/*

 * Copyright (C) the libgit2 contributors. All rights reserved.

 *

 * This file is part of libgit2, distributed under the GNU GPL v2 with

 * a Linking Exception. For full terms see the included COPYING file.

 */

#include "common.h"

#include "odb.h"

#include "pack.h"

#include "delta.h"

#include "sha1_lookup.h"

#include "mwindow.h"

#include "fileops.h"

#include "oid.h"

#include <zlib.h>

static int packfile_open(struct git_pack_file *p);

static git_off_t nth_packed_object_offset(const struct git_pack_file *p, uint32_t n);

static int packfile_unpack_compressed(

		git_rawobj *obj,

		struct git_pack_file *p,

		git_mwindow **w_curs,

		git_off_t *curpos,

		size_t size,

		git_otype type);

/* Can find the offset of an object given

 * a prefix of an identifier.

 * Throws GIT_EAMBIGUOUSOIDPREFIX if short oid

 * is ambiguous within the pack.

 * This method assumes that len is between

 * GIT_OID_MINPREFIXLEN and GIT_OID_HEXSZ.

 */

static int pack_entry_find_offset(

		git_off_t *offset_out,

		git_oid *found_oid,

		struct git_pack_file *p,

		const git_oid *short_oid,

		size_t len);

static int packfile_error(const char *message)

{

	giterr_set(GITERR_ODB, "invalid pack file - %s", message);

	return -1;

}

/********************

 * Delta base cache

 ********************/

static git_pack_cache_entry *new_cache_object(git_rawobj *source)

{

	git_pack_cache_entry *e = git__calloc(1, sizeof(git_pack_cache_entry));

	if (!e)

		return NULL;

	git_atomic_inc(&e->refcount);

	memcpy(&e->raw, source, sizeof(git_rawobj));

	return e;

}

static void free_cache_object(void *o)

{

	git_pack_cache_entry *e = (git_pack_cache_entry *)o;

	if (e != NULL) {

		assert(e->refcount.val == 0);

		git__free(e->raw.data);

		git__free(e);

	}

}

static void cache_free(git_pack_cache *cache)

{

	git_pack_cache_entry *entry;

	if (cache->entries) {

		git_offmap_foreach_value(cache->entries, entry, {

			free_cache_object(entry);

		});

		git_offmap_free(cache->entries);

		cache->entries = NULL;

	}

}

static int cache_init(git_pack_cache *cache)

{

	cache->entries = git_offmap_alloc();

	GITERR_CHECK_ALLOC(cache->entries);

	cache->memory_limit = GIT_PACK_CACHE_MEMORY_LIMIT;

	if (git_mutex_init(&cache->lock)) {

		giterr_set(GITERR_OS, "failed to initialize pack cache mutex");

		git__free(cache->entries);

		cache->entries = NULL;

		return -1;

	}

	return 0;

}

static git_pack_cache_entry *cache_get(git_pack_cache *cache, git_off_t offset)

{

	khiter_t k;

	git_pack_cache_entry *entry = NULL;

	if (git_mutex_lock(&cache->lock) < 0)

		return NULL;

	k = git_offmap_lookup_index(cache->entries, offset);

	if (git_offmap_valid_index(cache->entries, k)) { /* found it */

		entry = git_offmap_value_at(cache->entries, k);

		git_atomic_inc(&entry->refcount);

		entry->last_usage = cache->use_ctr++;

	}

	git_mutex_unlock(&cache->lock);

	return entry;

}

/* Run with the cache lock held */

static void free_lowest_entry(git_pack_cache *cache)

{

	git_off_t offset;

	git_pack_cache_entry *entry;

	git_offmap_foreach(cache->entries, offset, entry, {

		if (entry && entry->refcount.val == 0) {

			cache->memory_used -= entry->raw.len;

			git_offmap_delete(cache->entries, offset);

			free_cache_object(entry);

		}

	});

}

static int cache_add(

		git_pack_cache_entry **cached_out,

		git_pack_cache *cache,

		git_rawobj *base,

		git_off_t offset)

{

	git_pack_cache_entry *entry;

	int error, exists = 0;

	khiter_t k;

	if (base->len > GIT_PACK_CACHE_SIZE_LIMIT)

		return -1;

	entry = new_cache_object(base);

	if (entry) {

		if (git_mutex_lock(&cache->lock) < 0) {

			giterr_set(GITERR_OS, "failed to lock cache");

			git__free(entry);

			return -1;

		}

		/* Add it to the cache if nobody else has */

		exists = git_offmap_exists(cache->entries, offset);

		if (!exists) {

			while (cache->memory_used + base->len > cache->memory_limit)

				free_lowest_entry(cache);

			k = git_offmap_put(cache->entries, offset, &error);

			assert(error != 0);

			git_offmap_set_value_at(cache->entries, k, entry);

			cache->memory_used += entry->raw.len;

			*cached_out = entry;

		}

		git_mutex_unlock(&cache->lock);

		/* Somebody beat us to adding it into the cache */

		if (exists) {

			git__free(entry);

			return -1;

		}

	}

	return 0;

}

/***********************************************************

 *

 * PACK INDEX METHODS

 *

 ***********************************************************/

static void pack_index_free(struct git_pack_file *p)

{

	if (p->oids) {

		git__free(p->oids);

		p->oids = NULL;

	}

	if (p->index_map.data) {

		git_futils_mmap_free(&p->index_map);

		p->index_map.data = NULL;

	}

}

static int pack_index_check(const char *path, struct git_pack_file *p)

{

	struct git_pack_idx_header *hdr;

	uint32_t version, nr, i, *index;

	void *idx_map;

	size_t idx_size;

	struct stat st;

	int error;

	/* TODO: properly open the file without access time using O_NOATIME */

	git_file fd = git_futils_open_ro(path);

	if (fd < 0)

		return fd;

	if (p_fstat(fd, &st) < 0) {

		p_close(fd);

		giterr_set(GITERR_OS, "unable to stat pack index '%s'", path);

		return -1;

	}

	if (!S_ISREG(st.st_mode) ||

		!git__is_sizet(st.st_size) ||

		(idx_size = (size_t)st.st_size) < 4 * 256 + 20 + 20)

	{

		p_close(fd);

		giterr_set(GITERR_ODB, "invalid pack index '%s'", path);

		return -1;

	}

	error = git_futils_mmap_ro(&p->index_map, fd, 0, idx_size);

	p_close(fd);

	if (error < 0)

		return error;

	hdr = idx_map = p->index_map.data;

	if (hdr->idx_signature == htonl(PACK_IDX_SIGNATURE)) {

		version = ntohl(hdr->idx_version);

		if (version < 2 || version > 2) {

			git_futils_mmap_free(&p->index_map);

			return packfile_error("unsupported index version");

		}

	} else

		version = 1;

	nr = 0;

	index = idx_map;

	if (version > 1)

		index += 2; /* skip index header */

	for (i = 0; i < 256; i++) {

		uint32_t n = ntohl(index[i]);

		if (n < nr) {

			git_futils_mmap_free(&p->index_map);

			return packfile_error("index is non-monotonic");

		}

		nr = n;

	}

	if (version == 1) {

		/*

		 * Total size:

		 * - 256 index entries 4 bytes each

		 * - 24-byte entries * nr (20-byte sha1 + 4-byte offset)

		 * - 20-byte SHA1 of the packfile

		 * - 20-byte SHA1 file checksum

		 */

		if (idx_size != 4*256 + nr * 24 + 20 + 20) {

			git_futils_mmap_free(&p->index_map);

			return packfile_error("index is corrupted");

		}

	} else if (version == 2) {

		/*

		 * Minimum size:

		 * - 8 bytes of header

		 * - 256 index entries 4 bytes each

		 * - 20-byte sha1 entry * nr

		 * - 4-byte crc entry * nr

		 * - 4-byte offset entry * nr

		 * - 20-byte SHA1 of the packfile

		 * - 20-byte SHA1 file checksum

		 * And after the 4-byte offset table might be a

		 * variable sized table containing 8-byte entries

		 * for offsets larger than 2^31.

		 */

		unsigned long min_size = 8 + 4*256 + nr*(20 + 4 + 4) + 20 + 20;

		unsigned long max_size = min_size;

		if (nr)

			max_size += (nr - 1)*8;

		if (idx_size < min_size || idx_size > max_size) {

			git_futils_mmap_free(&p->index_map);

			return packfile_error("wrong index size");

		}

	}

	p->num_objects = nr;

	p->index_version = version;

	return 0;

}

static int pack_index_open(struct git_pack_file *p)

{

	int error = 0;

	size_t name_len;

	git_buf idx_name;

	if (p->index_version > -1)

		return 0;

	name_len = strlen(p->pack_name);

	assert(name_len > strlen(".pack")); /* checked by git_pack_file alloc */

	if (git_buf_init(&idx_name, name_len) < 0)

		return -1;

	git_buf_put(&idx_name, p->pack_name, name_len - strlen(".pack"));

	git_buf_puts(&idx_name, ".idx");

	if (git_buf_oom(&idx_name)) {

		git_buf_free(&idx_name);

		return -1;

	}

	if ((error = git_mutex_lock(&p->lock)) < 0) {

		git_buf_free(&idx_name);

		return error;

	}

	if (p->index_version == -1)

		error = pack_index_check(idx_name.ptr, p);

	git_buf_free(&idx_name);

	git_mutex_unlock(&p->lock);

	return error;

}

static unsigned char *pack_window_open(

		struct git_pack_file *p,

		git_mwindow **w_cursor,

		git_off_t offset,

		unsigned int *left)

{

	if (p->mwf.fd == -1 && packfile_open(p) < 0)

		return NULL;

	/* Since packfiles end in a hash of their content and it's

	 * pointless to ask for an offset into the middle of that

	 * hash, and the pack_window_contains function above wouldn't match

	 * don't allow an offset too close to the end of the file.

	 *

	 * Don't allow a negative offset, as that means we've wrapped

	 * around.

	 */

	if (offset > (p->mwf.size - 20))

		return NULL;

	if (offset < 0)

		return NULL;

	return git_mwindow_open(&p->mwf, w_cursor, offset, 20, left);

 }

/*

 * The per-object header is a pretty dense thing, which is

 *  - first byte: low four bits are "size",

 *    then three bits of "type",

 *    with the high bit being "size continues".

 *  - each byte afterwards: low seven bits are size continuation,

 *    with the high bit being "size continues"

 */

size_t git_packfile__object_header(unsigned char *hdr, size_t size, git_otype type)

{

	unsigned char *hdr_base;

	unsigned char c;

	assert(type >= GIT_OBJ_COMMIT && type <= GIT_OBJ_REF_DELTA);

	/* TODO: add support for chunked objects; see git.git 6c0d19b1 */

	c = (unsigned char)((type << 4) | (size & 15));

	size >>= 4;

	hdr_base = hdr;

	while (size) {

		*hdr++ = c | 0x80;

		c = size & 0x7f;

		size >>= 7;

	}

	*hdr++ = c;

	return (hdr - hdr_base);

}

static int packfile_unpack_header1(

		unsigned long *usedp,

		size_t *sizep,

		git_otype *type,

		const unsigned char *buf,

		unsigned long len)

{

	unsigned shift;

	unsigned long size, c;

	unsigned long used = 0;

	c = buf[used++];

	*type = (c >> 4) & 7;

	size = c & 15;

	shift = 4;

	while (c & 0x80) {

		if (len <= used) {

			giterr_set(GITERR_ODB, "buffer too small");

			return GIT_EBUFS;

		}

		if (bitsizeof(long) <= shift) {

			*usedp = 0;

			giterr_set(GITERR_ODB, "packfile corrupted");

			return -1;

		}

		c = buf[used++];

		size += (c & 0x7f) << shift;

		shift += 7;

	}

	*sizep = (size_t)size;

	*usedp = used;

	return 0;

}

int git_packfile_unpack_header(

		size_t *size_p,

		git_otype *type_p,

		git_mwindow_file *mwf,

		git_mwindow **w_curs,

		git_off_t *curpos)

{

	unsigned char *base;

	unsigned int left;

	unsigned long used;

	int ret;

	/* pack_window_open() assures us we have [base, base + 20) available

	 * as a range that we can look at at. (Its actually the hash

	 * size that is assured.) With our object header encoding

	 * the maximum deflated object size is 2^137, which is just

	 * insane, so we know won't exceed what we have been given.

	 */

/*	base = pack_window_open(p, w_curs, *curpos, &left); */

	base = git_mwindow_open(mwf, w_curs, *curpos, 20, &left);

	if (base == NULL)

		return GIT_EBUFS;

	ret = packfile_unpack_header1(&used, size_p, type_p, base, left);

	git_mwindow_close(w_curs);

	if (ret == GIT_EBUFS)

		return ret;

	else if (ret < 0)

		return packfile_error("header length is zero");

	*curpos += used;

	return 0;

}

int git_packfile_resolve_header(

		size_t *size_p,

		git_otype *type_p,

		struct git_pack_file *p,

		git_off_t offset)

{

	git_mwindow *w_curs = NULL;

	git_off_t curpos = offset;

	size_t size;

	git_otype type;

	git_off_t base_offset;

	int error;

	error = git_packfile_unpack_header(&size, &type, &p->mwf, &w_curs, &curpos);

	if (error < 0)

		return error;

	if (type == GIT_OBJ_OFS_DELTA || type == GIT_OBJ_REF_DELTA) {

		size_t base_size;

		git_packfile_stream stream;

		base_offset = get_delta_base(p, &w_curs, &curpos, type, offset);

		git_mwindow_close(&w_curs);

		if ((error = git_packfile_stream_open(&stream, p, curpos)) < 0)

			return error;

		error = git_delta_read_header_fromstream(&base_size, size_p, &stream);

		git_packfile_stream_free(&stream);

		if (error < 0)

			return error;

	} else {

		*size_p = size;

		base_offset = 0;

	}

	while (type == GIT_OBJ_OFS_DELTA || type == GIT_OBJ_REF_DELTA) {

		curpos = base_offset;

		error = git_packfile_unpack_header(&size, &type, &p->mwf, &w_curs, &curpos);

		if (error < 0)

			return error;

		if (type != GIT_OBJ_OFS_DELTA && type != GIT_OBJ_REF_DELTA)

			break;

		base_offset = get_delta_base(p, &w_curs, &curpos, type, base_offset);

		git_mwindow_close(&w_curs);

	}

	*type_p = type;

	return error;

}

#define SMALL_STACK_SIZE 64

/**

 * Generate the chain of dependencies which we need to get to the

 * object at `off`. `chain` is used a stack, popping gives the right

 * order to apply deltas on. If an object is found in the pack's base

 * cache, we stop calculating there.

 */

static int pack_dependency_chain(git_dependency_chain *chain_out,

				 git_pack_cache_entry **cached_out, git_off_t *cached_off,

				 struct pack_chain_elem *small_stack, size_t *stack_sz,

				 struct git_pack_file *p, git_off_t obj_offset)

{

	git_dependency_chain chain = GIT_ARRAY_INIT;

	git_mwindow *w_curs = NULL;

	git_off_t curpos = obj_offset, base_offset;

	int error = 0, use_heap = 0;

	size_t size, elem_pos;

	git_otype type;

	elem_pos = 0;

	while (true) {

		struct pack_chain_elem *elem;

		git_pack_cache_entry *cached = NULL;

		/* if we have a base cached, we can stop here instead */

		if ((cached = cache_get(&p->bases, obj_offset)) != NULL) {

			*cached_out = cached;

			*cached_off = obj_offset;

			break;

		}

		/* if we run out of space on the small stack, use the array */

		if (elem_pos == SMALL_STACK_SIZE) {

			git_array_init_to_size(chain, elem_pos);

			GITERR_CHECK_ARRAY(chain);

			memcpy(chain.ptr, small_stack, elem_pos * sizeof(struct pack_chain_elem));

			chain.size = elem_pos;

			use_heap = 1;

		}

		curpos = obj_offset;

		if (!use_heap) {

			elem = &small_stack[elem_pos];

		} else {

			elem = git_array_alloc(chain);

			if (!elem) {

				error = -1;

				goto on_error;

			}

		}

		elem->base_key = obj_offset;

		error = git_packfile_unpack_header(&size, &type, &p->mwf, &w_curs, &curpos);

		if (error < 0)

			goto on_error;

		elem->offset = curpos;

		elem->size = size;

		elem->type = type;

		elem->base_key = obj_offset;

		if (type != GIT_OBJ_OFS_DELTA && type != GIT_OBJ_REF_DELTA)

			break;

		base_offset = get_delta_base(p, &w_curs, &curpos, type, obj_offset);

		git_mwindow_close(&w_curs);

		if (base_offset == 0) {

			error = packfile_error("delta offset is zero");

			goto on_error;

		}

		if (base_offset < 0) { /* must actually be an error code */

			error = (int)base_offset;

			goto on_error;

		}

		/* we need to pass the pos *after* the delta-base bit */

		elem->offset = curpos;

		/* go through the loop again, but with the new object */

		obj_offset = base_offset;

		elem_pos++;

	}

	*stack_sz = elem_pos + 1;

	*chain_out = chain;

	return error;

on_error:

	git_array_clear(chain);

	return error;

}

int git_packfile_unpack(

	git_rawobj *obj,

	struct git_pack_file *p,

	git_off_t *obj_offset)

{

	git_mwindow *w_curs = NULL;

	git_off_t curpos = *obj_offset;

	int error, free_base = 0;

	git_dependency_chain chain = GIT_ARRAY_INIT;

	struct pack_chain_elem *elem = NULL, *stack;

	git_pack_cache_entry *cached = NULL;

	struct pack_chain_elem small_stack[SMALL_STACK_SIZE];

	size_t stack_size = 0, elem_pos, alloclen;

	git_otype base_type;

	/*

	 * TODO: optionally check the CRC on the packfile

	 */

	error = pack_dependency_chain(&chain, &cached, obj_offset, small_stack, &stack_size, p, *obj_offset);

	if (error < 0)

		return error;

	obj->data = NULL;

	obj->len = 0;

	obj->type = GIT_OBJ_BAD;

	/* let's point to the right stack */

	stack = chain.ptr ? chain.ptr : small_stack;

	elem_pos = stack_size;

	if (cached) {

		memcpy(obj, &cached->raw, sizeof(git_rawobj));

		base_type = obj->type;

		elem_pos--;	/* stack_size includes the base, which isn't actually there */

	} else {

		elem = &stack[--elem_pos];

		base_type = elem->type;

	}

	switch (base_type) {

	case GIT_OBJ_COMMIT:

	case GIT_OBJ_TREE:

	case GIT_OBJ_BLOB:

	case GIT_OBJ_TAG:

		if (!cached) {

			curpos = elem->offset;

			error = packfile_unpack_compressed(obj, p, &w_curs, &curpos, elem->size, elem->type);

			git_mwindow_close(&w_curs);

			base_type = elem->type;

		}

		if (error < 0)

			goto cleanup;

		break;

	case GIT_OBJ_OFS_DELTA:

	case GIT_OBJ_REF_DELTA:

		error = packfile_error("dependency chain ends in a delta");

		goto cleanup;

	default:

		error = packfile_error("invalid packfile type in header");

		goto cleanup;

	}

	/*

	 * Finding the object we want a cached base element is

	 * problematic, as we need to make sure we don't accidentally

	 * give the caller the cached object, which it would then feel

	 * free to free, so we need to copy the data.

	 */

	if (cached && stack_size == 1) {

		void *data = obj->data;

		GITERR_CHECK_ALLOC_ADD(&alloclen, obj->len, 1);

		obj->data = git__malloc(alloclen);

		GITERR_CHECK_ALLOC(obj->data);

		memcpy(obj->data, data, obj->len + 1);

		git_atomic_dec(&cached->refcount);

		goto cleanup;

	}

	/* we now apply each consecutive delta until we run out */

	while (elem_pos > 0 && !error) {

		git_rawobj base, delta;

		/*

		 * We can now try to add the base to the cache, as

		 * long as it's not already the cached one.

		 */

		if (!cached)

			free_base = !!cache_add(&cached, &p->bases, obj, elem->base_key);

		elem = &stack[elem_pos - 1];

		curpos = elem->offset;

		error = packfile_unpack_compressed(&delta, p, &w_curs, &curpos, elem->size, elem->type);

		git_mwindow_close(&w_curs);

		if (error < 0) {

			/* We have transferred ownership of the data to the cache. */

			obj->data = NULL;

			break;

		}

		/* the current object becomes the new base, on which we apply the delta */

		base = *obj;

		obj->data = NULL;

		obj->len = 0;

		obj->type = GIT_OBJ_BAD;

		error = git_delta_apply(&obj->data, &obj->len, base.data, base.len, delta.data, delta.len);

		obj->type = base_type;

		/*

		 * We usually don't want to free the base at this

		 * point, as we put it into the cache in the previous

		 * iteration. free_base lets us know that we got the

		 * base object directly from the packfile, so we can free it.

		 */

		git__free(delta.data);

		if (free_base) {

			free_base = 0;

			git__free(base.data);

		}

		if (cached) {

			git_atomic_dec(&cached->refcount);

			cached = NULL;

		}

		if (error < 0)

			break;

		elem_pos--;

	}

cleanup:

	if (error < 0) {

		git__free(obj->data);

		if (cached)

			git_atomic_dec(&cached->refcount);

	}

	if (elem)

		*obj_offset = curpos;

	git_array_clear(chain);

	return error;

}

static void *use_git_alloc(void *opaq, unsigned int count, unsigned int size)

{

	GIT_UNUSED(opaq);

	return git__calloc(count, size);

}

static void use_git_free(void *opaq, void *ptr)

{

	GIT_UNUSED(opaq);

	git__free(ptr);

}

int git_packfile_stream_open(git_packfile_stream *obj, struct git_pack_file *p, git_off_t curpos)

{

	int st;

	memset(obj, 0, sizeof(git_packfile_stream));

	obj->curpos = curpos;

	obj->p = p;

	obj->zstream.zalloc = use_git_alloc;

	obj->zstream.zfree = use_git_free;

	obj->zstream.next_in = Z_NULL;

	obj->zstream.next_out = Z_NULL;

	st = inflateInit(&obj->zstream);

	if (st != Z_OK) {

		giterr_set(GITERR_ZLIB, "failed to init packfile stream");

		return -1;

	}

	return 0;

}

ssize_t git_packfile_stream_read(git_packfile_stream *obj, void *buffer, size_t len)

{

	unsigned char *in;

	size_t written;

	int st;

	if (obj->done)

		return 0;

	in = pack_window_open(obj->p, &obj->mw, obj->curpos, &obj->zstream.avail_in);

	if (in == NULL)

		return GIT_EBUFS;

	obj->zstream.next_out = buffer;

	obj->zstream.avail_out = (unsigned int)len;

	obj->zstream.next_in = in;

	st = inflate(&obj->zstream, Z_SYNC_FLUSH);

	git_mwindow_close(&obj->mw);

	obj->curpos += obj->zstream.next_in - in;

	written = len - obj->zstream.avail_out;

	if (st != Z_OK && st != Z_STREAM_END) {

		giterr_set(GITERR_ZLIB, "error reading from the zlib stream");

		return -1;

	}

	if (st == Z_STREAM_END)

		obj->done = 1;

	/* If we didn't write anything out but we're not done, we need more data */

	if (!written && st != Z_STREAM_END)

		return GIT_EBUFS;

	return written;

}

void git_packfile_stream_free(git_packfile_stream *obj)

{

	inflateEnd(&obj->zstream);

}

static int packfile_unpack_compressed(

	git_rawobj *obj,

	struct git_pack_file *p,

	git_mwindow **w_curs,

	git_off_t *curpos,

	size_t size,

	git_otype type)

{

	size_t buf_size;

	int st;

	z_stream stream;

	unsigned char *buffer, *in;

	GITERR_CHECK_ALLOC_ADD(&buf_size, size, 1);

	buffer = git__calloc(1, buf_size);

	GITERR_CHECK_ALLOC(buffer);

	memset(&stream, 0, sizeof(stream));

	stream.next_out = buffer;

	stream.avail_out = (uInt)buf_size;

	stream.zalloc = use_git_alloc;

	stream.zfree = use_git_free;

	st = inflateInit(&stream);

	if (st != Z_OK) {

		git__free(buffer);

		giterr_set(GITERR_ZLIB, "failed to init zlib stream on unpack");

		return -1;

	}

	do {

		in = pack_window_open(p, w_curs, *curpos, &stream.avail_in);