This section describes how FME reads and writes FME Feature Store (FFS) files.
FME Feature Store is a memory dump of FME features, so it supports the complete FME data model and can hold anything that FME features carry. This makes the format useful as a holding area for data that should persist between FME runs.
A spatial index may also be created and saved with the feature store which the FFS reader uses to quickly extract only those features within a specified area.
A logical FFS dataset consists of one or more files in the same folder with the extension .ffs. This extension is added to the base name of the FFS files. You can also use the FFS reader and writer to read and write only a single file.
File Size Limits
When writing a large amount of data to a single FFS file, file size limits may be encountered. If this occurs, the data is automatically split into multiple files of acceptable sizes. Reading in the first FFS file will automatically read in all files that were produced when the file was split.
FFS files contain all custom coordinate system definitions used on the features it contains, if any.
FME Raster Store Files
Rasters stored to an FFS file will have their data written to a corresponding .frs (FME Raster Store) file. One FRS file may hold the data for multiple raster features.
FRS files hold up to 2 GB of raster data; if the file surpasses this size, the data is automatically split across multiple files. If an FFS file containing raster features is opened and the corresponding FRS file cannot be opened, then the raster features will be restored as polygons with attributes indicating the properties of the raster (for example, number of rows/columns, spacing, etc.).
FFS and FME Version Compatibility
Because the FFS format is frequently updated with each new release of FME, any FFS dataset is not guaranteed to be compatible with a release of FME earlier than that with which it was created. If you do not have a current version of FME and you are using the FFS format for data exchange or storage, you may see an error message.
The FFS reader first reads the input file and passes the resultant features on to the rest of FME for further processing. If the FFS files have associated spatial indexes, then a spatial query can be used to limit the features returned.
Note: Compressed FFS files may take a longer time to generate, since the whole file is scanned for schema features.
The FFS writer dumps the memory representation of each FME feature into an FFS file, optionally creating a spatial index for the features.