settings: Rename MaxBigNumBits to MaxNumberBits

The idea of "big num" should not make it to the user interface.

Signed-off-by: Christophe de Dinechin <christophe@dinechin.org>

doc/5-Unimplemented.md

@@ -900,7 +900,7 @@ Hewlett-Packard RPL implementation.
 * [MathMenu](#mathmenu)
 * [MathModesMenu](#mathmodesmenu)
 * [MatrixMenu](#matrixmenu)
-* [MaxBigNumBits](#maxbignumbits): Maximum number of bits for a big integer
+* [MaxNumberBits](#maxnumberbits): Maximum number of bits used by a number
 * [MaxRewrites](#maxrewrites): Maximum number of equation rewrites
 * [MemMenu](#memmenu)
 * [MenuFirstPage](#menufirstpage)

doc/commands/binary.md

@@ -5,7 +5,7 @@ shifts. They operate on [based numbers](#based-numbers),
 [integers](#integers) or [big integers](#big-integers). When operating on based
 numbers, the operation happens on the number of bits defined by the
 [WordSize](#wordsize) setting. For integer values, the maximum number of bits is
-defined by the [MaxBigNumBits](#maxbignumbits) setting.
+defined by the [MaxNumberBits](#maxnumberbits) setting.
 
 ## ShiftLeft (SL)
 

doc/commands/settings.md

@@ -276,15 +276,21 @@ Defines the maximum number of rewrites in an equation.
 'B+A' rewrite` can never end, since it keeps rewriting terms. This setting
 indicates how many attempts at rewriting will be done before erroring out.
 
-## MaxBigNumBits
+## MaxNumberBits
 
-Define the maxmimum number of bits for a large integer.
+Define the maxmimum number of bits for numbers.
 
 Large integer operations can take a very long time, notably when displaying them
 on the stack. With the default value of 1024 bits, you can compute `100!` but
 computing `200!` will result in an error, `Number is too big`. You can however
-compute it seting a higher value for `MaxBigNumBits`, for example
-`2048 MaxBigNumBits`.
+compute it seting a higher value for `MaxNumberBits`, for example
+`2048 MaxNumberBits`.
+
+This setting applies to integer components in a number. In other words, it
+applies separately for the numerator and denominator in a fraction, or for the
+real and imaginary part in a complex number. A complex number made of two
+fractions can therefore take up to four times the number of bits specified by
+this setting.
 
 ## ToFractionIterations (→QIterations, →FracIterations)
 
@@ -353,6 +359,24 @@ When this setting is active, statistics functions that return sums, such as
 in special variable `ΣParameters`, in the same way as required for
 `LinearRegression`.
 
+## DetailedTypes
+
+The `Type` command returns detailed DB48X type values, which can distinguish
+between all DB48X object types, e.g. distinguish between polar and rectangular
+objects, or the three internal representations for decimal numbers. Returned
+values are all negative, which distinguishes them from RPL standard values, and
+makes it possible to write code that accepts both the compatible and detailed
+values.
+
+This is the opposite of [CompatibleTypes](#compatibletypes).
+
+## CompatibleTypes
+
+The `Type` command returns values as close to possible to the values documented
+on page 3-262 of the HP50G advanced reference manual. This is the opposite of
+[NativeTypes](#nativetypes).
+
+
 # States
 
 The calculator can save and restore state in files with extension `.48S`.

src/bignum.cc

@@ -506,7 +506,7 @@ bignum_g bignum::multiply(bignum_r yg, bignum_r xg, id ty)
     size_t wbits = wordsize(xt);
     size_t wbytes = (wbits + 7) / 8;
     size_t needed = xs + ys;
-    if (needed * 8 > Settings.MaxBigNumBits())
+    if (needed * 8 > Settings.MaxNumberBits())
     {
         rt.number_too_big_error();
         return nullptr;
@@ -820,7 +820,7 @@ bignum_p bignum::shift(bignum_r xg, int bits, bool rotate, bool arith)
     size_t   wbytes = (wbits + 7) / 8;
     size_t   abits  = bits < 0 ? 0 : bits;
     size_t   needed = std::max(xs + (abits + 7) / 8, wbytes) ;
-    if (needed * 8 > Settings.MaxBigNumBits())
+    if (needed * 8 > Settings.MaxNumberBits())
     {
         rt.number_too_big_error();
         return nullptr;

src/bignum.h

@@ -324,7 +324,7 @@ bignum_g bignum::binary(Op op, bignum_r xg, bignum_r yg, id ty)
     size_t   wbytes = (wbits + 7) / 8;
     uint16_t c      = 0;
     size_t   needed = std::max(xs, ys) + 1;
-    if (needed * 8 > Settings.MaxBigNumBits())
+    if (needed * 8 > Settings.MaxNumberBits())
     {
         rt.number_too_big_error();
         return nullptr;

src/ids.tbl

@@ -721,7 +721,7 @@ SETTING_BITS(BasedSeparatorCommand,  2, ID_BasedSpaces,  ID_BasedUnderscore,  ID
 SETTING(WordSize,               1U, 256U * 1024U,       512U)
 ALIAS(WordSize,                 "stws")
 OP(RecallWordSize,              "rcws")
-SETTING(MaxBigNumBits,          1U, 256U * 1024U,       512U)
+SETTING(MaxNumberBits,          1U, 256U * 1024U,       512U)
 SETTING(MaxFlags,               8U, 32768U,             128U)
 SETTING(MaxRewrites,            1U, 10000U,             100U)
 SETTING(FractionIterations,     1U, 10000U,             10U)

src/menu.cc

@@ -1382,7 +1382,7 @@ MENU(MathModesMenu,
 
      "0^0=1",                                   ID_ZeroPowerZeroIsOne,
      "0^0=?",                                   ID_ZeroPowerZeroIsUndefined,
-     MaxBigNumBits::label,                      ID_MaxBigNumBits,
+     MaxNumberBits::label,                      ID_MaxNumberBits,
      MaxRewrites::label,                        ID_MaxRewrites,
      FractionIterations::label,                 ID_FractionIterations,
      FractionDigits::label,                     ID_FractionDigits,

src/settings.cc

@@ -575,8 +575,8 @@ cstring setting::label(object::id ty)
         return printf("MLEd %u", s.MultilineEditorFont());
     case ID_CursorBlinkRate:
         return printf("Blink %u", s.CursorBlinkRate());
-    case ID_MaxBigNumBits:
-        return printf("Bits %u", s.MaxBigNumBits());
+    case ID_MaxNumberBits:
+        return printf("Bits %u", s.MaxNumberBits());
     case ID_MaxRewrites:
         return printf("Rwr %u", s.MaxRewrites());