rtutil: Add RT.parse method.

This method can parse the ticket and (optional) attachment IDs
out of our short-hand link formats, both for Ledger and Beancount.

conservancy_beancount/rtutil.py

@@ -16,6 +16,7 @@
 
 import functools
 import mimetypes
+import re
 import urllib.parse as urlparse
 
 import rt
@@ -31,6 +32,11 @@ AttachmentTuple = Tuple[str, str, str, str]
 RTId = Union[int, str]
 
 class RT:
+    PARSE_REGEXPS = [
+        re.compile(r'^rt:([0-9]+)(?:/([0-9]+))?/?$'),
+        re.compile(r'^rt://ticket/([0-9]+)(?:/attachments?/([0-9]+))?/?$'),
+    ]
+
     def __init__(self, rt_client: rt.Rt) -> None:
         self.rt = rt_client
         urlparts = urlparse.urlparse(rt_client.url)
@@ -94,6 +100,14 @@ class RT:
     def exists(self, ticket_id: RTId, attachment_id: Optional[RTId]=None) -> bool:
         return self.url(ticket_id, attachment_id) is not None
 
+    def parse(self, s: str) -> Optional[Tuple[str, Optional[str]]]:
+        for regexp in self.PARSE_REGEXPS:
+            match = regexp.match(s)
+            if match is not None:
+                ticket_id, attachment_id = match.groups()
+                return (ticket_id, attachment_id)
+        return None
+
     @functools.lru_cache()
     def ticket_url(self, ticket_id: RTId) -> Optional[str]:
         if self.rt.get_ticket(ticket_id) is None:

tests/test_rtutil.py

@@ -35,7 +35,7 @@ EXPECTED_URLS = [
     (9, None, None),
 ]
 
-@pytest.fixture
+@pytest.fixture(scope='module')
 def rt():
     client = testutil.RTClient()
     return rtutil.RT(client)
@@ -97,3 +97,27 @@ def test_exists_caches(new_client):
     assert rt.exists(2)
     assert not rt.exists(1, 9)
     assert not rt.exists(9)
+
+@pytest.mark.parametrize('link,expected', [
+    ('rt:1/2', ('1', '2')),
+    ('rt:123/456', ('123', '456')),
+    ('rt:12345', ('12345', None)),
+    ('rt:12346/', ('12346', None)),
+    ('rt:12346/789', ('12346', '789')),
+    ('rt:12346/780/', ('12346', '780')),
+    ('rt://ticket/1', ('1', None)),
+    ('rt://ticket/1/', ('1', None)),
+    ('rt://ticket/1234/attachments/5678', ('1234', '5678')),
+    ('rt://ticket/1234/attachments/5678/', ('1234', '5678')),
+    ('rt://ticket/1234/attachment/5678', ('1234', '5678')),
+    ('rt://ticket/1234/attachment/5678/', ('1234', '5678')),
+    ('rt:', None),
+    ('rt://', None),
+    ('rt:example.org', None),
+    ('rt:example.org/1', None),
+    ('rt://example.org', None),
+    ('rt://example.org/1', None),
+    ('https://example.org/rt/Ticket/Display.html?id=123', None),
+])
+def test_parse(rt, link, expected):
+    assert rt.parse(link) == expected