SQL DATABASE - README & ERROR MESSAGES ======================= Name: Helen Chac Name: Dehowe Feng Class: ECS 165B SQ 2016 To compile: Use the Makfile that is included in the project. On the terminal in the directory type in "make". To run: ./professorParser Functionality from the project: - Statistics for tables. -> Can be found in data/tables.stat -> Indentation is ugly, but correct. -> Did not implement DISTINCT VALUES for each indexed column. - WHERE clause: -> operations work. -> They need to have paranthesis in the correct areas (Order of operations). - Nested Loop Join -> This function assumes that the column names of the tables will not collide. - GROUP BY -> Group by works. -> Example query: select count(id) as b, name from student group by id; - ORDER BY -> Works with only by sorting by one thing. -> Example query: select id from student order by id; - IN operation -> Example query: select id from student where id IN (select id from student); -> If you want more operations in the WHERE: select id from student where (id IN (select id from student)) AND (id IN (select id from student)); - INSERT INTO operation -> Example query: insert into student values(9881840, "Helen"); - CREATE TABLE operation -> Example query: create table student(id, name); create table student(id key, name); Queries that work: - select count(id) as b from student group by id; - select count(id) as b, name from student group by id, name; - select id,name from student; - select id,name from student where id=2; - select id, name from student where (id < 2) AND (id =1); - select id, name from student where id IN (select id from student); Errors: (TO SEE THESE, you must be in -d mode) ============================================= Creation / Insertion -------------------- - Checks to see if it's a B-Tree Index. "DEBUG: tablename.columnname has B-Tree index" - Checks to see if that key already exists when inserting into the table. "DEBUG: Key already exists." - Checks to see if it's a valid OP_CODE in the expression tree. "Invalid OP_CODE in the expression tree -- need table_name" "Invalid OP_CODE in the expression tree -- create/insert" - Checks to see if the table is valid name. "Not valid Table name." - Checks to see if the table exists for both insert & create "Error: Table already exists." - Checks to see if the column_names of table are valid in creating table. "Error: Invalid column name" - Checks if there's enough memory to create the table. "Error: Cannot create table. Not enough memory." - Other creation errors: "Error: Unknown Server Error." - Checks the number of values to be inserted into the table. "Error: Too few values." "Error: Too many values." - Checks to see if the values entered are valid. "Error: Invalid Data." - Other insertion errors: "Error: Server Error." Loading the Database: - Checks to see if directory could be created if the data directory doesn't already exist. "Could not create directory" - Show that the file could be read into memory (not an error.) "DEBUG: Loading in file_name into database" - Show loading of the index files into memory (not an error.) "DEBUG: Loading in table_name.col_name." Projection ---------- - Checks if the projection table exists, if it doesn't exist, it prints out: "Table does not exist.\n"; - When doing select, it checks the projection: "Missing column names in the AS clause in the expression tree" "Missing AS in the expression tree" - Added checks for when function codes aren't correct, just in case you see a projection in a create table or something. Selection -------- - Checks the OP_CODE. "Function code does not match -- SELECTION" "Missing table in the expression tree" - Checks to see if the table exists. "Table does not exist" GROUP BY Operator ----------------- - Added checks for when we can't find column names in GROUP BY. "Could not find column name." - When doing function calls such as COUNT, SUM, AVG, MAX, MIN etc. Prints out error message if the expression tree is not correctly set up. - Checks if it's the right OP CODE in the expression tree: "OP COLUMN IS NOT OF RLIST" "OP COLUMN IS NOT OF SORTSPEC" "Function code does not match - GROUP BY" "Ordering not specified. Node in expression tree is NULL." "There was an error in parsing out the group names." "Foriegn OP_CODE in Groupby." - Checks to see if table exists "Could not find table in Group by" Aggregate Functions ------------------- - Checks if the OP Code is correct. "OP CODE does not match -- FUNC" "Operation code is not valid -- FuncCall" "OPCODE is not correct -- FuncCall aggregate functions" - Checks to see if there is an column_name in the aggregate function exists. --> Example: AVG(ID) -- checking if ID exists. "Column name in the aggregate function does not exist." - If the aggregate function does not exist, we state that we did not implement the function. "Did not implement this aggregate function" SORT Operator ------------- - When sorting, it checks all the OP_CODES to make sure that the expression tree is correct "Ordering not specified. Node in expression tree is NULL" "OP CODE IS NOT OF TYPE RLIST" "OP CODES IS NOT OF TYPE OP_SORTSPEC" "OP CODE IS NOT OP COLNAME" - When sorting, checks whether or not the column is within that table. "Could not find the column name in the table." IN Operator ----------- - When doing IN, it checks whether the subquery was valid - it needs to return a 1d array. "Invalid subquery in IN." - When doing IN, checks whether or not the column name exists inside that row. "Column does not exist in IN." - Checks to see if the OP_CODES match. "OP CODE IS NOT COLUMN NAME -- Checking IN." JOINING ------- **NOTE: Joins only work as a Cartesian Product (A X B) where they don't have the same column names. This is done using the nested loop join. - Checks to see if the OP_CODEs match "Went into the product function when op code is not product." "Error in calculating JOIN-ing" "OP Code's in Product is not TABLENAME" - Checks whether or not the database can find the table. "Database could not find table." ***THINGS THAT AREN'T IMPLEMENTED*** ==================================== - Does not account for large tables (loads the tables in memory - passed around through references). - Operations on the projections, however operations in the selection's where clause works. - Sorting with more than one column name. - Did not implement part iii, or vi. - For part ii, did not implement Hash Join or Index Join.