fix(accounts): match Beancount's DATE grammar in duplicate detection (libra-#48)

_open_directive_exists hardcoded '^YYYY-MM-DD open ' (dash-only, 2-digit,
single-space), but Beancount's DATE token (parser/lexer.l) is
(17|18|19|20)[0-9]{2}[-/][0-9]+[-/][0-9]+ and inter-token whitespace is any
[ \t\r] run. So a validly-formatted existing Open written as '2024/3/5 open X'
or '2020-01-01  open  X' escaped detection → duplicate Open appended →
bean-check rejects the file. Anchor on Beancount's actual date pattern and
[ \t]+ separators. Adds parametrized coverage for slash/single-digit/multi-
space/tab variants.

Found in a coherence pass over the Beancount source.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

fava_client.py

@@ -60,11 +60,21 @@ def _escape_beancount_string(value: str) -> str:
     )
 
 
+# Beancount's DATE token (parser/lexer.l): (17|18|19|20)[0-9]{2}[-/][0-9]+[-/][0-9]+
+# — '-' OR '/' separators, 1+ digit month/day. Inter-token whitespace is any
+# run of [ \t\r] (ignored by the lexer). The duplicate-detection regex must
+# mirror this, or a validly-formatted existing Open (e.g. '2024/3/5 open X' or
+# '2020-01-01  open  X') escapes detection and a duplicate Open is appended,
+# which bean-check then rejects — breaking every later write.
+_OPEN_DATE = r"(?:17|18|19|20)\d\d[-/]\d+[-/]\d+"
+
+
 def _open_directive_exists(source: str, account_name: str) -> bool:
     """Return True if `source` already contains an Open directive for exactly
     `account_name`.
 
-    Anchored to a real `YYYY-MM-DD open <account>` directive line (re.MULTILINE)
+    Anchored to a real `<date> open <account>` directive line (re.MULTILINE),
+    with `<date>` and the inter-token whitespace matching Beancount's grammar,
     so the account name can't match text inside another account's description
     metadata or a comment (false positive → spurious 409). The trailing
     negative-lookahead `(?![\\w:-])` requires the next char not to be an
@@ -72,12 +82,11 @@ def _open_directive_exists(source: str, account_name: str) -> bool:
       - a prefix (Expenses:Gas) does not match a longer sibling
         (Expenses:GasStation / Expenses:Gas:Vehicle), and
       - a real directive with an inline comment and no space
-        (`open Expenses:Gas;legacy`) is still detected (`;` ends the name),
-        which the previous `(?:\\s|$)` boundary missed → duplicate write.
+        (`open Expenses:Gas;legacy`) is still detected (`;` ends the name).
     """
     return bool(
         re.search(
-            rf"^\d{{4}}-\d{{2}}-\d{{2}} open {re.escape(account_name)}(?![\w:-])",
+            rf"^{_OPEN_DATE}[ \t]+open[ \t]+{re.escape(account_name)}(?![\w:-])",
             source,
             re.MULTILINE,
         )

tests/test_unit.py

@@ -84,6 +84,22 @@ def test_open_directive_exists_does_not_match_deeper_child():
     assert fc._open_directive_exists(src, "Expenses:Vehicle:Gas") is False
 
 
+@pytest.mark.parametrize(
+    "line",
+    [
+        "2024/3/5 open Expenses:Vehicle:Gas",      # slash date, single-digit M/D
+        "2020-1-1 open Expenses:Vehicle:Gas",       # dash date, single-digit M/D
+        "2020-01-01  open  Expenses:Vehicle:Gas",   # multiple spaces
+        "2020-01-01\topen\tExpenses:Vehicle:Gas",   # tab separators
+        "1970-01-01 open Expenses:Vehicle:Gas EUR", # currency-constrained
+    ],
+)
+def test_open_directive_exists_matches_beancount_date_and_whitespace_variants(line):
+    # All of these are valid Beancount Open directives per lexer.l's DATE token
+    # and ignored inter-token whitespace; each must be detected as existing.
+    assert fc._open_directive_exists(line + "\n", "Expenses:Vehicle:Gas") is True
+
+
 # ---------------------------------------------------------------------------
 # beancount_format.sanitize_link
 # ---------------------------------------------------------------------------