class
GdalInputFormat extends FileInputFormat[GdalRasterInfo, Tile]
Instance Constructors
-
new
GdalInputFormat()
Value Members
-
final
def
!=(arg0: AnyRef): Boolean
-
final
def
!=(arg0: Any): Boolean
-
final
def
##(): Int
-
final
def
==(arg0: AnyRef): Boolean
-
final
def
==(arg0: Any): Boolean
-
def
addInputPathRecursively(arg0: List[FileStatus], arg1: FileSystem, arg2: Path, arg3: PathFilter): Unit
-
final
def
asInstanceOf[T0]: T0
-
def
clone(): AnyRef
-
def
computeSplitSize(arg0: Long, arg1: Long, arg2: Long): Long
-
def
createRecordReader(split: InputSplit, context: TaskAttemptContext): GdalRecordReader
-
final
def
eq(arg0: AnyRef): Boolean
-
def
equals(arg0: Any): Boolean
-
def
finalize(): Unit
-
def
getBlockIndex(arg0: Array[BlockLocation], arg1: Long): Int
-
final
def
getClass(): Class[_]
-
def
getFormatMinSplitSize(): Long
-
def
getSplits(arg0: JobContext): List[InputSplit]
-
def
hashCode(): Int
-
final
def
isInstanceOf[T0]: Boolean
-
def
isSplitable(context: JobContext, fileName: Path): Boolean
-
def
listStatus(arg0: JobContext): List[FileStatus]
-
def
makeSplit(arg0: Path, arg1: Long, arg2: Long, arg3: Array[String], arg4: Array[String]): FileSplit
-
def
makeSplit(arg0: Path, arg1: Long, arg2: Long, arg3: Array[String]): FileSplit
-
final
def
ne(arg0: AnyRef): Boolean
-
final
def
notify(): Unit
-
final
def
notifyAll(): Unit
-
final
def
synchronized[T0](arg0: ⇒ T0): T0
-
def
toString(): String
-
final
def
wait(): Unit
-
final
def
wait(arg0: Long, arg1: Int): Unit
-
final
def
wait(arg0: Long): Unit
Inherited from AnyRef
Inherited from Any
Uses GDAL to attempt to read a raster file.
GDAL only supports reading files from local filesystem. In order for this InputFormat to work it must copy the entire input file to hadoop.tmp.dir, if it is not already on a local file system, and invoke GDAL JNI bindings.
If there the .so/.dylib files are not found in the class path, this will crash gloriously.
Also note that all the blocks need to be shuffled to a single machine for each file. If the file being ingested is much larger than HDFS block size this will be very inefficient.