OPENDJ-2792 Fix schema parsing to correctly read base64 encoded values

Fabio Pistolesi

14.06.2016 5e37a4a285508c0acefe9d771077a7ff48f83ddb

OPENDJ-2792 Fix schema parsing to correctly read base64 encoded values

Reading the concatenated schema is relying on pure string processing, instead of LDAP parsing,
it silently ignored definitions that could not be parsed for any reason.
Moreover, introduce a bit more error reporting.

 opendj-server-legacy/src/main/java/org/opends/server/types/Schema.java

@@ -23,6 +23,7 @@
import java.io.FileWriter;
import java.io.FilenameFilter;
import java.io.IOException;
import java.text.ParseException;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Collection;
@@ -63,6 +64,7 @@
import org.opends.server.schema.NameFormSyntax;
import org.opends.server.schema.ObjectClassSyntax;
import org.opends.server.schema.SomeSchemaElement;
import org.opends.server.util.Base64;
import org.opends.server.util.ServerConstants;
import org.opends.server.util.StaticUtils;

@@ -2257,27 +2259,8 @@
      // element list.
      for (StringBuilder buffer : lines)
      {
        // Get the line and add the X-SCHEMA-FILE extension to the end
        // of it.  All of them should end with " )" but some might
        // have the parenthesis crammed up against the last character
        // so deal with that as well.
        String line = buffer.toString().trim();
        if (line.endsWith(" )"))
        {
         line = line.substring(0, line.length()-1) +
                SCHEMA_PROPERTY_FILENAME + " '" + f.getName() + "' )";
        }
        else if (line.endsWith(")"))
        {
         line = line.substring(0, line.length()-1) + " " +
                SCHEMA_PROPERTY_FILENAME + " '" + f.getName() + "' )";
        }
        else
        {
          continue;
        }

        parseSchemaLine(line, attributeTypes, objectClasses,
        parseSchemaLine(line, f.getName(), attributeTypes, objectClasses,
            nameForms, ditContentRules, ditStructureRules, matchingRuleUses,
            ldapSyntaxes);
      }
@@ -2336,7 +2319,7 @@
      {
        break;
      }
      parseSchemaLine(line, attributeTypes, objectClasses,
      parseSchemaLine(line, null, attributeTypes, objectClasses,
          nameForms, ditContentRules, ditStructureRules, matchingRuleUses,
          ldapSyntaxes);
    }
@@ -2344,79 +2327,80 @@
    reader.close();
  }

  /**
   * Parse a line of a schema file into the provided sets.
   *
   * @param line                The current line of schema.
   * @param  attributeTypes     The set into which to place the
   *                            attribute type if the line represents
   *                            one.
   * @param  objectClasses      The set into which to place the object
   *                            class if the line represents one.
   * @param  nameForms          The set into which to place the name
   *                            form if the line represents one.
   * @param  ditContentRules    The set into which to place the DIT
   *                            content rule if the line represents one.
   * @param  ditStructureRules  The set into which to place the DIT
   *                            structure rule if the line represents one.
   * @param  matchingRuleUses   The set into which to place the
   *                            matching rule use if the line represents
   *                            one.
   * @param ldapSyntaxes        The set into which to place the ldap
   *                            syntax if the line represents one.
   */

  private static void parseSchemaLine(String line,
                               Set<String> attributeTypes,
                               Set<String> objectClasses,
                               Set<String> nameForms,
                               Set<String> ditContentRules,
                               Set<String> ditStructureRules,
                               Set<String> matchingRuleUses,
                               Set<String> ldapSyntaxes)
  private static void parseSchemaLine(String line, String fileName,
      Set<String> attributeTypes,
      Set<String> objectClasses,
      Set<String> nameForms,
      Set<String> ditContentRules,
      Set<String> ditStructureRules,
      Set<String> matchingRuleUses,
      Set<String> ldapSyntaxes)
  {
    String value;
    String lowerLine = toLowerCase(line);
    if (lowerLine.startsWith(ATTR_ATTRIBUTE_TYPES_LC))

    try
    {
      value =
          line.substring(ATTR_ATTRIBUTE_TYPES.length()+1).trim();
      attributeTypes.add(value);
    }
    else if (lowerLine.startsWith(ATTR_OBJECTCLASSES_LC))
      if (lowerLine.startsWith(ATTR_ATTRIBUTE_TYPES_LC))
      {
        attributeTypes.add(getSchemaDefinition(line.substring(ATTR_ATTRIBUTE_TYPES_LC.length()), fileName));
      }
      else if (lowerLine.startsWith(ATTR_OBJECTCLASSES_LC))
      {
        objectClasses.add(getSchemaDefinition(line.substring(ATTR_OBJECTCLASSES_LC.length()), fileName));
      }
      else if (lowerLine.startsWith(ATTR_NAME_FORMS_LC))
      {
        nameForms.add(getSchemaDefinition(line.substring(ATTR_NAME_FORMS_LC.length()), fileName));
      }
      else if (lowerLine.startsWith(ATTR_DIT_CONTENT_RULES_LC))
      {
        ditContentRules.add(getSchemaDefinition(line.substring(ATTR_DIT_CONTENT_RULES_LC.length()), fileName));
      }
      else if (lowerLine.startsWith(ATTR_DIT_STRUCTURE_RULES_LC))
      {
        ditStructureRules.add(getSchemaDefinition(line.substring(ATTR_DIT_STRUCTURE_RULES_LC.length()), fileName));
      }
      else if (lowerLine.startsWith(ATTR_MATCHING_RULE_USE_LC))
      {
        matchingRuleUses.add(getSchemaDefinition(line.substring(ATTR_MATCHING_RULE_USE_LC.length()), fileName));
      }
      else if (lowerLine.startsWith(ATTR_LDAP_SYNTAXES_LC))
      {
        ldapSyntaxes.add(getSchemaDefinition(line.substring(ATTR_LDAP_SYNTAXES_LC.length()), fileName));
      }
    } catch (ParseException pe)
    {
      value = line.substring(ATTR_OBJECTCLASSES.length()+1).trim();
      objectClasses.add(value);
      logger.error(ERR_SCHEMA_PARSE_LINE.get(line, pe.getLocalizedMessage()));
    }
    else if (lowerLine.startsWith(ATTR_NAME_FORMS_LC))
  }

  private static String getSchemaDefinition(String line, String fileName) throws ParseException
  {
    if (line.startsWith("::"))
    {
      value = line.substring(ATTR_NAME_FORMS.length()+1).trim();
      nameForms.add(value);
      line = ByteString.wrap(Base64.decode(line.substring(2).trim())).toString();
    }
    else if (lowerLine.startsWith(ATTR_DIT_CONTENT_RULES_LC))
    else if (line.startsWith(":"))
    {
      value = line.substring(
          ATTR_DIT_CONTENT_RULES.length()+1).trim();
      ditContentRules.add(value);
      line = line.substring(1).trim();
    }
    else if (lowerLine.startsWith(ATTR_DIT_STRUCTURE_RULES_LC))
    else
    {
      value = line.substring(
          ATTR_DIT_STRUCTURE_RULES.length()+1).trim();
      ditStructureRules.add(value);
      throw new ParseException(ERR_SCHEMA_COULD_NOT_PARSE_DEFINITION.get().toString(), 0);
    }
    else if (lowerLine.startsWith(ATTR_MATCHING_RULE_USE_LC))

    if (fileName != null)
    {
      value = line.substring(
          ATTR_MATCHING_RULE_USE.length()+1).trim();
      matchingRuleUses.add(value);
      if (line.endsWith(" )"))
      {
        line = line.substring(0, line.length() - 1) + SCHEMA_PROPERTY_FILENAME + " '" + fileName + "' )";
      }
      else if (line.endsWith(")"))
      {
        line = line.substring(0, line.length() - 1) + " " + SCHEMA_PROPERTY_FILENAME + " '" + fileName + "' )";
      }
    }
    else if (lowerLine.startsWith(ATTR_LDAP_SYNTAXES_LC))
    {
      value = line.substring(
          ATTR_LDAP_SYNTAXES.length()+1).trim();
      ldapSyntaxes.add(value);
    }
    return line;
  }

  /**

 opendj-server-legacy/src/messages/org/opends/messages/backend.properties

@@ -1078,6 +1078,8 @@
children for DN <%s> (got %d, expecting %d)
ERR_VERIFY_ID2COUNT_WRONG_ID_597=File id2ChildrenCount references non-existing EntryID <%d>.
NOTE_REBUILD_NOTHING_TO_REBUILD_598=Rebuilding index finished: no indexes to rebuild.
NOTE_IMPORT_LDIF_OFFHEAP_MEM_BUF_INFO_520=Setting DB cache size to %d bytes. \
NOTE_IMPORT_LDIF_OFFHEAP_MEM_BUF_INFO_599=Setting DB cache size to %d bytes. \
 Using %d Mb off-heap memory through %d phase one buffers of %d Kb.
 
ERR_SCHEMA_PARSE_LINE_600=Ignoring schema definition '%s' because the following error occurred while \
  it was being parsed: %s
ERR_SCHEMA_COULD_NOT_PARSE_DEFINITION_601=Schema definition could not be parsed as valid attribute value