Update known issue handling for DateTime Format (#518)

* Updating number format data generation and NodeJS number_fmt

* Updated C++ number format, fixing errors and failing tests

* Fix number format in ICU4J

* Remove cout

* Add halfOdd to possible test results

* Add dateTimeFormatType to datagen, schema, and CPP version

* Updating regex for unsupported test types.

* Updates as suggested

* Classify some datetime_fmt errors as known issues, fixing some ICU4C and ICU4J failures.

* Handle Node known issue for timezone only format

* responding to review suggestions

* DateTime checking known issues

* Removing schema change

* Removing change in cpp file

* Undoing unrelated change

* Remove extra check for comma

* Remove copy

* Reply to suggestions

* Restore check_list_differences

* Fix when result is missing

* Fixing types of comparisons for known issues

Author: Craig Cornelius

verifier/check_known_issues.py

@@ -57,11 +57,11 @@ class knownIssueType(Enum):
     # Datetime format
     datetime_fmt_at_inserted = 'Alternate formatting with "at" between time and date'
     datetime_fmt_arabic_comma = 'Arabic comma vs. ASCII comma'
-    datetime_unexpected_comma = 'Unexpected comma'
     datetime_inserted_comma = 'inserted comma'
 
     datetime_semantic_Z = 'NodeJS always includes date or time'
 
+    datetime_GMT_UTC = 'UTC instead of GMT'
 
     # Likely Subtags
     likely_subtags_sr_latn = "sr_latin becoming en"
@@ -108,9 +108,9 @@ def diff_ascii_space_vs_nbsp(actual, expected_value):
     if not expected_value or not actual:
         return None
 
-    # If replacing all the NBSP characdters in expected gives the actual result,
-    # then the only differences were with this type of space in formatted output.
-    if expected_value.replace(NBSP, SP) == actual:
+    # Normalizing all NBSP spaces to ASCII in both to check if the type of space
+    # is the only difference in formatted output.
+    if actual.replace(NBSP, SP) == expected_value.replace(NBSP, SP):
         return knownIssueType.known_issue_sp_nbsp
     else:
         return None
@@ -164,41 +164,35 @@ def unsupported_unit_quarter(test):
     return None
 
 
-def dt_check_for_alternate_long_form(test, actual, expected):
+def dt_check_for_alternate_long_form(actual, expected):
     # For datetime_fmt, is the format type "standard"?
     if actual == expected:
         return None
-    if actual.replace(' at', ',') == expected:
-        return knownIssueType.datetime_fmt_at_inserted
-
-    # Thai language difference with "time" inserted
-    if actual.replace('เวลา ', '') == expected:
-        return knownIssueType.datetime_fmt_at_inserted
-    # Arabic
-    if actual.replace(' في', '،') == expected:
-        return knownIssueType.datetime_fmt_at_inserted
-    # Vietnamese
-    if actual.replace('lúc ', '') == expected:
-        return knownIssueType.datetime_fmt_at_inserted
-    # Bengali
-    if actual.replace('এ ', '') == expected:
-        return knownIssueType.datetime_fmt_at_inserted
+    new_expected = expected.replace(NBSP, SP)
+    new_actual = actual.replace(NBSP, SP)
+
+    # TODO: Make this an array of replacements
+    replacements_to_try = [
+        (' at', ','),  # English
+        ('เวลา ', ''),  # Thai
+        (' في', '،'),  # Arabic
+        ('lúc ', ''),  # Vietnamese
+        ('এ ', ''),  # Bengali
+    ]
+
+    for replacement in replacements_to_try:
+        if new_actual.replace(replacement[0], replacement[1]) == new_expected:
+            return knownIssueType.datetime_fmt_at_inserted
     return None
 
 
-def dt_check_arabic_comma(test, actual, expected):
+def dt_check_arabic_comma(actual, expected):
     if expected.replace('\u002c', '\u060c') == actual:
         return knownIssueType.datetime_fmt_arabic_comma
     return None
 
 
-def dt_unexpected_comma(test, actual, expected):
-    if actual.replace('\u002c', '') == expected:
-        return knownIssueType.datetime_unexpected_comma
-    else:
-        return None
-
-def dt_inserted_comma(test, actual, expected):
+def dt_inserted_comma(actual, expected):
     sm = SequenceMatcher(None, expected, actual)
     sm_opcodes = sm.get_opcodes()
     # Look for one additional comma
@@ -211,10 +205,23 @@ def dt_inserted_comma(test, actual, expected):
     return None
 
 
-def check_datetime_known_issues(test):
+def dt_gmt_utc(actual, expected):
+    # The difference may also include NBSP vs ASCII space
+    new_expected = expected.replace(NBSP, SP)
+    new_actual = actual.replace(NBSP, SP)
+
+    if new_actual.replace('UTC', 'GMT') == new_expected or \
+            new_actual.replace('Coordinated Universal', 'Greenwich Mean') == new_expected:
+        return knownIssueType.datetime_GMT_UTC
+    return None
+
+def check_datetime_known_issues(test, platform_info):
     # Examine a single test for date/time isses
     # Returns known issues identified for this test in this category
     remove_this_one = False
+
+    all_matching_issues = []
+
     try:
         try:
             result = test['result']
@@ -224,50 +231,32 @@ def check_datetime_known_issues(test):
         expected = test['expected']
         input_data = test.get('input_data')
 
-        is_ki = diff_nbsp_vs_ascii_space(result, expected)
-        if is_ki:
-            # Mark the test with this issue
-            test['known_issue'] = is_ki.value
-            remove_this_one = True
-
-        is_ki = diff_ascii_space_vs_nbsp(result, expected)
-        if is_ki:
-            # Mark the test with this issue
-            test['known_issue'] = is_ki.value
-            remove_this_one = True
-
-        is_ki = numerals_replaced_by_another_numbering_system(result, expected)
-        if is_ki:
-            test['known_issue_id'] = is_ki.value
-            remove_this_one = True
-
-        is_ki = dt_check_for_alternate_long_form(test, result, expected)
-        if is_ki:
-            test['known_issue_id'] = is_ki.value
-            remove_this_one = True
+        # Perform each test, computing matches with known issues by means of the functions in this list
+        check_fns = [dt_gmt_utc, diff_nbsp_vs_ascii_space, diff_ascii_space_vs_nbsp, numerals_replaced_by_another_numbering_system,
+                  dt_check_arabic_comma, dt_inserted_comma, dt_check_for_alternate_long_form]
+        for check_fn in check_fns:
+            is_ki = check_fn(result, expected)
+            if is_ki:
+                test['known_issue_id'] = is_ki.value
+                remove_this_one = True
+                all_matching_issues.append(is_ki.value)
+
+        # Check if the semantic skeleton has "Z" for NodeJS
+        if platform_info['platform'] == 'NodeJS' and result:
+            if input_data and 'semanticSkeleton' in input_data['options'] and \
+               input_data['options']['semanticSkeleton'] == 'Z':
+                test['known_issue_id'] = knownIssueType.datetime_semantic_Z.value
+                remove_this_one = True
+                all_matching_issues.append(knownIssueType.datetime_semantic_Z.value)
 
-        is_ki = dt_check_arabic_comma(test, result, expected)
-        if is_ki:
-            test['known_issue_id'] = is_ki.value
-            remove_this_one = True
-
-        is_ki = dt_unexpected_comma(test, result, expected)
-        if is_ki:
-            test['known_issue_id'] = is_ki.value
-            remove_this_one = True
-
-        is_ki = dt_inserted_comma(test, result, expected)
-        if is_ki:
-            test['known_issue_id'] = is_ki.value
-            remove_this_one = True
 
     except BaseException as err:
         # Can't get the info
         pass
 
     return remove_this_one
 
-def check_rdt_known_issues(test):
+def check_rdt_known_issues(test, platform_info=None):
     # ??? Do wwe need platform ID and/or icu version?
     remove_this_one = False
     try:
@@ -292,7 +281,7 @@ def check_rdt_known_issues(test):
     return remove_this_one
 
 
-def check_likely_subtags_issues(test):
+def check_likely_subtags_issues(test, platform_info=None):
     remove_this_one = False
     try:
         result = test['result']
@@ -315,7 +304,7 @@ def sr_latin_likely_subtag(test):
         return None
 
 # Language names
-def check_langnames_issues(test):
+def check_langnames_issues(test, platform_info=None):
     remove_this_one = False
     try:
         result = test['result']
@@ -391,7 +380,7 @@ def check_number_fmt_issues(test, platform_info):
     return None
 
 
-def check_plural_rules_issues(test):
+def check_plural_rules_issues(test, platform_info=None):
     try:
         input_data = test['input_data']
         sample_string = input_data['sample']
@@ -406,7 +395,7 @@ def check_plural_rules_issues(test):
         return None
 
 
-def check_collation_issues(test):
+def check_collation_issues(test, platform_info=None):
     input_data = test.get('input_data', {})
 
     # Check for jsonc bug with surrogates
@@ -436,19 +425,19 @@ def compute_known_issues_for_single_test(test_type, test, platform_info):
     # Returns True if this single test is an example of one or more known issues,
     known_issue_found = False
     if test_type == ddt_data.testType.collation.value:
-        known_issue_found = check_collation_issues(test)
+        known_issue_found = check_collation_issues(test, platform_info)
     if test_type == ddt_data.testType.datetime_fmt.value:
-        known_issue_found = check_datetime_known_issues(test)
+        known_issue_found = check_datetime_known_issues(test, platform_info)
     elif test_type == ddt_data.testType.rdt_fmt.value:
-        known_issue_found = check_rdt_known_issues(test)
+        known_issue_found = check_rdt_known_issues(test, platform_info)
     elif test_type == ddt_data.testType.likely_subtags.value:
-        known_issue_found = check_likely_subtags_issues(test)
+        known_issue_found = check_likely_subtags_issues(test, platform_info)
     elif test_type == ddt_data.testType.lang_names.value:
-        known_issue_found = check_langnames_issues(test)
+        known_issue_found = check_langnames_issues(test, platform_info)
     elif test_type == ddt_data.testType.number_fmt.value:
         known_issue_found = check_number_fmt_issues(test, platform_info)
     elif test_type == ddt_data.testType.plural_rules.value:
-        known_issue_found = check_plural_rules_issues(test)
+        known_issue_found = check_plural_rules_issues(test, platform_info)
 
     # TODO: Add checks here for known issues in other test types