Add .curve_edges, use by default in table-detect.

Most of the groundwork was already there to add a PDF/Page.curve_edges
property. And, inspired, by
#858 and related issues,
we now include 0/90/180/270-degree oriented curve segments into the
default table-detection strategy. As before, you can still switch to the
"lines_strict" to use only lines defined as such (rather than also using
rect and curve edges).

README.md

@@ -230,7 +230,9 @@ my_char_rotation = my_char_ctm.skew_x
 |`non_stroking_color`|The curve’s fill color.|
 |`object_type`| "curve"|
 
-Additionally, both `pdfplumber.PDF` and `pdfplumber.Page` provide access to two derived lists of objects: `.rect_edges` (which decomposes each rectangle into its four lines) and `.edges` (which combines `.rect_edges` with `.lines`). 
+#### Derived properties
+
+Additionally, both `pdfplumber.PDF` and `pdfplumber.Page` provide access to several derived lists of objects: `.rect_edges` (which decomposes each rectangle into its four lines), `.curve_edges` (which does the same for `curve` objects), and `.edges` (which combines `.rect_edges`, `.curve_edges`, and `.lines`). 
 
 #### `image` properties
 

pdfplumber/container.py

@@ -10,7 +10,7 @@
 
 
 class Container(object):
-    cached_properties = ["_rect_edges", "_edges", "_objects"]
+    cached_properties = ["_rect_edges", "_curve_edges", "_edges", "_objects"]
 
     @property
     def pages(self) -> Optional[List[Any]]:
@@ -73,12 +73,20 @@ def rect_edges(self) -> T_obj_list:
         self._rect_edges: T_obj_list = list(chain(*rect_edges_gen))
         return self._rect_edges
 
+    @property
+    def curve_edges(self) -> T_obj_list:
+        if hasattr(self, "_curve_edges"):
+            return self._curve_edges
+        curve_edges_gen = (utils.curve_to_edges(r) for r in self.curves)
+        self._curve_edges: T_obj_list = list(chain(*curve_edges_gen))
+        return self._curve_edges
+
     @property
     def edges(self) -> T_obj_list:
         if hasattr(self, "_edges"):
             return self._edges
         line_edges = list(map(utils.line_to_edge, self.lines))
-        self._edges: T_obj_list = self.rect_edges + line_edges
+        self._edges: T_obj_list = line_edges + self.rect_edges + self.curve_edges
         return self._edges
 
     @property

pdfplumber/utils/geometry.py

@@ -193,6 +193,7 @@ def curve_to_edges(curve: T_obj) -> T_obj_list:
     point_pairs = zip(curve["pts"], curve["pts"][1:])
     return [
         {
+            "object_type": "curve_edge",
             "x0": min(p0[0], p1[0]),
             "x1": max(p0[0], p1[0]),
             "top": min(p0[1], p1[1]),
@@ -253,12 +254,13 @@ def line_to_edge(line: T_obj) -> T_obj:
 
 
 def obj_to_edges(obj: T_obj) -> T_obj_list:
-    return {
-        "line": lambda x: [line_to_edge(x)],
-        "rect": rect_to_edges,
-        "rect_edge": rect_to_edges,
-        "curve": curve_to_edges,
-    }[obj["object_type"]](obj)
+    t = obj["object_type"]
+    if "_edge" in t:
+        return [obj]
+    elif t == "line":
+        return [line_to_edge(obj)]
+    else:
+        return {"rect": rect_to_edges, "curve": curve_to_edges}[t](obj)
 
 
 def filter_edges(

tests/test_basics.py

@@ -17,10 +17,14 @@ class Test(unittest.TestCase):
     def setup_class(self):
         path = os.path.join(HERE, "pdfs/nics-background-checks-2015-11.pdf")
         self.pdf = pdfplumber.open(path)
+        # via http://www.pdfill.com/example/pdf_drawing_new.pdf
+        path_2 = os.path.join(HERE, "pdfs/pdffill-demo.pdf")
+        self.pdf_2 = pdfplumber.open(path_2)
 
     @classmethod
     def teardown_class(self):
         self.pdf.close()
+        self.pdf_2.close()
 
     def test_metadata(self):
         metadata = self.pdf.metadata
@@ -38,18 +42,18 @@ def test_objects(self):
         assert len(self.pdf.rects)
         assert len(self.pdf.lines)
         assert len(self.pdf.rect_edges)
+        assert len(self.pdf_2.curve_edges)
         # Ensure that caching is working:
         assert id(self.pdf._rect_edges) == id(self.pdf.rect_edges)
+        assert id(self.pdf_2._curve_edges) == id(self.pdf_2.curve_edges)
         assert id(self.pdf.pages[0]._layout) == id(self.pdf.pages[0].layout)
 
     def test_annots(self):
-        # via http://www.pdfill.com/example/pdf_drawing_new.pdf
-        path = os.path.join(HERE, "pdfs/pdffill-demo.pdf")
-        with pdfplumber.open(path) as pdf:
-            assert len(pdf.annots)
-            assert len(pdf.hyperlinks) == 17
-            uri = "http://www.pdfill.com/pdf_drawing.html"
-            assert pdf.hyperlinks[0]["uri"] == uri
+        pdf = self.pdf_2
+        assert len(pdf.annots)
+        assert len(pdf.hyperlinks) == 17
+        uri = "http://www.pdfill.com/pdf_drawing.html"
+        assert pdf.hyperlinks[0]["uri"] == uri
 
         path = os.path.join(HERE, "pdfs/annotations.pdf")
         with pdfplumber.open(path) as pdf:

tests/test_table.py

@@ -211,3 +211,16 @@ def test_discussion_539_null_value(self):
             }
             assert page.extract_table(table_settings)
             assert page.extract_tables(table_settings)
+
+    def test_table_curves(self):
+        # See https://github.com/jsvine/pdfplumber/discussions/808
+        path = os.path.join(HERE, "pdfs/table-curves-example.pdf")
+        with pdfplumber.open(path) as pdf:
+            page = pdf.pages[0]
+            assert len(page.curves)
+            tables = page.extract_tables()
+            assert len(tables) == 1
+            t = tables[0]
+            assert t[-2][-2] == "Uncommon"
+
+            assert len(page.extract_tables({"vertical_strategy": "lines_strict"})) == 0