libpsl/src/lookup_string_in_fixed_set.c

/* Copyright 2015 The Chromium Authors. All rights reserved.
 * Use of this source code is governed by a BSD-style license that can be
 * found in the LICENSE.chromium file.
 *
 * Converted to C89 2015 by Tim Rühsen
 */

#include <stddef.h>

#if defined(__GNUC__) && defined(__GNUC_MINOR__)
#       define _GCC_VERSION_AT_LEAST(major, minor) ((__GNUC__ > (major)) || (__GNUC__ == (major) && __GNUC_MINOR__ >= (minor)))
#else
#       define _GCC_VERSION_AT_LEAST(major, minor) 0
#endif

#if _GCC_VERSION_AT_LEAST(4,0)
#  define _HIDDEN __attribute__ ((visibility ("hidden")))
#else
#  define _HIDDEN
#endif

#define CHECK_LT(a, b) if ((a) >= b) return 0

/*
 * Read next offset from pos.
 * Returns true if an offset could be read, false otherwise.
 */

static int GetNextOffset(const unsigned char** pos,
	const unsigned char* end,
	const unsigned char** offset)
{
	size_t bytes_consumed;

	if (*pos == end)
		return 0;

	/* When reading an offset the byte array must always contain at least
	 * three more bytes to consume. First the offset to read, then a node
	 * to skip over and finally a destination node. No object can be smaller
	 * than one byte. */
	CHECK_LT(*pos + 2, end);
	switch (**pos & 0x60) {
	case 0x60: /* Read three byte offset */
		*offset += (((*pos)[0] & 0x1F) << 16) | ((*pos)[1] << 8) | (*pos)[2];
		bytes_consumed = 3;
		break;
	case 0x40: /* Read two byte offset */
		*offset += (((*pos)[0] & 0x1F) << 8) | (*pos)[1];
		bytes_consumed = 2;
		break;
	default:
		*offset += (*pos)[0] & 0x3F;
		bytes_consumed = 1;
	}
	if ((**pos & 0x80) != 0) {
		*pos = end;
	} else {
		*pos += bytes_consumed;
	}
	return 1;
}

/*
 * Check if byte at offset is last in label.
 */

static int IsEOL(const unsigned char* offset, const unsigned char* end)
{
	CHECK_LT(offset, end);
	return(*offset & 0x80) != 0;
}

/*
 * Check if byte at offset matches first character in key.
 * This version matches characters not last in label.
 */

static int IsMatch(const unsigned char* offset,
	const unsigned char* end,
	const char* key)
{
	CHECK_LT(offset, end);
	return *offset == *key;
}

/*
 * Check if byte at offset matches first character in key.
 * This version matches characters last in label.
 */

static int IsEndCharMatch(const unsigned char* offset,
	const unsigned char* end,
	const char* key)
{
	CHECK_LT(offset, end);
	return *offset == (*key | 0x80);
}

/*
 * Read return value at offset.
 * Returns true if a return value could be read, false otherwise.
 */

static int GetReturnValue(const unsigned char* offset,
	const unsigned char* end,
	int* return_value)
{
	CHECK_LT(offset, end);
	if ((*offset & 0xE0) == 0x80) {
		*return_value = *offset & 0x0F;
		return 1;
	}
	return 0;
}

/*
 *  Looks up the string |key| with length |key_length| in a fixed set of
 * strings. The set of strings must be known at compile time. It is converted to
 * a graph structure named a DAFSA (Deterministic Acyclic Finite State
 * Automaton) by the script make_dafsa.py during compilation. This permits
 * efficient (in time and space) lookup. The graph generated by make_dafsa.py
 * takes the form of a constant byte array which should be supplied via the
 * |graph| and |length| parameters.  The return value is kDafsaNotFound,
 * kDafsaFound, or a bitmap consisting of one or more of kDafsaExceptionRule,
 * kDafsaWildcardRule and kDafsaPrivateRule ORed together.
 * 
 * Lookup a domain key in a byte array generated by make_dafsa.py.
 */

/* prototype to skip warning with -Wmissing-prototypes */
int _HIDDEN LookupStringInFixedSet(const unsigned char*, size_t,const char*, size_t);

int _HIDDEN LookupStringInFixedSet(const unsigned char* graph,
	size_t length,
	const char* key,
	size_t key_length)
{
	const unsigned char* pos = graph;
	const unsigned char* end = graph + length;
	const unsigned char* offset = pos;
	const char* key_end = key + key_length;

	while (GetNextOffset(&pos, end, &offset)) {
		/*char <char>+ end_char offsets
		 * char <char>+ return value
		 * char end_char offsets
		 * char return value
		 * end_char offsets
		 * return_value
		 */
		int did_consume = 0;

		if (key != key_end && !IsEOL(offset, end)) {
			/* Leading <char> is not a match. Don't dive into this child */
			if (!IsMatch(offset, end, key))
				continue;
			did_consume = 1;
			++offset;
			++key;
			/* Possible matches at this point:
			 * <char>+ end_char offsets
			 * <char>+ return value
			 * end_char offsets
			 * return value
			 */

			/* Remove all remaining <char> nodes possible */
			while (!IsEOL(offset, end) && key != key_end) {
				if (!IsMatch(offset, end, key))
					return -1;
				++key;
				++offset;
			}
		}
		/* Possible matches at this point:
		 * end_char offsets
		 * return_value
		 * If one or more <char> elements were consumed, a failure
		 * to match is terminal. Otherwise, try the next node.
		 */
		if (key == key_end) {
			int return_value;

			if (GetReturnValue(offset, end, &return_value))
				return return_value;
			/* The DAFSA guarantees that if the first char is a match, all
			 * remaining char elements MUST match if the key is truly present.
			 */
			if (did_consume)
				return -1;
			continue;
		}
		if (!IsEndCharMatch(offset, end, key)) {
			if (did_consume)
				return -1; /* Unexpected */
			continue;
		}
		++key;
		pos = ++offset; /* Dive into child */
	}

	return -1; /* No match */
}