test: expand corpus with real-world Org documents

.gitmodules

@@ -2,3 +2,15 @@
 	path = testdata/external/org-mode
 	url = https://github.com/bzg/org-mode.git
 	shallow = true
+[submodule "external/sachac"]
+	path = testdata/external/sachac
+	url = https://github.com/sachac/.emacs.d.git
+	shallow = true
+[submodule "external/exobrain"]
+	path = testdata/external/exobrain
+	url = https://github.com/karlicoss/exobrain.git
+	shallow = true
+[submodule "external/nvim-orgmode"]
+	path = testdata/external/nvim-orgmode
+	url = https://github.com/nvim-orgmode/orgmode.git
+	shallow = true

README.md

@@ -122,7 +122,8 @@ For an existing checkout, run `git submodule update --init --recursive`.
 Run the tests with `uv tool run --with tox-uv tox -e tests`.
 This also checks the examples in this README.
 
-The normal test suite parses the `.org` files in `testdata/external/org-mode/testing/examples` and checks tree structure and attribute access.
+Corpus tests parse upstream Org examples, Sacha Chua’s Emacs configuration, exobrain notes, and nvim-orgmode documents from `testdata/external`.
+They check tree structure, source line ranges, and attribute access.
 
 Edit `README.qmd`, then regenerate `README.md` with `uv tool run --with tox-uv tox -e quarto`.
 Quarto computes links to source code and tests from their definitions, so line numbers are refreshed when rendering.

README.qmd

@@ -151,7 +151,8 @@ For an existing checkout, run `git submodule update --init --recursive`.
 Run the tests with `uv tool run --with tox-uv tox -e tests`.
 This also checks the examples in this README.
 
-The normal test suite parses the `.org` files in `testdata/external/org-mode/testing/examples` and checks tree structure and attribute access.
+Corpus tests parse upstream Org examples, Sacha Chua's Emacs configuration, exobrain notes, and nvim-orgmode documents from `testdata/external`.
+They check tree structure, source line ranges, and attribute access.
 
 Edit `README.qmd`, then regenerate `README.md` with `uv tool run --with tox-uv tox -e quarto`.
 Quarto computes links to source code and tests from their definitions, so line numbers are refreshed when rendering.

src/orgparse/tests/test_corpus.py

@@ -1,4 +1,4 @@
-"""Exercise the public parser API on Org mode's upstream example files."""
+"""Exercise the public parser API on upstream examples and real-world Org documents."""
 
 from pathlib import Path
 
@@ -8,18 +8,27 @@
 from ..extra import Table
 from ..node import OrgNode
 
-CORPUS = Path(__file__).resolve().parents[3] / 'testdata' / 'external' / 'org-mode' / 'testing' / 'examples'
+CORPUS = Path(__file__).resolve().parents[3] / 'testdata' / 'external'
+CORPUS_DIRS = (
+    'org-mode/testing/examples',
+    'sachac',
+    'exobrain',
+    'nvim-orgmode',
+)
 
 
 def corpus_files() -> list[Path]:
-    paths = sorted(CORPUS.rglob('*.org'))
-    assert len(paths) > 0, 'Missing Org corpus: run git submodule update --init --recursive'
+    paths = []
+    for directory in CORPUS_DIRS:
+        files = sorted((CORPUS / directory).rglob('*.org'))
+        assert len(files) > 0, f'Missing Org corpus {directory}: run git submodule update --init --recursive'
+        paths.extend(files)
     return paths
 
 
 @pytest.mark.parametrize('path', corpus_files(), ids=lambda path: path.relative_to(CORPUS).as_posix())
 def test_corpus(path: Path) -> None:
-    """Check parser robustness and tree consistency across upstream Org examples.
+    """Check parser robustness and tree consistency across examples, configurations, notes, and documentation.
 
     Exercise heading/body formatting, property lookups, table row/block iteration,
         and formatting of populated timestamps.
@@ -49,7 +58,7 @@ def test_corpus(path: Path) -> None:
                 list(part.rows)
                 list(part.blocks)
 
-        dates = node.get_timestamps(active=True, inactive=True)
+        dates = node.get_timestamps(active=True, inactive=True, point=True, range=True)
         if isinstance(node, OrgNode):
             dates = [*dates, node.scheduled, node.deadline, node.closed, *node.clock, *node.repeated_tasks]
         for date in dates: