Merge pull request 'unify-byte-counts' (#21) from unify-byte-counts into main

Reviewed-on: #21
This commit is contained in:
Sasha Koshka 2025-10-13 08:49:46 -06:00
commit 2194198693
9 changed files with 173 additions and 47 deletions

View File

@ -75,16 +75,16 @@ connection. Thus, the value may range from 0 to 31 if unsigned, and from -16 to
17 if signed.
#### Large Integer (LI)
LI encodes an integer of up to 256 bits, which are stored in the payload. The CN
determine the length of the payload in bytes. The integer is big-endian. Whether
LI encodes an integer of up to 256 bits, which are stored in the payload. The
length of the payload (in bytes) is CN + 1. The integer is big-endian. Whether
the payload is interpreted as unsigned or as signed two's complement is semantic
information and must be agreed upon by both sides of the connection. Thus, the
value may range from 0 to 31 if unsigned, and from -16 to 17 if signed.
#### Floating Point (FP)
FP encodes an IEEE 754 floating point number of up to 256 bits, which are stored
in the payload. The CN determines the length of the payload in bytes, and it may
only be one of these values: 16, 32, 64, 128, or 256.
in the payload. The length of the payload (in bytes) is CN + 1. The only
supported bit widths for floats are as follows: 16, 32, 64, 128, and 256.
#### Small Byte Array (SBA)
SBA encodes an array of up to 32 bytes, which are stored in the paylod. The
@ -98,15 +98,16 @@ in bytes is determined by the CN.
#### One-Tag Array (OTA)
OTA encodes an array of up to 2^256 items, which are stored in the payload after
the length field and the item tag, where the length field comes first. Each item
must be the same length, as they all share the same tag. The length of the data
length field in bytes is determined by the CN.
must be the same length, as they all share the same tag. The length of the
length field (in bytes) is CN + 1.
#### Key-Tag-Value Table (KTV)
KTV encodes a table of up to 2^256 key/value pairs, which are stored in the
payload after the length field. The pairs themselves consist of a 16-bit
unsigned big-endian key followed by a tag and then the payload. Pair values can
be of different types and sizes. The order of the pairs is not significant and
should never be treated as such.
should never be treated as such. The length of the length field (in bytes) is
CN + 1.
## Transports
A transport is a protocol that HOPP connections can run on top of. HOPP

View File

@ -24,7 +24,7 @@ const preamble = `
const static = `
// Table is a KTV table with an undefined schema.
type Table map[uint16] any
type Table = map[uint16] any
// Message is any message that can be sent along this protocol.
type Message interface {
@ -372,7 +372,7 @@ func (this *Generator) generateEncodeValue(typ Type, valueSource, tagSource stri
nn, err = this.iprintf("}\n")
n += nn; if err != nil { return n, err }
nn, err = this.iprintf(
"nn, err = encoder.WriteUintN(uint64(len(%s)), %s.CN())\n",
"nn, err = encoder.WriteUintN(uint64(len(%s)), %s.CN() + 1)\n",
valueSource, tagSource)
n += nn; if err != nil { return n, err }
nn, err = this.generateErrorCheck()
@ -407,8 +407,8 @@ func (this *Generator) generateEncodeValue(typ Type, valueSource, tagSource stri
n += nn; if err != nil { return n, err }
nn, err = this.iprintf("nn, err = encoder.WriteTag(itemTag)\n")
n += nn; if err != nil { return n, err }
n += nn; if err != nil { return n, err }
nn, err = this.generateErrorCheck()
n += nn; if err != nil { return n, err }
nn, err = this.iprintf("for _, item := range %s {\n", valueSource)
n += nn; if err != nil { return n, err }
this.push()
@ -445,7 +445,7 @@ func (this *Generator) generateEncodeValue(typ Type, valueSource, tagSource stri
nn, err = this.iprintf("}\n")
n += nn; if err != nil { return n, err }
nn, err = this.iprintf(
"nn, err = encoder.WriteUintN(%d, %s.CN())\n",
"nn, err = encoder.WriteUintN(%d, %s.CN() + 1)\n",
len(typ.Fields), tagSource)
n += nn; if err != nil { return n, err }
nn, err = this.generateErrorCheck()
@ -709,7 +709,7 @@ func (this *Generator) generateDecodeBranch(hash [16]byte, typ Type, typeName st
this.pop()
nn, err = this.iprintf("}\n")
n += nn; if err != nil { return n, err }
nn, err = this.iprintf("%s, nn, err = decoder.ReadUintN(int(tag.CN()))\n", lengthVar)
nn, err = this.iprintf("%s, nn, err = decoder.ReadUintN(int(tag.CN()) + 1)\n", lengthVar)
n += nn; if err != nil { return n, err }
nn, err = this.generateErrorCheck()
n += nn; if err != nil { return n, err }
@ -785,7 +785,7 @@ func (this *Generator) generateDecodeBranch(hash [16]byte, typ Type, typeName st
this.pop()
nn, err = this.iprintf("}\n")
n += nn; if err != nil { return n, err }
nn, err = this.iprintf("%s, nn, err = decoder.ReadUintN(int(tag.CN()))\n", lengthVar)
nn, err = this.iprintf("%s, nn, err = decoder.ReadUintN(int(tag.CN()) + 1)\n", lengthVar)
n += nn; if err != nil { return n, err }
nn, err = this.generateErrorCheck()
n += nn; if err != nil { return n, err }
@ -952,13 +952,13 @@ func (this *Generator) generateTag(typ Type, source string) (tagVar string, n in
nn, err := this.iprintf("%s := tape.BufferTag([]byte(%s))\n", tagVar, source)
n += nn; if err != nil { return tagVar, n, err }
case TypeArray:
nn, err := this.iprintf("%s := tape.OTA.WithCN(tape.IntBytes(uint64(len(%s))))\n", tagVar, source)
nn, err := this.iprintf("%s := tape.OTA.WithCN(tape.IntBytes(uint64(len(%s))) - 1)\n", tagVar, source)
n += nn; if err != nil { return tagVar, n, err }
case TypeTable:
nn, err := this.iprintf("%s := tape.KTV.WithCN(tape.IntBytes(uint64(len(%s))))\n", tagVar, source)
nn, err := this.iprintf("%s := tape.KTV.WithCN(tape.IntBytes(uint64(len(%s))) - 1)\n", tagVar, source)
n += nn; if err != nil { return tagVar, n, err }
case TypeTableDefined:
nn, err := this.iprintf("%s := tape.KTV.WithCN(%d)\n", tagVar, tape.IntBytes(uint64(len(typ.Fields))))
nn, err := this.iprintf("%s := tape.KTV.WithCN(%d)\n", tagVar, tape.IntBytes(uint64(len(typ.Fields))) - 1)
n += nn; if err != nil { return tagVar, n, err }
case TypeNamed:
resolved, err := this.resolveTypeName(typ.Name)

View File

@ -103,7 +103,7 @@ func TestGenerateRunEncodeDecode(test *testing.T) {
}
testEncodeDecode(
&messageConnect,
tu.S(0xE1, 0x02).AddVar(
tu.S(0xE0, 0x02).AddVar(
[]byte { 0x00, 0x00, 0x86, 'r', 'a', 'r', 'i', 't', 'y' },
[]byte { 0x00, 0x01, 0x84, 'g', 'e', 'm', 's' },
))
@ -129,8 +129,8 @@ func TestGenerateRunEncodeDecode(test *testing.T) {
}
testEncodeDecode(
&messageUserList,
tu.S(0xE1, 0x01, 0x00, 0x00,
0xC1, 0x03, 0xE1,
tu.S(0xE0, 0x01, 0x00, 0x00,
0xC0, 0x03, 0xE0,
).Add(0x03).AddVar(
[]byte { 0x00, 0x00, 0x86, 'r', 'a', 'r', 'i', 't', 'y' },
[]byte { 0x00, 0x01, 0x87, 'a', 's', 'd', 'j', 'a', 'd', 's' },
@ -155,7 +155,7 @@ func TestGenerateRunEncodeDecode(test *testing.T) {
}
testEncodeDecode(
&messagePulse,
tu.S(0xE1, 0x05).AddVar(
tu.S(0xE0, 0x05).AddVar(
[]byte { 0x00, 0x00, 0x09 },
[]byte { 0x00, 0x01, 0x41, 0xCA, 0xDF },
[]byte { 0x00, 0x02, 0x61, 0x51, 0xAC },
@ -176,7 +176,7 @@ func TestGenerateRunEncodeDecode(test *testing.T) {
}
testEncodeDecode(
&messageNestedArray,
tu.S(0xC1, 0x02, 0xC1,
tu.S(0xC0, 0x02, 0xC0,
0x06, 0x20, 0xF1, 0xF2, 0xF3, 0xF4, 0xF5, 0xF6,
35, 0x20, 0xF1, 0xF2, 0xF3, 0xF4, 0xF5, 0xF6,
0xF7, 0xF8, 0xF9, 0xFA, 0xFB, 0xFC,
@ -202,7 +202,7 @@ func TestGenerateRunEncodeDecode(test *testing.T) {
}
testEncodeDecode(
&messageIntegers,
tu.S(0xE1, 13).AddVar(
tu.S(0xE0, 13).AddVar(
[]byte { 0x00, 0x00, 0x13 },
[]byte { 0x00, 0x01, 0x20, 0xC9 },
[]byte { 0x00, 0x02, 0x21, 0x34, 0xC9 },
@ -243,7 +243,7 @@ func TestGenerateRunEncodeDecode(test *testing.T) {
}
testEncodeDecode(
&messageDynamic,
tu.S(0xE1, 14).AddVar(
tu.S(0xE0, 14).AddVar(
[]byte { 0x00, 0x00, 0x20, 0x23 },
[]byte { 0x00, 0x01, 0x21, 0x32, 0x47 },
[]byte { 0x00, 0x02, 0x23, 0x87, 0x32, 0x45, 0x23 },
@ -255,11 +255,15 @@ func TestGenerateRunEncodeDecode(test *testing.T) {
[]byte { 0x00, 0x08, 0x63, 0x45, 0x12, 0x63, 0xCE },
[]byte { 0x00, 0x09, 0x67, 0x40, 0x74, 0x4E, 0x3D, 0x6F, 0xCD, 0x17, 0x75 },
[]byte { 0x00, 0x0A, 0x87, 'f', 'o', 'x', ' ', 'b', 'e', 'd' },
[]byte { 0x00, 0x0B, 0xC4, 0x00, 0x07, 0x00, 0x06, 0x00, 0x05, 0x00, 0x04 },
[]byte { 0x00, 0x0C, 0xE1, 0x02,
0x00, 0x01, 0x20, 0x08,
[]byte { 0x00, 0x0B, 0xC0, 0x04, 0x41,
0x00, 0x07,
0x00, 0x06,
0x00, 0x05,
0x00, 0x04 },
[]byte { 0x00, 0x0C, 0xE0, 0x02,
0x00, 0x01, 0x40, 0x08,
0x00, 0x02, 0x67, 0x40, 0x11, 0x99, 0x99, 0x99, 0x99, 0x99, 0x9A },
[]byte { 0x00, 0x0D, 0xE1, 0x03,
[]byte { 0x00, 0x0D, 0xE0, 0x03, // ERR
0x00, 0x01, 0x63, 0x43, 0xF4, 0xC0, 0x00,
0x00, 0x02, 0x82, 'h', 'i',
0x00, 0x03, 0x21, 0x39, 0x92 },

View File

@ -100,12 +100,12 @@ func testGenerateRun(test *testing.T, protocol *Protocol, title, imports, testCa
log.Println("decoding:")
destination := reflect.New(reflect.ValueOf(message).Elem().Type()).Interface().(Message)
flat := data.Flatten()
log.Println("before: ", destination)
log.Println("before: ", tu.Describe(destination))
decoder := tape.NewDecoder(bytes.NewBuffer(flat))
n, err = destination.Decode(decoder)
if err != nil { log.Fatalf("at %d: %v\n", n, err) }
log.Println("got: ", destination)
log.Println("correct:", message)
log.Println("got: ", tu.Describe(destination))
log.Println("correct:", tu.Describe(message))
if n != len(flat) {
log.Fatalf("n incorrect: %d != %d\n", n, len(flat))
}

View File

@ -2,8 +2,9 @@ package testutil
import "fmt"
import "slices"
import "strings"
import "reflect"
import "strings"
import "unicode"
// Snake lets you compare blocks of data where the ordering of certain parts may
// be swapped every which way. It is designed for comparing the encoding of
@ -124,6 +125,10 @@ func (this *describer) describe(value reflect.Value) {
return
}
value = reflect.ValueOf(value.Interface())
if !value.IsValid() {
this.printf("<invalid>")
return
}
switch value.Kind() {
case reflect.Array, reflect.Slice:
this.printf("[\n")
@ -141,12 +146,20 @@ func (this *describer) describe(value reflect.Value) {
typ := value.Type()
for index := range typ.NumField() {
indexBuffer := [1]int { index }
this.iprintf("%s: ", typ.Field(index).Name)
this.describe(value.FieldByIndex(indexBuffer[:]))
field := typ.Field(index)
this.iprintf("%s: ", field.Name)
for _, char := range field.Name {
if unicode.IsUpper(char) {
this.describe(value.FieldByIndex(indexBuffer[:]))
} else {
this.printf("<private>")
}
break
}
this.iprintf("\n")
}
this.indent -= 1
this.iprintf("}\n")
this.iprintf("}")
case reflect.Map:
this.printf("map {\n")
this.indent += 1
@ -159,7 +172,7 @@ func (this *describer) describe(value reflect.Value) {
this.iprintf("\n")
}
this.indent -= 1
this.iprintf("}\n")
this.iprintf("}")
case reflect.Pointer:
this.printf("& ")
this.describe(value.Elem())

View File

@ -116,9 +116,9 @@ func DecodeAnyInto(decoder *Decoder, destination any, tag Tag) (n int, err error
func DecodeAny(decoder *Decoder, tag Tag) (value any, n int, err error) {
destination, err := skeletonPointer(decoder, tag)
if err != nil { return nil, n, err }
nn, err := DecodeAnyInto(decoder, destination, tag)
nn, err := decodeAny(decoder, destination.Elem(), tag)
n += nn; if err != nil { return nil, n, err }
return destination, n, err
return destination.Elem().Interface(), n, err
}
// unknownSlicePlaceholder is inserted by skeletonValue and informs the program
@ -242,12 +242,22 @@ func decodeAnyOrError(decoder *Decoder, destination reflect.Value, tag Tag) (n i
}
lengthCast, err := Uint64ToIntSafe(length)
if err != nil { return n, err }
if isTypeAny(destination.Type()) {
// need a skeleton value if we are assigning to any.
value := reflect.MakeMapWithSize(reflect.TypeOf(dummyMap), lengthCast)
destination.Set(value)
destination = value
}
// im fucking so done dude. im so fucking done. this was
// supposed to only run when we need it but i guess it runs all
// the time, because when we get a map destination (a valid,
// allocated one) we break apart on SetMapIndex because of a nil
// map. yeah thats right. a fucking nil map panic. on the map we
// just allocated. but running this unconditionally (whether or
// not we receive an empty any value) actually makes it fucking
// work. go figure().
//
// (the map allocation functionality in skeletonPointer has been
// removed)
value := reflect.MakeMapWithSize(reflect.TypeOf(dummyMap), lengthCast)
destination.Set(value)
destination = value
destination.Clear()
for _ = range lengthCast {
key, nn, err := decoder.ReadUint16()
@ -387,7 +397,11 @@ func canSet(destination reflect.Type, tag Tag) error {
return errCantAssignf("cannot assign array to %v", destination)
}
case KTV:
if destination != reflect.TypeOf(dummyMap) {
cantAssign :=
destination.Kind() != reflect.Map ||
destination.Key().Kind() != reflect.Uint16 ||
!isTypeAny(destination.Elem())
if cantAssign {
return errCantAssignf("cannot assign table to %v", destination)
}
default:
@ -496,7 +510,8 @@ func skeletonValue(decoder *Decoder, tag Tag) (reflect.Value, error) {
func skeletonPointer(decoder *Decoder, tag Tag) (reflect.Value, error) {
typ, err := typeOf(decoder, tag)
if err != nil { return reflect.Value { }, err }
return reflect.New(typ), nil
value := reflect.New(typ)
return value, nil
}
// typeOf returns the type of the current tag being decoded. It does not use up

View File

@ -27,6 +27,12 @@ var samplePayloads = [][]byte {
0x02, 0x23, byte(LSI.WithCN(1)), 0x45, 0x67,
0x02, 0x24, byte(LI.WithCN(3)), 0x45, 0x67, 0x89, 0xAB,
},
/* map[uint16] any */ []byte {
byte(KTV.WithCN(0)), 3,
0x00, 0x01, 0x63, 0x43, 0xF4, 0xC0, 0x00,
0x00, 0x02, 0x82, 'h', 'i',
0x00, 0x03, 0x21, 0x39, 0x92,
},
}
var sampleValues = []any {
@ -49,6 +55,11 @@ var sampleValues = []any {
0x0223: int16(0x4567),
0x0224: uint32(0x456789AB),
},
/* map[uint16] any */ map[uint16] any {
0x0001: float32(489.5),
0x0002: "hi",
0x0003: uint16(0x3992),
},
}
type userDefinedInteger int16
@ -195,7 +206,7 @@ func TestDecodeWrongType(test *testing.T) {
{ var dest []string; arrayCase(&dest) }
}
// tables should only assign to other tables
if index != 12 {
if index != 12 && index != 13 {
test.Log("- map[uint16] any")
{ var dest = map[uint16] any { }; arrayCase(&dest) }
}
@ -296,3 +307,63 @@ func TestPeekSliceOnce(test *testing.T) {
test.Fatalf("wrong n: %d != %d", got, correct)
}
}
func TestTagAny(test *testing.T) {
cases := [][2]any {
[2]any { LSI.WithCN(3), int(9) },
[2]any { LSI.WithCN(0), int8(9) },
[2]any { LSI.WithCN(1), int16(9) },
[2]any { LSI.WithCN(3), int32(9) },
[2]any { LSI.WithCN(7), int64(9) },
[2]any { LI.WithCN(3), uint(9) },
[2]any { LI.WithCN(0), uint8(9) },
[2]any { LI.WithCN(1), uint16(9) },
[2]any { LI.WithCN(3), uint32(9) },
[2]any { LI.WithCN(7), uint64(9) },
[2]any { FP.WithCN(3), float32(9) },
[2]any { FP.WithCN(7), float64(9) },
[2]any { SBA.WithCN(12), "small string" },
[2]any { SBA.WithCN(12), []byte("small string") },
[2]any { LBA.WithCN(0), "this is a very long string that is long" },
[2]any { LBA.WithCN(0), []byte("this is a very long string that is long") },
[2]any { LBA.WithCN(1), lipsum },
[2]any { OTA.WithCN(0), []int { 1, 2, 3, 4, 5 } },
[2]any { OTA.WithCN(0), []string { "1, 2, 3, 4, 5" } },
[2]any { KTV.WithCN(0), map[uint16] any {
0: 1,
1: "wow",
2: 10.238,
45: -9,
9: map[uint16] any { },
}},
}
for _, cas := range cases {
test.Log(cas)
got, err := TagAny(cas[1])
if err != nil { test.Fatal(err) }
if correct := cas[0].(Tag); correct != got {
test.Fatalf("wrong tag: %v != %v", got, correct)
}
}
}
func TestDecodeAny(test *testing.T) {
for index, payload := range samplePayloads {
correctValue := sampleValues[index]
data := payload[1:]
decoder := NewDecoder(bytes.NewBuffer(data))
tag := Tag(payload[0])
decoded, n, err := DecodeAny(decoder, tag)
test.Log("n: ", n)
test.Log("tag: ", tag)
test.Log("got: ", tu.Describe(decoded))
test.Log("correct:", tu.Describe(correctValue))
if err != nil { test.Fatal(err) }
if !reflect.DeepEqual(decoded, correctValue) {
test.Fatal("values not equal")
}
if n != len(data) {
test.Fatalf("n not equal: %d != %d", n, len(data))
}
}
}

11
tape/strings.go Normal file
View File

@ -0,0 +1,11 @@
package tape
const lipsum = `Lorem ipsum dolor sit amet consectetur adipiscing elit. Quisque faucibus ex sapien vitae pellentesque sem placerat. In id cursus mi pretium tellus duis convallis. Tempus leo eu aenean sed diam urna tempor. Pulvinar vivamus fringilla lacus nec metus bibendum egestas. Iaculis massa nisl malesuada lacinia integer nunc posuere. Ut hendrerit semper vel class aptent taciti sociosqu. Ad litora torquent per conubia nostra inceptos himenaeos.
Lorem ipsum dolor sit amet consectetur adipiscing elit. Quisque faucibus ex sapien vitae pellentesque sem placerat. In id cursus mi pretium tellus duis convallis. Tempus leo eu aenean sed diam urna tempor. Pulvinar vivamus fringilla lacus nec metus bibendum egestas. Iaculis massa nisl malesuada lacinia integer nunc posuere. Ut hendrerit semper vel class aptent taciti sociosqu. Ad litora torquent per conubia nostra inceptos himenaeos.
Lorem ipsum dolor sit amet consectetur adipiscing elit. Quisque faucibus ex sapien vitae pellentesque sem placerat. In id cursus mi pretium tellus duis convallis. Tempus leo eu aenean sed diam urna tempor. Pulvinar vivamus fringilla lacus nec metus bibendum egestas. Iaculis massa nisl malesuada lacinia integer nunc posuere. Ut hendrerit semper vel class aptent taciti sociosqu. Ad litora torquent per conubia nostra inceptos himenaeos.
Lorem ipsum dolor sit amet consectetur adipiscing elit. Quisque faucibus ex sapien vitae pellentesque sem placerat. In id cursus mi pretium tellus duis convallis. Tempus leo eu aenean sed diam urna tempor. Pulvinar vivamus fringilla lacus nec metus bibendum egestas. Iaculis massa nisl malesuada lacinia integer nunc posuere. Ut hendrerit semper vel class aptent taciti sociosqu. Ad litora torquent per conubia nostra inceptos himenaeos.
Lorem ipsum dolor sit amet consectetur adipiscing elit. Quisque faucibus ex sapien vitae pellentesque sem placerat. In id cursus mi pretium tellus duis convallis. Tempus leo eu aenean sed diam urna tempor. Pulvinar vivamus fringilla lacus nec metus bibendum egestas. Iaculis massa nisl malesuada lacinia integer nunc posuere. Ut hendrerit semper vel class aptent taciti sociosqu. Ad litora torquent per conubia nostra inceptos himenaeos.`

View File

@ -18,6 +18,17 @@ type Tag byte; const (
CNLimit Tag = 32 // All valid CNs are < CNLimit
)
// what the first nybble of a tag means:
//
// 0-1 : SI
// 2-3 : LI
// 4-5 : LSI
// 6-7 : FP
// 8-9 : SBA
// A-B : LBA
// C-D : OTA
// E-F : KTV
func (tag Tag) TN() int {
return int(tag >> 5)
}
@ -67,6 +78,6 @@ func bufferLenTag(length int) Tag {
if length < int(CNLimit) {
return SBA.WithCN(length)
} else {
return LBA.WithCN(IntBytes(uint64(length)))
return LBA.WithCN(IntBytes(uint64(length)) - 1)
}
}