Overview
Dekereke PA Export bridges the gap between Dekereke Phonology databases and
SIL Phonology Assistant.
Rather than going through FLEx (Fieldworks Language Explorer), this tool reads your
Dekereke XML export directly and produces a Toolbox .db file that
Phonology Assistant can open immediately.
XML → Toolbox DB
Reads Dekereke .xml exports and writes standard Toolbox SFM .db files.
No FLEx Needed
Skip the Fieldworks Language Explorer round-trip entirely — open your data in Phonology Assistant directly.
Flexible Mappings
Ships with sensible defaults; supply your own CSV mapping file to match any field layout.
Cross-platform GUI
Native-feeling app bundles for Windows and macOS — no Python installation required.
How to Use
Export your Dekereke database as XML
From Dekereke, export your phonology database. The root element of the XML file must be <phon_data>, with child <data_form> elements — one per lexical entry.
Launch the app and select your XML file
Windows: download and double-click sfm_convert_advanced.exe — no installation required.
macOS: unzip the download, then open sfm_convert_advanced.app.
On first launch, right-click → Open to bypass Gatekeeper if needed.
Choose a field mapping
When prompted, click Yes to use the built-in default mappings (see table below), or No to supply your own CSV mapping file.
Open the output .db file in Phonology Assistant
The tool writes <your-filename>.db alongside the XML file. Add it as a data source in Phonology Assistant — no further conversion needed.
Default Field Mappings
These mappings are built into the application. The SFM code column is written into the
Toolbox .db file; the XML Field Name column is the element name inside
each <data_form> block in your Dekereke XML.
| SFM Code | XML Field Name | Description |
|---|---|---|
\ref | Reference | Unique record reference / ID |
\ge | Gloss | English gloss |
\gn | IndonesianGloss | Indonesian gloss |
\sf | SoundFile | Linked audio file name |
\ph | Phonetic | Phonetic transcription |
\ps | Category | Part of speech / grammatical category |
\var1 | Variant1 | First phonological variant |
\var2 | Variant2 | Second phonological variant |
\n | Notes | Free-form notes |
Custom CSV Mappings
If your Dekereke XML uses different field names, provide a .csv mapping file.
Export a spreadsheet from Excel or a text editor in CSV format with these two columns:
| Column A — SFM Code | Column B — XML Field Name |
|---|---|
\ref | Reference |
\ge | Gloss |
| …additional rows… |
\) in SFM codes must be included. Rows with an empty SFM code column are
ignored — you can use these to document unmapped fields.
The output .db file begins with the Toolbox header line and separates
records with blank lines:
\_sh v3.0 400 PhoneticData \ref 001 \ge dog \ph dɔg \ps noun \ref 002 \ge house \ph haʊs \ps noun
Command-Line Usage
You can bypass the GUI by passing arguments directly. This is useful for scripting or batch processing.
sfm_convert_advanced.exe <input.xml> <mappings.csv>
On macOS, run the app from the command line with:
./sfm_convert_advanced.app/Contents/MacOS/sfm_convert_advanced input.xml mappings.csv
Source & License
Dekereke PA Export is free and open-source software, released under the GNU Affero General Public License v3.0 (AGPL-3.0). Source code is hosted on GitHub.