| 1 | #ifndef URL_MATCH_H |
| 2 | #define URL_MATCH_H |
| 3 | |
| 4 | #include "string-list.h" |
| 5 | #include "config.h" |
| 6 | |
| 7 | struct url_info { |
| 8 | /* normalized url on success, must be freed, otherwise NULL */ |
| 9 | char *url; |
| 10 | /* if !url, a brief reason for the failure, otherwise NULL */ |
| 11 | const char *err; |
| 12 | |
| 13 | /* the rest of the fields are only set if url != NULL */ |
| 14 | |
| 15 | size_t url_len; /* total length of url (which is now normalized) */ |
| 16 | size_t scheme_len; /* length of scheme name (excluding final :) */ |
| 17 | size_t user_off; /* offset into url to start of user name (0 => none) */ |
| 18 | size_t user_len; /* length of user name; if user_off != 0 but |
| 19 | user_len == 0, an empty user name was given */ |
| 20 | size_t passwd_off; /* offset into url to start of passwd (0 => none) */ |
| 21 | size_t passwd_len; /* length of passwd; if passwd_off != 0 but |
| 22 | passwd_len == 0, an empty passwd was given */ |
| 23 | size_t host_off; /* offset into url to start of host name (0 => none) */ |
| 24 | size_t host_len; /* length of host name; |
| 25 | * file urls may have host_len == 0 */ |
| 26 | size_t port_off; /* offset into url to start of port number (0 => none) */ |
| 27 | size_t port_len; /* if a portnum is present (port_off != 0), it has |
| 28 | * this length (excluding the leading ':') starting |
| 29 | * from port_off (always 0 for file urls) */ |
| 30 | size_t path_off; /* offset into url to the start of the url path; |
| 31 | * this will always point to a '/' character |
| 32 | * after the url has been normalized */ |
| 33 | size_t path_len; /* length of path portion excluding any trailing |
| 34 | * '?...' and '#...' portion; will always be >= 1 */ |
| 35 | }; |
| 36 | |
| 37 | char *url_normalize(const char *, struct url_info *); |
| 38 | char *url_parse(const char *, struct url_info *); |
| 39 | |
| 40 | /* |
| 41 | * Like url_normalize(), but also allows '*' glob characters in the host |
| 42 | * portion. Use this when normalizing URL patterns from user configuration. |
| 43 | * |
| 44 | * Note that '*' is a valid path character per RFC 3986 (as a sub-delim), |
| 45 | * so glob patterns using '*' in the path are also accepted. |
| 46 | * |
| 47 | * Returns a newly allocated normalized string and fills out_info if |
| 48 | * non-NULL, or NULL if the pattern is invalid. |
| 49 | */ |
| 50 | char *url_normalize_pattern(const char *url, struct url_info *out_info); |
| 51 | |
| 52 | struct urlmatch_item { |
| 53 | size_t hostmatch_len; |
| 54 | size_t pathmatch_len; |
| 55 | char user_matched; |
| 56 | }; |
| 57 | |
| 58 | struct urlmatch_config { |
| 59 | struct string_list vars; |
| 60 | struct url_info url; |
| 61 | const char *section; |
| 62 | const char *key; |
| 63 | |
| 64 | void *cb; |
| 65 | config_fn_t collect_fn; |
| 66 | config_fn_t cascade_fn; |
| 67 | /* |
| 68 | * Compare the two matches, the one just discovered and the existing |
| 69 | * best match and return a negative value if the found item is to be |
| 70 | * rejected or a non-negative value if it is to be accepted. If this |
| 71 | * field is set to NULL, use the default comparison technique, which |
| 72 | * checks to ses if found is better (according to the urlmatch |
| 73 | * specificity rules) than existing. |
| 74 | */ |
| 75 | int (*select_fn)(const struct urlmatch_item *found, const struct urlmatch_item *existing); |
| 76 | /* |
| 77 | * An optional callback to allow e.g. for partial URLs; it shall |
| 78 | * return 1 or 0 depending whether `url` matches or not. |
| 79 | */ |
| 80 | int (*fallback_match_fn)(const char *url, void *cb); |
| 81 | }; |
| 82 | |
| 83 | #define URLMATCH_CONFIG_INIT { \ |
| 84 | .vars = STRING_LIST_INIT_DUP, \ |
| 85 | } |
| 86 | |
| 87 | int urlmatch_config_entry(const char *var, const char *value, |
| 88 | const struct config_context *ctx, void *cb); |
| 89 | void urlmatch_config_release(struct urlmatch_config *config); |
| 90 | |
| 91 | #endif /* URL_MATCH_H */ |