# Queries ## Built in query types These can be used in searches, as in ```python c.search(FullText("hello")) ``` Follow the links in the table below for examples specific to each query. ```{eval-rst} .. autosummary:: :toctree: generated tiled.queries.Comparison tiled.queries.Contains tiled.queries.Eq tiled.queries.NotEq tiled.queries.FullText tiled.queries.In tiled.queries.NotIn tiled.queries.Regex tiled.queries.SpecQuery tiled.queries.SpecsQuery tiled.queries.StructureFamilyQuery ``` (Some have the word `Query` at the end of their name to avoid confusion with other objects in the Tiled codebase.) ## Query expressions The `Key` object can be used to construct queries in a readable way using standard Python comparison operators, as in ```python Key("color") == "red" Key("shape") != "circle" Key("temperature") > 300 ``` used in searches like ```python c.search(Key("color") == "red").search(Key("shape") != "circle").search(Key("temperature") > 300) ``` Notice that, to progressively narrow results, you may use repeated calls to `.search()` as in ```python c.search(...).search(...).search(...) ``` The above achieves logical `AND`. To achieve logical `OR`, you must perform separate queries and combine them yourself. It may be convenient to combine them as shown in a `dict`, as long as the results are not too numerous to fit in memory. ```python a = c.search(...) b = c.search(...) c = c.search(...) combined_results = {**a, **b, **c} # to handle repeats, must use {...} syntax not dict(...) ``` ````{warning} **You cannot use queries with the Python keywords `not`, `in`, `and`, or `or`.** In Python, `and` and `or` have a particular behavior: ```python >>> 3 or 5 3 >>> 3 and 5 5 ``` which would result in the first or last query being used, respectively, ignoring all others. This is an unavoidable consequence of Python semantics. Likewise, `in X` and ``not X` must return `True` or `False`; they cannot return a query. To avoid confusion, Tiled raises a `TypeError` if you attempt to use a query with `not`, `in`, `and`, or `or`. ```` ```{eval-rst} .. autosummary:: :toctree: generated tiled.queries.Key ``` ## Convenience Functions We recommend building convenience functions to provide for succinct usage for common searches. For example: ```py def Sample(sample_name): return Key("sample_name") == sample_name ``` This reduces ```py c.search(Key("sample_name") == "stuff") ``` to ```py c.search(Sample("stuff")) ``` ## Custom queries Not all queries can be expressed as combinations of the built in ones. External libraries (like databroker) may register custom query types in addition to those built in to Tiled.