summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorMattias Andrée <maandree@kth.se>2021-09-07 23:16:18 +0200
committerMattias Andrée <maandree@kth.se>2021-09-07 23:16:18 +0200
commitd2ca51992555eedcd140a6e7e6ed5e5d2996ca67 (patch)
treed47680f9e7e840ded78eaf34fd49e0c1fda13d3c
parentRemove leftover code (diff)
downloadlibnumtext-d2ca51992555eedcd140a6e7e6ed5e5d2996ca67.tar.gz
libnumtext-d2ca51992555eedcd140a6e7e6ed5e5d2996ca67.tar.bz2
libnumtext-d2ca51992555eedcd140a6e7e6ed5e5d2996ca67.tar.xz
Simplify code
Signed-off-by: Mattias Andrée <maandree@kth.se>
-rw-r--r--common.h2
-rw-r--r--swedish.c317
2 files changed, 157 insertions, 162 deletions
diff --git a/common.h b/common.h
index 15801eb..e3c9912 100644
--- a/common.h
+++ b/common.h
@@ -24,5 +24,7 @@
#define UNICODE_MINUS "−"
+#define IS_UTF8_CHAR_CONTINUATION(B) (((B) & 0xC0) == 0x80)
+
ssize_t libnumtext_num2text_swedish__(char *outbuf, size_t outbuf_size, const char *num, size_t num_len, uint32_t flags);
diff --git a/swedish.c b/swedish.c
index ac080a0..5e6c194 100644
--- a/swedish.c
+++ b/swedish.c
@@ -2,6 +2,9 @@
#include "common.h"
+#define TYPE_INDEX(F) (ORDINAL(F) | DENOMINATOR(F))
+#define FORM_INDEX(F) ((PLURAL_FORM(F) | DEFINITE_FORM(F)) / LIBNUMTEXT_N2T_SWEDISH_PLURAL_FORM)
+
#define CARDINAL(F) (!ORDINAL(F))
#define ORDINAL(F) ((F) & LIBNUMTEXT_N2T_SWEDISH_ORDINAL)
@@ -79,7 +82,7 @@ static struct digit {
static struct ten {
const char *cardinal;
const char *ordinal;
-} tens[] = {
+} tens[10] = {
{NULL, NULL},
{NULL, NULL},
{"Tju¦go", "Tju¦gon"},
@@ -92,17 +95,38 @@ static struct ten {
{"Nit¦tio", "Nit¦ti¦on"}
};
-static const char *wholes_and_halves[][5] = {
- {"Hel", "He¦la", "Hel¦an", "Hel¦or¦na", "Hel¦te"},
- {"Halv", "Hal¦vor", "Halv¦an", "Halv¦or¦na", "Half¦te"}
+static struct special_denominator {
+ const char *cardinal[4 /* form */];
+ const char *ordinal;
+} wholes_and_halves[] = {
+ {{"Hel", "He¦la", "Hel¦an", "Hel¦or¦na"}, "Hel¦te"},
+ {{"Halv", "Hal¦vor", "Halv¦an", "Halv¦or¦na"}, "Half¦te"}
+};
+
+static struct denominator_suffix {
+ const char *cardinal[4 /* form */];
+ const char *ordinal;
+} denominator_suffixes = {{"||del", "||del¦ar", "||del¦en", "||del¦ar¦na"}, "||del¦te"};
+
+static const char *signs[2 /* whether + */][4 /* type */] = {
+ {"Min¦us ", "Min¦us-", "Min¦us-", "Min¦us-"},
+ {"Plus ", "Plus-", "Plus-", "Plus-"}
};
-static const char *great_suffixes[] = {
+static const char *great_suffixes[2] = {
"il¦jon",
"il¦jard"
};
-static const char *greats[][7] = {
+static struct great {
+ const char *single_digit;
+ const char *ones;
+ const char *ones_suffixes;
+ const char *tens_prefixes;
+ const char *tens;
+ const char *hundreds_prefixes;
+ const char *hundreds;
+} greats[] = {
{NULL, NULL, NULL, NULL, NULL, NULL, NULL},
{"||m", "||un", "", "n", "||de¦ci", "nx*", "||cen¦ti"},
{"||b", "||duo", "", "ms", "||vi|gin¦ti", "n", "||du|cen¦ti"},
@@ -122,6 +146,8 @@ struct state {
size_t len;
uint32_t flags;
char double_char;
+ char first;
+ const char *append_for_ordinal;
};
@@ -146,15 +172,13 @@ append(struct state *state, const char *appendage)
shift = appendage[0] == '<';
appendage = &appendage[shift];
appendage = &appendage[sizeof("¦") - 1];
- if (SECONDARY_HYPHENATION(flags)) {
+ if (SECONDARY_HYPHENATION(flags))
p = stpcpy(hyphen, "¦");
- } else if (SYLLABLE_HYPHENATION(flags)) {
+ else if (SYLLABLE_HYPHENATION(flags))
p = stpcpy(hyphen, "|");
- } else {
- *hyphen = 0;
- p = NULL; /* silence clang */
- }
- if (*hyphen) {
+ else
+ p = NULL;
+ if (p) {
if (shift && state->len <= state->outbuf_size) {
*p++ = state->outbuf[--state->len];
*p = '\0';
@@ -178,7 +202,7 @@ append(struct state *state, const char *appendage)
}
}
- if (state->len && (HYPHENATED(flags) | HYPHENATION(flags))) {
+ if (state->len && (HYPHENATED(flags) || HYPHENATION(flags))) {
if (isupper(appendage[0]) || (appendage[0] == "Å"[0] && appendage[1] == "Å"[1])) {
if (state->len < state->outbuf_size)
state->outbuf[state->len] = HYPHENATED(flags) ? '-' : '|';
@@ -196,41 +220,85 @@ append(struct state *state, const char *appendage)
static void
-suffix(struct state *state)
+append_final_digit(struct state *state, int digit)
{
- uint32_t flags = state->flags;
- const char *appendage;
+ uint32_t f = state->flags;
+ struct digit *d = &digits[digit];
- if (ORDINAL(flags)) {
- if (DENOMINATOR(flags))
- appendage = "||delte";
+ if (ORDINAL(f) || DENOMINATOR(f)) {
+ if (MASCULINE_GENDER(f) && d->ordinal_masculine)
+ append(state, d->ordinal_masculine);
else
- return;
- } else if (DENOMINATOR(flags)) {
- if (PLURAL_FORM(flags))
- appendage = DEFINITE_FORM(flags) ? "||del¦ar¦na" : "||del¦ar";
+ append(state, d->ordinal_other);
+ state->append_for_ordinal = d->ordinal_suffix;
+ } else {
+ if (COMMON_GENDER(f) || !d->cardinal_other)
+ append(state, d->cardinal_common);
else
- appendage = DEFINITE_FORM(flags) ? "||del¦en" : "||del";
+ append(state, d->cardinal_other);
+ state->append_for_ordinal = NULL;
+ }
+}
+
+
+static void
+append_hundreds(struct state *state, int hundreds)
+{
+ if (hundreds) {
+ if (state->first && hundreds == 1 && IMPLICIT_ONE(state->flags)) {
+ append(state, "Hun¦dra");
+ } else {
+ append(state, digits[hundreds].cardinal_common);
+ append(state, "||hun¦dra");
+ }
+ state->append_for_ordinal = "¦de";
+ state->first = 0;
+ }
+}
+
+
+static void
+append_thousands(struct state *state, int ten_thousands, int one_thousands)
+{
+ if (state->first && ten_thousands == 0 && one_thousands == 1 && IMPLICIT_ONE(state->flags)) {
+ append(state, "Tu¦sen");
} else {
- return;
+ if (tens[ten_thousands].cardinal)
+ append(state, tens[ten_thousands].cardinal);
+ else
+ one_thousands += ten_thousands * 10;
+ if (one_thousands)
+ append(state, digits[one_thousands].cardinal_common);
+ append(state, "||tu¦sen");
}
+ state->append_for_ordinal = "¦de";
+ state->first = 0;
+}
- append(state, appendage);
+
+static char
+get_common_affix(const char *suffixes, const char *prefixes)
+{
+ const char *s, *p;
+ if (suffixes && prefixes)
+ for (p = prefixes; *p; p++)
+ for (s = suffixes; *s; s++)
+ if (*p == *s)
+ return *p == '*' ? 's' : *p;
+ return '\0';
}
ssize_t
libnumtext_num2text_swedish__(char *outbuf, size_t outbuf_size, const char *num, size_t num_len, uint32_t flags)
{
- int first = 1;
- int hundred_thousands, thousands, orig_thousands, hundreds, ones;
+ int hundred_thousands, ten_thousands, one_thousands, hundreds, ones;
int32_t small_num;
- const char *great_1, *great_1_suffix, *great_last;
- const char *great_10, *great_10_prefix, *gsuffix;
- const char *great_100, *great_100_prefix, *gprefix;
+ const char *great_1, *great_1_suffixes, *great_last;
+ const char *great_10, *great_10_prefixes;
+ const char *great_100, *great_100_prefixes;
char affix[2] = {[1] = 0};
size_t great_order, small_order, great_order_suffix, i;
- const char *append_for_ordinal = NULL;
size_t trailing_zeroes;
struct state state;
@@ -245,16 +313,15 @@ libnumtext_num2text_swedish__(char *outbuf, size_t outbuf_size, const char *num,
state.len = 0;
state.flags = flags;
state.double_char = 0;
+ state.first = 1;
+ state.append_for_ordinal = NULL;
if (!isdigit(num[0])) {
- if (ORDINAL(flags) || DENOMINATOR(flags))
- append(&state, num[0] == '+' ? "Plus-" : "Min¦us-");
- else
- append(&state, num[0] == '+' ? "Plus " : "Min¦us ");
+ append(&state, signs[num[0] == '+'][TYPE_INDEX(flags)]);
do {
num++;
num_len--;
- } while ((*num & 0xC0) == 0x80);
+ } while (IS_UTF8_CHAR_CONTINUATION(*num));
}
while (num_len > 1 && num[0] == '0') {
@@ -264,18 +331,13 @@ libnumtext_num2text_swedish__(char *outbuf, size_t outbuf_size, const char *num,
if (num_len == 1) {
if (num[0] == '0') {
- if (ORDINAL(flags) || DENOMINATOR(flags))
- append(&state, digits[0].ordinal_other);
- else
- append(&state, digits[0].cardinal_common);
- suffix(&state);
- goto out;
+ append_final_digit(&state, 0);
+ goto out_and_suffix;
} else if (num[0] <= '2' && DENOMINATOR(flags)) {
if (ORDINAL(flags))
- i = 4;
+ append(&state, wholes_and_halves[num[0] - '1'].ordinal);
else
- i = (size_t)((flags / LIBNUMTEXT_N2T_SWEDISH_PLURAL_FORM) & 3);
- append(&state, wholes_and_halves[num[0] - '1'][i]);
+ append(&state, wholes_and_halves[num[0] - '1'].cardinal[FORM_INDEX(flags)]);
goto out;
}
}
@@ -294,102 +356,65 @@ libnumtext_num2text_swedish__(char *outbuf, size_t outbuf_size, const char *num,
small_order = num_len % 6;
great_order_suffix = 0;
- hundred_thousands = thousands = hundreds = ones = 0;
- small_num = 0;
+ hundred_thousands = ten_thousands = one_thousands = hundreds = ones = 0;
if (great_order && small_order >= 3) {
small_order -= 3;
great_order_suffix = 1;
}
- orig_thousands = 0;
+ small_num = 0;
switch (small_order) {
case 5:
hundred_thousands = *num++ - '0';
small_num = (int32_t)hundred_thousands;
num_len--;
- if (hundred_thousands) {
- if (first && hundred_thousands == 1 && IMPLICIT_ONE(flags)) {
- append(&state, "Hun¦dra");
- } else {
- append(&state, digits[hundred_thousands].cardinal_common);
- append(&state, "||hun¦dra");
- }
- append_for_ordinal = "¦de";
- first = 0;
- }
+ append_hundreds(&state, hundred_thousands);
FALL_THROUGH
/* fall through */
case 4:
- thousands = *num++ - '0';
- orig_thousands = thousands;
+ ten_thousands = *num++ - '0';
+ small_num *= 10;
+ small_num += (int32_t)ten_thousands;
num_len--;
- if (tens[thousands].cardinal) {
- append(&state, tens[thousands].cardinal);
- thousands = 0;
- first = 0;
- } else {
- thousands *= 10;
- }
FALL_THROUGH
/* fall through */
case 3:
+ one_thousands = *num++ - '0';
small_num *= 10;
- small_num += (int32_t)(*num - '0');
- thousands += *num++ - '0';
+ small_num += (int32_t)one_thousands;
num_len--;
- if (thousands) {
- if (first && thousands == 1 && IMPLICIT_ONE(flags)) {
- append(&state, "Tu¦sen");
- } else {
- append(&state, digits[thousands].cardinal_common);
- append(&state, "||tu¦sen");
- }
- append_for_ordinal = "¦de";
- first = 0;
- } else if (hundred_thousands || orig_thousands) {
- append(&state, "||tu¦sen");
- append_for_ordinal = "¦de";
- first = 0;
- }
+ if (small_num)
+ append_thousands(&state, ten_thousands, one_thousands);
FALL_THROUGH
/* fall through */
case 2:
- small_num *= 10;
- small_num += (int32_t)(*num - '0');
hundreds = *num++ - '0';
+ small_num *= 10;
+ small_num += (int32_t)hundreds;
num_len--;
- if (hundreds) {
- if (first && hundreds == 1 && IMPLICIT_ONE(flags)) {
- append(&state, "Hun¦dra");
- } else {
- append(&state, digits[hundreds].cardinal_common);
- append(&state, "||hun¦dra");
- }
- append_for_ordinal = "¦de";
- first = 0;
- }
+ append_hundreds(&state, hundreds);
FALL_THROUGH
/* fall through */
case 1:
- small_num *= 10;
- small_num += (int32_t)(*num - '0');
ones = *num++ - '0';
+ small_num *= 10;
+ small_num += (int32_t)ones;
num_len--;
if (tens[ones].cardinal) {
if (!great_order && (DENOMINATOR(flags) || ORDINAL(flags)) && *num == '0') {
append(&state, tens[ones].ordinal);
- append_for_ordinal = "¦de";
+ state.append_for_ordinal = "¦de";
} else {
append(&state, tens[ones].cardinal);
- append_for_ordinal = NULL;
+ state.append_for_ordinal = NULL;
}
ones = 0;
- first = 0;
+ state.first = 0;
} else {
ones *= 10;
}
@@ -401,83 +426,46 @@ libnumtext_num2text_swedish__(char *outbuf, size_t outbuf_size, const char *num,
small_num += (int32_t)(*num - '0');
ones += *num++ - '0';
if (ones) {
- append_for_ordinal = NULL;
- if (!great_order && (DENOMINATOR(flags) || ORDINAL(flags))) {
- if (MASCULINE_GENDER(flags) && digits[ones].ordinal_masculine)
- append(&state, digits[ones].ordinal_masculine);
- else
- append(&state, digits[ones].ordinal_other);
- append_for_ordinal = digits[ones].ordinal_suffix;
- } else if (!digits[ones].cardinal_other) {
- append(&state, digits[ones].cardinal_common);
- } else if (great_order) {
+ state.append_for_ordinal = NULL;
+ if (!great_order)
+ append_final_digit(&state, ones);
+ else if (digits[ones].cardinal_other)
append(&state, digits[ones].cardinal_other);
- } else if (COMMON_GENDER(flags)) {
+ else
append(&state, digits[ones].cardinal_common);
- } else {
- append(&state, digits[ones].cardinal_other);
- }
- first = 0;
+ state.first = 0;
}
break;
}
if (great_order && small_num) {
if (great_order < 10) {
- append(&state, greats[great_order][0]);
+ append(&state, greats[great_order].single_digit);
} else if (great_order > 999) {
errno = EDOM;
return -1;
} else {
- great_1 = greats[(great_order / 1) % 10][1];
- great_1_suffix = greats[(great_order / 1) % 10][2];
- great_10_prefix = greats[(great_order / 10) % 10][3];
- great_10 = greats[(great_order / 10) % 10][4];
- great_100_prefix = greats[(great_order / 100) % 10][5];
- great_100 = greats[(great_order / 100) % 10][6];
+ great_1 = greats[(great_order / 1) % 10].ones;
+ great_1_suffixes = greats[(great_order / 1) % 10].ones_suffixes;
+ great_10_prefixes = greats[(great_order / 10) % 10].tens_prefixes;
+ great_10 = greats[(great_order / 10) % 10].tens;
+ great_100_prefixes = greats[(great_order / 100) % 10].hundreds_prefixes;
+ great_100 = greats[(great_order / 100) % 10].hundreds;
great_last = NULL;
if (great_1) {
append(&state, great_1);
great_last = great_1;
}
if (great_10) {
- if (great_1_suffix && great_10_prefix) {
- gsuffix = NULL; /* silence clang */
- for (gprefix = great_10_prefix; *gprefix; gprefix++) {
- for (gsuffix = great_1_suffix; *gsuffix; gsuffix++)
- if (*gprefix == *gsuffix)
- break;
- if (*gsuffix)
- break;
- }
- if (*gprefix && *gprefix == *gsuffix) {
- affix[0] = *gprefix;
- if (affix[0] == '*')
- affix[0] = 's';
- append(&state, affix);
- }
- }
+ if ((affix[0] = get_common_affix(great_1_suffixes, great_10_prefixes)))
+ append(&state, affix);
append(&state, great_10);
great_last = great_10;
- great_1_suffix = NULL;
+ great_1_suffixes = NULL;
}
if (great_100) {
- if (great_1_suffix && great_100_prefix) {
- gsuffix = NULL; /* silence clang */
- for (gprefix = great_100_prefix; *gprefix; gprefix++) {
- for (gsuffix = great_1_suffix; *gsuffix; gsuffix++)
- if (*gprefix == *gsuffix)
- break;
- if (*gsuffix)
- break;
- }
- if (*gprefix && *gprefix == *gsuffix) {
- affix[0] = *gprefix;
- if (affix[0] == '*')
- affix[0] = 's';
- append(&state, affix);
- }
- }
+ if ((affix[0] = get_common_affix(great_1_suffixes, great_100_prefixes)))
+ append(&state, affix);
append(&state, great_100);
great_last = great_100;
}
@@ -487,16 +475,21 @@ libnumtext_num2text_swedish__(char *outbuf, size_t outbuf_size, const char *num,
state.len -= 1;
}
append(&state, great_suffixes[great_order_suffix]);
- append_for_ordinal = great_order_suffix == 0 ? "¦te" : "<¦e";
+ state.append_for_ordinal = great_order_suffix == 0 ? "¦te" : "<¦e";
if (small_num != 1)
if (num_len > trailing_zeroes || !(ORDINAL(flags) || DENOMINATOR(flags)))
append(&state, "¦er");
}
}
- if (ORDINAL(flags) && !DENOMINATOR(flags) && append_for_ordinal)
- append(&state, append_for_ordinal);
- suffix(&state);
+ if (ORDINAL(flags) && !DENOMINATOR(flags) && state.append_for_ordinal)
+ append(&state, state.append_for_ordinal);
+
+out_and_suffix:
+ if (DENOMINATOR(state.flags)) {
+ append(&state, ORDINAL(state.flags) ? denominator_suffixes.ordinal
+ : denominator_suffixes.cardinal[FORM_INDEX(state.flags)]);
+ }
out:
if (state.len < outbuf_size)
@@ -524,7 +517,7 @@ out:
i = 0;
if (SENTENCE_CASE(flags)) {
i = 1;
- while ((outbuf[i] & 0xC0) == 0x80)
+ while (IS_UTF8_CHAR_CONTINUATION(outbuf[i]))
i += 1;
goto lower_case;