diff options
author | John MacFarlane <jgm@berkeley.edu> | 2014-09-29 22:59:46 -0700 |
---|---|---|
committer | John MacFarlane <jgm@berkeley.edu> | 2014-09-29 22:59:46 -0700 |
commit | d9f7581443bd786e7d17532f6678efd2ee77c26f (patch) | |
tree | 831a33dc8990e3b85ea068b361750702f8e3f40a /src/html/houdini_href_e.c | |
parent | de1e28217f0da80b928bca0ca09541c0401314ee (diff) | |
parent | c006aececef112f61dd44cad43f0596221f29700 (diff) |
Merge branch 'master' into newemphasis
Conflicts:
Makefile
js/stmd.js
Diffstat (limited to 'src/html/houdini_href_e.c')
-rw-r--r-- | src/html/houdini_href_e.c | 107 |
1 files changed, 107 insertions, 0 deletions
diff --git a/src/html/houdini_href_e.c b/src/html/houdini_href_e.c new file mode 100644 index 0000000..12456ce --- /dev/null +++ b/src/html/houdini_href_e.c @@ -0,0 +1,107 @@ +#include <assert.h> +#include <stdio.h> +#include <string.h> + +#include "html/houdini.h" + +/* + * The following characters will not be escaped: + * + * -_.+!*'(),%#@?=;:/,+&$ alphanum + * + * Note that this character set is the addition of: + * + * - The characters which are safe to be in an URL + * - The characters which are *not* safe to be in + * an URL because they are RESERVED characters. + * + * We asume (lazily) that any RESERVED char that + * appears inside an URL is actually meant to + * have its native function (i.e. as an URL + * component/separator) and hence needs no escaping. + * + * There are two exceptions: the chacters & (amp) + * and ' (single quote) do not appear in the table. + * They are meant to appear in the URL as components, + * yet they require special HTML-entity escaping + * to generate valid HTML markup. + * + * All other characters will be escaped to %XX. + * + */ +static const char HREF_SAFE[] = { + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 1, 0, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 1, + 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, +}; + +int +houdini_escape_href(strbuf *ob, const uint8_t *src, size_t size) +{ + static const uint8_t hex_chars[] = "0123456789ABCDEF"; + size_t i = 0, org; + uint8_t hex_str[3]; + + hex_str[0] = '%'; + + while (i < size) { + org = i; + while (i < size && HREF_SAFE[src[i]] != 0) + i++; + + if (likely(i > org)) + strbuf_put(ob, src + org, i - org); + + /* escaping */ + if (i >= size) + break; + + switch (src[i]) { + /* amp appears all the time in URLs, but needs + * HTML-entity escaping to be inside an href */ + case '&': + strbuf_puts(ob, "&"); + break; + + /* the single quote is a valid URL character + * according to the standard; it needs HTML + * entity escaping too */ + case '\'': + strbuf_puts(ob, "'"); + break; + + /* the space can be escaped to %20 or a plus + * sign. we're going with the generic escape + * for now. the plus thing is more commonly seen + * when building GET strings */ +#if 0 + case ' ': + strbuf_putc(ob, '+'); + break; +#endif + + /* every other character goes with a %XX escaping */ + default: + hex_str[1] = hex_chars[(src[i] >> 4) & 0xF]; + hex_str[2] = hex_chars[src[i] & 0xF]; + strbuf_put(ob, hex_str, 3); + } + + i++; + } + + return 1; +} |