Gitweb links:
...log
http://git.netsurf-browser.org/netsurf.git/shortlog/50044f51623fecbb51dbb5f7ef496d45c09b2bca
...commit
http://git.netsurf-browser.org/netsurf.git/commit/50044f51623fecbb51dbb5f7ef496d45c09b2bca
...tree
http://git.netsurf-browser.org/netsurf.git/tree/50044f51623fecbb51dbb5f7ef496d45c09b2bca
The branch, jmb/uniopt has been created
at 50044f51623fecbb51dbb5f7ef496d45c09b2bca (commit)
- Log -----------------------------------------------------------------
commitdiff
http://git.netsurf-browser.org/netsurf.git/commit/?id=50044f51623fecbb51dbb5f7ef496d45c09b2bca
commit 50044f51623fecbb51dbb5f7ef496d45c09b2bca
Author: John-Mark Bell <[email protected]>
Commit: John-Mark Bell <[email protected]>
RISC OS: optimisations for UTF-8<->local encoding
diff --git a/frontends/riscos/ucstables.c b/frontends/riscos/ucstables.c
index 3e31c99..a94e434 100644
--- a/frontends/riscos/ucstables.c
+++ b/frontends/riscos/ucstables.c
@@ -398,49 +398,47 @@ static const char *localencodings[] = {
"ISO-8859-10//TRANSLIT",
"ISO-8859-13//TRANSLIT",
"ISO-8859-14//TRANSLIT",
- "ISO-8859-16//TRANSLIT",
-#define CONT_ENC_END 116 /* RISC OS alphabet numbers lie in a
- * contiguous range [100,CONT_ENC_END]
- * _except_ for Cyrillic2, which doesn't.
- */
- "CP866//TRANSLIT" /* Cyrillic2 - 120 */
+ "ISO-8859-16//TRANSLIT", /* Latin10 - 116 */
+ NULL, /* UTF-16, if you believe HdrSrc (Unused) */
+ NULL, /* Unused */
+ NULL, /* Unused */
+ "CP866//TRANSLIT" /* Cyrillic2 - 120 */
};
-static const struct special {
- char local; /**< Local 8bit representation */
- char len; /**< Length (in bytes) of UTF-8 character */
- const char *utf; /**< UTF-8 representation */
-} special_chars[] = {
- { 0x80, 3, "\xE2\x82\xAC" }, /* EURO SIGN */
- { 0x81, 2, "\xC5\xB4" }, /* LATIN CAPITAL LETTER W WITH
CIRCUMFLEX */
- { 0x82, 2, "\xC5\xB5" }, /* LATIN SMALL LETTER W WITH CIRCUMFLEX
*/
- { 0x84, 3, "\xE2\x9C\x98" }, /* HEAVY BALLOT X */
- { 0x85, 2, "\xC5\xB6" }, /* LATIN CAPITAL LETTER Y WITH
CIRCUMFLEX */
- { 0x86, 2, "\xC5\xB7" }, /* LATIN SMALL LETTER Y WITH CIRCUMFLEX
*/
- { 0x88, 3, "\xE2\x87\x90" }, /* LEFTWARDS DOUBLE ARROW */
- { 0x89, 3, "\xE2\x87\x92" }, /* RIGHTWARDS DOUBLE ARROW */
- { 0x8a, 3, "\xE2\x87\x93" }, /* DOWNWARDS DOUBLE ARROW */
- { 0x8b, 3, "\xE2\x87\x91" }, /* UPWARDS DOUBLE ARROW */
- { 0x8c, 3, "\xE2\x80\xA6" }, /* HORIZONTAL ELLIPSIS */
- { 0x8d, 3, "\xE2\x84\xA2" }, /* TRADE MARK SIGN */
- { 0x8e, 3, "\xE2\x80\xB0" }, /* PER MILLE SIGN */
- { 0x8f, 3, "\xE2\x80\xA2" }, /* BULLET */
- { 0x90, 3, "\xE2\x80\x98" }, /* LEFT SINGLE QUOTATION MARK */
- { 0x91, 3, "\xE2\x80\x99" }, /* RIGHT SINGLE QUOTATION MARK */
- { 0x92, 3, "\xE2\x80\xB9" }, /* SINGLE LEFT-POINTING ANGLE QUOTATION
MARK */
- { 0x93, 3, "\xE2\x80\xBA" }, /* SINGLE RIGHT-POINTING ANGLE
QUOTATION MARK */
- { 0x94, 3, "\xE2\x80\x9C" }, /* LEFT DOUBLE QUOTATION MARK */
- { 0x95, 3, "\xE2\x80\x9D" }, /* RIGHT DOUBLE QUOTATION MARK */
- { 0x96, 3, "\xE2\x80\x9E" }, /* DOUBLE LOW-9 QUOTATION MARK */
- { 0x97, 3, "\xE2\x80\x93" }, /* EN DASH */
- { 0x98, 3, "\xE2\x80\x94" }, /* EM DASH */
- { 0x99, 3, "\xE2\x88\x92" }, /* MINUS SIGN */
- { 0x9a, 2, "\xC5\x92" }, /* LATIN CAPITAL LIGATURE OE */
- { 0x9b, 2, "\xC5\x93" }, /* LATIN SMALL LIGATURE OE */
- { 0x9c, 3, "\xE2\x80\xA0" }, /* DAGGER */
- { 0x9d, 3, "\xE2\x80\xA1" }, /* DOUBLE DAGGER */
- { 0x9e, 3, "\xEF\xAC\x81" }, /* LATIN SMALL LIGATURE FI */
- { 0x9f, 3, "\xEF\xAC\x82" } /* LATIN SMALL LIGATURE FL */
+/* These are the Acorn Latin1 C1 block between [0x80,0x9f] */
+static const char *special_chars[] = {
+ "\xE2\x82\xAC", /* EURO SIGN */
+ "\xC5\xB4", /* LATIN CAPITAL LETTER W WITH CIRCUMFLEX */
+ "\xC5\xB5", /* LATIN SMALL LETTER W WITH CIRCUMFLEX */
+ NULL, /* unused */
+ "\xE2\x9C\x98", /* HEAVY BALLOT X */
+ "\xC5\xB6", /* LATIN CAPITAL LETTER Y WITH CIRCUMFLEX */
+ "\xC5\xB7", /* LATIN SMALL LETTER Y WITH CIRCUMFLEX */
+ NULL, /* unused */
+ "\xE2\x87\x90", /* LEFTWARDS DOUBLE ARROW */
+ "\xE2\x87\x92", /* RIGHTWARDS DOUBLE ARROW */
+ "\xE2\x87\x93", /* DOWNWARDS DOUBLE ARROW */
+ "\xE2\x87\x91", /* UPWARDS DOUBLE ARROW */
+ "\xE2\x80\xA6", /* HORIZONTAL ELLIPSIS */
+ "\xE2\x84\xA2", /* TRADE MARK SIGN */
+ "\xE2\x80\xB0", /* PER MILLE SIGN */
+ "\xE2\x80\xA2", /* BULLET */
+ "\xE2\x80\x98", /* LEFT SINGLE QUOTATION MARK */
+ "\xE2\x80\x99", /* RIGHT SINGLE QUOTATION MARK */
+ "\xE2\x80\xB9", /* SINGLE LEFT-POINTING ANGLE QUOTATION MARK */
+ "\xE2\x80\xBA", /* SINGLE RIGHT-POINTING ANGLE QUOTATION MARK */
+ "\xE2\x80\x9C", /* LEFT DOUBLE QUOTATION MARK */
+ "\xE2\x80\x9D", /* RIGHT DOUBLE QUOTATION MARK */
+ "\xE2\x80\x9E", /* DOUBLE LOW-9 QUOTATION MARK */
+ "\xE2\x80\x93", /* EN DASH */
+ "\xE2\x80\x94", /* EM DASH */
+ "\xE2\x88\x92", /* MINUS SIGN */
+ "\xC5\x92", /* LATIN CAPITAL LIGATURE OE */
+ "\xC5\x93", /* LATIN SMALL LIGATURE OE */
+ "\xE2\x80\xA0", /* DAGGER */
+ "\xE2\x80\xA1", /* DOUBLE DAGGER */
+ "\xEF\xAC\x81", /* LATIN SMALL LIGATURE FI */
+ "\xEF\xAC\x82" /* LATIN SMALL LIGATURE FL */
};
@@ -470,20 +468,22 @@ nserror utf8_to_local_encoding(const char *string, size_t
len, char **result)
/* read system alphabet */
error = xosbyte1(osbyte_ALPHABET_NUMBER, 127, 0, &alphabet);
- if (error)
+ /* Assume Latin1 for anything we know nothing about */
+ if (error || alphabet < territory_ALPHABET_BFONT ||
+ alphabet > territory_ALPHABET_CYRILLIC2)
alphabet = territory_ALPHABET_LATIN1;
/* UTF-8 -> simply copy string */
- if (alphabet == 111 /* UTF-8 */) {
+ if (alphabet == territory_ALPHABET_UTF8) {
*result = strndup(string, len);
return NSERROR_OK;
}
/* get encoding name */
- enc = (alphabet <= CONT_ENC_END ? localencodings[alphabet - 100]
- : (alphabet == 120 ?
- localencodings[CONT_ENC_END - 100 + 1]
- : localencodings[0]));
+ enc = localencodings[alphabet - territory_ALPHABET_BFONT];
+ /* Assume Latin1 for any that are unused */
+ if (enc == NULL)
+ enc = localencodings[0];
/* create output buffer */
*(result) = malloc(len + 1);
@@ -498,13 +498,32 @@ nserror utf8_to_local_encoding(const char *string, size_t
len, char **result)
* characters and inserting appropriate output for characters
* that iconv can't handle. */
for (off = 0; off < len; off = utf8_next(string, len, off)) {
- if (string[off] != 0xE2 &&
- string[off] != 0xC5 && string[off] != 0xEF)
+ /* Specials only start with C5/E2/EF */
+ if (string[off] != 0xC5 &&
+ string[off] != 0xE2 && string[off] != 0xEF)
continue;
+ /* Ignore truncated input */
+ if (off + 2 + (string[off] == 0xC5 ? 0 : 1) >= len)
+ continue;
+
+ /* Search to see if this character is special */
for (i = 0; i != NOF_ELEMENTS(special_chars); i++) {
- if (strncmp(string + off, special_chars[i].utf,
- special_chars[i].len) != 0)
+ /* Skip unused special char */
+ if (special_chars[i] == NULL)
+ continue;
+
+ /* Skip 2-byte non-match */
+ if (string[off] == 0xC5 &&
+ (string[off] != special_chars[i][0] ||
+ string[off+1] != special_chars[i][1]))
+ continue;
+
+ /* Skip 3-byte non-match */
+ if (string[off] != 0xC5 &&
+ (string[off] != special_chars[i][0] ||
+ string[off+1] != special_chars[i][1] ||
+ string[off+2] != special_chars[i][2]))
continue;
/* 0 length has a special meaning to utf8_to_enc */
@@ -524,9 +543,13 @@ nserror utf8_to_local_encoding(const char *string, size_t
len, char **result)
free(temp);
}
- *cur_pos = special_chars[i].local;
+ /* Emit conversion for this special character */
+ *cur_pos = 0x80 + i;
*(++cur_pos) = '\0';
- prev_off = off + special_chars[i].len;
+ prev_off = off + 2 + (string[off] == 0xC5 ? 0 : 1);
+
+ /* Return to outer loop to process remaining input */
+ break;
}
}
@@ -561,7 +584,7 @@ nserror utf8_to_local_encoding(const char *string, size_t
len, char **result)
nserror utf8_from_local_encoding(const char *string, size_t len, char **result)
{
os_error *error;
- int alphabet, i, num_specials = 0, result_alloc;
+ int alphabet, num_specials = 0, result_alloc;
#define SPECIAL_CHUNK_SIZE 255
size_t off, prev_off, cur_off;
char *temp;
@@ -576,11 +599,13 @@ nserror utf8_from_local_encoding(const char *string,
size_t len, char **result)
/* read system alphabet */
error = xosbyte1(osbyte_ALPHABET_NUMBER, 127, 0, &alphabet);
- if (error)
+ /* Assume Latin1 for anything we know nothing about */
+ if (error || alphabet < territory_ALPHABET_BFONT ||
+ alphabet > territory_ALPHABET_CYRILLIC2)
alphabet = territory_ALPHABET_LATIN1;
/* UTF-8 -> simply copy string */
- if (alphabet == 111 /* UTF-8 */) {
+ if (alphabet == territory_ALPHABET_UTF8) {
temp = strndup(string, len);
if (!temp)
return NSERROR_NOMEM;
@@ -590,10 +615,10 @@ nserror utf8_from_local_encoding(const char *string,
size_t len, char **result)
}
/* get encoding name */
- enc = (alphabet <= CONT_ENC_END ? localencodings[alphabet - 100]
- : (alphabet == 120 ?
- localencodings[CONT_ENC_END - 100 + 1]
- : localencodings[0]));
+ enc = localencodings[alphabet - territory_ALPHABET_BFONT];
+ /* Assume Latin1 for any that are unused */
+ if (enc == NULL)
+ enc = localencodings[0];
/* create output buffer (oversized) */
result_alloc = (len * 4) + (3 * SPECIAL_CHUNK_SIZE) + 1;
@@ -610,52 +635,54 @@ nserror utf8_from_local_encoding(const char *string,
size_t len, char **result)
* characters and inserting appropriate output for characters
* that iconv can't handle. */
for (off = 0; off < len; off++) {
+ /* Skip non-special characters */
if (string[off] < 0x80 || string[off] > 0x9f)
continue;
- for (i = 0; i != NOF_ELEMENTS(special_chars); i++) {
- if (string[off] != special_chars[i].local)
- continue;
-
- /* 0 length has a special meaning to utf8_from_enc */
- if (off - prev_off > 0) {
- err = utf8_from_enc(string + prev_off, enc,
- off - prev_off, &temp,
NULL);
- if (err != NSERROR_OK) {
- assert(err != NSERROR_BAD_ENCODING);
- NSLOG(netsurf, INFO,
- "utf8_from_enc failed");
- free(*result);
- return NSERROR_NOMEM;
- }
-
- strcat((*result) + cur_off, temp);
-
- cur_off += strlen(temp);
-
- free(temp);
+ /* 0 length has a special meaning to utf8_from_enc */
+ if (off - prev_off > 0) {
+ err = utf8_from_enc(string + prev_off, enc,
+ off - prev_off, &temp, NULL);
+ if (err != NSERROR_OK) {
+ assert(err != NSERROR_BAD_ENCODING);
+ NSLOG(netsurf, INFO, "utf8_from_enc failed");
+ free(*result);
+ return NSERROR_NOMEM;
}
- strcat((*result) + cur_off, special_chars[i].utf);
+ strcat((*result) + cur_off, temp);
- cur_off += special_chars[i].len;
+ cur_off += strlen(temp);
- prev_off = off + 1;
+ free(temp);
+ }
- num_specials++;
- if (num_specials % SPECIAL_CHUNK_SIZE ==
- SPECIAL_CHUNK_SIZE - 1) {
- char *temp = realloc((*result),
- result_alloc +
- (3 * SPECIAL_CHUNK_SIZE));
- if (!temp) {
- free(*result);
- return NSERROR_NOMEM;
- }
+ /* Append UTF-8 encoded special character or U+FFFD if none */
+ if (special_chars[string[off]-0x80] != NULL) {
+ const char *special = special_chars[string[off]-0x80];
+ strcat((*result) + cur_off, special);
+ cur_off += 2 + (special[0] == 0xC5 ? 0 : 1);
+ } else {
+ strcat((*result) + cur_off, "\xef\xbf\xbd");
+ cur_off += 3;
+ }
- *result = temp;
- result_alloc += (3 * SPECIAL_CHUNK_SIZE);
+ prev_off = off + 1;
+
+ /* Resize output buffer if necessary */
+ num_specials++;
+ if (num_specials % SPECIAL_CHUNK_SIZE ==
+ SPECIAL_CHUNK_SIZE - 1) {
+ char *temp = realloc((*result),
+ result_alloc +
+ (3 * SPECIAL_CHUNK_SIZE));
+ if (!temp) {
+ free(*result);
+ return NSERROR_NOMEM;
}
+
+ *result = temp;
+ result_alloc += (3 * SPECIAL_CHUNK_SIZE);
}
}
-----------------------------------------------------------------------
--
NetSurf Browser
_______________________________________________
netsurf-commits mailing list -- [email protected]
To unsubscribe send an email to [email protected]
