Improved CLDR parser.

Added support for "defaultNumberingSystem" element to get the right numeric
information (decimal point, group separator, zero digit etc).

Also fixed xpaths to the AM/PM text for CLDR 1.8.0

Reviewed-by: trustme

2 files changed, 39 insertions, 17 deletions

diff --git a/util/local_database/cldr2qlocalexml.py b/util/local_database/cldr2qlocalexml.py
index 71d14e8..ede7763 100755
--- a/util/local_database/cldr2qlocalexml.py
+++ b/util/local_database/cldr2qlocalexml.py
@@ -102,6 +102,9 @@ def generateLocaleInfo(path):
     country = ""
     if country_id != -1:
         country = enumdata.country_list[country_id][0]
+    if country == "":
+        sys.stderr.write("unnknown country code \"" + country_code + "\"\n")
+        return {}
 
     # So we say we accept only those values that have "contributed" or
     # "approved" resolution. see http://www.unicode.org/cldr/process.html
@@ -118,16 +121,29 @@ def generateLocaleInfo(path):
     result['variant_code'] = variant_code
     result['language_id'] = language_id
     result['country_id'] = country_id
-    result['decimal'] = findEntry(path, "numbers/symbols/decimal")
-    result['group'] = findEntry(path, "numbers/symbols/group")
-    result['list'] = findEntry(path, "numbers/symbols/list")
-    result['percent'] = findEntry(path, "numbers/symbols/percentSign")
-    result['zero'] = findEntry(path, "numbers/symbols/nativeZeroDigit")
-    result['minus'] = findEntry(path, "numbers/symbols/minusSign")
-    result['plus'] = findEntry(path, "numbers/symbols/plusSign")
-    result['exp'] = findEntry(path, "numbers/symbols/exponential").lower()
-    result['am'] = findEntry(path, "dates/calendars/calendar[gregorian]/am", draft)
-    result['pm'] = findEntry(path, "dates/calendars/calendar[gregorian]/pm", draft)
+
+    numberingSystem = None
+    try:
+        numbering_system = findEntry(path, "numbers/defaultNumberingSystem")
+    except:
+        pass
+    def get_number_in_system(path, xpath, numbering_system):
+        if numbering_system:
+            try:
+                return findEntry(path, xpath + "[numberSystem=" + numbering_system + "]")
+            except xpathlite.Error:
+                pass
+        return findEntry(path, xpath)
+    result['decimal'] = get_number_in_system(path, "numbers/symbols/decimal", numbering_system)
+    result['group'] = get_number_in_system(path, "numbers/symbols/group", numbering_system)
+    result['list'] = get_number_in_system(path, "numbers/symbols/list", numbering_system)
+    result['percent'] = get_number_in_system(path, "numbers/symbols/percentSign", numbering_system)
+    result['zero'] = get_number_in_system(path, "numbers/symbols/nativeZeroDigit", numbering_system)
+    result['minus'] = get_number_in_system(path, "numbers/symbols/minusSign", numbering_system)
+    result['plus'] = get_number_in_system(path, "numbers/symbols/plusSign", numbering_system)
+    result['exp'] = get_number_in_system(path, "numbers/symbols/exponential", numbering_system).lower()
+    result['am'] = findEntry(path, "dates/calendars/calendar[gregorian]/dayPeriods/dayPeriodContext[format]/dayPeriodWidth[wide]/dayPeriod[am]", draft)
+    result['pm'] = findEntry(path, "dates/calendars/calendar[gregorian]/dayPeriods/dayPeriodContext[format]/dayPeriodWidth[wide]/dayPeriod[pm]", draft)
     result['longDateFormat'] = findEntry(path, "dates/calendars/calendar[gregorian]/dateFormats/dateFormatLength[full]/dateFormat/pattern")
     result['shortDateFormat'] = findEntry(path, "dates/calendars/calendar[gregorian]/dateFormats/dateFormatLength[short]/dateFormat/pattern")
     result['longTimeFormat'] = findEntry(path, "dates/calendars/calendar[gregorian]/timeFormats/timeFormatLength[full]/timeFormat/pattern")
diff --git a/util/local_database/xpathlite.py b/util/local_database/xpathlite.py
index 0f21a19..7d569f3 100644
--- a/util/local_database/xpathlite.py
+++ b/util/local_database/xpathlite.py
@@ -64,16 +64,16 @@ class Error:
     def __str__(self):
         return self.msg
 
-def findChild(parent, tag_name, arg_value, draft=None):
+def findChild(parent, tag_name, arg_name=None, arg_value=None, draft=None):
     for node in parent.childNodes:
         if node.nodeType != node.ELEMENT_NODE:
             continue
         if node.nodeName != tag_name:
             continue
         if arg_value:
-            if not node.attributes.has_key('type'):
+            if not node.attributes.has_key(arg_name):
                 continue
-            if node.attributes['type'].nodeValue != arg_value:
+            if node.attributes[arg_name].nodeValue != arg_value:
                 continue
         if draft:
             if not node.attributes.has_key('draft'):
@@ -101,12 +101,18 @@ def _findEntryInFile(file, path, draft=None, attribute=None):
     for i in range(len(tag_spec_list)):
         tag_spec = tag_spec_list[i]
         tag_name = tag_spec
+        arg_name = 'type'
         arg_value = ''
         left_bracket = tag_spec.find('[')
         if left_bracket != -1:
             tag_name = tag_spec[:left_bracket]
-            arg_value = tag_spec[left_bracket+1:-1]
-        alias = findChild(elt, 'alias', None)
+            arg_value = tag_spec[left_bracket+1:-1].split("=")
+            if len(arg_value) == 2:
+                arg_name = arg_value[0]
+                arg_value = arg_value[1]
+            else:
+                arg_value = arg_value[0]
+        alias = findChild(elt, 'alias')
         if alias and alias.attributes['source'].nodeValue == 'locale':
             path = alias.attributes['path'].nodeValue
             aliaspath = tag_spec_list[:i] + path.split("/")
@@ -123,7 +129,7 @@ def _findEntryInFile(file, path, draft=None, attribute=None):
             aliaspath = "/".join(aliaspath)
             # "locale" aliases are special - we need to start lookup from scratch
             return (None, aliaspath)
-        elt = findChild(elt, tag_name, arg_value, draft)
+        elt = findChild(elt, tag_name, arg_name, arg_value, draft)
         if not elt:
             return ("", None)
     if attribute is not None:
@@ -136,7 +142,7 @@ def findAlias(file):
     if not doc_cache.has_key(file):
         return False
     doc = doc_cache[file]
-    alias_elt = findChild(doc.documentElement, "alias", "")
+    alias_elt = findChild(doc.documentElement, "alias")
     if not alias_elt:
         return False
     if not alias_elt.attributes.has_key('source'):