zfchen95 / querycsv Goto Github PK

Target in the scenarios that users want to acquire information from CSV files without putting data into a dataset, we developed this python-based software to ask questions over huge amount of data(300 attributes, millions of rows).

• Install an IDE for Python, e.g. Spyder or PyCharm
• Make sure you have Python files and CSV files in the same directory
• Open 'test_query.py'
• Modify file_path to current folder path (We don't have a user interface yet)
• Declare input query
• Build index first (Call build_index(), check instructions in build_index.py)
• Run 'test_query.py'

• The input query should be a string ended with a semicolon (;).
• The logic operators (SELECT, FROM, WHERE, AND, OR, NOT, LIKE) are case-sensitive and expected to be uppercase.
• Use comma (,) to separate attributes in SELECT clause, files in FROM clause, conditions in WHERE clause.
• Leave at least ONE space on both sides for operator(>, >=, <, <=, =, <>, AND, OR, NOT, LIKE).
• For LIKE, the pattern should be in string format with single quotes (' ').
• Rename required for querying two or more tables

"SELECT movie_title, title_year, imdb_score FROM movies.csv WHERE ( movie_title LIKE '%Kevin%' AND imdb_score > 7 );

['We Need to Talk About Kevin', '2011', '7.5'] ['Kevin Hart: Laugh at My Pain', '2011', '7.5']

"SELECT M.movie_title, A.Award FROM movies.csv M, oscars.csv A WHERE A.Name LIKE '%Kevin%' AND A.Winner = 1 AND A.Award = 'Actor in a Leading Role' AND A.Name = M.actor_1_name AND M.title_year > 2000 AND M.budget > M.gross AND M.movie_title LIKE '%Superman%' ;"

['Superman Returns', 'Actor in a Leading Role']

Some queries are provided in the 'test_query.py'.

Feel free to contact me: [email protected], if you have any issue in running the program.

• Handling query with at least two pairs of parentheses
• Handling query with arithmatic expression
• User interface
• Querying 4 tables

Copyright 2017 GSON - Zhenfeng CHEN, Yekai YU, Lu JIN, Xi LI