From cff486dda7352395feb3687d274f57c945c0b522 Mon Sep 17 00:00:00 2001
From: Marcel <marcel@familienarchiv>
Date: Mon, 25 May 2026 13:35:19 +0200
Subject: [PATCH] =?UTF-8?q?fix(normalizer):=20treat=20leading=20date=20qua?=
 =?UTF-8?q?lifiers=20(nach/vor/=E2=80=A6)=20as=20APPROX?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

_preprocess now sets approx=True when a leading marker is stripped; add
_match_year_only so bare years (e.g. "nach 1900" -> "1900") resolve to
1900-01-01/YEAR before being upgraded to APPROX. Strengthen
test_parse_approx_marker_upgrades_precision and add
test_parse_leading_qualifier_is_approx (11 tests, all pass).

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
---
 tools/import-normalizer/dates.py            | 19 +++++++++++++++----
 tools/import-normalizer/tests/test_dates.py |  7 ++++++-
 2 files changed, 21 insertions(+), 5 deletions(-)

diff --git a/tools/import-normalizer/dates.py b/tools/import-normalizer/dates.py
index a34bd357..65449a52 100644
--- a/tools/import-normalizer/dates.py
+++ b/tools/import-normalizer/dates.py
@@ -72,7 +72,7 @@ _LEADING_MARKERS = re.compile(
 
 
 def _preprocess(raw: str):
-    """Return (cleaned_string, approx_flag)."""
+    """Return (cleaned_string, approx_flag). Any uncertainty/qualifier marker -> approx."""
     s = (raw or "").strip()
     if not s:
         return "", False
@@ -82,8 +82,10 @@ def _preprocess(raw: str):
     s = re.sub(r"\(\s*\?\s*\)", " ", s)   # remove "(?)"
     s = s.replace("?", " ")
     s = re.sub(r",.*$", "", s)            # drop trailing editorial note (", 2. Brief")
-    s = _LEADING_MARKERS.sub("", s)
-    s = re.sub(r"\s+", " ", s).strip(" .,")
+    stripped = _LEADING_MARKERS.sub("", s)
+    if stripped != s:                     # a leading qualifier (um/ca/nach/vor/anfang/…) signals approximation
+        approx = True
+    s = re.sub(r"\s+", " ", stripped).strip(" .,")
     return s, approx
 
 
@@ -114,8 +116,17 @@ def _match_numeric(s):
         return None
 
 
+_YEAR_ONLY_RE = re.compile(r"\d{4}")
+
+
+def _match_year_only(s):
+    if _YEAR_ONLY_RE.fullmatch(s):
+        return datetime.date(int(s), 1, 1).isoformat(), Precision.YEAR
+    return None
+
+
 # Matchers are tried in order. Later tasks append to this list.
-_MATCHERS = [_match_iso, _match_numeric]
+_MATCHERS = [_match_iso, _match_numeric, _match_year_only]
 
 
 def parse_date(raw: str, date_overrides: dict | None = None) -> ParsedDate:
diff --git a/tools/import-normalizer/tests/test_dates.py b/tools/import-normalizer/tests/test_dates.py
index d8e06e22..5a2a6f5b 100644
--- a/tools/import-normalizer/tests/test_dates.py
+++ b/tools/import-normalizer/tests/test_dates.py
@@ -56,5 +56,10 @@ def test_parse_numeric_unparseable():
 
 def test_parse_approx_marker_upgrades_precision():
     r = dates.parse_date("17.Nov (?) 1887")  # month-name handled in a later task; here just the marker path
-    # after the marker is detected, a parsed date becomes APPROX (verified fully in Task 8)
     assert r.raw == "17.Nov (?) 1887"
+    assert r.precision == Precision.UNKNOWN  # no month-name matcher until Task 7; full APPROX check in Task 8
+
+def test_parse_leading_qualifier_is_approx():
+    r = dates.parse_date("nach 1900")  # "after 1900" -> year salvaged, but precision is APPROX not exact
+    assert r.iso == "1900-01-01"
+    assert r.precision == Precision.APPROX