FileMaker-DDR-Splitter

A Python script to split a FileMaker Database Design Report (DDR) into it's components.

The script has been used on OSX with V10, V11, V15, V16, V18 and V19 DDRs. Windows and DDRs V12-14 may or may not work.

It's very fast. The longest run I have measured was 2 min. for a 420MB / 40 file DDR, creating more than 12000 files. A 275MB single file DDR was split in 19s.

Why would I like to do that?

Because it's the cheapest and one of the fastest ways to find all occurences of a field for example.

• Split your DDR

• Use a text editor that can search directories (TextWrangler and BBEdit on OSX come to mind) and search for a fieldname.

• The granularity of the found set is now down to scripts, layouts etc. instead of one big DDR file.

Because it makes versions of a database compaparable

• Create a DDR and split it.

• check the export into a git repository

• make changes in the database

• redo DDR and split

• git status and git diff are very telling what changed in the database.

Download an OSX app for OSX 10.9 and above.

Releases

Prerequisites

For the Python script

Python 3.11 - Higher version should work. Version 2.x probably not. All libraries needed are included with Python.

For building the OSX app the following additional libraries are needed:

PyObjC

py2app and it's dependencies

Usage

python3 ddrsplit.py /PATH/TO/Summary.xml

...or run the OSX app.

This will create a folder named "Exports" at the Summary.xml level.

Inside Exports for each database file this will create a folder with the database name. More precise: the xml filename without the extension.

Inside the database folder (drumroll) there will be up to 13 subfolders, namely Accounts, Assets, Basetables, CustomFunctions, CustomMenus, CustomMenuSets, Privileges, ExtendedPrivileges, Filereferences, Layouts, Relationships, Scripts and ValueLists.

• Accounts - The name pattern for each account is "ID NAME.xml"

• Assets - contains images and other stuff found in layouts. The filename will be it's SHA-1 hexdigest. The file extension is either derived from it's type (JPEG -> .jpg) for known types or ".TYPE" for unknown types. Using the SHA1 hecdigest as the filename has the advantage that identical files are stored only once.

• Basetables - The name pattern is "ID NAME.xml"

• CustomFunctions - The name pattern is "ID NAME.xml"

• CustomMenus - The name pattern is "ID NAME.xml"

• CustomMenuSets - The name pattern is "ID NAME.xml"

• ExtendedPrivileges - The name pattern is "ID NAME.xml"

• Filereferences - The name pattern is "ID NAME.xml"

• Layouts - The name pattern is "SORT ID NAME.xml". SORT is a 5 digit number to preserve the layout order. Groups (layout folders) will be presented as folders.

• Relationships - has 2 subfolders: Relationship, TableList. The name pattern is "ID NAME.xml"

• Scripts - for each script there will be it's xml from the DDR. The name pattern is "SORT-ID_NAME.xml". SORT is a 5 digit number to preserve the script order.

• ValueLists - The name pattern is "ID NAME.xml"

New 2017-07-12

• AuthFiles - The name pattern is "ID NAME.xml"

• ExternalDataSources - The name pattern is "ID NAME.xml"

• Themes - The name pattern is "ID NAME.xml"

• BaseDirectory - The name pattern is "ID NAME.xml"

The application

Configuration options

• Assets - This was the original inspiration for this program: to extract images from layouts. Somewhere between Versions 12 to 15 FileMaker stopped including images in the DDR.

• Asset options "per Solution", "per File", "per Layout" are disabled because they were never implemented.

• Layouts - include Layouts in split.

• Layout option "Create Layout Folders" - recreate layout folders in export.

• Layout option "Keep Layout Order" - recreate layout sorting in export. This is accomplished by prepending a sortindex in the filename.

• Scripts - include Scripts in split.

• Script option "Create Script Folders" - recreate script folders in export.

• Script option "Keep Script Order" - recreate script sorting in export. This is accomplished by prepending a sortindex in the filename.

• cross references - This function is in development and can't be turned off at the moment. The result is in the folder "references".

• Drop filename IDs - Omit IDs from filenames. This is useful when comparing two exports with git.

• OPML - disabled because they don't exist yet and remind me of writing them one day. The idea is to have a OPML file for scripts, basetables and layouts which can be opened with any outliner.