Gitweb links: ...log http://git.netsurf-browser.org/netsurf.git/shortlog/5b6344182634b08e15866... ...commit http://git.netsurf-browser.org/netsurf.git/commit/5b6344182634b08e1586682... ...tree http://git.netsurf-browser.org/netsurf.git/tree/5b6344182634b08e1586682c1... The branch, vince/fix-gcc12-warnings has been updated via 5b6344182634b08e1586682c1c76a0a88d49a24c (commit) via 109527d02fd3aeef09bf82a450ecb0e3ed86f8ea (commit) from 3e14db3dc84dbf38898a6efcb5b180b47b4308f5 (commit) Those revisions listed above that are new to this repository have not appeared on any other notification email; so we list those revisions in full, below. - Log ----------------------------------------------------------------- commitdiff http://git.netsurf-browser.org/netsurf.git/commit/?id=5b6344182634b08e158... commit 5b6344182634b08e1586682c1c76a0a88d49a24c Author: Vincent Sanders <vince(a)kyllikki.org&gt; Commit: Vincent Sanders <vince(a)kyllikki.org&gt; remove the ambiguity around the reallocation in utf8_to_html() diff --git a/utils/utf8.c b/utils/utf8.c index 7aa7d93..84918cc 100644 --- a/utils/utf8.c +++ b/utils/utf8.c @@ -365,11 +365,11 @@ utf8_convert_html_chunk(iconv_t cd, /* exported interface documented in utils/utf8.h */ nserror -utf8_to_html(const char *string, const char *encname, size_t len, char **result) +utf8_to_html(const char *string, const char *encname, size_t len, char **result_out) { iconv_t cd; const char *in; - char *out, *origout; + char *out, *origout, *result; size_t off, prev_off, inlen, outlen, origoutlen, esclen; nserror ret; char *pescape, escape[11]; @@ -452,11 +452,12 @@ utf8_to_html(const char *string, const char *encname, size_t len, char **result) outlen -= 4; /* Shrink-wrap */ - *result = realloc(origout, origoutlen - outlen); - if (*result == NULL) { + result = realloc(origout, origoutlen - outlen); + if (result == NULL) { free(origout); return NSERROR_NOMEM; } + *result_out = result; return NSERROR_OK; } commitdiff http://git.netsurf-browser.org/netsurf.git/commit/?id=109527d02fd3aeef09b... commit 109527d02fd3aeef09bf82a450ecb0e3ed86f8ea Author: Vincent Sanders <vince(a)kyllikki.org&gt; Commit: Vincent Sanders <vince(a)kyllikki.org&gt; consolodate duplicated conversion descriptor cache code diff --git a/utils/utf8.c b/utils/utf8.c index 84cacd7..7aa7d93 100644 --- a/utils/utf8.c +++ b/utils/utf8.c @@ -44,7 +44,7 @@ uint32_t utf8_to_ucs4(const char *s_in, size_t l) parserutils_error perror; perror = parserutils_charset_utf8_to_ucs4((const uint8_t *) s_in, l, - &ucs4, &len); + &ucs4, &len); if (perror != PARSERUTILS_OK) ucs4 = 0xfffd; @@ -106,7 +106,7 @@ size_t utf8_char_byte_length(const char *s) parserutils_error perror; perror = parserutils_charset_utf8_char_byte_length((const uint8_t *) s, - &len); + &len); assert(perror == PARSERUTILS_OK); return len; @@ -131,7 +131,7 @@ size_t utf8_next(const char *s, size_t l, size_t o) parserutils_error perror; perror = parserutils_charset_utf8_next((const uint8_t *) s, l, o, - &next); + &next); assert(perror == PARSERUTILS_OK); return next; @@ -151,6 +151,47 @@ static inline void utf8_clear_cd_cache(void) last_cd.cd = 0; } +/** + * obtain a cached conversion descriptor + * + * either return the cached conversion descriptor or create one if required + */ +static nserror +get_cached_cd(const char *enc_from, const char *enc_to, iconv_t *cd_out) +{ + iconv_t cd; + /* we cache the last used conversion descriptor, + * so check if we're trying to use it here */ + if (strncasecmp(last_cd.from, enc_from, sizeof(last_cd.from)) == 0 && + strncasecmp(last_cd.to, enc_to, sizeof(last_cd.to)) == 0 && + last_cd.cd != 0) { + *cd_out = last_cd.cd; + return NSERROR_OK; + } + + /* no match, so create a new cd */ + cd = iconv_open(enc_to, enc_from); + if (cd == (iconv_t) -1) { + if (errno == EINVAL) { + return NSERROR_BAD_ENCODING; + } + /* default to no memory */ + return NSERROR_NOMEM; + } + + /* close the last cd - we don't care if this fails */ + if (last_cd.cd) { + iconv_close(last_cd.cd); + } + + /* and safely copy the to/from/cd data into last_cd */ + snprintf(last_cd.from, sizeof(last_cd.from), enc_from); + snprintf(last_cd.to, sizeof(last_cd.to), "%s", enc_to); + *cd_out = last_cd.cd = cd; + + return NSERROR_OK; +} + /* exported interface documented in utils/utf8.h */ nserror utf8_finalise(void) { @@ -187,6 +228,7 @@ utf8_convert(const char *string, iconv_t cd; char *temp, *out, *in, *result; size_t result_len; + nserror res; assert(string && from && to && result_out); @@ -215,29 +257,9 @@ utf8_convert(const char *string, in = (char *)string; - /* we cache the last used conversion descriptor, - * so check if we're trying to use it here */ - if (strncasecmp(last_cd.from, from, sizeof(last_cd.from)) == 0 && - strncasecmp(last_cd.to, to, sizeof(last_cd.to)) == 0) { - cd = last_cd.cd; - } else { - /* no match, so create a new cd */ - cd = iconv_open(to, from); - if (cd == (iconv_t)-1) { - if (errno == EINVAL) - return NSERROR_BAD_ENCODING; - /* default to no memory */ - return NSERROR_NOMEM; - } - - /* close the last cd - we don't care if this fails */ - if (last_cd.cd) - iconv_close(last_cd.cd); - - /* and copy the to/from/cd data into last_cd */ - snprintf(last_cd.from, sizeof(last_cd.from), "%s", from); - snprintf(last_cd.to, sizeof(last_cd.to), "%s", to); - last_cd.cd = cd; + res = get_cached_cd(from, to, &cd); + if (res != NSERROR_OK) { + return res; } /* Worst case = ASCII -> UCS4, so allocate an output buffer @@ -289,14 +311,14 @@ utf8_convert(const char *string, /* exported interface documented in utils/utf8.h */ nserror utf8_to_enc(const char *string, const char *encname, - size_t len, char **result) + size_t len, char **result) { return utf8_convert(string, len, "UTF-8", encname, result, NULL); } /* exported interface documented in utils/utf8.h */ nserror utf8_from_enc(const char *string, const char *encname, - size_t len, char **result, size_t *result_len) + size_t len, char **result, size_t *result_len) { return utf8_convert(string, len, encname, "UTF-8", result, result_len); } @@ -327,7 +349,7 @@ utf8_convert_html_chunk(iconv_t cd, esclen = snprintf(escape, sizeof(escape), "&#x%06x;", ucs4); pescape = escape; ret = iconv(cd, (void *) &pescape, &esclen, - (void *) out, outlen); + (void *) out, outlen); if (ret == (size_t) -1) return NSERROR_NOMEM; @@ -339,6 +361,8 @@ utf8_convert_html_chunk(iconv_t cd, return NSERROR_OK; } + + /* exported interface documented in utils/utf8.h */ nserror utf8_to_html(const char *string, const char *encname, size_t len, char **result) @@ -349,35 +373,14 @@ utf8_to_html(const char *string, const char *encname, size_t len, char **result) size_t off, prev_off, inlen, outlen, origoutlen, esclen; nserror ret; char *pescape, escape[11]; + nserror res; if (len == 0) len = strlen(string); - /* we cache the last used conversion descriptor, - * so check if we're trying to use it here */ - if (strncasecmp(last_cd.from, "UTF-8", sizeof(last_cd.from)) == 0 && - strncasecmp(last_cd.to, encname, - sizeof(last_cd.to)) == 0 && - last_cd.cd != 0) { - cd = last_cd.cd; - } else { - /* no match, so create a new cd */ - cd = iconv_open(encname, "UTF-8"); - if (cd == (iconv_t) -1) { - if (errno == EINVAL) - return NSERROR_BAD_ENCODING; - /* default to no memory */ - return NSERROR_NOMEM; - } - - /* close the last cd - we don't care if this fails */ - if (last_cd.cd) - iconv_close(last_cd.cd); - - /* and safely copy the to/from/cd data into last_cd */ - snprintf(last_cd.from, sizeof(last_cd.from), "UTF-8"); - snprintf(last_cd.to, sizeof(last_cd.to), "%s", encname); - last_cd.cd = cd; + res = get_cached_cd("UTF-8", encname, &cd); + if (res != NSERROR_OK) { + return res; } /* Worst case is ASCII -> UCS4, with all characters escaped: @@ -397,13 +400,13 @@ utf8_to_html(const char *string, const char *encname, size_t len, char **result) while (off < len) { /* Must escape '&', '<', and '>' */ if (string[off] == '&' || string[off] == '<' || - string[off] == '>') { + string[off] == '>') { if (off - prev_off > 0) { /* Emit chunk */ in = string + prev_off; inlen = off - prev_off; ret = utf8_convert_html_chunk(cd, in, inlen, - &out, &outlen); + &out, &outlen); if (ret != NSERROR_OK) { free(origout); iconv_close(cd); @@ -414,10 +417,10 @@ utf8_to_html(const char *string, const char *encname, size_t len, char **result) /* Emit mandatory escape */ esclen = snprintf(escape, sizeof(escape), - "&#x%06x;", string[off]); + "&#x%06x;", string[off]); pescape = escape; ret = utf8_convert_html_chunk(cd, pescape, esclen, - &out, &outlen); + &out, &outlen); if (ret != NSERROR_OK) { free(origout); iconv_close(cd); ----------------------------------------------------------------------- Summary of changes: utils/utf8.c | 128 ++++++++++++++++++++++++++++++---------------------------- 1 file changed, 66 insertions(+), 62 deletions(-) diff --git a/utils/utf8.c b/utils/utf8.c index 84cacd7..84918cc 100644 --- a/utils/utf8.c +++ b/utils/utf8.c @@ -44,7 +44,7 @@ uint32_t utf8_to_ucs4(const char *s_in, size_t l) parserutils_error perror; perror = parserutils_charset_utf8_to_ucs4((const uint8_t *) s_in, l, - &ucs4, &len); + &ucs4, &len); if (perror != PARSERUTILS_OK) ucs4 = 0xfffd; @@ -106,7 +106,7 @@ size_t utf8_char_byte_length(const char *s) parserutils_error perror; perror = parserutils_charset_utf8_char_byte_length((const uint8_t *) s, - &len); + &len); assert(perror == PARSERUTILS_OK); return len; @@ -131,7 +131,7 @@ size_t utf8_next(const char *s, size_t l, size_t o) parserutils_error perror; perror = parserutils_charset_utf8_next((const uint8_t *) s, l, o, - &next); + &next); assert(perror == PARSERUTILS_OK); return next; @@ -151,6 +151,47 @@ static inline void utf8_clear_cd_cache(void) last_cd.cd = 0; } +/** + * obtain a cached conversion descriptor + * + * either return the cached conversion descriptor or create one if required + */ +static nserror +get_cached_cd(const char *enc_from, const char *enc_to, iconv_t *cd_out) +{ + iconv_t cd; + /* we cache the last used conversion descriptor, + * so check if we're trying to use it here */ + if (strncasecmp(last_cd.from, enc_from, sizeof(last_cd.from)) == 0 && + strncasecmp(last_cd.to, enc_to, sizeof(last_cd.to)) == 0 && + last_cd.cd != 0) { + *cd_out = last_cd.cd; + return NSERROR_OK; + } + + /* no match, so create a new cd */ + cd = iconv_open(enc_to, enc_from); + if (cd == (iconv_t) -1) { + if (errno == EINVAL) { + return NSERROR_BAD_ENCODING; + } + /* default to no memory */ + return NSERROR_NOMEM; + } + + /* close the last cd - we don't care if this fails */ + if (last_cd.cd) { + iconv_close(last_cd.cd); + } + + /* and safely copy the to/from/cd data into last_cd */ + snprintf(last_cd.from, sizeof(last_cd.from), enc_from); + snprintf(last_cd.to, sizeof(last_cd.to), "%s", enc_to); + *cd_out = last_cd.cd = cd; + + return NSERROR_OK; +} + /* exported interface documented in utils/utf8.h */ nserror utf8_finalise(void) { @@ -187,6 +228,7 @@ utf8_convert(const char *string, iconv_t cd; char *temp, *out, *in, *result; size_t result_len; + nserror res; assert(string && from && to && result_out); @@ -215,29 +257,9 @@ utf8_convert(const char *string, in = (char *)string; - /* we cache the last used conversion descriptor, - * so check if we're trying to use it here */ - if (strncasecmp(last_cd.from, from, sizeof(last_cd.from)) == 0 && - strncasecmp(last_cd.to, to, sizeof(last_cd.to)) == 0) { - cd = last_cd.cd; - } else { - /* no match, so create a new cd */ - cd = iconv_open(to, from); - if (cd == (iconv_t)-1) { - if (errno == EINVAL) - return NSERROR_BAD_ENCODING; - /* default to no memory */ - return NSERROR_NOMEM; - } - - /* close the last cd - we don't care if this fails */ - if (last_cd.cd) - iconv_close(last_cd.cd); - - /* and copy the to/from/cd data into last_cd */ - snprintf(last_cd.from, sizeof(last_cd.from), "%s", from); - snprintf(last_cd.to, sizeof(last_cd.to), "%s", to); - last_cd.cd = cd; + res = get_cached_cd(from, to, &cd); + if (res != NSERROR_OK) { + return res; } /* Worst case = ASCII -> UCS4, so allocate an output buffer @@ -289,14 +311,14 @@ utf8_convert(const char *string, /* exported interface documented in utils/utf8.h */ nserror utf8_to_enc(const char *string, const char *encname, - size_t len, char **result) + size_t len, char **result) { return utf8_convert(string, len, "UTF-8", encname, result, NULL); } /* exported interface documented in utils/utf8.h */ nserror utf8_from_enc(const char *string, const char *encname, - size_t len, char **result, size_t *result_len) + size_t len, char **result, size_t *result_len) { return utf8_convert(string, len, encname, "UTF-8", result, result_len); } @@ -327,7 +349,7 @@ utf8_convert_html_chunk(iconv_t cd, esclen = snprintf(escape, sizeof(escape), "&#x%06x;", ucs4); pescape = escape; ret = iconv(cd, (void *) &pescape, &esclen, - (void *) out, outlen); + (void *) out, outlen); if (ret == (size_t) -1) return NSERROR_NOMEM; @@ -339,45 +361,26 @@ utf8_convert_html_chunk(iconv_t cd, return NSERROR_OK; } + + /* exported interface documented in utils/utf8.h */ nserror -utf8_to_html(const char *string, const char *encname, size_t len, char **result) +utf8_to_html(const char *string, const char *encname, size_t len, char **result_out) { iconv_t cd; const char *in; - char *out, *origout; + char *out, *origout, *result; size_t off, prev_off, inlen, outlen, origoutlen, esclen; nserror ret; char *pescape, escape[11]; + nserror res; if (len == 0) len = strlen(string); - /* we cache the last used conversion descriptor, - * so check if we're trying to use it here */ - if (strncasecmp(last_cd.from, "UTF-8", sizeof(last_cd.from)) == 0 && - strncasecmp(last_cd.to, encname, - sizeof(last_cd.to)) == 0 && - last_cd.cd != 0) { - cd = last_cd.cd; - } else { - /* no match, so create a new cd */ - cd = iconv_open(encname, "UTF-8"); - if (cd == (iconv_t) -1) { - if (errno == EINVAL) - return NSERROR_BAD_ENCODING; - /* default to no memory */ - return NSERROR_NOMEM; - } - - /* close the last cd - we don't care if this fails */ - if (last_cd.cd) - iconv_close(last_cd.cd); - - /* and safely copy the to/from/cd data into last_cd */ - snprintf(last_cd.from, sizeof(last_cd.from), "UTF-8"); - snprintf(last_cd.to, sizeof(last_cd.to), "%s", encname); - last_cd.cd = cd; + res = get_cached_cd("UTF-8", encname, &cd); + if (res != NSERROR_OK) { + return res; } /* Worst case is ASCII -> UCS4, with all characters escaped: @@ -397,13 +400,13 @@ utf8_to_html(const char *string, const char *encname, size_t len, char **result) while (off < len) { /* Must escape '&', '<', and '>' */ if (string[off] == '&' || string[off] == '<' || - string[off] == '>') { + string[off] == '>') { if (off - prev_off > 0) { /* Emit chunk */ in = string + prev_off; inlen = off - prev_off; ret = utf8_convert_html_chunk(cd, in, inlen, - &out, &outlen); + &out, &outlen); if (ret != NSERROR_OK) { free(origout); iconv_close(cd); @@ -414,10 +417,10 @@ utf8_to_html(const char *string, const char *encname, size_t len, char **result) /* Emit mandatory escape */ esclen = snprintf(escape, sizeof(escape), - "&#x%06x;", string[off]); + "&#x%06x;", string[off]); pescape = escape; ret = utf8_convert_html_chunk(cd, pescape, esclen, - &out, &outlen); + &out, &outlen); if (ret != NSERROR_OK) { free(origout); iconv_close(cd); @@ -449,11 +452,12 @@ utf8_to_html(const char *string, const char *encname, size_t len, char **result) outlen -= 4; /* Shrink-wrap */ - *result = realloc(origout, origoutlen - outlen); - if (*result == NULL) { + result = realloc(origout, origoutlen - outlen); + if (result == NULL) { free(origout); return NSERROR_NOMEM; } + *result_out = result; return NSERROR_OK; } -- NetSurf Browser