Binary pointer frames
Test vectors: Binary pointer frames. Shown here as it is in conformance/vectors/pointer.json.
Generated at build time from conformance/vectors/pointer.json.
Protocol 0.1. Covers: SPEC.md section 11, "Pointer stream" (layout table and encoding of fractions).
Download pointer.json: the file itself, as the conformance suite reads it.
How to use it #
Encode frame and compare with hex; decode hex and compare with frame; clamp and rounding cases check the encoding of fractions and dt; invalid must be refused.
Layout #
offset 0 u8 kind 0x01 (BIN_POINTER)
offset 1 u8 flags 0x01 PTR_FLAG_DROPPABLE, 0x02 PTR_FLAG_END
offset 2 u16 ch
offset 4 u32 seq of the first point
offset 8 u16 n, number of points
offset 10 n x { u16 x, u16 y, u16 dt }, all little-endian
a fraction f is sent as round(f x 65535) after clamping to 0..1 (non-finite as 0), halves up; read back as u / 65535; dt is rounded and clamped to 0..65535Notes #
The layout and the rounding rule are in SPEC 11. cases use exact multiples of 1/65535; rounding checks fractions between two steps; clamp checks out-of-range values; invalid frames are not pointer frames.
cases #
4 cases.
[
{
"name": "one point, no flags",
"frame": {
"ch": 1,
"seq": 0,
"flags": 0,
"points": [
{
"x": 0,
"y": 1,
"dt": 0
}
]
},
"hex": "010001000000000001000000ffff0000"
},
{
"name": "three points, droppable",
"frame": {
"ch": 7,
"seq": 1000,
"flags": 1,
"points": [
{
"x": 0.000015259021896696422,
"y": 0.000030518043793392844,
"dt": 16
},
{
"x": 0.49999237048905165,
"y": 0.5000076295109483,
"dt": 17
},
{
"x": 0.9999847409781033,
"y": 1,
"dt": 65535
}
]
},
"hex": "01010700e80300000300010002001000ff7f00801100feffffffffff"
},
{
"name": "end of stroke, no points",
"frame": {
"ch": 65535,
"seq": 4294967295,
"flags": 2,
"points": []
},
"hex": "0102ffffffffffff0000"
},
{
"name": "end and droppable bits together",
"frame": {
"ch": 2,
"seq": 5,
"flags": 3,
"points": [
{
"x": 0.2,
"y": 0.8,
"dt": 8
}
]
},
"hex": "010302000500000001003333cccc0800"
}
]clamp #
3 cases.
[
{
"name": "x below 0 and y above 1",
"point": {
"x": -0.25,
"y": 1.5,
"dt": 10
},
"expect": {
"xu": 0,
"yu": 65535,
"dt": 10
}
},
{
"name": "dt above 65535",
"point": {
"x": 0,
"y": 0,
"dt": 70000
},
"expect": {
"xu": 0,
"yu": 0,
"dt": 65535
}
},
{
"name": "negative dt",
"point": {
"x": 1,
"y": 0,
"dt": -5
},
"expect": {
"xu": 65535,
"yu": 0,
"dt": 0
}
}
]rounding #
3 cases.
[
{
"name": "halves round up",
"point": {
"x": 0.5,
"y": 0.25,
"dt": 16.5
},
"expect": {
"xu": 32768,
"yu": 16384,
"dt": 17
}
},
{
"name": "0.1 and 0.9 land on halves and round up; dt rounds to nearest",
"point": {
"x": 0.1,
"y": 0.9,
"dt": 16.4
},
"expect": {
"xu": 6554,
"yu": 58982,
"dt": 16
}
},
{
"name": "between steps",
"point": {
"x": 0.123,
"y": 0.876,
"dt": 0
},
"expect": {
"xu": 8061,
"yu": 57409,
"dt": 0
}
}
]invalid #
4 entries.
name |
hex |
|---|---|
shorter than the header |
0100010000000000 |
wrong kind byte |
02000100000000000000 |
n says one point, none present |
01000100000000000100 |
one trailing byte |
0100010000000000000000 |
Generated by conformance/tools/make-vectors.mjs from the formulas in the specification, with node:crypto only.