[svn r3304] r3524@delle: sbehnel | 2008-02-15 13:39:53 +0100

 integrate all error type enums: parser, XPath, schema, relaxng

--HG--
branch : trunk

1 files changed, 31 insertions, 14 deletions

diff --git a/update-error-constants.py b/update-error-constants.py
index 873c23d2..5cc6de40 100644
--- a/update-error-constants.py
+++ b/update-error-constants.py
@@ -11,20 +11,24 @@ if len(sys.argv) < 2 or sys.argv[1].lower() in ('-h', '--help'):
     print sys.argv[0], "/path/to/libxml2-doc-dir"
     sys.exit(len(sys.argv) > 1)
 
-HTML_FILE = os.path.join(sys.argv[1], 'html', 'libxml-xmlerror.html')
-os.stat(HTML_FILE) # raise an error if we can't find it
+HTML_DIR = os.path.join(sys.argv[1], 'html')
+os.stat(HTML_DIR) # raise an error if we can't find it
 
 sys.path.insert(0, 'src')
 from lxml import etree
 
 # map enum name to Python variable name and alignment for constant name
 ENUM_MAP = {
-    'xmlErrorLevel'   : ('__ERROR_LEVELS',  'XML_ERR_'),
-    'xmlErrorDomain'  : ('__ERROR_DOMAINS', 'XML_FROM_'),
-    'xmlParserErrors' : ('__ERROR_TYPES',   'XML_')
+    'xmlErrorLevel'       : ('__ERROR_LEVELS',  'XML_ERR_'),
+    'xmlErrorDomain'      : ('__ERROR_DOMAINS', 'XML_FROM_'),
+    'xmlParserErrors'     : ('__PARSER_ERROR_TYPES',   'XML_'),
+    'xmlXPathError'       : ('__XPATH_ERROR_TYPES',   ''),
+    'xmlSchemaValidError' : ('__XMLSCHEMA_ERROR_TYPES',   'XML_'),
+    'xmlRelaxNGValidErr'  : ('__RELAXNG_ERROR_TYPES',   'XML_'),
     }
 
-ENUM_ORDER = ('xmlErrorLevel', 'xmlErrorDomain', 'xmlParserErrors')
+ENUM_ORDER = ('xmlErrorLevel', 'xmlErrorDomain', 'xmlParserErrors',
+              'xmlXPathError', 'xmlSchemaValidError', 'xmlRelaxNGValidErr')
 
 COMMENT = """
 # This section is generated by the script '%s'.
@@ -61,27 +65,40 @@ def regenerate_file(filename, result):
     f.write(''.join(post))
     f.close()
 
-def parse_enums(html_file):
+def parse_enums(html_dir, html_filename, enum_dict):
     PARSE_ENUM_NAME  = re.compile('\s*enum\s+(\w+)\s*{', re.I).match
     PARSE_ENUM_VALUE = re.compile('\s*=\s+([0-9]+)\s*(?::\s*(.*))?').match
-    tree = etree.parse(html_file)
+    tree = etree.parse(os.path.join(html_dir, html_filename))
     xpath = etree.XPathEvaluator(
         tree, namespaces={'html' : 'http://www.w3.org/1999/xhtml'})
+    collect_text = etree.XPath("string()")
 
-    enum_dict = {}
-    enums = xpath.evaluate("//html:pre[@class = 'programlisting' and contains(text(), 'Enum') and html:a[@name]]")
+    enums = xpath.evaluate("//html:pre[@class = 'programlisting' and contains(text(), 'Enum')]")
     for enum in enums:
-        enum_name = PARSE_ENUM_NAME(enum.text).group(1)
+        enum_name = PARSE_ENUM_NAME(collect_text(enum))
+        if not enum_name or enum_name not in ENUM_MAP:
+            continue
+        enum_name = enum_name.group(1)
         print "Found enum", enum_name
         entries = []
-        enum_dict[enum_name] = entries
         for child in enum:
             name = child.text
-            value, descr = PARSE_ENUM_VALUE(child.tail).groups()
+            match = PARSE_ENUM_VALUE(child.tail)
+            if not match:
+                print("Ignoring enum %s (failed to parse field '%s')" % (
+                        enum_name, name))
+                break
+            value, descr = match.groups()
             entries.append((name, int(value), descr))
+        else:
+            enum_dict[enum_name] = entries
     return enum_dict
 
-enum_dict = parse_enums(HTML_FILE)
+enum_dict = {}
+parse_enums(HTML_DIR, 'libxml-xmlerror.html',   enum_dict)
+parse_enums(HTML_DIR, 'libxml-xpath.html',      enum_dict)
+parse_enums(HTML_DIR, 'libxml-xmlschemas.html', enum_dict)
+parse_enums(HTML_DIR, 'libxml-relaxng.html',    enum_dict)
 
 # regenerate source files
 pxi_result = []