Adding annotations to the PDF to link back its content to its source.

weasyprint/__init__.py

@@ -45,6 +45,9 @@
 #:     Whether custom HTML metadata should be stored in the generated PDF.
 #: :param bool presentational_hints:
 #:     Whether HTML presentational hints are followed.
+#: :param bool srgb:
+#:     Whether sRGB color profile should be included and set as default for
+#:     device-dependant RGB colors.
 #: :param bool optimize_images:
 #:     Whether size of embedded images should be optimized, with no quality
 #:     loss.
@@ -71,6 +74,7 @@
     'uncompressed_pdf': False,
     'custom_metadata': False,
     'presentational_hints': False,
+    'srgb': False,
     'optimize_images': False,
     'jpeg_quality': None,
     'dpi': None,
@@ -210,6 +214,8 @@ def render(self, font_config=None, counter_style=None, **options):
         :returns: A :class:`document.Document` object.
 
         """
+        for unknown in set(options) - set(DEFAULT_OPTIONS):
+            LOGGER.warning('Unknown rendering option: %s.', unknown)
         new_options = DEFAULT_OPTIONS.copy()
         new_options.update(options)
         options = new_options

weasyprint/__main__.py

@@ -98,6 +98,9 @@ def docstring(self):
 PARSER.add_argument(
     '-p', '--presentational-hints', action='store_true',
     help='follow HTML presentational hints')
+PARSER.add_argument(
+    '--srgb', action='store_true',
+    help='include sRGB color profile')
 PARSER.add_argument(
     '--optimize-images', action='store_true',
     help='optimize size of embedded images with no quality loss')
@@ -165,7 +168,8 @@ def main(argv=None, stdout=None, stdin=None, HTML=HTML):  # noqa: N803
     if args.timeout is not None:
         url_fetcher = partial(default_url_fetcher, timeout=args.timeout)
 
-    options = vars(args)
+    options = {
+        key: value for key, value in vars(args).items() if key in DEFAULT_OPTIONS}
 
     # Default to logging to stderr.
     if args.debug:
@@ -174,7 +178,14 @@ def main(argv=None, stdout=None, stdin=None, HTML=HTML):  # noqa: N803
         LOGGER.setLevel(logging.INFO)
     if not args.quiet:
         handler = logging.StreamHandler()
-        handler.setFormatter(logging.Formatter('%(levelname)s: %(message)s'))
+        if args.debug:
+            # Add extra information when debug logging
+            handler.setFormatter(
+                logging.Formatter(
+                    '%(levelname)s: %(filename)s:%(lineno)d '
+                    '(%(funcName)s): %(message)s'))
+        else:
+            handler.setFormatter(logging.Formatter('%(levelname)s: %(message)s'))
         LOGGER.addHandler(handler)
 
     html = HTML(

weasyprint/anchors.py

@@ -113,13 +113,17 @@ def gather_anchors(box, anchors, links, bookmarks, forms, parent_matrix=None,
             links.append((link_type, target, rectangle, box))
         if is_input:
             forms[parent_form].append((box.element, box.style, rectangle))
-        if matrix and (has_bookmark or has_anchor):
-            pos_x, pos_y = matrix.transform_point(pos_x, pos_y)
         if has_bookmark:
+            if matrix:
+                pos_x, pos_y = matrix.transform_point(pos_x, pos_y)
             bookmark = (bookmark_level, bookmark_label, (pos_x, pos_y), state)
             bookmarks.append(bookmark)
         if has_anchor:
-            anchors[anchor_name] = pos_x, pos_y
+            pos_x1, pos_y1, pos_x2, pos_y2 = pos_x, pos_y, pos_x + width, pos_y + height
+            if matrix:
+                pos_x1, pos_y1 = matrix.transform_point(pos_x1, pos_y1)
+                pos_x2, pos_y2 = matrix.transform_point(pos_x2, pos_y2)
+            anchors[anchor_name] = (pos_x1, pos_y1, pos_x2, pos_y2)
 
     for child in box.all_children():
         gather_anchors(child, anchors, links, bookmarks, forms, matrix, parent_form)

weasyprint/css/html5_ua.css

@@ -161,14 +161,14 @@ input[type="reset"]:not([value])::before {
 }
 input[type="checkbox"],
 input[type="radio"] {
-  height: 1.2em;
-  width: 1.2em;
+  height: 0.7em;
+  vertical-align: -0.2em;
+  width: 0.7em;
 }
 input[type="checkbox"][checked]:before,
 input[type="radio"][checked]:before {
   background: black;
   content: "";
-  display: block;
   height: 100%;
 }
 input[type="radio"][checked]:before {
@@ -179,7 +179,6 @@ input[type="hidden"] {
 }
 input[type="radio"] {
   border-radius: 50%;
-  margin: 0.2em 0.2em 0 0.4em;
 }
 input[value]::before {
   content: attr(value);
@@ -191,6 +190,7 @@ input[value=""]::before,
 input[type="checkbox"]::before,
 input[type="radio"]::before {
   content: "";
+  display: block;
 }
 select {
   background: lightgrey;

weasyprint/css/html5_ua_form.css

@@ -3,3 +3,13 @@
 button, input, select, textarea {
   appearance: auto;
 }
+
+select option,
+select:not([multiple])::before,
+input:not([type="submit"])::before {
+  visibility: hidden;
+}
+
+textarea {
+  color: transparent;
+}

weasyprint/draw/__init__.py

@@ -326,7 +326,7 @@ def draw_background_image(stream, layer, image_rendering):
     with stacked(stream):
         layer.image.draw(group, image_width, image_height, image_rendering)
         pattern.draw_x_object(group.id)
-        stream.color_space('Pattern')
+        stream.set_color_space('Pattern')
         stream.set_color_special(pattern.id)
         if layer.unbounded:
             x1, y1, x2, y2 = stream.page_rectangle

weasyprint/draw/border.py

@@ -51,7 +51,10 @@ def draw_border(stream, box):
             with stacked(stream):
                 rule_width = box.style['column_rule_width']
                 rule_style = box.style['column_rule_style']
-                gap = percentage(box.style['column_gap'], box.width)
+                if box.style['column_gap'] == 'normal':
+                    gap = box.style['font_size']  # normal equals 1em
+                else:
+                    gap = percentage(box.style['column_gap'], box.width)
                 position_x = (
                     child.position_x - (box.style['column_rule_width'] + gap) / 2)
                 border_box = (position_x, child.position_y, rule_width, child.height)

weasyprint/draw/text.py

@@ -126,7 +126,7 @@ def draw_first_line(stream, textbox, text_overflow, block_ellipsis, matrix):
 
     utf8_text = textbox.pango_layout.text.encode()
     previous_utf8_position = 0
-    stream.text_matrix(*matrix.values)
+    stream.set_text_matrix(*matrix.values)
     last_font = None
     string = ''
     x_advance = 0

weasyprint/formatting_structure/build.py

@@ -223,7 +223,7 @@ def element_to_box(element, style_for, get_image_from_uri, base_url,
         if not counter_values[name]:
             counter_values.pop(name)
 
-    box.children = children if style['appearance'] == 'none' else []
+    box.children = children
     process_whitespace(box)
     set_content_lists(
         element, box, style, counter_values, target_collector, counter_style)

weasyprint/images.py

@@ -539,7 +539,7 @@ def draw(self, stream, concrete_width, concrete_height, _image_rendering):
             alpha_stream.transform(d=scale_y)
             alpha_stream.stream = [f'/{alpha_shading.id} sh']
 
-        stream.shading(shading.id)
+        stream.paint_shading(shading.id)
 
     def layout(self, width, height):
         """Get layout information about the gradient.

weasyprint/layout/flex.py

@@ -5,6 +5,7 @@
 
 from ..css.properties import Dimension
 from ..formatting_structure import boxes
+from .absolute import AbsolutePlaceholder, absolute_layout
 from .percent import resolve_one_percentage, resolve_percentages
 from .preferred import max_content_width, min_content_width
 from .table import find_in_flow_baseline
@@ -21,6 +22,10 @@ def flex_layout(context, box, bottom_space, skip_stack, containing_block,
     context.create_block_formatting_context()
     resume_at = None
 
+    if box.style['position'] == 'relative':
+        # New containing block, use a new absolute list
+        absolute_boxes = []
+
     # Step 1 is done in formatting_structure.boxes
     # Step 2
     if box.style['flex_direction'].startswith('row'):
@@ -111,8 +116,19 @@ def flex_layout(context, box, bottom_space, skip_stack, containing_block,
     else:
         skip_stack = None
     child_skip_stack = skip_stack
-    for child in children:
+    for index, child in enumerate(children):
         if not child.is_flex_item:
+            # Absolute child layout: create placeholder.
+            if child.is_absolutely_positioned():
+                child.position_x = box.content_box_x()
+                child.position_y = box.content_box_y()
+                new_child = placeholder = AbsolutePlaceholder(child)
+                placeholder.index = index
+                children[index] = placeholder
+                if child.style['position'] == 'absolute':
+                    absolute_boxes.append(placeholder)
+                else:
+                    fixed_boxes.append(placeholder)
             continue
 
         # See https://www.w3.org/TR/css-flexbox-1/#min-size-auto
@@ -839,8 +855,8 @@ def flex_layout(context, box, bottom_space, skip_stack, containing_block,
 
     # TODO: don't use block_box_layout, see TODOs in Step 14 and
     # build.flex_children.
-    box = box.copy()
-    box.children = []
+    box = box.copy_with_children(
+        [child for child in children if not child.is_flex_item])
     child_skip_stack = skip_stack
     for line in flex_lines:
         for i, child in line:
@@ -876,6 +892,13 @@ def flex_layout(context, box, bottom_space, skip_stack, containing_block,
         if resume_at:
             break
 
+    if box.style['position'] == 'relative':
+        # New containing block, resolve the layout of the absolute descendants
+        for absolute_box in absolute_boxes:
+            absolute_layout(
+                context, absolute_box, box, fixed_boxes, bottom_space,
+                skip_stack=None)
+
     # Set box height
     # TODO: this is probably useless because of step #15
     if axis == 'width' and box.height == 'auto':

weasyprint/layout/preferred.py

@@ -71,7 +71,7 @@ def max_content_width(context, box, outer=True):
     if box.is_table_wrapper:
         return table_and_columns_preferred_widths(context, box, outer)[1]
     elif isinstance(box, boxes.TableCellBox):
-        return table_cell_max_content_width(context, box, outer)
+        return table_cell_min_max_content_width(context, box, outer)[1]
     elif isinstance(box, (
             boxes.BlockContainerBox, boxes.TableColumnBox, boxes.FlexBox)):
         return block_max_content_width(context, box, outer)
@@ -265,22 +265,18 @@ def table_cell_min_content_width(context, box, outer):
     return children_min_width
 
 
-def table_cell_max_content_width(context, box, outer):
-    """Return the max-content width for a ``TableCellBox``."""
-    return max(
-        table_cell_min_content_width(context, box, outer),
-        block_max_content_width(context, box, outer))
+def table_cell_min_max_content_width(context, box, outer=True):
+    """Return the min- and max-content width for a ``TableCellBox``."""
+    # This is much faster than calling min and max separately.
+    min_width = table_cell_min_content_width(context, box, outer)
+    max_width = max(min_width, block_max_content_width(context, box, outer))
+    return min_width, max_width
 
 
-def inline_line_widths(context, box, outer, is_line_start, minimum,
-                       skip_stack=None, first_line=False):
-    if isinstance(box, boxes.LineBox):
-        if box.style['text_indent'].unit == '%':
-            # TODO: this is wrong, text-indent percentages should be resolved
-            # before calling this function.
-            text_indent = 0
-        else:
-            text_indent = box.style['text_indent'].value
+def inline_line_widths(context, box, outer, is_line_start, minimum, skip_stack=None,
+                       first_line=False):
+    if isinstance(box, boxes.LineBox) and box.style['text_indent'].unit != '%':
+        text_indent = box.style['text_indent'].value
     else:
         text_indent = 0
 
@@ -459,37 +455,31 @@ def table_and_columns_preferred_widths(context, box, outer=True):
     colspan_cells = []
 
     # Define the intermediate content widths
-    min_content_widths = [0 for i in range(grid_width)]
-    max_content_widths = [0 for i in range(grid_width)]
-    intrinsic_percentages = [0 for i in range(grid_width)]
+    min_content_widths = [0] * grid_width
+    max_content_widths = [0] * grid_width
+    intrinsic_percentages = [0] * grid_width
 
     # Intermediate content widths for span 1
     for i in range(grid_width):
         for groups in (column_groups, columns):
-            if groups[i]:
+            if group := groups[i]:
                 min_content_widths[i] = max(
-                    min_content_widths[i],
-                    min_content_width(context, groups[i]))
+                    min_content_widths[i], min_content_width(context, group))
                 max_content_widths[i] = max(
-                    max_content_widths[i],
-                    max_content_width(context, groups[i]))
+                    max_content_widths[i], max_content_width(context, group))
                 intrinsic_percentages[i] = max(
-                    intrinsic_percentages[i],
-                    _percentage_contribution(groups[i]))
+                    intrinsic_percentages[i], _percentage_contribution(group))
         for cell in zipped_grid[i]:
-            if cell:
-                if cell.colspan == 1:
-                    min_content_widths[i] = max(
-                        min_content_widths[i],
-                        min_content_width(context, cell))
-                    max_content_widths[i] = max(
-                        max_content_widths[i],
-                        max_content_width(context, cell))
-                    intrinsic_percentages[i] = max(
-                        intrinsic_percentages[i],
-                        _percentage_contribution(cell))
-                else:
-                    colspan_cells.append(cell)
+            if not cell:
+                continue
+            if cell.colspan == 1:
+                min_width, max_width = table_cell_min_max_content_width(context, cell)
+                min_content_widths[i] = max(min_content_widths[i], min_width)
+                max_content_widths[i] = max(max_content_widths[i], max_width)
+                intrinsic_percentages[i] = max(
+                    intrinsic_percentages[i], _percentage_contribution(cell))
+            else:
+                colspan_cells.append(cell)
 
     # Intermediate content widths for span > 1 is wrong in the 4.1 section, as
     # explained in its third issue. Min- and max-content widths are handled by
@@ -501,7 +491,7 @@ def table_and_columns_preferred_widths(context, box, outer=True):
         percentage_contributions = []
         for i in range(grid_width):
             percentage_contribution = intrinsic_percentages[i]
-            for j, cell in enumerate(zipped_grid[i]):
+            for j in range(len(zipped_grid[i])):
                 indexes = [k for k in range(i + 1) if grid[j][k]]
                 if not indexes:
                     continue
@@ -520,8 +510,7 @@ def table_and_columns_preferred_widths(context, box, outer=True):
                         baseline_percentage)
                     other_columns_contributions = [
                         max_content_widths[j]
-                        for j in range(
-                            origin, origin + origin_cell.colspan)
+                        for j in range(origin, origin + origin_cell.colspan)
                         if intrinsic_percentages[j] == 0]
                     other_columns_contributions_sum = sum(
                         other_columns_contributions)

weasyprint/layout/table.py

@@ -190,9 +190,7 @@ def group_layout(group, position_y, bottom_space, page_is_empty, skip_stack):
                 else:
                     cell = new_cell
 
-                cell.remove_decoration(
-                    start=cell_skip_stack is not None,
-                    end=cell_resume_at is not None)
+                cell.remove_decoration(start=cell_skip_stack is not None, end=False)
                 if cell_resume_at:
                     if resume_at is None:
                         resume_at = {index_row: {}}
@@ -206,12 +204,23 @@ def group_layout(group, position_y, bottom_space, page_is_empty, skip_stack):
                 new_row_children.append(cell)
 
             if resume_at and not page_is_empty:
-                if avoid_page_break(row.style['break_inside'], context):
+                # Avoid break when "break-inside: avoid" is set on row or any
+                # on its cells.
+                avoid_break = (
+                    avoid_page_break(row.style['break_inside'], context) or any(
+                        avoid_page_break(cell.style['break_inside'], context)
+                        for cell in row.children))
+                if avoid_break:
                     resume_at = {index_row: {}}
                     remove_placeholders(
                         context, new_row_children, absolute_boxes, fixed_boxes)
                     break
 
+            if resume_at:
+                # Remove bottom decoration if row is split.
+                for cell in new_row_children:
+                    cell.remove_decoration(start=False, end=True)
+
             row = row.copy_with_children(new_row_children)
 
             # Table height algorithm