#ifndef RFC822_PARSER_H
#define RFC822_PARSER_H

#include "unichar.h"

/* Maximum number of parameters to parse. After this the rest of the parameters
   are skipped. This is to avoid excessive CPU usage that can be caused by
   merging of these parameters. */
#define RFC2231_MAX_PARAMS 128

/* This can be used as a common NUL replacement character */
#define RFC822_NUL_REPLACEMENT_STR UNICODE_REPLACEMENT_CHAR_UTF8

struct rfc822_parser_context {
	const unsigned char *data, *end;
	string_t *last_comment;

	/* Replace NUL characters with this string */
	const char *nul_replacement_str;
};

#define IS_ATEXT(c) \
	(rfc822_atext_chars[(int)(unsigned char)(c)] != 0)
#define IS_ATEXT_NON_TSPECIAL(c) \
	((rfc822_atext_chars[(int)(unsigned char)(c)] & 3) != 0)
extern unsigned char rfc822_atext_chars[256];

/* Parse given data using RFC 822 token parser. */
void rfc822_parser_init(struct rfc822_parser_context *ctx,
			const unsigned char *data, size_t size,
			string_t *last_comment) ATTR_NULL(4);
static inline void rfc822_parser_deinit(struct rfc822_parser_context *ctx)
{
	/* make sure the parsing didn't trigger a bug that caused reading
	   past the end pointer. */
	i_assert(ctx->data <= ctx->end);
	/* make sure the parser is no longer accessed */
	ctx->data = ctx->end = NULL;
}

/* The functions below return 1 = more data available, 0 = no more data
   available (but a value might have been returned now), -1 = invalid input.

   LWSP is automatically skipped after value, but not before it. So typically
   you begin with skipping LWSP and then start using the parse functions. */

/* Parse comment. Assumes parser's data points to '(' */
int rfc822_skip_comment(struct rfc822_parser_context *ctx);
/* Skip LWSP if there is any */
int ATTR_NOWARN_UNUSED_RESULT
rfc822_skip_lwsp(struct rfc822_parser_context *ctx);
/* Stop at next non-atext char */
int rfc822_parse_atom(struct rfc822_parser_context *ctx, string_t *str);
/* Like parse_atom() but don't stop at '.' */
int rfc822_parse_dot_atom(struct rfc822_parser_context *ctx, string_t *str);
/* Like parse_dot_atom() but stops for '/', '?' and '='.
   Also it doesn't allow LWSP around '.' chars. */
int rfc822_parse_mime_token(struct rfc822_parser_context *ctx, string_t *str);
/* "quoted string" */
int rfc822_parse_quoted_string(struct rfc822_parser_context *ctx,
			       string_t *str);
/* atom or quoted-string */
int rfc822_parse_phrase(struct rfc822_parser_context *ctx, string_t *str);
/* dot-atom / domain-literal */
int rfc822_parse_domain(struct rfc822_parser_context *ctx, string_t *str);

/* Parse Content-Type header's type/subtype. */
int rfc822_parse_content_type(struct rfc822_parser_context *ctx, string_t *str);
/* For Content-Type style parameter parsing. Expect ";" key "=" value.
   value is unescaped if needed. The returned key is allocated from data
   stack. The value string is truncated for each call. Returns 1 = key/value
   set, 0 = no more data, -1 = invalid input. */
int rfc822_parse_content_param(struct rfc822_parser_context *ctx,
			       const char **key_r, string_t *value);

/* Decode a punycode-encoded domain name and return the UTF8
   form in result. Returns 0 on success and -1 on failure. */
void rfc822_decode_punycode(const char *input, size_t len, string_t *result);

#endif