Merge pull request #8 from dhdaines/playa05

Update for PLAYA 0.5

Author: dhdaines

pyproject.toml

@@ -30,7 +30,7 @@ classifiers = [
     "Programming Language :: Python :: Implementation :: PyPy",
 ]
 dependencies = [
-    "playa-pdf>=0.3.0,<=0.4.2",
+    "playa-pdf>=0.5.0",
     "pillow",
 ]
 
@@ -58,7 +58,7 @@ exclude = [
 extra-dependencies = [ "pdfminer.six", "pandas", "polars-lts-cpu", "pypdfium2" ]
 
 [tool.hatch.envs.default]
-dependencies = [ "pytest", "mypy", "pdfminer.six", "pandas", "polars-lts-cpu", "pypdfium2" ]
+dependencies = [ "pytest", "pytest-xdist", "mypy", "pdfminer.six", "pandas", "polars-lts-cpu", "pypdfium2" ]
 
 [tool.hatch.envs.hatch-static-analysis]
 config-path = "none"  # Disable hatch's unreasonable ruff defaults

src/paves/bears.py

@@ -22,7 +22,8 @@
     TextObject,
     XObjectObject,
 )
-from playa.utils import Point, apply_matrix_norm, apply_matrix_pt, get_bound
+from playa.utils import Point, apply_matrix_pt, get_bound
+from paves.compat import subpaths
 
 # Stub out Polars if not present
 try:
@@ -251,7 +252,7 @@ def make_path(
 
 @process_object.register
 def _(obj: PathObject) -> Iterator[LayoutDict]:
-    for path in obj:
+    for path in subpaths(obj):
         ops = []
         pts: List[Point] = []
         for seg in path.raw_segments:
@@ -363,18 +364,16 @@ def _(obj: ImageObject) -> Iterator[LayoutDict]:
 def _(obj: TextObject) -> Iterator[LayoutDict]:
     for glyph in obj:
         x0, y0, x1, y1 = glyph.bbox
-        tstate = glyph.textstate
         gstate = glyph.gstate
-        # apparently we can assert this?
-        font = tstate.font
-        assert font is not None
-        glyph_x, glyph_y = apply_matrix_norm(glyph.ctm, tstate.glyph_offset)
+        font = glyph.font
+        glyph_origin_x, glyph_origin_y = glyph.origin
+        line_origin_x, line_origin_y = obj.line_matrix[-2:]
         (a, b, c, d, e, f) = glyph.matrix
         if font.vertical:
-            size = abs(tstate.fontsize * a)
+            size = abs(gstate.fontsize * a)
         else:
-            size = abs(tstate.fontsize * d)
-        scaling = tstate.scaling * 0.01  # FIXME: unnecessary?
+            size = abs(gstate.fontsize * d)
+        scaling = gstate.scaling * 0.01  # FIXME: unnecessary?
         upright = a * d * scaling > 0 and b * c <= 0
 
         yield LayoutDict(
@@ -388,9 +387,9 @@ def _(obj: TextObject) -> Iterator[LayoutDict]:
             text=glyph.text,
             cid=glyph.cid,
             fontname=font.fontname,
-            glyph_offset_x=glyph_x,
-            glyph_offset_y=glyph_y,
-            render_mode=tstate.render_mode,
+            glyph_offset_x=glyph_origin_x - line_origin_x,
+            glyph_offset_y=glyph_origin_y - line_origin_y,
+            render_mode=gstate.render_mode,
             dash_pattern=gstate.dash.dash,
             dash_phase=gstate.dash.phase,
             stroking_colorspace=gstate.scs.name,

src/paves/compat.py

@@ -0,0 +1,46 @@
+"""
+Compatibility functions.
+"""
+
+from typing import Iterator, List
+
+from playa.content import PathObject, PathSegment
+
+
+def subpaths(path: PathObject) -> Iterator[PathObject]:
+    """Iterate over "subpaths".
+
+    Note: subpaths inherit the values of `fill` and `evenodd` from
+    the parent path, but these values are no longer meaningful
+    since the winding rules must be applied to the composite path
+    as a whole (this is not a bug, just don't rely on them to know
+    which regions are filled or not).
+
+    """
+    # FIXME: Is there an itertool or a more_itertool for this?
+    segs: List[PathSegment] = []
+    for seg in path.raw_segments:
+        if seg.operator == "m" and segs:
+            yield PathObject(
+                _pageref=path._pageref,
+                gstate=path.gstate,
+                ctm=path.ctm,
+                mcstack=path.mcstack,
+                raw_segments=segs,
+                stroke=path.stroke,
+                fill=path.fill,
+                evenodd=path.evenodd,
+            )
+            segs = []
+        segs.append(seg)
+    if segs:
+        yield PathObject(
+            _pageref=path._pageref,
+            gstate=path.gstate,
+            ctm=path.ctm,
+            mcstack=path.mcstack,
+            raw_segments=segs,
+            stroke=path.stroke,
+            fill=path.fill,
+            evenodd=path.evenodd,
+        )

src/paves/image.py

@@ -3,7 +3,6 @@
 models and/or visualisation.`
 """
 
-import itertools
 import functools
 import subprocess
 import tempfile
@@ -24,8 +23,7 @@
 from playa.document import Document, PageList
 from playa.page import ContentObject, Page, Annotation
 from playa.structure import Element
-from playa.utils import Rect, get_transformed_bound, get_bound
-from playa import resolve
+from playa.utils import Rect, get_transformed_bound
 
 if TYPE_CHECKING:
     import pypdfium2  # types: ignore
@@ -331,7 +329,8 @@ def get_box_rect(obj: Rect) -> Rect:
 
 
 @get_box.register(ContentObject)
-def get_box_content(obj: ContentObject) -> Rect:
+@get_box.register(Element)
+def get_box_content(obj: Union[ContentObject, Element]) -> Rect:
     """Get the bounding box of a ContentObject"""
     return obj.bbox
 
@@ -342,47 +341,6 @@ def get_box_annotation(obj: Annotation) -> Rect:
     return get_transformed_bound(obj.page.ctm, obj.rect)
 
 
-@get_box.register(Element)
-def get_box_element(obj: Element) -> Rect:
-    """Get the bounding box for a structural Element"""
-    # It might just *have* a BBox already
-    page = obj.page
-    if page is None:
-        raise ValueError("Has no page, has no content! No box for you!")
-    if "BBox" in obj.props:
-        return get_transformed_bound(page.ctm, resolve(obj.props["BBox"]))
-    else:
-        return _get_marked_content_box(obj)
-
-
-def _get_marked_content_box(el: Element) -> Rect:
-    """Get the bounding box of an Element's marked content.
-
-    This will be superseded in PLAYA 0.5 so do not use!
-    """
-    page = el.page
-    if page is None:
-        raise ValueError("Has no page, has no content! No box for you!")
-
-    def get_mcids(k):
-        k = resolve(k)
-        if isinstance(k, int):
-            yield k
-        elif isinstance(k, list):
-            for kk in k:
-                yield from get_mcids(kk)
-        elif isinstance(k, dict):
-            if "K" in k:
-                yield from get_mcids(k["K"])
-
-    mcids = set(get_mcids(el.props["K"]))
-    pts = itertools.chain.from_iterable(
-        ((x0, y0), (x1, y1))
-        for x0, y0, x1, y1 in (obj.bbox for obj in page if obj.mcid in mcids)
-    )
-    return get_bound(pts)
-
-
 @functools.singledispatch
 def get_label(obj: Union[Annotation, ContentObject, Element, Rect]) -> str:
     """Default function to get the label text for an object."""

src/paves/miner.py

@@ -31,7 +31,7 @@
 from playa.data_structures import NameTree, NumberTree
 from playa.document import Document as PDFDocument
 from playa.exceptions import PDFException
-from playa.page import (
+from playa.content import (
     ContentObject,
     GlyphObject,
     GraphicState,
@@ -52,7 +52,9 @@
     apply_matrix_pt,
     decode_text,
     get_bound,
+    transform_bbox,
 )
+from paves.compat import subpaths
 
 PSException = Exception
 __all__ = [
@@ -519,33 +521,45 @@ def __init__(
         glyph: GlyphObject,
     ) -> None:
         LTText.__init__(self)
-        textstate = glyph.textstate
         gstate = glyph.gstate
         matrix = glyph.matrix
+        font = glyph.font
         if glyph.text is None:
-            logger.debug("undefined: %r, %r", textstate.font, glyph.cid)
+            logger.debug("undefined: %r, %r", font, glyph.cid)
             # Horrible awful pdfminer.six behaviour
             self._text = "(cid:%d)" % glyph.cid
         else:
             self._text = glyph.text
-        self.matrix = matrix
         self.mcstack = glyph.mcstack
-        font = textstate.font
-        assert font is not None
         self.fontname = font.fontname
-        self.render_mode = textstate.render_mode
         self.graphicstate = gstate
+        self.render_mode = gstate.render_mode
         self.stroking_color = gstate.scolor
         self.non_stroking_color = gstate.ncolor
         self.scs = gstate.scs
         self.ncs = gstate.ncs
-        self.adv = glyph.adv
+        scaling = gstate.scaling * 0.01
+        fontsize = gstate.fontsize
         (a, b, c, d, e, f) = matrix
-        scaling = textstate.scaling
         # FIXME: Still really not sure what this means
         self.upright = a * d * scaling > 0 and b * c <= 0
-        LTComponent.__init__(self, glyph.bbox, glyph.mcstack)
-        # FIXME: This is now quite wrong for rotated glyphs
+        # Unscale the matrix to match pdfminer.six
+        xscale = 1 / (fontsize * scaling)
+        yscale = 1 / fontsize
+        self.matrix = (a * xscale, b * yscale, c * xscale, d * yscale, e, f)
+        # Recreate pdfminer.six's bogus bboxes
+        if font.vertical:
+            vdisp = font.vdisp(glyph.cid)
+            self.adv = vdisp * fontsize
+            vx, vy = font.position(glyph.cid)
+            textbox = (-vx, vy + vdisp, -vx + 1, vy)
+        else:
+            textwidth = font.char_width(glyph.cid)
+            self.adv = textwidth * fontsize * scaling
+            textbox = (0, font.get_descent(), textwidth, font.get_descent() + 1)
+        miner_box = transform_bbox(glyph.matrix, textbox)
+        LTComponent.__init__(self, miner_box, glyph.mcstack)
+        # FIXME: This is quite wrong for rotated glyphs, but so is pdfminer.six
         if font.vertical:
             self.size = self.width
         else:
@@ -1157,7 +1171,7 @@ def process_object(obj: ContentObject) -> Iterator[LTComponent]:
 
 @process_object.register
 def _(obj: PathObject) -> Iterator[LTComponent]:
-    for path in obj:
+    for path in subpaths(obj):
         ops = []
         pts: List[Point] = []
         for seg in path.raw_segments: