Break clashing-names test function out of CldrAccess.__checkEnum()

Moving it makes it easier to document what it's up to and why, while
leaving __checkEnum() easier to read; and I'm going to need it
elsewhere anyway. This makes no difference to generated data.

Task-number: QTBUG-94460
Change-Id: I684375bc926d5d54928fbf5b5e08978528aef487
Reviewed-by: Ievgenii Meshcheriakov <[email protected]>

2 files changed, 40 insertions, 17 deletions

diff --git a/util/locale_database/cldr.py b/util/locale_database/cldr.py
index 0bbdad39f9b..d1ad90295ef 100644
--- a/util/locale_database/cldr.py
+++ b/util/locale_database/cldr.py
@@ -16,6 +16,7 @@ from weakref import WeakValueDictionary as CacheDict
 from pathlib import Path
 
 from ldml import Error, Node, XmlScanner, Supplement, LocaleScanner
+from localetools import names_clash
 from qlocalexml import Locale
 
 class CldrReader (object):
@@ -353,10 +354,7 @@ class CldrAccess (object):
                     language, script, territory, variant)
 
     @staticmethod
-    def __checkEnum(given, proper, scraps,
-                    remap = { 'å': 'a', 'ã': 'a', 'ç': 'c', 'é': 'e', 'í': 'i', 'ü': 'u'},
-                    prefix = { 'St.': 'Saint', 'U.S.': 'United States' },
-                    skip = '\u02bc'):
+    def __checkEnum(given, proper, scraps):
         # Each is a { code: full name } mapping
         for code, name in given.items():
             try: right = proper[code]
@@ -366,19 +364,9 @@ class CldrAccess (object):
                 if code not in scraps:
                     yield name, f'[Found no CLDR name for code {code}]'
                 continue
-            if name == right: continue
-            ok = right.replace('&', 'And')
-            for k, v in prefix.items():
-                if ok.startswith(k + ' '):
-                    ok = v + ok[len(k):]
-            while '(' in ok:
-                try: f, t = ok.index('('), ok.index(')')
-                except ValueError: break
-                ok = ok[:f].rstrip() + ' ' + ok[t:].lstrip()
-            if ''.join(ch for ch in name.lower() if not ch.isspace()) in ''.join(
-                remap.get(ch, ch) for ch in ok.lower() if ch.isalpha() and ch not in skip):
-                continue
-            yield name, ok
+            cleaned = names_clash(right, name)
+            if cleaned:
+                yield name, cleaned
 
     def checkEnumData(self, grumble):
         scraps = set()
diff --git a/util/locale_database/localetools.py b/util/locale_database/localetools.py
index ae4a50c09ab..a0ad5f397d7 100644
--- a/util/locale_database/localetools.py
+++ b/util/locale_database/localetools.py
@@ -48,6 +48,41 @@ def wrap_list(lst, perline=20):
             yield head
     return ",\n".join(", ".join(x) for x in split(lst, perline))
 
+def names_clash(cldr, enum):
+    """True if the reader might not recognize cldr as the name of enum
+
+    First argument, cldr, is the name CLDR gives for some language,
+    script or territory; second, enum, is the name enumdata.py gives
+    for it. If these are enough alike, returns None; otherwise, a
+    non-empty string that results from adapting cldr to be more like
+    how enumdata.py would express it."""
+    if cldr == enum:
+        return None
+
+    # Some common substitutions:
+    cldr = cldr.replace('&', 'And')
+    prefix = { 'St.': 'Saint', 'U.S.': 'United States' }
+    for k, v in prefix.items():
+        if cldr.startswith(k + ' '):
+            cldr = v + cldr[len(k):]
+
+    # Chop out any parenthesised part, e.g. (Burma):
+    while '(' in cldr:
+        try:
+            f, t = cldr.index('('), cldr.rindex(')')
+        except ValueError:
+            break
+        cldr = cldr[:f].rstrip() + ' ' + cldr[t + 1:].lstrip()
+
+    # Various accented letters:
+    remap = { 'å': 'a', 'ã': 'a', 'ç': 'c', 'é': 'e', 'í': 'i', 'ô': 'o', 'ü': 'u'}
+    skip = '\u02bc' # Punctuation for which .isalpha() is true.
+    # Let cldr match (ignoring non-letters and case) any substring as enum:
+    if ''.join(enum.lower().split()) in ''.join(
+            remap.get(ch, ch) for ch in cldr.lower() if ch.isalpha() and ch not in skip):
+        return None
+    return cldr
+
 
 @contextmanager
 def AtomicRenameTemporaryFile(originalLocation: Path, *, prefix: str, dir: Path):