Skip to content

Latest commit

 

History

History
65 lines (41 loc) · 2.07 KB

README.md

File metadata and controls

65 lines (41 loc) · 2.07 KB
Raw

pg-schema-diff

Compares two schemas in postgresql databases and prints SQL commands to modify the first one in order to match the second one.

It does NOT execute the statements. It only prints the statements to the standard output.

Installing

Install globally with npm

npm install pg-schema-diff -g

Usage

pg-schema-diff \
  postgres://user:pass@host[:port]/dbname1 \
  schemaname1 \
  postgres://user:pass@host[:port]/dbname2 \
  schemaname2

The first database connection string and schema name describe the ‘current schema’. The second pair describe the ‘future schema’. The output describes the commands required to move from the current schema to the future schema.

Caveats

Some statements may fail or may produce data loss depending on the data stored in the target schema. For example:

## Dropping tables and columns

pg-schema-diff will generate DROP TABLE and DROP COLUMN statements. Make sure you want to drop those tables / columns.

## Changing the data type of existing columns

Postgresql is not able to change the existing data to the new data type. In that case you will get an error similar to this:

ERROR:  column "column_name" cannot be cast automatically to type integer
HINT:  Specify a USING expression to perform the conversion.

So you will need to specify a USING expression to perform de conversion. For example to convert text to integers:

ALTER TABLE table_name
  ALTER column_name TYPE data_type USING column_name::integer

## Table Column CONSTRAINTs

At this version table column constraints are compared but the SQL statements are not generated. pg-schema-diff will output comments describing differences between the schemas.

## NOT NULL violations

If an existing column needs to be changed from nullable to not nullable the statement may fail if there are existing rows with a NULL value in that column. In that case you will get an error like:

ERROR:  column "column_name" contains null values.

You should fill the existing rows with not null values before making the column not nullable.