Skip to content

Commit

Permalink
Merge pull request #976 from rkingsbury/docs
Browse files Browse the repository at this point in the history 
Store Documentation update
  • Loading branch information
@rkingsbury
rkingsbury authored Jul 1, 2024
2 parents 5c1d875 + d099a1c commit fe2eb65
Show file tree
Hide file tree
Showing 7 changed files with 484 additions and 48 deletions.
2 changes: 1 addition & 1 deletion README.md
View file
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@

# ![Maggma](docs/logo_w_text.svg)

[![Static Badge](https://img.shields.io/badge/documentation-blue?logo=github)](https://materialsproject.github.io/maggma) [![testing](https://github.com/materialsproject/maggma/workflows/testing/badge.svg)](https://github.com/materialsproject/maggma/actions?query=workflow%3Atesting) [![codecov](https://codecov.io/gh/materialsproject/maggma/branch/main/graph/badge.svg)](https://codecov.io/gh/materialsproject/maggma) [![python](https://img.shields.io/badge/Python-3.8+-blue.svg?logo=python&logoColor=white)]()
[![Static Badge](https://img.shields.io/badge/documentation-blue?logo=github)](https://materialsproject.github.io/maggma) [![testing](https://github.com/materialsproject/maggma/workflows/testing/badge.svg)](https://github.com/materialsproject/maggma/actions?query=workflow%3Atesting) [![codecov](https://codecov.io/gh/materialsproject/maggma/branch/main/graph/badge.svg)](https://codecov.io/gh/materialsproject/maggma) [![python](https://img.shields.io/badge/Python-3.9+-blue.svg?logo=python&logoColor=white)]()

## What is Maggma

Expand Down
180 changes: 180 additions & 0 deletions docs/getting_started/query_101.md
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,180 @@
# Understanding Queries

Putting your data into a `maggma` `Store` gives you powerful search, summary,
and analytical capabilities. All are based on "queries", which specify how
you want to search your data, and which parts of it you want to get in return.

`maggma` query syntax closely follows [MongoDB Query syntax](https://www.mongodb.com/docs/manual/tutorial/query-documents/). In this tutorial, we'll cover the syntax of the most common query operations. You can refer to the
[MongoDB](https://www.mongodb.com/docs/manual/tutorial/query-documents/) or [pymongo](https://pymongo.readthedocs.io/en/stable/tutorial.html) (python interface to MongoDB) documentation for examples of more advanced use cases.

Let's create an example dataset describing the [Teenage Mutant Ninja Turtles](https://en.wikipedia.org/wiki/Teenage_Mutant_Ninja_Turtles).

```python
>>> turtles = [{"name": "Leonardo",
"color": "blue",
"tool": "sword",
"occupation": "ninja"
},
{"name": "Donatello",
"color": "purple",
"tool": "staff",
"occupation": "ninja"
},
{"name": "Michelangelo",
"color": "orange",
"tool": "nunchuks",
"occupation": "ninja"
},
{"name":"Raphael",
"color": "red",
"tool": "sai",
"occupation": "ninja"
},
{"name":"Splinter",
"occupation": "sensei"
}
]
```

Notice how this data follows the principles described in [Structuring `Store` data](stores.md/#structuring-store-data):
- every document (`dict`) has a `name` key with a unique value
- every document has a common set of keys (`name`,
`occupation`).
- Note that SOME documents also share the keys `tool` and `color`, but not all. This is OK.

For the rest of this tutorial, we will assume that this data has already been
added to a `Store` called `tmnt_store`, which we are going to query.

## The `query` method

`Store.query()` is the primary method you will use to search your data.

- `query`
always returns a generator yielding any and all documents that match the query
you provide.
- There are no mandatory arguments. If you run `query()` you will get a generator
containing all documents in the `Store`
- The first (optional) argument is `criteria`, which is a query formatted as a `dict` as described in the next section.
- You can also specify `properties`, which is a list of fields from the documents you want to return. This is useful when working with large documents because then you only have to download the data you need rather than the entire document.
- You can also `skip` every N documents, `limit` the number of documents returned, and `sort` the result by some field.

Since `query` returns a generator, you will typically want to turn the results into a list, or use them in a `for` loop.

Turn into a list
```python
results = [d for d in store.query()]
```

Use in a `for` loop
```python
for doc in store.query():
print(doc)
```

## The structure of a query

A query is also a `dict`. Each key in the dict corresponds to a fjeld in the
documents you want to query (such as `name`, `color`, etc.), and the value
is the value of that key that you want to match. For example, a query to
select all documents where `occupation` is `ninja`, would look like

```python
{"occupation": "ninja"}
```

This query will be passed as an argument to `Store` methods like `query_one`,
`query`, and `count`, as demonstrated next.


## Example queries

### Match a single value

To select all records where a field matches a single value, set the key to
the field you want to match and its value to the value you are looking for.

Return all records where 'occupation' is 'ninja'
```python
>>> with tmnt_store as store:
... results = list(store.query({"occupation": "ninja"}))
>>> len(results)
4
```

Return all records where 'name' is 'Splinter'

```python
>>> with tmnt_store as store:
... results = list(store.query({"name": "Splinter"}))
>>> len(results)
1
```

### Match any value in a list: `$in`

To find all documents where a field matches one of several different
values, use `$in` with a list of the value you want to search.

```python
>>> with tmnt_store as store:
... results = list(store.query({"color": {"$in": ["red", "blue"]}}))
>>> len(results)
2
```

`$in` is an example of a "query operator". Others include:

- `$nin`: a value is NOT in a list (the inverse of the above example)
- `$gt`, `$gte`: greater than, greater than or equal to a value
- `$lt`, `$lte`: greater than, greater than or equal to a value
- `$ne`: not equal to a value
- `$not`: inverts the effect of a query expression, returning results that
do NOT match.

See the [MongoDB docs](https://www.mongodb.com/docs/manual/reference/operator/query/#query-selectors) for a complete list.

!!! Note

When using query operators like `$in`, you must include a nested `dict` in
your query, where the operator is the key and the search parameters are
the value, e.g., the dictionary `{"$in": ["red", "blue"]}` is the **value**
associated with the search field (`color`) in the parent dictionary.

### Nested fields

Suppose that our documents had a nested structure, for example, by having
separate fields for first and last name:

```python
>>> turtles = [{"name":
{"first": "Leonardo",
"last": "turtle"
},
"color": "blue",
"tool": "sword",
"occupation": "ninja"
},
...
]
```

You can query nested fields by placing a period `.` between each level in the
hierarchy. For example:

```python
>>> with tmnt_store as store:
... results = list(store.query({"name.first": "Splinter"}))
>>> len(results)
1
```

### Numerical Values

You can query numerical values in analogous fashion to the examples given above.

!!! Note
When querying on numerical values, be mindful of the `type` of the data.
Data stored in `json` format is often converted entirely to `str`, so if
you use a numerical query operator like `$gte`, you might not get the
results you expect unless you first verify that the numerical data
in the `Store` is a `float` or `int` .
136 changes: 106 additions & 30 deletions docs/getting_started/stores.md
View file
Original file line number Diff line number Diff line change
@@ -1,55 +1,131 @@
# Using `Store`

A `Store` is just a wrapper to access data from a data source. That data source is typically a MongoDB collection, but it could also be an Amazon S3 bucket, a GridFS collection, or folder of files on disk. `maggma` makes interacting with all of these data sources feel the same (see the `Store` interface, below). `Store` can also perform logic, concatenating two or more `Store` together to make them look like one data source for instance.
A `Store` is just a wrapper to access data from a data source. That data source is typically a MongoDB collection, but it could also be an Amazon S3 bucket, a GridFS collection, or folder of files on disk. `maggma` makes interacting with all of these data sources feel the same (see the [`Store` interface](#the-store-interface), below). `Store` can also perform logic, concatenating two or more `Store` together to make them look like one data source for instance.

The benefit of the `Store` interface is that you only have to write a `Builder` once. As your data moves or evolves, you simply point it to different `Store` without having to change your processing code.

## List of Stores
## Structuring `Store` data

Current working and tested `Store` include:

- `MongoStore`: interfaces to a MongoDB Collection using port and hostname.
- `MongoURIStore`: interfaces to a MongoDB Collection using a "mongodb+srv://" URI.
- `MemoryStore`: just a Store that exists temporarily in memory
- `JSONStore`: builds a MemoryStore and then populates it with the contents of the given JSON files
- `FileStore`: query and add metadata to files stored on disk as if they were in a database
- `GridFSStore`: interfaces to GridFS collection in MongoDB using port and hostname.
- `GridFSURIStore`: interfaces to GridFS collection in MongoDB using a "mongodb+srv://" URI.
- `S3Store`: provides an interface to an S3 Bucket either on AWS or self-hosted solutions ([additional documentation](advanced_stores.md))
- `ConcatStore`: concatenates several Stores together so they look like one Store
- `VaultStore`: uses Vault to get credentials for a MongoDB database
- `AliasingStore`: aliases keys from the underlying store to new names
- `SandboxStore: provides permission control to documents via a `_sbxn` sandbox key
- `JointStore`: joins several MongoDB collections together, merging documents with the same `key`, so they look like one collection
- `AzureBlobStore`: provides an interface to Azure Blobs for the storage of large amount of data
- `MontyStore`: provides an interface to [montydb](https://github.com/davidlatwe/montydb) for in-memory or filesystem-based storage
- `MongograntStore`: (DEPRECATED) uses Mongogrant to get credentials for MongoDB database
Because `Store` is built around a MongoDB-like query syntax, data that goes into `Store` needs to be structured similarly to MongoDB data. In python terms,
that means **the data in a `Store` must be structured as a `list` of `dict`**,
where each `dict` represents a single record (called a 'document').

```python
data = [{"AM": "sunrise"}, {"PM": "sunset"} ... ]
```

Note that this structure is very similar to the widely-used [JSON](https://en.wikipedia.org/wiki/JSON) format. So structuring your data in this manner
enables highly flexible storage options -- you can easily write it to a `.json`
file, place it in a `Store`, insert it into a Mongo database, etc. `maggma` is
designed to facilitate this.

In addition to being structured as a `list` of `dict`, **every document (`dict`)
must have a key that uniquely identifies it.** By default, this key is the `task_id`, but it can be set to any value you
like using the `key` argument when you instantiate a `Store`.

```python
data = [{"task_id": 1, "AM": "sunrise"}, {"task_id: 2, "PM": "sunset"} ... ]
```

Just to emphasize - **every document must have a `task_id`, and the value of `task_id` must be unique for every document**. The rest of the document structure
is up to you, but `maggma` works best when every document follows a pre-defined
schema (i.e., all `dict` have the same set of keys / same structure).

## The `Store` interface

All `Store` provide a number of basic methods that facilitate querying, updating, and removing data:

- `query`: Standard mongo style `find` method that lets you search the store.
- `query`: Standard mongo style `find` method that lets you search the store. See [Understanding Queries](query_101.md) for more details about the query syntax.
- `query_one`: Same as above but limits returned results to just the first document that matches your query. Very useful for understanding the structure of the returned data.
- `update`: Update the documents into the collection. This will override documents if the key field matches.
- `ensure_index`: This creates an index for the underlying data-source for fast querying.
- `distinct`: Gets distinct values of a field.
- `count`: Counts documents in the `Store`
- `distinct`: Returns a list of distinct values of a field.
- `groupby`: Similar to query but performs a grouping operation and returns sets of documents.
- `update`: Update (insert) documents into the `Store`. This will overwrite documents if the key field matches.
- `remove_docs`: Removes documents from the underlying data source.
- `last_updated`: Finds the most recently updated `last_updated_field` value and returns that. Useful for knowing how old a data-source is.
- `newer_in`: Finds all documents that are newer in the target collection and returns their `key`s. This is a very useful way of performing incremental processing.
- `ensure_index`: Creates an index for the underlying data-source for fast querying.
- `last_updated`: Finds the most recently updated `last_updated_field` value and returns that. Useful for knowing how old a data-source is.

!!! Note
If you are familiar with `pymongo`, you may find the comparison table below
helpful. This table illustrates how `maggma` method and argument names map
onto `pymongo` concepts.


### Initializing a Store
| `maggma` | `pymongo` equivalent |
| -------- | ------- |
| **methods** |
| `query_one` | `find_one` |
| `query` | `find` |
| `count` | `count_documents` |
| `distinct` | `distinct` |
| `groupby` | `group` |
| `update` | `insert` |
| **arguments** |
| `criteria={}` | `filter={}` |
| `properties=[]` | `projection=[]` |

All `Store`s have a few basic arguments that are critical for basic usage. Every `Store` has two attributes that the user should customize based on the data contained in that store: `key` and `last_updated_field`. The `key` defines how the `Store` tells documents apart. Typically this is `_id` in MongoDB, but you could use your own field (be sure all values under the key field can be used to uniquely identify documents). `last_updated_field` tells `Store` how to order the documents by a date, which is typically in the `datetime` format, but can also be an ISO 8601-format (ex: `2009-05-28T16:15:00`) `Store`s can also take a `Validator` object to make sure the data going into it obeys some schema.

### Using a Store
## Creating a Store

All `Store`s have a few basic arguments that are critical for basic usage. Every `Store` has two attributes that the user should customize based on the data contained in that store: `key` and `last_updated_field`.

The `key` defines how the `Store` tells documents apart. Typically this is `_id` in MongoDB, but you could use your own field (be sure all values under the key field can be used to uniquely identify documents).

`last_updated_field` tells `Store` how to order the documents by a date, which is typically in the `datetime` format, but can also be an ISO 8601-format (ex: `2009-05-28T16:15:00`) `Store`s can also take a `Validator` object to make sure the data going into it obeys some schema.

In the example below, we create a `MongoStore`, which connects to a MongoDB database.
To create this store, we have to provide `maggma` the connection details to the
database like the hostname, collection name, and authentication info. Note that
we've set `key='name'` because we want to use that `name` as our unique identifier.

```python
>>> store = MongoStore(database="my_db_name",
collection_name="my_collection_name",
username="my_username",
password="my_password",
host="my_hostname",
port=27017,
key="name",
)
```

The specific arguments required to create a `Store` depend on the underlying
format. For example, the `MemoryStore`, which just loads data into memory,
requires no arguments to instantiate. Refer to the [list of Stores](#list-of-stores)
below (and their associated documentation) for specific details.

## Connecting to a `Store`

You must connect to a store by running `store.connect()` before querying or updating the store.
If you are operating on the stores inside of another code it is recommended to use the built-in context manager,
which will take care of the `connect()` automatically, e.g.:
If you are operating on the stores inside of another code it is recommended to use the built-in context manager, e.g.:

```python
with MongoStore(...) as store:
store.query()
```

This will take care of the `connect()` automatically while ensuring that the
connection is closed properly after the store tasks are complete.

## List of Stores

Current working and tested `Store` include the following. Click the name of
each store for more detailed documentation.

- [`MongoStore`](/maggma/reference/stores/#maggma.stores.mongolike.MongoStore): interfaces to a MongoDB Collection using port and hostname.
- [`MongoURIStore`](/maggma/reference/stores/#maggma.stores.mongolike.MongoURIStore): interfaces to a MongoDB Collection using a "mongodb+srv://" URI.
- [`MemoryStore`](/maggma/reference/stores/#maggma.stores.mongolike.MemoryStore): just a Store that exists temporarily in memory
- [`JSONStore`](/maggma/reference/stores/#maggma.stores.mongolike.JSONStore): builds a MemoryStore and then populates it with the contents of the given JSON files
- [`FileStore`](/maggma/reference/stores/#maggma.stores.file_store.FileStore): query and add metadata to files stored on disk as if they were in a database
- [`GridFSStore`](/maggma/reference/stores/#maggma.stores.gridfs.GridFSStore): interfaces to GridFS collection in MongoDB using port and hostname.
- [`GridFSURIStore`](/maggma/reference/stores/#maggma.stores.gridfs.GridFSURIStore): interfaces to GridFS collection in MongoDB using a "mongodb+srv://" URI.
- [`S3Store`](/maggma/reference/stores/#maggma.stores.aws.S3Store): provides an interface to an S3 Bucket either on AWS or self-hosted solutions ([additional documentation](advanced_stores.md))
- [`ConcatStore`](/maggma/reference/stores/#maggma.stores.compound_stores.ConcatStore): concatenates several Stores together so they look like one Store
- [`VaultStore`](/maggma/reference/stores/#maggma.stores.advanced_stores.VaultStore): uses Vault to get credentials for a MongoDB database
- [`AliasingStore`](/maggma/reference/stores/#maggma.stores.advanced_stores.AliasingStore): aliases keys from the underlying store to new names
- `SandboxStore: provides permission control to documents via a `_sbxn` sandbox key
- [`JointStore`](/maggma/reference/stores/#maggma.stores.compound_stores.JointStore): joins several MongoDB collections together, merging documents with the same `key`, so they look like one collection
- [`AzureBlobStore`](/maggma/reference/stores/#maggma.stores.azure.AzureBlobStore): provides an interface to Azure Blobs for the storage of large amount of data
- [`MontyStore`](/maggma/reference/stores/#maggma.stores.mongolike.MontyStore): provides an interface to [montydb](https://github.com/davidlatwe/montydb) for in-memory or filesystem-based storage
- [`MongograntStore`](/maggma/reference/stores/#maggma.stores.advanced_stores.MongograntStore): (DEPRECATED) uses Mongogrant to get credentials for MongoDB database
Loading

0 comments on commit fe2eb65

Please sign in to comment.