Skip to content

Commit

Permalink
dbfile: rename the table "hashes" to "blocks"
Browse files Browse the repository at this point in the history 
Signed-off-by: Alexandre Bruyelles <[email protected]>
  • Loading branch information
@JackSlateur
JackSlateur committed Nov 17, 2023
1 parent d6cca3d commit 54d4eaf
Show file tree
Hide file tree
Showing 2 changed files with 37 additions and 37 deletions.
70 changes: 35 additions & 35 deletions dbfile.c
View file
Original file line number Diff line number Diff line change
Expand Up @@ -162,12 +162,12 @@ static int create_tables(sqlite3 *db)
if (ret)
goto out;

#define CREATE_TABLE_HASHES \
"CREATE TABLE IF NOT EXISTS hashes(digest BLOB KEY NOT NULL, " \
#define CREATE_TABLE_BLOCKS \
"CREATE TABLE IF NOT EXISTS blocks(digest BLOB KEY NOT NULL, " \
"fileid INTEGER, loff INTEGER, " \
"UNIQUE(fileid, loff) " \
"FOREIGN KEY(fileid) REFERENCES files(id) ON DELETE CASCADE);"
ret = sqlite3_exec(db, CREATE_TABLE_HASHES, NULL, NULL, NULL);
ret = sqlite3_exec(db, CREATE_TABLE_BLOCKS, NULL, NULL, NULL);

out:
if (ret)
Expand All @@ -180,15 +180,15 @@ static int create_indexes(sqlite3 *db)
{
int ret;

#define CREATE_HASHES_DIGEST_INDEX \
"create index if not exists idx_digest on hashes(digest);"
ret = sqlite3_exec(db, CREATE_HASHES_DIGEST_INDEX, NULL, NULL, NULL);
#define CREATE_BLOCKS_DIGEST_INDEX \
"create index if not exists idx_digest on blocks(digest);"
ret = sqlite3_exec(db, CREATE_BLOCKS_DIGEST_INDEX, NULL, NULL, NULL);
if (ret)
goto out;

#define CREATE_HASHES_INOSUB_INDEX \
"create index if not exists idx_hashes_inosub on hashes(fileid);"
ret = sqlite3_exec(db, CREATE_HASHES_INOSUB_INDEX, NULL, NULL, NULL);
#define CREATE_BLOCKS_INOSUB_INDEX \
"create index if not exists idx_blocks_inosub on blocks(fileid);"
ret = sqlite3_exec(db, CREATE_BLOCKS_INOSUB_INDEX, NULL, NULL, NULL);
if (ret)
goto out;

Expand Down Expand Up @@ -380,9 +380,9 @@ struct dbhandle *dbfile_open_handle(char *filename)
if (!result->db)
goto err;

#define INSERT_HASH \
"INSERT INTO hashes (fileid, loff, digest) VALUES (?1, ?2, ?3);"
dbfile_prepare_stmt(insert_hash, INSERT_HASH);
#define INSERT_BLOCK \
"INSERT INTO blocks (fileid, loff, digest) VALUES (?1, ?2, ?3);"
dbfile_prepare_stmt(insert_block, INSERT_BLOCK);

#define INSERT_EXTENTS \
"INSERT INTO extents (fileid, loff, poff, len, digest) " \
Expand All @@ -394,13 +394,13 @@ struct dbhandle *dbfile_open_handle(char *filename)
dbfile_prepare_stmt(update_scanned_file, UPDATE_SCANNED_FILE);

#define FIND_BLOCKS \
"select files.filename, hashes.loff from files " \
"INNER JOIN hashes " \
"on hashes.digest = ?1 AND files.id = hashes.fileid;"
"select files.filename, blocks.loff from files " \
"INNER JOIN blocks " \
"on blocks.digest = ?1 AND files.id = blocks.fileid;"
dbfile_prepare_stmt(find_blocks, FIND_BLOCKS);

#define FIND_TOP_B_HASHES \
"select digest, count(digest) from hashes " \
"select digest, count(digest) from blocks " \
"group by digest having (count(digest) > 1) " \
"order by (count(digest)) desc;"
dbfile_prepare_stmt(find_top_b_hashes, FIND_TOP_B_HASHES);
Expand All @@ -413,8 +413,8 @@ struct dbhandle *dbfile_open_handle(char *filename)

#define FIND_B_FILES_COUNT \
"select count (distinct files.filename) from files " \
"INNER JOIN hashes " \
"on hashes.digest = ?1 AND files.id = hashes.fileid;"
"INNER JOIN blocks " \
"on blocks.digest = ?1 AND files.id = blocks.fileid;"
dbfile_prepare_stmt(find_b_files_count, FIND_B_FILES_COUNT);

#define FIND_E_FILES_COUNT \
Expand All @@ -435,7 +435,7 @@ struct dbhandle *dbfile_open_handle(char *filename)
dbfile_prepare_stmt(write_file, WRITE_FILE);

#define REMOVE_BLOCK_HASHES \
"delete from hashes where fileid = (select id from files " \
"delete from blocks where fileid = (select id from files " \
"where ino = ?1 and subvol = ?2);"
dbfile_prepare_stmt(remove_block_hashes, REMOVE_BLOCK_HASHES);

Expand All @@ -448,22 +448,22 @@ struct dbhandle *dbfile_open_handle(char *filename)
"select filename, size from files where ino = ?1 and subvol = ?2;"
dbfile_prepare_stmt(load_filerec, LOAD_FILEREC);

#define GET_DUPLICATE_HASHES \
"WITH without_future_hashes as (" \
" select * from hashes " \
" join files on files.id = hashes.fileid " \
#define GET_DUPLICATE_BLOCKS \
"WITH without_future_blocks as (" \
" select * from blocks " \
" join files on files.id = blocks.fileid " \
" and files.dedupe_seq <= ?1), " \
"current_hashes as (" \
" select * from hashes " \
" join files on files.id = hashes.fileid " \
"current_blocks as (" \
" select * from blocks " \
" join files on files.id = blocks.fileid " \
" and files.dedupe_seq = ?1) " \
"SELECT without_future_hashes.digest, ino, subvol, loff " \
"FROM without_future_hashes " \
"JOIN (SELECT DISTINCT digest FROM current_hashes " \
"SELECT without_future_blocks.digest, ino, subvol, loff " \
"FROM without_future_blocks " \
"JOIN (SELECT DISTINCT digest FROM current_blocks " \
"GROUP BY digest " \
"HAVING count(*) > 1) AS duplicate_hashes " \
"on without_future_hashes.digest = duplicate_hashes.digest;"
dbfile_prepare_stmt(get_duplicate_hashes, GET_DUPLICATE_HASHES);
"HAVING count(*) > 1) AS duplicate_blocks " \
"on without_future_blocks.digest = duplicate_blocks.digest;"
dbfile_prepare_stmt(get_duplicate_blocks, GET_DUPLICATE_BLOCKS);

/*
* We need to select on both digest and len, otherwise we
Expand Down Expand Up @@ -525,7 +525,7 @@ struct dbhandle *dbfile_open_handle(char *filename)
"select mtime, size from files where ino = ?1 and subvol = ?2;"
dbfile_prepare_stmt(select_file_changes, SELECT_FILE_CHANGES);

#define COUNT_B_HASHES "select COUNT(*) from hashes;"
#define COUNT_B_HASHES "select COUNT(*) from blocks;"
dbfile_prepare_stmt(count_b_hashes, COUNT_B_HASHES);

#define COUNT_E_HASHES "select COUNT(*) from extents;"
Expand Down Expand Up @@ -1074,7 +1074,7 @@ int dbfile_store_block_hashes(struct dbhandle *db, int64_t fileid,
{
int ret;
uint64_t i;
_cleanup_(sqlite3_reset_stmt) sqlite3_stmt *stmt = db->stmts.insert_hash;
_cleanup_(sqlite3_reset_stmt) sqlite3_stmt *stmt = db->stmts.insert_block;

for (i = 0; i < nb_hash; i++) {
ret = sqlite3_bind_int64(stmt, 1, fileid);
Expand Down Expand Up @@ -1243,7 +1243,7 @@ int dbfile_load_block_hashes(struct hash_tree *hash_tree, unsigned int seq)
if (!db)
return ENOENT;

stmt = db->stmts.get_duplicate_hashes;
stmt = db->stmts.get_duplicate_blocks;

ret = sqlite3_bind_int64(stmt, 1, seq);
if (ret) {
Expand Down
4 changes: 2 additions & 2 deletions dbfile.h
View file
Original file line number Diff line number Diff line change
Expand Up @@ -18,7 +18,7 @@ struct dbfile_config;
#define DB_FILE_MINOR 0

struct stmts {
sqlite3_stmt *insert_hash;
sqlite3_stmt *insert_block;
sqlite3_stmt *insert_extent;
sqlite3_stmt *update_scanned_file;
sqlite3_stmt *find_blocks;
Expand All @@ -31,7 +31,7 @@ struct stmts {
sqlite3_stmt *remove_block_hashes;
sqlite3_stmt *remove_extent_hashes;
sqlite3_stmt *load_filerec;
sqlite3_stmt *get_duplicate_hashes;
sqlite3_stmt *get_duplicate_blocks;
sqlite3_stmt *get_duplicate_extents;
sqlite3_stmt *get_duplicate_files;
sqlite3_stmt *get_file_extent;
Expand Down

0 comments on commit 54d4eaf

Please sign in to comment.