feat: add FillMode and stroke/fill color attributes to Line and Rect

Author: monchin

CHANGELOG.md

@@ -9,6 +9,9 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### Added
 
+- Add `is_stroked` attribute to `Rect` and `Line`: indicates whether the PDF path is stroked (from `path.is_stroked()`)
+- Add `fill_mode` attribute to `Rect` and `Line`: fill rule from pdfium-render `PdfPathFillMode` (NONE, WINDING, or EVEN_ODD); expose `FillMode` enum to Python with same three values
+- Replace `Line.color` with `Line.stroke_color` and `Line.fill_color` (aligned with `Rect`)
 - Add `Table.to_list()` returning `list[list[TableCellValue]]`: each cell has `text` (or `None` when merged), `merged_left`, and `merged_top` so merge direction (left vs above) is explicit
 - Add `TableCellValue` class with attributes `text`, `merged_left`, and `merged_top` for use with `to_list()`
 - Add `get_intersections_from_edges(h_edges, v_edges, ...)` function: given horizontal and vertical edges (as returned by `get_edges`), returns a mapping from every `(x, y)` intersection point to the edges that pass through it; accepts the same tolerance kwargs as `get_edges`
@@ -21,6 +24,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### Changed
 
+- **Breaking:** `Line.color` has been removed. Use `Line.stroke_color` and `Line.fill_color` instead (aligned with `Rect`).
 - `Page` is now a Python-level wrapper that holds a `doc` back-reference; Rust-side type is `Pyo3Page`
 
 ### Deprecated

docs/reference/api.md

@@ -469,6 +469,16 @@ Represents a rectangle extracted from a PDF page.
 | `fill_color` | `tuple[int, int, int, int]` | Fill color (RGBA) |
 | `stroke_color` | `tuple[int, int, int, int]` | Stroke color (RGBA) |
 | `stroke_width` | `float` | Stroke width |
+| `is_stroked` | `bool` | Whether the path is stroked |
+| `fill_mode` | `FillMode` | Fill rule (NONE, WINDING, or EVEN_ODD) |
+
+---
+
+### FillMode
+
+PDF path fill rule: winding (nonzero) or even-odd.
+
+**Values:** `FillMode.NONE`, `FillMode.WINDING`, `FillMode.EVEN_ODD` (mirrors pdfium-render `PdfPathFillMode`)
 
 ---
 
@@ -482,8 +492,11 @@ Represents a line segment extracted from a PDF page.
 |-----------|------|-------------|
 | `line_type` | `Literal["straight", "polyline", "curve"]` | Type of line |
 | `points` | `list[tuple[float, float]]` | Points defining the line path |
-| `color` | `tuple[int, int, int, int]` | Color (RGBA) |
+| `stroke_color` | `tuple[int, int, int, int]` | Stroke color (RGBA) |
+| `fill_color` | `tuple[int, int, int, int]` | Fill color (RGBA) |
 | `width` | `float` | Line width |
+| `is_stroked` | `bool` | Whether the line is stroked |
+| `fill_mode` | `FillMode` | Fill rule (NONE, WINDING, or EVEN_ODD) |
 
 ---
 

python/tablers/tablers.pyi

@@ -351,12 +351,25 @@ class Rect:
         The stroke (border) color as an RGBA tuple.
     stroke_width : float
         The stroke width of the rectangle border.
+    is_stroked : bool
+        Whether the path is stroked.
+    fill_mode : FillMode
+        Fill rule for the path (NONE, WINDING, or EVEN_ODD; mirrors pdfium-render PdfPathFillMode).
     """
 
     bbox: BBox
     fill_color: Color
     stroke_color: Color
     stroke_width: float
+    is_stroked: bool
+    fill_mode: FillMode
+
+class FillMode:
+    """PDF path fill rule: mirrors pdfium-render PdfPathFillMode (NONE, WINDING, EVEN_ODD)."""
+
+    NONE: FillMode  # Path not filled
+    WINDING: FillMode
+    EVEN_ODD: FillMode
 
 class Line:
     """
@@ -370,16 +383,25 @@ class Line:
         The type of line segment.
     points : list of Point
         The points that define the line path.
-    color : Color
-        The color of the line as an RGBA tuple.
+    stroke_color : Color
+        The stroke color of the line as an RGBA tuple.
+    fill_color : Color
+        The fill color of the line as an RGBA tuple.
     width : float
         The width of the line stroke.
+    is_stroked : bool
+        Whether the line is stroked.
+    fill_mode : FillMode
+        Fill rule for the path (NONE, WINDING, or EVEN_ODD; mirrors pdfium-render PdfPathFillMode).
     """
 
     line_type: Literal["straight", "polyline", "curve"]
     points: list[Point]
-    color: Color
+    stroke_color: Color
+    fill_color: Color
     width: float
+    is_stroked: bool
+    fill_mode: FillMode
 
 class Char:
     """

src/edges.rs

@@ -601,7 +601,7 @@ pub(crate) fn make_edges(
                         x2: p1.0,
                         y2: cmp::max(p1.1, p2.1),
                         width: line.width,
-                        color: line.color,
+                        color: line.stroke_color,
                     });
                 } else if ((h_strat & 0b11u8) != 0)
                     && ((p1.1 - p2.1).abs() < snap_y_tol.into_inner())
@@ -613,7 +613,7 @@ pub(crate) fn make_edges(
                         x2: cmp::max(p1.0, p2.0),
                         y2: p1.1,
                         width: line.width,
-                        color: line.color,
+                        color: line.stroke_color,
                     })
                 }
             } else if line.line_type == LineType::Polyline
@@ -633,7 +633,7 @@ pub(crate) fn make_edges(
                         x2: x_min,
                         y2: y_max,
                         width: x_max - x_min,
-                        color: line.color,
+                        color: line.fill_color,
                     });
                 } else if ((h_strat & 0b11u8) != 0) && ((y_max - y_min) < *snap_y_tol) {
                     edges.get_mut(&Orientation::Horizontal).unwrap().push(Edge {
@@ -643,7 +643,7 @@ pub(crate) fn make_edges(
                         x2: x_max,
                         y2: y_min,
                         width: y_max - y_min,
-                        color: line.color,
+                        color: line.fill_color,
                     });
                 }
             }
@@ -743,7 +743,7 @@ mod tests {
     use crate::test_utils::load_pdfium;
     use crate::words::Word;
     use ordered_float::OrderedFloat;
-    use pdfium_render::prelude::PdfColor;
+    use pdfium_render::prelude::{PdfColor, PdfPathFillMode};
 
     fn of(v: f32) -> OrderedFloat<f32> {
         OrderedFloat(v)
@@ -1140,8 +1140,11 @@ mod tests {
                 .into_iter()
                 .map(|(x, y)| (OrderedFloat(x), OrderedFloat(y)))
                 .collect(),
-            color: PdfColor::new(0, 0, 0, 255),
+            stroke_color: PdfColor::new(0, 0, 0, 255),
+            fill_color: PdfColor::new(0, 0, 0, 255),
             width: OrderedFloat(1.0),
+            is_stroked: false,
+            fill_mode: PdfPathFillMode::Winding,
         }
     }
 
@@ -1153,8 +1156,11 @@ mod tests {
                 .into_iter()
                 .map(|(x, y)| (OrderedFloat(x), OrderedFloat(y)))
                 .collect(),
-            color: PdfColor::new(0, 0, 0, 255),
+            stroke_color: PdfColor::new(0, 0, 0, 255),
+            fill_color: PdfColor::new(0, 0, 0, 255),
             width: OrderedFloat(1.0),
+            is_stroked: true,
+            fill_mode: PdfPathFillMode::None,
         }
     }
 

src/lib.rs

@@ -961,6 +961,7 @@ fn tablers(_py: Python<'_>, m: &Bound<PyModule>) -> PyResult<()> {
     m.add_class::<Table>()?;
     m.add_class::<PyCellGroup>()?;
     m.add_class::<PyTableCellValue>()?;
+    m.add_class::<crate::objects::FillMode>()?;
     m.add_class::<TfSettings>()?;
     m.add_class::<WordsExtractSettings>()?;
     m.add_function(pyo3::wrap_pyfunction!(py_find_all_cells_bboxes, m)?)?;

src/objects.rs

@@ -1,7 +1,31 @@
 use ordered_float::OrderedFloat;
-use pdfium_render::prelude::{PdfColor, PdfPathSegmentType};
+use pdfium_render::prelude::{PdfColor, PdfPathFillMode, PdfPathSegmentType};
 use pyo3::prelude::*;
 
+/// PDF path fill rule: mirrors pdfium-render's PdfPathFillMode for Python exposure.
+#[pyclass]
+#[derive(Clone, Copy, PartialEq, Eq)]
+#[allow(non_camel_case_types)]
+#[allow(clippy::upper_case_acronyms)]
+pub enum FillMode {
+    /// Path is not filled.
+    NONE,
+    /// Nonzero winding rule.
+    WINDING,
+    /// Even-odd rule.
+    EVEN_ODD,
+}
+
+impl From<PdfPathFillMode> for FillMode {
+    fn from(m: PdfPathFillMode) -> Self {
+        match m {
+            PdfPathFillMode::None => FillMode::NONE,
+            PdfPathFillMode::Winding => FillMode::WINDING,
+            PdfPathFillMode::EvenOdd => FillMode::EVEN_ODD,
+        }
+    }
+}
+
 /// Container for all extracted objects from a PDF page.
 ///
 /// This struct holds all rectangles, lines, and characters found in a page.
@@ -49,6 +73,11 @@ pub struct Rect {
     /// The stroke width of the rectangle border.
     #[pyo3(get)]
     pub stroke_width: f32,
+    /// Whether the path is stroked.
+    #[pyo3(get)]
+    pub is_stroked: bool,
+    /// Fill rule for the path (from pdfium-render PdfPathFillMode).
+    pub fill_mode: PdfPathFillMode,
 }
 
 #[pymethods]
@@ -85,6 +114,12 @@ impl Rect {
             self.stroke_color.alpha(),
         )
     }
+
+    /// Returns the fill mode (NONE, WINDING, or EVEN_ODD).
+    #[getter]
+    fn fill_mode(&self) -> FillMode {
+        self.fill_mode.into()
+    }
 }
 
 /// Represents a line segment extracted from a PDF page.
@@ -97,10 +132,17 @@ pub struct Line {
     pub line_type: LineType,
     /// The points that define the line path.
     pub points: Vec<Point>,
-    /// The color of the line.
-    pub color: PdfColor,
+    /// The stroke color of the line.
+    pub stroke_color: PdfColor,
+    /// The fill color of the line.
+    pub fill_color: PdfColor,
     /// The width of the line stroke.
     pub width: OrderedFloat<f32>,
+    /// Whether the line is stroked.
+    #[pyo3(get)]
+    pub is_stroked: bool,
+    /// Fill rule for the path (from pdfium-render PdfPathFillMode).
+    pub fill_mode: PdfPathFillMode,
 }
 
 #[pymethods]
@@ -124,14 +166,25 @@ impl Line {
             .collect()
     }
 
-    /// Returns the line color as an RGBA tuple.
+    /// Returns the stroke color as an RGBA tuple.
+    #[getter]
+    fn stroke_color(&self) -> (u8, u8, u8, u8) {
+        (
+            self.stroke_color.red(),
+            self.stroke_color.green(),
+            self.stroke_color.blue(),
+            self.stroke_color.alpha(),
+        )
+    }
+
+    /// Returns the fill color as an RGBA tuple.
     #[getter]
-    fn color(&self) -> (u8, u8, u8, u8) {
+    fn fill_color(&self) -> (u8, u8, u8, u8) {
         (
-            self.color.red(),
-            self.color.green(),
-            self.color.blue(),
-            self.color.alpha(),
+            self.fill_color.red(),
+            self.fill_color.green(),
+            self.fill_color.blue(),
+            self.fill_color.alpha(),
         )
     }
 
@@ -140,6 +193,12 @@ impl Line {
     fn width(&self) -> f32 {
         self.width.into_inner()
     }
+
+    /// Returns the fill mode (NONE, WINDING, or EVEN_ODD).
+    #[getter]
+    fn fill_mode(&self) -> FillMode {
+        self.fill_mode.into()
+    }
 }
 
 /// Represents a text character extracted from a PDF page.

src/pages.rs

@@ -1,7 +1,7 @@
 use crate::objects::*;
 use ordered_float::OrderedFloat;
 use pdfium_render::prelude::PdfPage as PdfiumPage;
-use pdfium_render::prelude::*;
+use pdfium_render::prelude::{PdfPathFillMode, *};
 use std::cell::RefCell;
 use std::cmp;
 
@@ -248,13 +248,18 @@ impl Page {
                     fill_color: obj.fill_color().unwrap(),
                     stroke_color: obj.stroke_color().unwrap(),
                     stroke_width: obj.stroke_width().unwrap().value,
+                    is_stroked: obj.is_stroked().unwrap(),
+                    fill_mode: obj.fill_mode().unwrap_or(PdfPathFillMode::None),
                 });
             } else if points.len() == 2 && points[1].1 == PdfPathSegmentType::LineTo {
                 objects.lines.push(Line {
                     line_type: LineType::Straight,
                     points: points.iter().map(|p| p.0).collect(),
-                    color: obj.stroke_color().unwrap(),
+                    stroke_color: obj.stroke_color().unwrap(),
+                    fill_color: obj.fill_color().unwrap(),
                     width: OrderedFloat(obj.stroke_width().unwrap().value * 2.0),
+                    is_stroked: obj.is_stroked().unwrap(),
+                    fill_mode: obj.fill_mode().unwrap_or(PdfPathFillMode::None),
                 });
             } else {
                 // Check if all points except points[0] have LineTo as their second element
@@ -269,8 +274,11 @@ impl Page {
                 objects.lines.push(Line {
                     line_type,
                     points: points.iter().map(|p| p.0).collect(),
-                    color: obj.stroke_color().unwrap(),
+                    stroke_color: obj.stroke_color().unwrap(),
+                    fill_color: obj.fill_color().unwrap(),
                     width: OrderedFloat(obj.stroke_width().unwrap().value * 2.0),
+                    is_stroked: obj.is_stroked().unwrap(),
+                    fill_mode: obj.fill_mode().unwrap_or(PdfPathFillMode::None),
                 });
             }
         }