filters: Add filter_meta_match function.

conservancy_beancount/filters.py

@@ -14,10 +14,14 @@
 # You should have received a copy of the GNU Affero General Public License
 # along with this program.  If not, see <https://www.gnu.org/licenses/>.
 
+import re
+
 from . import data
 
 from typing import (
     Iterable,
+    Pattern,
+    Union,
 )
 from .beancount_types import (
     MetaKey,
@@ -25,6 +29,7 @@ from .beancount_types import (
 )
 
 Postings = Iterable[data.Posting]
+Regexp = Union[str, Pattern]
 
 def filter_meta_equal(postings: Postings, key: MetaKey, value: MetaValue) -> Postings:
     for post in postings:
@@ -33,3 +38,11 @@ def filter_meta_equal(postings: Postings, key: MetaKey, value: MetaValue) -> Pos
                 yield post
         except KeyError:
             pass
+
+def filter_meta_match(postings: Postings, key: MetaKey, regexp: Regexp) -> Postings:
+    for post in postings:
+        try:
+            if re.search(regexp, post.meta[key]):
+                yield post
+        except (KeyError, TypeError):
+            pass

tests/test_filters.py

@@ -81,3 +81,17 @@ def test_filter_meta_equal(cc_txn_pair, key, value, expected_indexes):
     postings = data.Posting.from_entries(cc_txn_pair)
     actual = filters.filter_meta_equal(postings, key, value)
     check_filter(actual, cc_txn_pair, expected_indexes)
+
+@pytest.mark.parametrize('key,regexp,expected_indexes', [
+    ('entity', '^Smith-', range(5)),
+    ('receipt', r'\.pdf$', range(5)),
+    ('receipt', 'Receipt', range(3)),
+    ('statement', '.', [4]),
+    ('metadate', 'foo', ()),
+    ('BadKey', '.', ()),
+    ('emptykey', '.', ()),
+])
+def test_filter_meta_match(cc_txn_pair, key, regexp, expected_indexes):
+    postings = data.Posting.from_entries(cc_txn_pair)
+    actual = filters.filter_meta_match(postings, key, regexp)
+    check_filter(actual, cc_txn_pair, expected_indexes)