_x = first_x
prev_highlight = "transparent"
prev_ul_str, prev_ul = "none", None
prev_ol_str, prev_ol = "none", None
prev_lt_str, prev_lt = "none", None
for idx, item in enumerate(queue):
tspan = self.dwg.tspan(item["text"], style=item["style"])
if idx == 0:
first_span = tspan
tspan.__setitem__('dy', line_height)
tspan.__setitem__('x', span_xs[idx])
bs_adj = 0
if item["stack-frame"]["style:text-position"] != "normal":
# Adjust baseline for superscript or subscript text
if item["stack-frame"]["style:text-position"].split()[0] == "super":
bs_adj = item["height"] * -0.6
elif item["stack-frame"]["style:text-position"].split()[0] == "sub":
bs_adj = item["height"] * 0.33
else:
bs_adj = item["height"] * \
-units_to_float(item["stack-frame"]["style:text-position"].split()[0])/100
if bs_adj != 0:
tspan.__setitem__('dy', bs_adj)
self.textbox.add(tspan)
if bs_adj != 0:
# Reset baseline after superscript or subscript
tspan = self.dwg.tspan('\u200B', style=item["style"])
tspan.__setitem__('dy', -float(bs_adj))
self.textbox.add(tspan)
# Record highlight
if item["stack-frame"]["fo:background-color"] != "transparent":
if item["stack-frame"]["fo:background-color"] != prev_highlight:
highlights.append({
"baseline": self.cur_y,
"x": hl_x,
"descent": line_descent,
"width": item["width"]/DPCM,
"height": line_height,
"color": item["stack-frame"]["fo:background-color"]})
prev_highlight = item["stack-frame"]["fo:background-color"]
else:
highlights[-1]["width"] = (item["width"] /DPCM) + hl_x - highlights[-1]["x"]
else:
prev_highlight = "transparent"
if item["stack-frame"]["fo:text-shadow"] == "none":
shadow_color = "none"
elif item["stack-frame"]["fo:color"] == "#000000":
shadow_color = "#cccccc"
else:
shadow_color = "#000000"
# Record underline
if item["stack-frame"]["style:text-underline-style"] != "none":
if item["stack-frame"]["style:text-underline-color"] == "font-color":
line_color = item["stack-frame"]["fo:color"]
else:
line_color = item["stack-frame"]["style:text-underline-color"]
ul_string = item["stack-frame"]["style:text-underline-style"] + "_" + \
item["stack-frame"]["style:text-underline-width"] + "_" + \
item["stack-frame"]["style:text-underline-type"] + "_" + \
line_color + "_" + shadow_color
if ul_string != prev_ul_str:
decor_lines.append({
"x": hl_x,
"y": self.cur_y + line_descent/2,
"width": item["width"]/DPCM,
"font-size": item["font-size"],
"decor-type": "ul-" + item["stack-frame"]["style:text-underline-type"],
"color": line_color,
"style": item["stack-frame"]["style:text-underline-style"],
"line-width": item["stack-frame"]["style:text-underline-width"],
"shadow-color": shadow_color
})
prev_ul_str = ul_string
prev_ul = decor_lines[-1]
else:
prev_ul["width"] = (item["width"] /DPCM) + hl_x - prev_ul["x"]
else:
prev_ul_str = "none"
# Record overline
if item["stack-frame"]["style:text-overline-style"] != "none":
if item["stack-frame"]["style:text-overline-color"] == "font-color":
line_color = item["stack-frame"]["fo:color"]
else:
line_color = item["stack-frame"]["style:text-overline-color"]
ol_string = item["stack-frame"]["style:text-overline-style"] + "_" + \
item["stack-frame"]["style:text-overline-width"] + "_" + \
item["stack-frame"]["style:text-overline-type"] + "_" + \
line_color + "_" + shadow_color
if ol_string != prev_ol_str:
decor_lines.append({
"x": hl_x,
"y": self.cur_y - line_ascent + line_descent/2,
"width": item["width"]/DPCM,
"font-size": item["font-size"],
"decor-type": "ol-" + item["stack-frame"]["style:text-overline-type"],
"color": line_color,
"style": item["stack-frame"]["style:text-overline-style"],
"line-width": item["stack-frame"]["style:text-overline-width"],
"shadow-color": shadow_color
})
prev_ol_str = ol_string
prev_ol = decor_lines[-1]
else:
prev_ol["width"] = (item["width"] /DPCM) + hl_x - prev_ol["x"]
else:
prev_ol_str = "none"
# Record line through
if item["stack-frame"]["style:text-line-through-style"] != "none":
if item["stack-frame"]["style:text-line-through-color"] == "font-color":
line_color = item["stack-frame"]["fo:color"]
else:
line_color = item["stack-frame"]["style:text-line-through-color"]
lt_string = item["stack-frame"]["style:text-line-through-style"] + "_" + \
item["stack-frame"]["style:text-line-through-width"] + "_" + \
item["stack-frame"]["style:text-line-through-type"] + "_" + \
line_color + "_" + shadow_color
if lt_string != prev_lt_str:
decor_lines.append({
"x": hl_x,
"y": self.cur_y + line_descent - (line_height * 0.4),
"width": item["width"]/DPCM,
"font-size": item["font-size"],
"decor-type": "lt-" + item["stack-frame"]["style:text-line-through-type"],
"color": line_color,
"style": item["stack-frame"]["style:text-line-through-style"],
"line-width": item["stack-frame"]["style:text-line-through-width"],
"shadow-color": shadow_color
})
prev_lt_str = lt_string
prev_lt = decor_lines[-1]
else:
prev_lt["width"] = (item["width"] /DPCM) + hl_x - prev_lt["x"]
else:
prev_lt_str = "none"
if idx < len(space_lens):
hl_x += ((item["width"] + space_lens[idx]) / DPCM)
return line_height, line_descent, first_span, highlights, decor_lines