py/objstr: Add support for the :_b/o/x specifier in str.format.

This groups non-decimal values by fours, such as bbb_bbbb_bbbb.  It also
supports `{:_d}` to use underscore for decimal numbers (grouped in threes).

Use of incorrect ":,b" is not diagnosed.

Thanks to @dpgeorge for the suggestion to reduce code size.

Signed-off-by: Jeff Epler <[email protected]>

Author: jepler

py/mpprint.c

@@ -58,7 +58,7 @@ int mp_print_str(const mp_print_t *print, const char *str) {
     return len;
 }
 
-int mp_print_strn(const mp_print_t *print, const char *str, size_t len, int flags, char fill, int width) {
+int mp_print_strn(const mp_print_t *print, const char *str, size_t len, unsigned int flags, char fill, int width) {
     int left_pad = 0;
     int right_pad = 0;
     int pad = width - len;
@@ -201,7 +201,7 @@ static int mp_print_int(const mp_print_t *print, mp_uint_t x, int sgn, int base,
     return len;
 }
 
-int mp_print_mp_int(const mp_print_t *print, mp_obj_t x, int base, int base_char, int flags, char fill, int width, int prec) {
+int mp_print_mp_int(const mp_print_t *print, mp_obj_t x, unsigned int base, int base_char, int flags, char fill, int width, int prec) {
     // These are the only values for "base" that are required to be supported by this
     // function, since Python only allows the user to format integers in these bases.
     // If needed this function could be generalised to handle other values.
@@ -248,10 +248,7 @@ int mp_print_mp_int(const mp_print_t *print, mp_obj_t x, int base, int base_char
     int prefix_len = prefix - prefix_buf;
     prefix = prefix_buf;
 
-    char comma = '\0';
-    if (flags & PF_FLAG_SHOW_COMMA) {
-        comma = ',';
-    }
+    char comma = flags >> PF_FLAG_SEP_POS;
 
     // The size of this buffer is rather arbitrary. If it's not large
     // enough, a dynamic one will be allocated.
@@ -340,7 +337,7 @@ int mp_print_mp_int(const mp_print_t *print, mp_obj_t x, int base, int base_char
 }
 
 #if MICROPY_PY_BUILTINS_FLOAT
-int mp_print_float(const mp_print_t *print, mp_float_t f, char fmt, int flags, char fill, int width, int prec) {
+int mp_print_float(const mp_print_t *print, mp_float_t f, char fmt, unsigned int flags, char fill, int width, int prec) {
     char buf[32];
     char sign = '\0';
     int chrs = 0;

py/mpprint.h

@@ -33,11 +33,11 @@
 #define PF_FLAG_SPACE_SIGN        (0x004)
 #define PF_FLAG_NO_TRAILZ         (0x008)
 #define PF_FLAG_SHOW_PREFIX       (0x010)
-#define PF_FLAG_SHOW_COMMA        (0x020)
-#define PF_FLAG_PAD_AFTER_SIGN    (0x040)
-#define PF_FLAG_CENTER_ADJUST     (0x080)
-#define PF_FLAG_ADD_PERCENT       (0x100)
-#define PF_FLAG_SHOW_OCTAL_LETTER (0x200)
+#define PF_FLAG_PAD_AFTER_SIGN    (0x020)
+#define PF_FLAG_CENTER_ADJUST     (0x040)
+#define PF_FLAG_ADD_PERCENT       (0x080)
+#define PF_FLAG_SHOW_OCTAL_LETTER (0x100)
+#define PF_FLAG_SEP_POS           (9) // must be above all the above PF_FLAGs
 
 #if MICROPY_PY_IO && MICROPY_PY_SYS_STDFILES
 #define MP_PYTHON_PRINTER &mp_sys_stdout_print
@@ -69,9 +69,9 @@ extern const mp_print_t mp_sys_stdout_print;
 #endif
 
 int mp_print_str(const mp_print_t *print, const char *str);
-int mp_print_strn(const mp_print_t *print, const char *str, size_t len, int flags, char fill, int width);
+int mp_print_strn(const mp_print_t *print, const char *str, size_t len, unsigned int flags, char fill, int width);
 #if MICROPY_PY_BUILTINS_FLOAT
-int mp_print_float(const mp_print_t *print, mp_float_t f, char fmt, int flags, char fill, int width, int prec);
+int mp_print_float(const mp_print_t *print, mp_float_t f, char fmt, unsigned int flags, char fill, int width, int prec);
 #endif
 
 int mp_printf(const mp_print_t *print, const char *fmt, ...);

py/mpz.c

@@ -1672,6 +1672,8 @@ size_t mpz_as_str_inpl(const mpz_t *i, unsigned int base, const char *prefix, ch
 
     size_t ilen = i->len;
 
+    int n_comma = (base == 10) ? 3 : 4;
+
     char *s = str;
     if (ilen == 0) {
         if (prefix) {
@@ -1717,7 +1719,7 @@ size_t mpz_as_str_inpl(const mpz_t *i, unsigned int base, const char *prefix, ch
                 break;
             }
         }
-        if (!done && comma && (s - last_comma) == 3) {
+        if (!done && comma && (s - last_comma) == n_comma) {
             *s++ = comma;
             last_comma = s;
         }

py/objint.c

@@ -209,7 +209,7 @@ static const uint8_t log_base2_floor[] = {
 size_t mp_int_format_size(size_t num_bits, int base, const char *prefix, char comma) {
     assert(2 <= base && base <= 16);
     size_t num_digits = num_bits / log_base2_floor[base - 1] + 1;
-    size_t num_commas = comma ? num_digits / 3 : 0;
+    size_t num_commas = comma ? (base == 10 ? num_digits / 3 : num_digits / 4): 0;
     size_t prefix_len = prefix ? strlen(prefix) : 0;
     return num_digits + num_commas + prefix_len + 2; // +1 for sign, +1 for null byte
 }
@@ -251,6 +251,7 @@ char *mp_obj_int_formatted(char **buf, size_t *buf_size, size_t *fmt_size, mp_co
         sign = '-';
     }
 
+    int n_comma = (base == 10) ? 3 : 4;
     size_t needed_size = mp_int_format_size(sizeof(fmt_int_t) * 8, base, prefix, comma);
     if (needed_size > *buf_size) {
         *buf = m_new(char, needed_size);
@@ -275,7 +276,7 @@ char *mp_obj_int_formatted(char **buf, size_t *buf_size, size_t *fmt_size, mp_co
                 c += '0';
             }
             *(--b) = c;
-            if (comma && num != 0 && b > str && (last_comma - b) == 3) {
+            if (comma && num != 0 && b > str && (last_comma - b) == n_comma) {
                 *(--b) = comma;
                 last_comma = b;
             }

py/objstr.c

@@ -1184,7 +1184,7 @@ static vstr_t mp_obj_str_format_helper(const char *str, const char *top, int *ar
         int width = -1;
         int precision = -1;
         char type = '\0';
-        int flags = 0;
+        unsigned int flags = 0;
 
         if (format_spec) {
             // The format specifier (from http://docs.python.org/2/library/string.html#formatspec)
@@ -1229,8 +1229,9 @@ static vstr_t mp_obj_str_format_helper(const char *str, const char *top, int *ar
                 }
             }
             s = str_to_int(s, stop, &width);
-            if (*s == ',') {
-                flags |= PF_FLAG_SHOW_COMMA;
+            if (*s == ',' || *s == '_') {
+                MP_STATIC_ASSERT((unsigned)'_' << PF_FLAG_SEP_POS >> PF_FLAG_SEP_POS == '_');
+                flags |= (unsigned)*s << PF_FLAG_SEP_POS;
                 s++;
             }
             if (*s == '.') {

py/runtime.h

@@ -128,7 +128,7 @@ void mp_event_wait_indefinite(void);
 void mp_event_wait_ms(mp_uint_t timeout_ms);
 
 // extra printing method specifically for mp_obj_t's which are integral type
-int mp_print_mp_int(const mp_print_t *print, mp_obj_t x, int base, int base_char, int flags, char fill, int width, int prec);
+int mp_print_mp_int(const mp_print_t *print, mp_obj_t x, unsigned base, int base_char, int flags, char fill, int width, int prec);
 
 void mp_arg_check_num_sig(size_t n_args, size_t n_kw, uint32_t sig);
 static inline void mp_arg_check_num(size_t n_args, size_t n_kw, size_t n_args_min, size_t n_args_max, bool takes_kw) {

tests/basics/string_format.py

@@ -22,7 +22,17 @@ def test(fmt, *args):
 test("{:4x}", 123)
 test("{:4X}", 123)
 
+test("{:4,d}", 1)
+test("{:4_d}", 1)
+test("{:4_o}", 1)
+test("{:4_b}", 1)
+test("{:4_x}", 1)
+
 test("{:4,d}", 12345678)
+test("{:4_d}", 12345678)
+test("{:4_o}", 12345678)
+test("{:4_b}", 12345678)
+test("{:4_x}", 12345678)
 
 test("{:#4b}", 10)
 test("{:#4o}", 123)