simplify further

adbar · Nov 3, 2023 · 0a4128e · 0a4128e
1 parent dbf474c
commit 0a4128e
Show file tree

Hide file tree

Showing 2 changed files with 19 additions and 39 deletions.
diff --git a/htmldate/core.py b/htmldate/core.py
@@ -355,9 +355,7 @@ def select_candidate(
     occurrences: Counter_Type[str],
     catch: Pattern[str],
     yearpat: Pattern[str],
-    original_date: bool,
-    min_date: datetime,
-    max_date: datetime,
+    options: Extractor,
 ) -> Optional[Match[str]]:
     """Select a candidate among the most frequent matches"""
     if not occurrences or len(occurrences) > MAX_POSSIBLE_CANDIDATES:
@@ -372,7 +370,7 @@ def select_candidate(
     firstselect = occurrences.most_common(10)
     LOGGER.debug("firstselect: %s", firstselect)
     # sort and find probable candidates
-    bestones = sorted(firstselect, reverse=not original_date)[:2]
+    bestones = sorted(firstselect, reverse=not options.original)[:2]
     LOGGER.debug("bestones: %s", bestones)
 
     # plausibility heuristics
@@ -385,7 +383,7 @@ def select_candidate(
             years[i] = year_match[1]
             dateobject = datetime(int(year_match[1]), 1, 1)
             validation[i] = is_valid_date(
-                dateobject, "%Y", earliest=min_date, latest=max_date
+                dateobject, "%Y", earliest=options.min, latest=options.max
             )
 
     # safety net: plausibility
@@ -422,9 +420,7 @@ def search_pattern(
         earliest=options.min,
         latest=options.max,
     )
-    return select_candidate(
-        candidates, catch, yearpat, options.original, options.min, options.max
-    )
+    return select_candidate(candidates, catch, yearpat, options)
 
 
 @lru_cache(maxsize=CACHE_SIZE)
@@ -651,9 +647,7 @@ def search_page(htmlstring: str, options: Extractor) -> Optional[str]:
         replacement[candidate] = candidates[item]
     candidates = Counter(replacement)
     # select
-    bestmatch = select_candidate(
-        candidates, YMD_PATTERN, YMD_YEAR, options.original, options.min, options.max
-    )
+    bestmatch = select_candidate(candidates, YMD_PATTERN, YMD_YEAR, options)
     result = filter_ymd_candidate(
         bestmatch,
         SELECT_YMD_PATTERN,
@@ -702,9 +696,7 @@ def search_page(htmlstring: str, options: Extractor) -> Optional[str]:
         candidate = normalize_match(match)
         replacement[candidate] = candidates[item]
     candidates = Counter(replacement)
-    bestmatch = select_candidate(
-        candidates, YMD_PATTERN, YMD_YEAR, options.original, options.min, options.max
-    )
+    bestmatch = select_candidate(candidates, YMD_PATTERN, YMD_YEAR, options)
     result = filter_ymd_candidate(
         bestmatch,
         SLASHES_PATTERN,
@@ -760,9 +752,7 @@ def search_page(htmlstring: str, options: Extractor) -> Optional[str]:
         replacement[candidate] = candidates[item]
     candidates = Counter(replacement)
     # select
-    bestmatch = select_candidate(
-        candidates, YMD_PATTERN, YMD_YEAR, options.original, options.min, options.max
-    )
+    bestmatch = select_candidate(candidates, YMD_PATTERN, YMD_YEAR, options)
     result = filter_ymd_candidate(
         bestmatch,
         MMYYYY_PATTERN,

diff --git a/tests/unit_tests.py b/tests/unit_tests.py
@@ -860,9 +860,9 @@ def test_compare_reference():
     assert compare_reference(1517500000, "2018-02-01", options) == 1517500000
 
 
-def test_candidate_selection(min_date=MIN_DATE, max_date=LATEST_POSSIBLE):
+def test_candidate_selection():
     """test the algorithm for several candidates"""
-    original_date = False
+    options = Extractor(False, LATEST_POSSIBLE, MIN_DATE, False, OUTPUTFORMAT)
     # patterns
     catch = re.compile(r"([0-9]{4})-([0-9]{2})-([0-9]{2})")
     yearpat = re.compile(r"^([0-9]{4})")
@@ -878,9 +878,7 @@ def test_candidate_selection(min_date=MIN_DATE, max_date=LATEST_POSSIBLE):
             "2-28",
         ]
     )
-    result = select_candidate(
-        occurrences, catch, yearpat, original_date, min_date, max_date
-    )
+    result = select_candidate(occurrences, catch, yearpat, options)
     assert result is None
     # plausible
     occurrences = Counter(
@@ -894,38 +892,30 @@ def test_candidate_selection(min_date=MIN_DATE, max_date=LATEST_POSSIBLE):
             "2017-11-28",
         ]
     )
-    result = select_candidate(
-        occurrences, catch, yearpat, original_date, min_date, max_date
-    )
+    result = select_candidate(occurrences, catch, yearpat, options)
     assert result.group(0) == "2017-11-28"
-    original_date = True
-    result = select_candidate(
-        occurrences, catch, yearpat, original_date, min_date, max_date
-    )
+
+    options = Extractor(False, LATEST_POSSIBLE, MIN_DATE, True, OUTPUTFORMAT)
+    result = select_candidate(occurrences, catch, yearpat, options)
     assert result.group(0) == "2016-07-12"
     # mix plausible/implausible
     occurrences = Counter(
         ["2116-12-23", "2116-12-23", "2116-12-23", "2017-08-11", "2017-08-11"]
     )
-    result = select_candidate(
-        occurrences, catch, yearpat, original_date, min_date, max_date
-    )
+    result = select_candidate(occurrences, catch, yearpat, options)
     assert result.group(0) == "2017-08-11"
-    original_date = False
+
+    options = Extractor(False, LATEST_POSSIBLE, MIN_DATE, False, OUTPUTFORMAT)
     occurrences = Counter(
         ["2116-12-23", "2116-12-23", "2116-12-23", "2017-08-11", "2017-08-11"]
     )
-    result = select_candidate(
-        occurrences, catch, yearpat, original_date, min_date, max_date
-    )
+    result = select_candidate(occurrences, catch, yearpat, options)
     assert result.group(0) == "2017-08-11"
     # taking date present twice, corner case
     occurrences = Counter(
         ["2016-12-23", "2016-12-23", "2017-08-11", "2017-08-11", "2017-08-11"]
     )
-    result = select_candidate(
-        occurrences, catch, yearpat, original_date, min_date, max_date
-    )
+    result = select_candidate(occurrences, catch, yearpat, options)
     assert result.group(0) == "2016-12-23"