Date: 2005-02-17T22:21:22 Editor: AlexKarasulu Wiki: Apache Directory Project Wiki Page: TLVPageInfo URL: http://wiki.apache.org/directory/TLVPageInfo
no comment Change Log: ------------------------------------------------------------------------------ @@ -19,11 +19,13 @@ Labels are numbers used to identify elements in a '''SET''' (see Asn.1 grammar), for instance. Generally, we don't have to deal with label above 30, which can be encoded in 5 bits (so this kind of '''Tag''' will be 1 byte long), and never above 1024. In the LDAP ASN.1 grammar, no label exceed 19 (in ''LdapMessage'', the ''ExtendedResponse'' label is 19), so we can focus on 1 byte tags. Whatever, it could be interesting to accept longer labels to be able to support any LDAP evolution (or other protocols, as this '''Tag''' decoder is not specifically written for LDAP) -''In the current setup we use a Java primitive int to store a maximum of 4 Tag field bytes. The Tags within a protocol can be pre-fabricated. An enumeration constant can be reused. This way all we need is a 4 byte reference to the constant to represent the tag value. This works well especially for using the enumeration type's value for switch statments. Today here's how we pack Tag bytes into a Java primitive int:'' +''In the current setup we use a Java primitive int to store a maximum of 4 Tag field bytes. The Tags within a protocol can be pre-fabricated. An enumeration constant can be reused. This way all we need is a 4 byte reference to the constant to represent the tag value. This works well especially for using the enumeration type's value for switch statments. Today here's how we pack Tag bytes into a Java primitive int.'' [http://incubator.apache.org/directory/subprojects/asn1/images/tag-integer-encoding.png] -''Looking at the automaton below it occurred to me that we can create two different Tag decoders. A stateless one for protocols that only requires a single byte Tag. You get whole bytes delivered in buffers so no need to keep state right? The stateless Tag really simplifies the automaton ;). For multibyte protocols we do collect bytes in a TagOctetCollector which is just a wrapper around a Java int. Do you think this needs to be revamped?'' +''Looking at the automaton below it occurred to me that we can create two different Tag decoders. A stateless one for protocols that only requires a single byte Tag. You get whole bytes delivered in buffers so no need to keep state right? The stateless Tag really simplifies the automaton ;). For multibyte protocols we do collect bytes in a TagOctetCollector which is just a wrapper around a Java int. Do you think this needs to be revamped? In the case of a protocol where the number of Tag octets can be greater than 1 we can use a less efficient stateful Tag decoder which needs the full logic of the automaton. I know this is just some extra if-then-else logic but that adds up for every byte. So we can swap out the Tag decoder used based on the protocol. Most hopefully will take the simple tag decoder and not pay too much of a penalty.'' + +''BTW note that if we have a protocol with more that 30 tags we only need one multibyte tag decoder per stream. This is a per connection setup cost paid once and not for every Tag that comes through.'' Decoding a Tag has to follow the finite state automaton showed on this picture :
