Structured data organizes information into predictable fields and records, such as customer, order date, delivery date, item, and quantity. In a Snowflake pipeline, it can be loaded into tables for querying, analysis, and downstream processing. section 7.0 section 1.1.2
A practical structured-data workflow can:
- Convert unstructured email text into fields representing customer orders. section 7.0
- Store extracted records in a Snowflake table for further analysis. section 7.0
- Ingest structured files such as CSV, JSON, XML, and Parquet from cloud storage. section 1.1.2
Orders extracted from an email
An email requesting six loaves of white bread on September 5 and sixteen bagels on September 6 becomes two table rows. Each row contains the customer, order date, delivery date, item, and quantity, making the order available to query as tabular data. section 7.4.4
Table 7.3 Output after calling the READ_EMAIL_PROC stored procedure that returns structured data from a sample customer email
|
Customer
|
Order date
|
Delivery date
|
Item
|
Quantity
|
|---|---|---|---|---|
|
Lilys Coffee
|
2023-09-05
|
2023-09-05
|
White loaf
|
6
|
|
Lilys Coffee
|
2023-09-06
|
2023-09-06
|
Bagel
|
16
|
Browse the sample table to see how prose from a customer email is represented as structured rows and fixed columns. section 7.4.4
