zbrookle / sql_to_ibis Goto Github PK

Make this meta object so that things like expression / literal count are not shared

Make function to extract a certain part of a date

Kindly please add support for recent ibis-framework which is compatible with recent SQLAlchemy>=1.4.
Without this, I can't use sql-to-ibis with recent SQLAlchemy, because in version 1.4 some of data types were renamed (and probably more breaking changes were introduced).

Add mypy types EVERYWHERE

mergequery = """
select  distinct * 
from mergeTable
where 
case when Status='52wkHigh' then strike < lastPrice
else strike > lastPrice end 
"""
rr = query(mergequery)

Getting below error

I got the following error when I tried to execute the code as is

Traceback (most recent call last):
  File "/Users/pavel/Projects/sql_to_ibis/mytest.py", line 1, in <module>
    import ibis.pandas.api
ModuleNotFoundError: No module named 'ibis.pandas'

After I adjusted ibis packages

import ibis
import pandas
import sql_to_ibis

df = pandas.DataFrame({"column1": [1, 2, 3], "column2": ["4", "5", "6"]})
ibis_table = ibis.backends.pandas.from_dataframe(
    df, name="my_table", client=ibis.backends.pandas.PandasClient({})
)
sql_to_ibis.register_temp_table(ibis_table, "my_table")
sql_to_ibis.query(
    "select column1, cast(column2 as integer) + 1 as my_col2 from my_table"
).execute()

I got the following error:

Traceback (most recent call last):
  File "~/miniconda3/envs/sql_to_ibis/lib/python3.9/site-packages/lark/lexer.py", line 376, in lex
    for x in l.lex(stream, self.root_lexer.newline_types, self.root_lexer.ignore_types):
  File "~/miniconda3/envs/sql_to_ibis/lib/python3.9/site-packages/lark/lexer.py", line 182, in lex
    raise UnexpectedCharacters(stream, line_ctr.char_pos, line_ctr.line, line_ctr.column, allowed=allowed, state=self.state, token_history=last_token and [last_token])
lark.exceptions.UnexpectedCharacters: No terminal defined for 'e' at line 1 col 36

select column1, cast(column2 as integer) + 1 as my_col2 from my_table
                                   ^

Expecting: {'RPAR'}

Previous tokens: Token(TYPENAME, 'int')

Queries aren't working when using != ''. Relates to zbrookle/dataframe_sql#41

Example:

select column1 as alias1, column2 as alias2, other_column1 as other_alias1, other_column2 as other_alias2
from table1 join table2
on table1.something = table2.something
where
alias1 != other_alias1 and alias2 != other_alias2

Support syntax highlighting in VSCode for sql_to_ibis's "universal" SQL dialect

ValueError: mutable default <class 'sql_to_ibis.sql.sql_clause_objects.Preceding'> for field preceding is not allowed: use default_factory

I'm getting the above error simply importing the module, in python 3.11.5. It is the 0.4.1 release of sql_to_ibis, with ibis 3.2.0

Make date function which extracts the date from an object

SELECT * FROM mytable WHERE mycolumn IN ('value1', 'value2') works fine, but if I have more than two values within the parentheses, InvalidQueryException is raised. The typical use of IN should allow for n comparison values. Am I misunderstanding something in composing my query or is there a workaround for this operator to use more than two values? Thanks!

Drop support for python 3.6 because ibis is no longer supporting 3.6

can you add some examples of expressions in README and/or docs?

"SELECT * FROM mytable WHERE column1='x' AND (column2='y' OR column3='z') works but
"SELECT * FROM mytable WHERE (column1='x' AND (column2='y' OR column3='z')) raises InvalidQueryException.

Sure, the outer parens around the where clause are extraneous, but the expression should still evaluate. I'm compiling where clauses from thousands of overlapping geographic features, and it is helpful as a default setting to enclose each sub-clause in parens, even if that means a lone feature ends up wrapped in these extraneous characters.

Happy to provide additional information/clarification if needed.

For example "FULL\nOUTER JOIN" is valid sql but with the current grammar, it can only be "FULL\sOUTER JOIN"