Ptolemy, an AWS DMS table mapping generator
Write terse AWS DMS table mappings.
Background
Amazon Web Services provides the Database Migration Service (DMS) tool for migrating data to, from or between SQL databases. When running DMS, users can supply a table mapping, which allows the user to specify what data is sent from the source database to the target database. A full list of table mapping options can be found here.
Table mappings are written as JSON documents, which can grow to be long and complex. ptolemy
allows the user to write terse YAML source
files, which can be compiled to valid JSON table mappings using the ptolemy
cli tool.
Usage
$ # Display an example source file:
$ cat migrate_all_tables_in_a_schema.yaml
selection:
include:
-
object-locators:
schema-names:
- Test
table-names:
- "%"
$ # Compile it to a DMS table mapping:
$ ptolemy migrate_all_tables_in_a_schema.yaml
{
"rules": [
{
"object-locator": {
"schema-name": "Test",
"table-name": "%"
},
"rule-action": "include",
"rule-id": "1",
"rule-name": "1",
"rule-type": "selection"
}
]
}
API
$ ptolemy -h
ptolemy [-h] [-d] [-v] source
positional arguments:
source path to the source file
optional arguments:
-h, --help show this help message and exit
-d, --debug enable debug logs
-v, --version show program's version number and exit
Install
Install via pip (recommended):
$ pip install ptolemy
Install from source:
$ git clone git@github.com:cloudreach/ptolemy.git
$ cd ptolemy
$ make install
License
ptolemy is licensed under the Apache Software License 2.0.
Source Syntax
The following sections describe the source syntax. It is intended to show users who have working knowledge of DMS JSON mapping files how to write their YAML equivalents. For an overview of the JSON mapping files, see the documentation. Most items are the same as those in JSON mapping files, with the exception of object-locators
, which are explained in the section Object Locators
The descriptions are written in pseudo-yaml, where the syntax ( option_a|option_b )
indicates that an item could take the value option_a
or option_b
.
For working examples, see the examples directory.
Selection Rules and Actions
selection:
( include|exclude ):
-
object-locators:
schema-names:
- <schema-name-1>
table-names:
- <table-name-1>
filters:
-
filter-type: source
column-name: <column-name-1>
filter-conditions:
-
filter-operator: ( ste|gte|eq|between )
value: <value>
Transformation Rules and Actions
transformation:
( rename|remove-column|convert-lowercase|convert-uppercase|add-prefix|remove-prefix|replace-prefix|add-suffix|remove-suffix|replace-suffix ):
-
object-locators:
schema-names:
- <schema-name-1>
table-names:
- <table-name-1>
column-names:
- <column-name-1>
rule-target: ( schema|table|column )
value: <value>
old-value: <old-value>
Object Locators
object-locators
offer a powerful way to apply selection and transformation rules to large numbers of objects. The singular schema-name
, table-name
and column-name
parameters of the native JSON table mapping object-locator
have been replaced by their plurals. These new parameters each accept a list of objects. The rule is then applied to each column listed, for each table listed, for each schema listed.
Multiple Source File Compilation
Multiple source files can be compiled at once with the following Bash snippet. The snippet recursively finds all YAML files under the directory src/
, compiles the source to a DMS mapping file, and saves it to a file with the same name and path under a directory named mappings/
, with the extension .json
.
source_files="$(find src -type f -name '*.yaml')"
for source_file in $source_files; do
source_file_without_extension=${source_file%.*}
source_file_with_json_extension=${source_file_without_extension}.json
destination_file=mappings${source_file_with_json_extension#src}
mkdir -p "$(dirname $destination_file)"
ptolemy $source_file > $destination_file
done
Running the code from the following directory:
.
└── src
├── db-a
│ ├── schema-1.yaml
│ └── schema-2.yaml
└── db-b
├── schema-1.yaml
└── schema-2.yaml
would result in:
.
├── mappings
│ ├── db-a
│ │ ├── schema-1.json
│ │ └── schema-2.json
│ └── db-b
│ ├── schema-1.json
│ └── schema-2.json
└── src
├── db-a
│ ├── schema-1.yaml
│ └── schema-2.yaml
└── db-b
├── schema-1.yaml
└── schema-2.yaml