quantumcoin 6.14.2 → 6.14.5

package/src.ts/utils/fixednumber.ts

@@ -1,643 +1,643 @@
-/**
- *  The **FixedNumber** class permits using values with decimal places,
- *  using fixed-pont math.
- *
- *  Fixed-point math is still based on integers under-the-hood, but uses an
- *  internal offset to store fractional components below, and each operation
- *  corrects for this after each operation.
- *
- *  @_section: api/utils/fixed-point-math:Fixed-Point Maths  [about-fixed-point-math]
- */
-import { getBytes } from "./data.js";
-import { assert, assertArgument, assertPrivate } from "./errors.js";
-import {
-    getBigInt, getNumber, fromTwos, mask, toBigInt
-} from "./maths.js";
-import { defineProperties } from "./properties.js";
-
-import type { BigNumberish, BytesLike, Numeric } from "./index.js";
-
-const BN_N1 = BigInt(-1);
-const BN_0 = BigInt(0);
-const BN_1 = BigInt(1);
-const BN_5 = BigInt(5);
-
-const _guard = { };
-
-
-// Constant to pull zeros from for multipliers
-let Zeros = "0000";
-while (Zeros.length < 80) { Zeros += Zeros; }
-
-// Returns a string "1" followed by decimal "0"s
-function getTens(decimals: number): bigint {
-    let result = Zeros;
-    while (result.length < decimals) { result += result; }
-    return BigInt("1" + result.substring(0, decimals));
-}
-
-
-
-    /*
-     *  Returns a new FixedFormat for %%value%%.
-     *
-     *  If %%value%% is specified as a ``number``, the bit-width is
-     *  128 bits and %%value%% is used for the ``decimals``.
-     *
-     *  A string %%value%% may begin with ``fixed`` or ``ufixed``
-     *  for signed and unsigned respectfully. If no other properties
-     *  are specified, the bit-width is 128-bits with 18 decimals.
-     *
-     *  To specify the bit-width and demicals, append them separated
-     *  by an ``"x"`` to the %%value%%.
-     *
-     *  For example, ``ufixed128x18`` describes an unsigned, 128-bit
-     *  wide format with 18 decimals.
-     *
-     *  If %%value%% is an other object, its properties for ``signed``,
-     *  ``width`` and ``decimals`` are checked.
-     */
-
-/**
- *  A description of a fixed-point arithmetic field.
- *
- *  When specifying the fixed format, the values override the default of
- *  a ``fixed128x18``, which implies a signed 128-bit value with 18
- *  decimals of precision.
- *
- *  The alias ``fixed`` and ``ufixed`` can be used for ``fixed128x18`` and
- *  ``ufixed128x18`` respectively.
- *
- *  When a fixed format string begins with a ``u``, it indicates the field
- *  is unsigned, so any negative values will overflow. The first number
- *  indicates the bit-width and the second number indicates the decimal
- *  precision.
- *
- *  When a ``number`` is used for a fixed format, it indicates the number
- *  of decimal places, and the default width and signed-ness will be used.
- *
- *  The bit-width must be byte aligned and the decimals can be at most 80.
- */
-export type FixedFormat = number | string | {
-    signed?: boolean,
-    width?: number,
-    decimals?: number
-};
-
-function checkValue(val: bigint, format: _FixedFormat, safeOp?: string): bigint {
-    const width = BigInt(format.width);
-    if (format.signed) {
-        const limit = (BN_1 << (width - BN_1));
-        assert(safeOp == null || (val >= -limit  && val < limit), "overflow", "NUMERIC_FAULT", {
-            operation: <string>safeOp, fault: "overflow", value: val
-        });
-
-        if (val > BN_0) {
-            val = fromTwos(mask(val, width), width);
-        } else {
-            val = -fromTwos(mask(-val, width), width);
-        }
-
-    } else {
-        const limit = (BN_1 << width);
-        assert(safeOp == null || (val >= 0 && val < limit), "overflow", "NUMERIC_FAULT", {
-            operation: <string>safeOp, fault: "overflow", value: val
-        });
-        val = (((val % limit) + limit) % limit) & (limit - BN_1);
-    }
-
-    return val;
-}
-
-type _FixedFormat = { signed: boolean, width: number, decimals: number, name: string }
-
-function getFormat(value?: FixedFormat): _FixedFormat {
-    if (typeof(value) === "number") { value = `fixed128x${value}` }
-
-    let signed = true;
-    let width = 128;
-    let decimals = 18;
-
-    if (typeof(value) === "string") {
-        // Parse the format string
-        if (value === "fixed") {
-            // defaults...
-        } else if (value === "ufixed") {
-            signed = false;
-        } else {
-            const match = value.match(/^(u?)fixed([0-9]+)x([0-9]+)$/);
-            assertArgument(match, "invalid fixed format", "format", value);
-            signed = (match[1] !== "u");
-            width = parseInt(match[2]);
-            decimals = parseInt(match[3]);
-        }
-    } else if (value) {
-        // Extract the values from the object
-        const v: any = value;
-        const check = (key: string, type: string, defaultValue: any): any => {
-            if (v[key] == null) { return defaultValue; }
-            assertArgument(typeof(v[key]) === type,
-                "invalid fixed format (" + key + " not " + type +")", "format." + key, v[key]);
-            return v[key];
-        }
-        signed = check("signed", "boolean", signed);
-        width = check("width", "number", width);
-        decimals = check("decimals", "number", decimals);
-    }
-
-    assertArgument((width % 8) === 0, "invalid FixedNumber width (not byte aligned)", "format.width", width);
-    assertArgument(decimals <= 80, "invalid FixedNumber decimals (too large)", "format.decimals", decimals);
-
-    const name = (signed ? "": "u") + "fixed" + String(width) + "x" + String(decimals);
-
-    return { signed, width, decimals, name };
-}
-
-function toString(val: bigint, decimals: number) {
-    let negative = "";
-    if (val < BN_0) {
-        negative = "-";
-        val *= BN_N1;
-    }
-
-    let str = val.toString();
-
-    // No decimal point for whole values
-    if (decimals === 0) { return (negative + str); }
-
-    // Pad out to the whole component (including a whole digit)
-    while (str.length <= decimals) { str = Zeros + str; }
-
-    // Insert the decimal point
-    const index = str.length - decimals;
-    str = str.substring(0, index) + "." + str.substring(index);
-
-    // Trim the whole component (leaving at least one 0)
-    while (str[0] === "0" && str[1] !== ".") {
-        str = str.substring(1);
-    }
-
-    // Trim the decimal component (leaving at least one 0)
-    while (str[str.length - 1] === "0" && str[str.length - 2] !== ".") {
-        str = str.substring(0, str.length - 1);
-    }
-
-    return (negative + str);
-}
-
-
-/**
- *  A FixedNumber represents a value over its [[FixedFormat]]
- *  arithmetic field.
- *
- *  A FixedNumber can be used to perform math, losslessly, on
- *  values which have decmial places.
- *
- *  A FixedNumber has a fixed bit-width to store values in, and stores all
- *  values internally by multiplying the value by 10 raised to the power of
- *  %%decimals%%.
- *
- *  If operations are performed that cause a value to grow too high (close to
- *  positive infinity) or too low (close to negative infinity), the value
- *  is said to //overflow//.
- *
- *  For example, an 8-bit signed value, with 0 decimals may only be within
- *  the range ``-128`` to ``127``; so ``-128 - 1`` will overflow and become
- *  ``127``. Likewise, ``127 + 1`` will overflow and become ``-127``.
- *
- *  Many operation have a normal and //unsafe// variant. The normal variant
- *  will throw a [[NumericFaultError]] on any overflow, while the //unsafe//
- *  variant will silently allow overflow, corrupting its value value.
- *
- *  If operations are performed that cause a value to become too small
- *  (close to zero), the value loses precison and is said to //underflow//.
- *
- *  For example, a value with 1 decimal place may store a number as small
- *  as ``0.1``, but the value of ``0.1 / 2`` is ``0.05``, which cannot fit
- *  into 1 decimal place, so underflow occurs which means precision is lost
- *  and the value becomes ``0``.
- *
- *  Some operations have a normal and //signalling// variant. The normal
- *  variant will silently ignore underflow, while the //signalling// variant
- *  will thow a [[NumericFaultError]] on underflow.
- */
-export class FixedNumber {
-
-    /**
-     *  The specific fixed-point arithmetic field for this value.
-     */
-    readonly format!: string;
-
-    readonly #format: _FixedFormat;
-
-    // The actual value (accounting for decimals)
-    #val: bigint;
-
-    // A base-10 value to multiple values by to maintain the magnitude
-    readonly #tens: bigint;
-
-    /**
-     *  This is a property so console.log shows a human-meaningful value.
-     *
-     *  @private
-     */
-    readonly _value!: string;
-
-    // Use this when changing this file to get some typing info,
-    // but then switch to any to mask the internal type
-    //constructor(guard: any, value: bigint, format: _FixedFormat) {
-
-    /**
-     *  @private
-     */
-    constructor(guard: any, value: bigint, format: any) {
-        assertPrivate(guard, _guard, "FixedNumber");
-
-        this.#val = value;
-
-        this.#format = format;
-
-        const _value = toString(value, format.decimals);
-
-        defineProperties<FixedNumber>(this, { format: format.name, _value });
-
-        this.#tens = getTens(format.decimals);
-    }
-
-    /**
-     *  If true, negative values are permitted, otherwise only
-     *  positive values and zero are allowed.
-     */
-    get signed(): boolean { return this.#format.signed; }
-
-    /**
-     *  The number of bits available to store the value.
-     */
-    get width(): number { return this.#format.width; }
-
-    /**
-     *  The number of decimal places in the fixed-point arithment field.
-     */
-    get decimals(): number { return this.#format.decimals; }
-
-    /**
-     *  The value as an integer, based on the smallest unit the
-     *  [[decimals]] allow.
-     */
-    get value(): bigint { return this.#val; }
-
-    #checkFormat(other: FixedNumber): void {
-        assertArgument(this.format === other.format,
-            "incompatible format; use fixedNumber.toFormat", "other", other);
-    }
-
-    #checkValue(val: bigint, safeOp?: string): FixedNumber {
-/*
-        const width = BigInt(this.width);
-        if (this.signed) {
-            const limit = (BN_1 << (width - BN_1));
-            assert(safeOp == null || (val >= -limit  && val < limit), "overflow", "NUMERIC_FAULT", {
-                operation: <string>safeOp, fault: "overflow", value: val
-            });
-
-            if (val > BN_0) {
-                val = fromTwos(mask(val, width), width);
-            } else {
-                val = -fromTwos(mask(-val, width), width);
-            }
-
-        } else {
-            const masked = mask(val, width);
-            assert(safeOp == null || (val >= 0 && val === masked), "overflow", "NUMERIC_FAULT", {
-                operation: <string>safeOp, fault: "overflow", value: val
-            });
-            val = masked;
-        }
-*/
-        val = checkValue(val, this.#format, safeOp);
-        return new FixedNumber(_guard, val, this.#format);
-    }
-
-    #add(o: FixedNumber, safeOp?: string): FixedNumber {
-        this.#checkFormat(o);
-        return this.#checkValue(this.#val + o.#val, safeOp);
-    }
-
-    /**
-     *  Returns a new [[FixedNumber]] with the result of %%this%% added
-     *  to %%other%%, ignoring overflow.
-     */
-    addUnsafe(other: FixedNumber): FixedNumber { return this.#add(other); }
-
-    /**
-     *  Returns a new [[FixedNumber]] with the result of %%this%% added
-     *  to %%other%%. A [[NumericFaultError]] is thrown if overflow
-     *  occurs.
-     */
-    add(other: FixedNumber): FixedNumber { return this.#add(other, "add"); }
-
-    #sub(o: FixedNumber, safeOp?: string): FixedNumber {
-        this.#checkFormat(o);
-        return this.#checkValue(this.#val - o.#val, safeOp);
-    }
-
-    /**
-     *  Returns a new [[FixedNumber]] with the result of %%other%% subtracted
-     *  from %%this%%, ignoring overflow.
-     */
-    subUnsafe(other: FixedNumber): FixedNumber { return this.#sub(other); }
-
-    /**
-     *  Returns a new [[FixedNumber]] with the result of %%other%% subtracted
-     *  from %%this%%. A [[NumericFaultError]] is thrown if overflow
-     *  occurs.
-     */
-    sub(other: FixedNumber): FixedNumber { return this.#sub(other, "sub"); }
-
-    #mul(o: FixedNumber, safeOp?: string): FixedNumber {
-        this.#checkFormat(o);
-        return this.#checkValue((this.#val * o.#val) / this.#tens, safeOp);
-    }
-
-    /**
-     *  Returns a new [[FixedNumber]] with the result of %%this%% multiplied
-     *  by %%other%%, ignoring overflow and underflow (precision loss).
-     */
-    mulUnsafe(other: FixedNumber): FixedNumber { return this.#mul(other); }
-
-    /**
-     *  Returns a new [[FixedNumber]] with the result of %%this%% multiplied
-     *  by %%other%%. A [[NumericFaultError]] is thrown if overflow
-     *  occurs.
-     */
-    mul(other: FixedNumber): FixedNumber { return this.#mul(other, "mul"); }
-
-    /**
-     *  Returns a new [[FixedNumber]] with the result of %%this%% multiplied
-     *  by %%other%%. A [[NumericFaultError]] is thrown if overflow
-     *  occurs or if underflow (precision loss) occurs.
-     */
-    mulSignal(other: FixedNumber): FixedNumber {
-        this.#checkFormat(other);
-        const value = this.#val * other.#val;
-        assert((value % this.#tens) === BN_0, "precision lost during signalling mul", "NUMERIC_FAULT", {
-            operation: "mulSignal", fault: "underflow", value: this
-        });
-        return this.#checkValue(value / this.#tens, "mulSignal");
-    }
-
-    #div(o: FixedNumber, safeOp?: string): FixedNumber {
-        assert(o.#val !== BN_0, "division by zero", "NUMERIC_FAULT", {
-            operation: "div", fault: "divide-by-zero", value: this
-        });
-        this.#checkFormat(o);
-        return this.#checkValue((this.#val * this.#tens) / o.#val, safeOp);
-    }
-
-    /**
-     *  Returns a new [[FixedNumber]] with the result of %%this%% divided
-     *  by %%other%%, ignoring underflow (precision loss). A
-     *  [[NumericFaultError]] is thrown if overflow occurs.
-     */
-    divUnsafe(other: FixedNumber): FixedNumber { return this.#div(other); }
-
-    /**
-     *  Returns a new [[FixedNumber]] with the result of %%this%% divided
-     *  by %%other%%, ignoring underflow (precision loss). A
-     *  [[NumericFaultError]] is thrown if overflow occurs.
-     */
-    div(other: FixedNumber): FixedNumber { return this.#div(other, "div"); }
-
-
-    /**
-     *  Returns a new [[FixedNumber]] with the result of %%this%% divided
-     *  by %%other%%. A [[NumericFaultError]] is thrown if underflow
-     *  (precision loss) occurs.
-     */
-    divSignal(other: FixedNumber): FixedNumber {
-        assert(other.#val !== BN_0, "division by zero", "NUMERIC_FAULT", {
-            operation: "div", fault: "divide-by-zero", value: this
-        });
-        this.#checkFormat(other);
-        const value = (this.#val * this.#tens);
-        assert((value % other.#val) === BN_0, "precision lost during signalling div", "NUMERIC_FAULT", {
-            operation: "divSignal", fault: "underflow", value: this
-        });
-        return this.#checkValue(value / other.#val, "divSignal");
-    }
-
-    /**
-     *  Returns a comparison result between %%this%% and %%other%%.
-     *
-     *  This is suitable for use in sorting, where ``-1`` implies %%this%%
-     *  is smaller, ``1`` implies %%this%% is larger and ``0`` implies
-     *  both are equal.
-     */
-     cmp(other: FixedNumber): number {
-         let a = this.value, b = other.value;
-
-         // Coerce a and b to the same magnitude
-         const delta = this.decimals - other.decimals;
-         if (delta > 0) {
-             b *= getTens(delta);
-         } else if (delta < 0) {
-             a *= getTens(-delta);
-         }
-
-         // Comnpare
-         if (a < b) { return -1; }
-         if (a > b) { return 1; }
-         return 0;
-     }
-
-    /**
-     *  Returns true if %%other%% is equal to %%this%%.
-     */
-     eq(other: FixedNumber): boolean { return this.cmp(other) === 0; }
-
-    /**
-     *  Returns true if %%other%% is less than to %%this%%.
-     */
-     lt(other: FixedNumber): boolean { return this.cmp(other) < 0; }
-
-    /**
-     *  Returns true if %%other%% is less than or equal to %%this%%.
-     */
-     lte(other: FixedNumber): boolean { return this.cmp(other) <= 0; }
-
-    /**
-     *  Returns true if %%other%% is greater than to %%this%%.
-     */
-     gt(other: FixedNumber): boolean { return this.cmp(other) > 0; }
-
-    /**
-     *  Returns true if %%other%% is greater than or equal to %%this%%.
-     */
-     gte(other: FixedNumber): boolean { return this.cmp(other) >= 0; }
-
-    /**
-     *  Returns a new [[FixedNumber]] which is the largest **integer**
-     *  that is less than or equal to %%this%%.
-     *
-     *  The decimal component of the result will always be ``0``.
-     */
-    floor(): FixedNumber {
-        let val = this.#val;
-        if (this.#val < BN_0) { val -= this.#tens - BN_1; }
-        val = (this.#val / this.#tens) * this.#tens;
-        return this.#checkValue(val, "floor");
-    }
-
-    /**
-     *  Returns a new [[FixedNumber]] which is the smallest **integer**
-     *  that is greater than or equal to %%this%%.
-     *
-     *  The decimal component of the result will always be ``0``.
-     */
-    ceiling(): FixedNumber {
-        let val = this.#val;
-        if (this.#val > BN_0) { val += this.#tens - BN_1; }
-        val = (this.#val / this.#tens) * this.#tens;
-        return this.#checkValue(val, "ceiling");
-    }
-
-    /**
-     *  Returns a new [[FixedNumber]] with the decimal component
-     *  rounded up on ties at %%decimals%% places.
-     */
-    round(decimals?: number): FixedNumber {
-        if (decimals == null) { decimals = 0; }
-
-        // Not enough precision to not already be rounded
-        if (decimals >= this.decimals) { return this; }
-
-        const delta = this.decimals - decimals;
-        const bump = BN_5 * getTens(delta - 1);
-
-        let value = this.value + bump;
-        const tens = getTens(delta);
-        value = (value / tens) * tens;
-
-        checkValue(value, this.#format, "round");
-
-        return new FixedNumber(_guard, value, this.#format);
-    }
-
-    /**
-     *  Returns true if %%this%% is equal to ``0``.
-     */
-    isZero(): boolean { return (this.#val === BN_0); }
-
-    /**
-     *  Returns true if %%this%% is less than ``0``.
-     */
-    isNegative(): boolean { return (this.#val < BN_0); }
-
-    /**
-     *  Returns the string representation of %%this%%.
-     */
-    toString(): string { return this._value; }
-
-    /**
-     *  Returns a float approximation.
-     *
-     *  Due to IEEE 754 precission (or lack thereof), this function
-     *  can only return an approximation and most values will contain
-     *  rounding errors.
-     */
-    toUnsafeFloat(): number { return parseFloat(this.toString()); }
-
-    /**
-     *  Return a new [[FixedNumber]] with the same value but has had
-     *  its field set to %%format%%.
-     *
-     *  This will throw if the value cannot fit into %%format%%.
-     */
-    toFormat(format: FixedFormat): FixedNumber {
-        return FixedNumber.fromString(this.toString(), format);
-    }
-
-    /**
-     *  Creates a new [[FixedNumber]] for %%value%% divided by
-     *  %%decimal%% places with %%format%%.
-     *
-     *  This will throw a [[NumericFaultError]] if %%value%% (once adjusted
-     *  for %%decimals%%) cannot fit in %%format%%, either due to overflow
-     *  or underflow (precision loss).
-     */
-    static fromValue(_value: BigNumberish, _decimals?: Numeric, _format?: FixedFormat): FixedNumber {
-        const decimals = (_decimals == null) ? 0: getNumber(_decimals);
-        const format = getFormat(_format);
-
-        let value = getBigInt(_value, "value");
-        const delta = decimals - format.decimals;
-        if (delta > 0) {
-            const tens = getTens(delta);
-            assert((value % tens) === BN_0, "value loses precision for format", "NUMERIC_FAULT", {
-                operation: "fromValue", fault: "underflow", value: _value
-            });
-            value /= tens;
-        } else if (delta < 0) {
-            value *= getTens(-delta);
-        }
-
-        checkValue(value, format, "fromValue");
-
-        return new FixedNumber(_guard, value, format);
-    }
-
-    /**
-     *  Creates a new [[FixedNumber]] for %%value%% with %%format%%.
-     *
-     *  This will throw a [[NumericFaultError]] if %%value%% cannot fit
-     *  in %%format%%, either due to overflow or underflow (precision loss).
-     */
-    static fromString(_value: string, _format?: FixedFormat): FixedNumber {
-        const match = _value.match(/^(-?)([0-9]*)\.?([0-9]*)$/);
-        assertArgument(match && (match[2].length + match[3].length) > 0, "invalid FixedNumber string value", "value", _value);
-
-        const format = getFormat(_format);
-
-        let whole = (match[2] || "0"), decimal = (match[3] || "");
-
-        // Pad out the decimals
-        while (decimal.length < format.decimals) { decimal += Zeros; }
-
-        // Check precision is safe
-        assert(decimal.substring(format.decimals).match(/^0*$/), "too many decimals for format", "NUMERIC_FAULT", {
-            operation: "fromString", fault: "underflow", value: _value
-        });
-
-        // Remove extra padding
-        decimal = decimal.substring(0, format.decimals);
-
-        const value = BigInt(match[1] + whole + decimal)
-
-        checkValue(value, format, "fromString");
-
-        return new FixedNumber(_guard, value, format);
-    }
-
-    /**
-     *  Creates a new [[FixedNumber]] with the big-endian representation
-     *  %%value%% with %%format%%.
-     *
-     *  This will throw a [[NumericFaultError]] if %%value%% cannot fit
-     *  in %%format%% due to overflow.
-     */
-    static fromBytes(_value: BytesLike, _format?: FixedFormat): FixedNumber {
-        let value = toBigInt(getBytes(_value, "value"));
-        const format = getFormat(_format);
-
-        if (format.signed) { value = fromTwos(value, format.width); }
-
-        checkValue(value, format, "fromBytes");
-
-        return new FixedNumber(_guard, value, format);
-    }
-}
-
-//const f1 = FixedNumber.fromString("12.56", "fixed16x2");
-//const f2 = FixedNumber.fromString("0.3", "fixed16x2");
-//console.log(f1.divSignal(f2));
-//const BUMP = FixedNumber.from("0.5");
+/**
+ *  The **FixedNumber** class permits using values with decimal places,
+ *  using fixed-pont math.
+ *
+ *  Fixed-point math is still based on integers under-the-hood, but uses an
+ *  internal offset to store fractional components below, and each operation
+ *  corrects for this after each operation.
+ *
+ *  @_section: api/utils/fixed-point-math:Fixed-Point Maths  [about-fixed-point-math]
+ */
+import { getBytes } from "./data.js";
+import { assert, assertArgument, assertPrivate } from "./errors.js";
+import {
+    getBigInt, getNumber, fromTwos, mask, toBigInt
+} from "./maths.js";
+import { defineProperties } from "./properties.js";
+
+import type { BigNumberish, BytesLike, Numeric } from "./index.js";
+
+const BN_N1 = BigInt(-1);
+const BN_0 = BigInt(0);
+const BN_1 = BigInt(1);
+const BN_5 = BigInt(5);
+
+const _guard = { };
+
+
+// Constant to pull zeros from for multipliers
+let Zeros = "0000";
+while (Zeros.length < 80) { Zeros += Zeros; }
+
+// Returns a string "1" followed by decimal "0"s
+function getTens(decimals: number): bigint {
+    let result = Zeros;
+    while (result.length < decimals) { result += result; }
+    return BigInt("1" + result.substring(0, decimals));
+}
+
+
+
+    /*
+     *  Returns a new FixedFormat for %%value%%.
+     *
+     *  If %%value%% is specified as a ``number``, the bit-width is
+     *  128 bits and %%value%% is used for the ``decimals``.
+     *
+     *  A string %%value%% may begin with ``fixed`` or ``ufixed``
+     *  for signed and unsigned respectfully. If no other properties
+     *  are specified, the bit-width is 128-bits with 18 decimals.
+     *
+     *  To specify the bit-width and demicals, append them separated
+     *  by an ``"x"`` to the %%value%%.
+     *
+     *  For example, ``ufixed128x18`` describes an unsigned, 128-bit
+     *  wide format with 18 decimals.
+     *
+     *  If %%value%% is an other object, its properties for ``signed``,
+     *  ``width`` and ``decimals`` are checked.
+     */
+
+/**
+ *  A description of a fixed-point arithmetic field.
+ *
+ *  When specifying the fixed format, the values override the default of
+ *  a ``fixed128x18``, which implies a signed 128-bit value with 18
+ *  decimals of precision.
+ *
+ *  The alias ``fixed`` and ``ufixed`` can be used for ``fixed128x18`` and
+ *  ``ufixed128x18`` respectively.
+ *
+ *  When a fixed format string begins with a ``u``, it indicates the field
+ *  is unsigned, so any negative values will overflow. The first number
+ *  indicates the bit-width and the second number indicates the decimal
+ *  precision.
+ *
+ *  When a ``number`` is used for a fixed format, it indicates the number
+ *  of decimal places, and the default width and signed-ness will be used.
+ *
+ *  The bit-width must be byte aligned and the decimals can be at most 80.
+ */
+export type FixedFormat = number | string | {
+    signed?: boolean,
+    width?: number,
+    decimals?: number
+};
+
+function checkValue(val: bigint, format: _FixedFormat, safeOp?: string): bigint {
+    const width = BigInt(format.width);
+    if (format.signed) {
+        const limit = (BN_1 << (width - BN_1));
+        assert(safeOp == null || (val >= -limit  && val < limit), "overflow", "NUMERIC_FAULT", {
+            operation: <string>safeOp, fault: "overflow", value: val
+        });
+
+        if (val > BN_0) {
+            val = fromTwos(mask(val, width), width);
+        } else {
+            val = -fromTwos(mask(-val, width), width);
+        }
+
+    } else {
+        const limit = (BN_1 << width);
+        assert(safeOp == null || (val >= 0 && val < limit), "overflow", "NUMERIC_FAULT", {
+            operation: <string>safeOp, fault: "overflow", value: val
+        });
+        val = (((val % limit) + limit) % limit) & (limit - BN_1);
+    }
+
+    return val;
+}
+
+type _FixedFormat = { signed: boolean, width: number, decimals: number, name: string }
+
+function getFormat(value?: FixedFormat): _FixedFormat {
+    if (typeof(value) === "number") { value = `fixed128x${value}` }
+
+    let signed = true;
+    let width = 128;
+    let decimals = 18;
+
+    if (typeof(value) === "string") {
+        // Parse the format string
+        if (value === "fixed") {
+            // defaults...
+        } else if (value === "ufixed") {
+            signed = false;
+        } else {
+            const match = value.match(/^(u?)fixed([0-9]+)x([0-9]+)$/);
+            assertArgument(match, "invalid fixed format", "format", value);
+            signed = (match[1] !== "u");
+            width = parseInt(match[2]);
+            decimals = parseInt(match[3]);
+        }
+    } else if (value) {
+        // Extract the values from the object
+        const v: any = value;
+        const check = (key: string, type: string, defaultValue: any): any => {
+            if (v[key] == null) { return defaultValue; }
+            assertArgument(typeof(v[key]) === type,
+                "invalid fixed format (" + key + " not " + type +")", "format." + key, v[key]);
+            return v[key];
+        }
+        signed = check("signed", "boolean", signed);
+        width = check("width", "number", width);
+        decimals = check("decimals", "number", decimals);
+    }
+
+    assertArgument((width % 8) === 0, "invalid FixedNumber width (not byte aligned)", "format.width", width);
+    assertArgument(decimals <= 80, "invalid FixedNumber decimals (too large)", "format.decimals", decimals);
+
+    const name = (signed ? "": "u") + "fixed" + String(width) + "x" + String(decimals);
+
+    return { signed, width, decimals, name };
+}
+
+function toString(val: bigint, decimals: number) {
+    let negative = "";
+    if (val < BN_0) {
+        negative = "-";
+        val *= BN_N1;
+    }
+
+    let str = val.toString();
+
+    // No decimal point for whole values
+    if (decimals === 0) { return (negative + str); }
+
+    // Pad out to the whole component (including a whole digit)
+    while (str.length <= decimals) { str = Zeros + str; }
+
+    // Insert the decimal point
+    const index = str.length - decimals;
+    str = str.substring(0, index) + "." + str.substring(index);
+
+    // Trim the whole component (leaving at least one 0)
+    while (str[0] === "0" && str[1] !== ".") {
+        str = str.substring(1);
+    }
+
+    // Trim the decimal component (leaving at least one 0)
+    while (str[str.length - 1] === "0" && str[str.length - 2] !== ".") {
+        str = str.substring(0, str.length - 1);
+    }
+
+    return (negative + str);
+}
+
+
+/**
+ *  A FixedNumber represents a value over its [[FixedFormat]]
+ *  arithmetic field.
+ *
+ *  A FixedNumber can be used to perform math, losslessly, on
+ *  values which have decmial places.
+ *
+ *  A FixedNumber has a fixed bit-width to store values in, and stores all
+ *  values internally by multiplying the value by 10 raised to the power of
+ *  %%decimals%%.
+ *
+ *  If operations are performed that cause a value to grow too high (close to
+ *  positive infinity) or too low (close to negative infinity), the value
+ *  is said to //overflow//.
+ *
+ *  For example, an 8-bit signed value, with 0 decimals may only be within
+ *  the range ``-128`` to ``127``; so ``-128 - 1`` will overflow and become
+ *  ``127``. Likewise, ``127 + 1`` will overflow and become ``-127``.
+ *
+ *  Many operation have a normal and //unsafe// variant. The normal variant
+ *  will throw a [[NumericFaultError]] on any overflow, while the //unsafe//
+ *  variant will silently allow overflow, corrupting its value value.
+ *
+ *  If operations are performed that cause a value to become too small
+ *  (close to zero), the value loses precison and is said to //underflow//.
+ *
+ *  For example, a value with 1 decimal place may store a number as small
+ *  as ``0.1``, but the value of ``0.1 / 2`` is ``0.05``, which cannot fit
+ *  into 1 decimal place, so underflow occurs which means precision is lost
+ *  and the value becomes ``0``.
+ *
+ *  Some operations have a normal and //signalling// variant. The normal
+ *  variant will silently ignore underflow, while the //signalling// variant
+ *  will thow a [[NumericFaultError]] on underflow.
+ */
+export class FixedNumber {
+
+    /**
+     *  The specific fixed-point arithmetic field for this value.
+     */
+    readonly format!: string;
+
+    readonly #format: _FixedFormat;
+
+    // The actual value (accounting for decimals)
+    #val: bigint;
+
+    // A base-10 value to multiple values by to maintain the magnitude
+    readonly #tens: bigint;
+
+    /**
+     *  This is a property so console.log shows a human-meaningful value.
+     *
+     *  @private
+     */
+    readonly _value!: string;
+
+    // Use this when changing this file to get some typing info,
+    // but then switch to any to mask the internal type
+    //constructor(guard: any, value: bigint, format: _FixedFormat) {
+
+    /**
+     *  @private
+     */
+    constructor(guard: any, value: bigint, format: any) {
+        assertPrivate(guard, _guard, "FixedNumber");
+
+        this.#val = value;
+
+        this.#format = format;
+
+        const _value = toString(value, format.decimals);
+
+        defineProperties<FixedNumber>(this, { format: format.name, _value });
+
+        this.#tens = getTens(format.decimals);
+    }
+
+    /**
+     *  If true, negative values are permitted, otherwise only
+     *  positive values and zero are allowed.
+     */
+    get signed(): boolean { return this.#format.signed; }
+
+    /**
+     *  The number of bits available to store the value.
+     */
+    get width(): number { return this.#format.width; }
+
+    /**
+     *  The number of decimal places in the fixed-point arithment field.
+     */
+    get decimals(): number { return this.#format.decimals; }
+
+    /**
+     *  The value as an integer, based on the smallest unit the
+     *  [[decimals]] allow.
+     */
+    get value(): bigint { return this.#val; }
+
+    #checkFormat(other: FixedNumber): void {
+        assertArgument(this.format === other.format,
+            "incompatible format; use fixedNumber.toFormat", "other", other);
+    }
+
+    #checkValue(val: bigint, safeOp?: string): FixedNumber {
+/*
+        const width = BigInt(this.width);
+        if (this.signed) {
+            const limit = (BN_1 << (width - BN_1));
+            assert(safeOp == null || (val >= -limit  && val < limit), "overflow", "NUMERIC_FAULT", {
+                operation: <string>safeOp, fault: "overflow", value: val
+            });
+
+            if (val > BN_0) {
+                val = fromTwos(mask(val, width), width);
+            } else {
+                val = -fromTwos(mask(-val, width), width);
+            }
+
+        } else {
+            const masked = mask(val, width);
+            assert(safeOp == null || (val >= 0 && val === masked), "overflow", "NUMERIC_FAULT", {
+                operation: <string>safeOp, fault: "overflow", value: val
+            });
+            val = masked;
+        }
+*/
+        val = checkValue(val, this.#format, safeOp);
+        return new FixedNumber(_guard, val, this.#format);
+    }
+
+    #add(o: FixedNumber, safeOp?: string): FixedNumber {
+        this.#checkFormat(o);
+        return this.#checkValue(this.#val + o.#val, safeOp);
+    }
+
+    /**
+     *  Returns a new [[FixedNumber]] with the result of %%this%% added
+     *  to %%other%%, ignoring overflow.
+     */
+    addUnsafe(other: FixedNumber): FixedNumber { return this.#add(other); }
+
+    /**
+     *  Returns a new [[FixedNumber]] with the result of %%this%% added
+     *  to %%other%%. A [[NumericFaultError]] is thrown if overflow
+     *  occurs.
+     */
+    add(other: FixedNumber): FixedNumber { return this.#add(other, "add"); }
+
+    #sub(o: FixedNumber, safeOp?: string): FixedNumber {
+        this.#checkFormat(o);
+        return this.#checkValue(this.#val - o.#val, safeOp);
+    }
+
+    /**
+     *  Returns a new [[FixedNumber]] with the result of %%other%% subtracted
+     *  from %%this%%, ignoring overflow.
+     */
+    subUnsafe(other: FixedNumber): FixedNumber { return this.#sub(other); }
+
+    /**
+     *  Returns a new [[FixedNumber]] with the result of %%other%% subtracted
+     *  from %%this%%. A [[NumericFaultError]] is thrown if overflow
+     *  occurs.
+     */
+    sub(other: FixedNumber): FixedNumber { return this.#sub(other, "sub"); }
+
+    #mul(o: FixedNumber, safeOp?: string): FixedNumber {
+        this.#checkFormat(o);
+        return this.#checkValue((this.#val * o.#val) / this.#tens, safeOp);
+    }
+
+    /**
+     *  Returns a new [[FixedNumber]] with the result of %%this%% multiplied
+     *  by %%other%%, ignoring overflow and underflow (precision loss).
+     */
+    mulUnsafe(other: FixedNumber): FixedNumber { return this.#mul(other); }
+
+    /**
+     *  Returns a new [[FixedNumber]] with the result of %%this%% multiplied
+     *  by %%other%%. A [[NumericFaultError]] is thrown if overflow
+     *  occurs.
+     */
+    mul(other: FixedNumber): FixedNumber { return this.#mul(other, "mul"); }
+
+    /**
+     *  Returns a new [[FixedNumber]] with the result of %%this%% multiplied
+     *  by %%other%%. A [[NumericFaultError]] is thrown if overflow
+     *  occurs or if underflow (precision loss) occurs.
+     */
+    mulSignal(other: FixedNumber): FixedNumber {
+        this.#checkFormat(other);
+        const value = this.#val * other.#val;
+        assert((value % this.#tens) === BN_0, "precision lost during signalling mul", "NUMERIC_FAULT", {
+            operation: "mulSignal", fault: "underflow", value: this
+        });
+        return this.#checkValue(value / this.#tens, "mulSignal");
+    }
+
+    #div(o: FixedNumber, safeOp?: string): FixedNumber {
+        assert(o.#val !== BN_0, "division by zero", "NUMERIC_FAULT", {
+            operation: "div", fault: "divide-by-zero", value: this
+        });
+        this.#checkFormat(o);
+        return this.#checkValue((this.#val * this.#tens) / o.#val, safeOp);
+    }
+
+    /**
+     *  Returns a new [[FixedNumber]] with the result of %%this%% divided
+     *  by %%other%%, ignoring underflow (precision loss). A
+     *  [[NumericFaultError]] is thrown if overflow occurs.
+     */
+    divUnsafe(other: FixedNumber): FixedNumber { return this.#div(other); }
+
+    /**
+     *  Returns a new [[FixedNumber]] with the result of %%this%% divided
+     *  by %%other%%, ignoring underflow (precision loss). A
+     *  [[NumericFaultError]] is thrown if overflow occurs.
+     */
+    div(other: FixedNumber): FixedNumber { return this.#div(other, "div"); }
+
+
+    /**
+     *  Returns a new [[FixedNumber]] with the result of %%this%% divided
+     *  by %%other%%. A [[NumericFaultError]] is thrown if underflow
+     *  (precision loss) occurs.
+     */
+    divSignal(other: FixedNumber): FixedNumber {
+        assert(other.#val !== BN_0, "division by zero", "NUMERIC_FAULT", {
+            operation: "div", fault: "divide-by-zero", value: this
+        });
+        this.#checkFormat(other);
+        const value = (this.#val * this.#tens);
+        assert((value % other.#val) === BN_0, "precision lost during signalling div", "NUMERIC_FAULT", {
+            operation: "divSignal", fault: "underflow", value: this
+        });
+        return this.#checkValue(value / other.#val, "divSignal");
+    }
+
+    /**
+     *  Returns a comparison result between %%this%% and %%other%%.
+     *
+     *  This is suitable for use in sorting, where ``-1`` implies %%this%%
+     *  is smaller, ``1`` implies %%this%% is larger and ``0`` implies
+     *  both are equal.
+     */
+     cmp(other: FixedNumber): number {
+         let a = this.value, b = other.value;
+
+         // Coerce a and b to the same magnitude
+         const delta = this.decimals - other.decimals;
+         if (delta > 0) {
+             b *= getTens(delta);
+         } else if (delta < 0) {
+             a *= getTens(-delta);
+         }
+
+         // Comnpare
+         if (a < b) { return -1; }
+         if (a > b) { return 1; }
+         return 0;
+     }
+
+    /**
+     *  Returns true if %%other%% is equal to %%this%%.
+     */
+     eq(other: FixedNumber): boolean { return this.cmp(other) === 0; }
+
+    /**
+     *  Returns true if %%other%% is less than to %%this%%.
+     */
+     lt(other: FixedNumber): boolean { return this.cmp(other) < 0; }
+
+    /**
+     *  Returns true if %%other%% is less than or equal to %%this%%.
+     */
+     lte(other: FixedNumber): boolean { return this.cmp(other) <= 0; }
+
+    /**
+     *  Returns true if %%other%% is greater than to %%this%%.
+     */
+     gt(other: FixedNumber): boolean { return this.cmp(other) > 0; }
+
+    /**
+     *  Returns true if %%other%% is greater than or equal to %%this%%.
+     */
+     gte(other: FixedNumber): boolean { return this.cmp(other) >= 0; }
+
+    /**
+     *  Returns a new [[FixedNumber]] which is the largest **integer**
+     *  that is less than or equal to %%this%%.
+     *
+     *  The decimal component of the result will always be ``0``.
+     */
+    floor(): FixedNumber {
+        let val = this.#val;
+        if (this.#val < BN_0) { val -= this.#tens - BN_1; }
+        val = (this.#val / this.#tens) * this.#tens;
+        return this.#checkValue(val, "floor");
+    }
+
+    /**
+     *  Returns a new [[FixedNumber]] which is the smallest **integer**
+     *  that is greater than or equal to %%this%%.
+     *
+     *  The decimal component of the result will always be ``0``.
+     */
+    ceiling(): FixedNumber {
+        let val = this.#val;
+        if (this.#val > BN_0) { val += this.#tens - BN_1; }
+        val = (this.#val / this.#tens) * this.#tens;
+        return this.#checkValue(val, "ceiling");
+    }
+
+    /**
+     *  Returns a new [[FixedNumber]] with the decimal component
+     *  rounded up on ties at %%decimals%% places.
+     */
+    round(decimals?: number): FixedNumber {
+        if (decimals == null) { decimals = 0; }
+
+        // Not enough precision to not already be rounded
+        if (decimals >= this.decimals) { return this; }
+
+        const delta = this.decimals - decimals;
+        const bump = BN_5 * getTens(delta - 1);
+
+        let value = this.value + bump;
+        const tens = getTens(delta);
+        value = (value / tens) * tens;
+
+        checkValue(value, this.#format, "round");
+
+        return new FixedNumber(_guard, value, this.#format);
+    }
+
+    /**
+     *  Returns true if %%this%% is equal to ``0``.
+     */
+    isZero(): boolean { return (this.#val === BN_0); }
+
+    /**
+     *  Returns true if %%this%% is less than ``0``.
+     */
+    isNegative(): boolean { return (this.#val < BN_0); }
+
+    /**
+     *  Returns the string representation of %%this%%.
+     */
+    toString(): string { return this._value; }
+
+    /**
+     *  Returns a float approximation.
+     *
+     *  Due to IEEE 754 precission (or lack thereof), this function
+     *  can only return an approximation and most values will contain
+     *  rounding errors.
+     */
+    toUnsafeFloat(): number { return parseFloat(this.toString()); }
+
+    /**
+     *  Return a new [[FixedNumber]] with the same value but has had
+     *  its field set to %%format%%.
+     *
+     *  This will throw if the value cannot fit into %%format%%.
+     */
+    toFormat(format: FixedFormat): FixedNumber {
+        return FixedNumber.fromString(this.toString(), format);
+    }
+
+    /**
+     *  Creates a new [[FixedNumber]] for %%value%% divided by
+     *  %%decimal%% places with %%format%%.
+     *
+     *  This will throw a [[NumericFaultError]] if %%value%% (once adjusted
+     *  for %%decimals%%) cannot fit in %%format%%, either due to overflow
+     *  or underflow (precision loss).
+     */
+    static fromValue(_value: BigNumberish, _decimals?: Numeric, _format?: FixedFormat): FixedNumber {
+        const decimals = (_decimals == null) ? 0: getNumber(_decimals);
+        const format = getFormat(_format);
+
+        let value = getBigInt(_value, "value");
+        const delta = decimals - format.decimals;
+        if (delta > 0) {
+            const tens = getTens(delta);
+            assert((value % tens) === BN_0, "value loses precision for format", "NUMERIC_FAULT", {
+                operation: "fromValue", fault: "underflow", value: _value
+            });
+            value /= tens;
+        } else if (delta < 0) {
+            value *= getTens(-delta);
+        }
+
+        checkValue(value, format, "fromValue");
+
+        return new FixedNumber(_guard, value, format);
+    }
+
+    /**
+     *  Creates a new [[FixedNumber]] for %%value%% with %%format%%.
+     *
+     *  This will throw a [[NumericFaultError]] if %%value%% cannot fit
+     *  in %%format%%, either due to overflow or underflow (precision loss).
+     */
+    static fromString(_value: string, _format?: FixedFormat): FixedNumber {
+        const match = _value.match(/^(-?)([0-9]*)\.?([0-9]*)$/);
+        assertArgument(match && (match[2].length + match[3].length) > 0, "invalid FixedNumber string value", "value", _value);
+
+        const format = getFormat(_format);
+
+        let whole = (match[2] || "0"), decimal = (match[3] || "");
+
+        // Pad out the decimals
+        while (decimal.length < format.decimals) { decimal += Zeros; }
+
+        // Check precision is safe
+        assert(decimal.substring(format.decimals).match(/^0*$/), "too many decimals for format", "NUMERIC_FAULT", {
+            operation: "fromString", fault: "underflow", value: _value
+        });
+
+        // Remove extra padding
+        decimal = decimal.substring(0, format.decimals);
+
+        const value = BigInt(match[1] + whole + decimal)
+
+        checkValue(value, format, "fromString");
+
+        return new FixedNumber(_guard, value, format);
+    }
+
+    /**
+     *  Creates a new [[FixedNumber]] with the big-endian representation
+     *  %%value%% with %%format%%.
+     *
+     *  This will throw a [[NumericFaultError]] if %%value%% cannot fit
+     *  in %%format%% due to overflow.
+     */
+    static fromBytes(_value: BytesLike, _format?: FixedFormat): FixedNumber {
+        let value = toBigInt(getBytes(_value, "value"));
+        const format = getFormat(_format);
+
+        if (format.signed) { value = fromTwos(value, format.width); }
+
+        checkValue(value, format, "fromBytes");
+
+        return new FixedNumber(_guard, value, format);
+    }
+}
+
+//const f1 = FixedNumber.fromString("12.56", "fixed16x2");
+//const f2 = FixedNumber.fromString("0.3", "fixed16x2");
+//console.log(f1.divSignal(f2));
+//const BUMP = FixedNumber.from("0.5");