Tables and items in DynamoDB: Free Lesson

Saving an order

DynamoDB is a cloud database that stores records, called items, in tables. Shops and apps use it to save orders and find them again in milliseconds.

Unlike a spreadsheet, every value in a DynamoDB item says what type it is. Your Python code has to read and write those typed values.

By the end of this lesson

Note

I can read and write a DynamoDB item in its typed form.

Key words

Word	What it means	Example
Table	a named collection of items
Item	one record in a table
Attribute	a named value inside an item
Type letter	the code that says what kind of value it is	S
String set	a group of different strings with no order

The cells students fill in are left blank in this preview.

Every value carries its type letter

A DynamoDB table holds items. An item is a set of attributes, each with a name and a value.

The value is a dictionary with one key, a type letter, such as {"S": "tea"}. S is a string and N is a number. BOOL is true or false, and NULL is no value.

Worked example: the type letters

Letter	Holds	Example
S	a string	{"S": "tea"}
N	a number, written as text	{"N": "4"}
BOOL	true or false	{"BOOL": false}
NULL	no value	{"NULL": true}
SS	a string set, in no fixed order	{"SS": ["rush", "fragile"]}

The cells students fill in are left blank in this preview.

Run it: read an item

Code: python

answer = {
    "Item": {
        "customer": {"S": "ana"},
        "gift": {"BOOL": False},
        "item": {"S": "tea"},
        "note": {"NULL": True},
        "order_id": {"N": "1"},
        "status": {"S": "paid"},
        "tags": {"SS": ["rush", "fragile"]},
        "total": {"N": "4"},
    }
}
item = answer["Item"]
for name in item:
    kind = list(item[name].keys())[0]
    print(name, "has type", kind)
print("item:", item["item"]["S"])
# Add a line that prints the status.

A number travels as text

DynamoDB sends every number as a string, such as {"N": "15"}, so that every digit arrives as written.

To use one, convert it with int("15") or float("2.5"). To store one, turn it back into text with str.

Worked example: using a number

Step	Python	Result
Read	total["N"]	"15"
Convert	int(total["N"])	15
Add one	int(total["N"]) + 1	16
Write back	{"N": str(16)}	{"N": "16"}

The cells students fill in are left blank in this preview.

Run it: text against number

Code: python

total = {"N": "15"}
print(total["N"] + "1")
print(int(total["N"]) + 1)
print("9" > "15", 9 > 15)
print({"N": str(16)})
# Change "15" to "40". What does each line print now?

A common mistake

Warning

Many people think Python reads {"N": "15"} as the number 15. In Python it is the text "15". Adding to it joins text, so "15" + "1" gives "151". Convert with int or float before any sum, and use str to write it back.

A missing item is an empty answer

Asking for an item that does not exist is not an error. get_item answers with no Item key at all.

A write with a condition, such as "only if this order is new", can be refused. The error has the Code ConditionalCheckFailedException and status 400. The captured file writes it as {"error": ...}. boto3 raises an exception that carries the same Code and status.

Run it: found, missing and refused

Code: python

found = {"Item": {"customer": {"S": "ana"}, "note": {"NULL": True}}}
missing = {}
refused = {"error": {"Code": "ConditionalCheckFailedException", "HTTPStatusCode": 400}}
for answer in [found, missing]:
    if "Item" in answer:
        print("found, note is null:", answer["Item"]["note"]["NULL"])
    else:
        print("no such item")
print(refused["error"]["Code"], refused["error"]["HTTPStatusCode"])
# Change the note to {"S": "gift card"}, then read it with ["S"].

Try it in the cloud lab

The cloud lab is a practice copy of S3, SQS and DynamoDB. Press Start the lab on the bar, then press Run on each command below. They use the real AWS command line, so the same words work on a real account. The last command prints nothing: a missing item is an empty answer here, not an error. This step is for practice. The assignment below stays plain Python, so you can do it without the lab.