org.das2.qds.util

Class AsciiHeadersParser

java.lang.Object

org.das2.qds.util.AsciiHeadersParser

public class AsciiHeadersParser
extends java.lang.Object

This additional support for parsing ascii data looks at the comment block in an ASCII file for a structured set of Dataset tags further describing the data. This is loosely formatted JSON that describes each column of the data file more abstractly than the AsciiParser. This is based on QDataSet metadata tags, as much as possible.

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	AsciiHeadersParser.BundleDescriptor

Field Summary

Fields

Modifier and Type	Field and Description
static java.lang.String	PROP_DIMENSION property for dimension of the data defining rank and qube dims.
static java.lang.String	PROP_ELEMENT_LABELS Human-readable label for each column of the parameter.
static java.lang.String	PROP_ELEMENT_NAMES NAME identifier to assign to each column of the parameter.

Constructor Summary

Constructors

Constructor and Description
AsciiHeadersParser()

Method Summary

Modifier and Type	Method and Description
static QDataSet	getInlineDataSet(QDataSet bds, java.lang.String name) allow inline dataset to be retrieved.
static java.lang.String[]	getInlineDataSetNames(QDataSet bds) return the list of inline dataset names.
static AsciiHeadersParser.BundleDescriptor	parseMetadata(java.lang.String header, java.lang.String[] columns, java.lang.String[] columnLabels) attempt to parse the JSON metadata stored in the header.
static AsciiHeadersParser.BundleDescriptor	parseMetadataHapi(JSONObject doc)
protected java.lang.String	prep(java.lang.String s) Preprocess the string to make more valid JSON.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

PROP_DIMENSION

public static final java.lang.String PROP_DIMENSION

property for dimension of the data defining rank and qube dims. For example, "[]" (default} "[1]" scalar--alternate form, and [] is preferred. "[3]" (three element vector) "[20,30]" (qube of 60 elements).

See Also:

Constant Field Values

PROP_ELEMENT_NAMES

public static final java.lang.String PROP_ELEMENT_NAMES

NAME identifier to assign to each column of the parameter. These should follow QDataSet NAME rules. This is always a 1-D array.

See Also:

Constant Field Values

PROP_ELEMENT_LABELS

public static final java.lang.String PROP_ELEMENT_LABELS

Human-readable label for each column of the parameter.

See Also:

Constant Field Values

Constructor Detail

AsciiHeadersParser

public AsciiHeadersParser()

Method Detail

prep

protected java.lang.String prep(java.lang.String s)

Preprocess the string to make more valid JSON. 1. pop off comment character (#) from line. 2. add leading and trailing braces (}) if the first char is not an opening brace. 3. add implicit comma at line breaks unless the next line starts with comma or closing bracket (]). 4. closing brace closes JSON. Note the Java JSON parser used is already pretty loose: 1. strings needn't be quoted if they don't contain whitespace. a. I like this for structure tag names "UNITS:" b. I'd discourage this for string values, because they might be interpreted as numbers. UNITS:"1e2 nT" 2. equals (and other characters) can be used instead of COLONs. UNITS="1e2 nT" TODO: We see that it's misguided to have all this preprocessing done, since it limits who can process these headers. Many if all of the features here are going to be removed.

Parameters:

s -

Returns:

parseMetadataHapi

public static AsciiHeadersParser.BundleDescriptor parseMetadataHapi(JSONObject doc)
                                                             throws JSONException,
                                                                    java.text.ParseException

Throws:

JSONException

java.text.ParseException

parseMetadata

public static AsciiHeadersParser.BundleDescriptor parseMetadata(java.lang.String header,
                                                                java.lang.String[] columns,
                                                                java.lang.String[] columnLabels)
                                                         throws java.text.ParseException

attempt to parse the JSON metadata stored in the header. The header lines must be prefixed with hash (#). Loosely formatted text is converted into nicely-formatted JSON and then parsed with a JSON parser. Note the Java JSON parser itself is pretty loose, for example allowing 1-word strings to go without quote delimiters. The scheme for this header is either:

• "rich ascii" or http://autoplot.org/richAscii
• "hapi" https://github.com/hapi-server/data-specification/blob/master/hapi-dev/HAPI-data-access-spec-dev.md#info

Parameters:

header - the JSON header

columns - identifiers for each column

columnLabels - labels for each column

Returns:

the BundleDescriptor

Throws:

java.text.ParseException

See Also:

http://autoplot.org/richAscii, https://github.com/hapi-server/data-specification/blob/master/hapi-dev/HAPI-data-access-spec-dev.md#info

getInlineDataSet

public static QDataSet getInlineDataSet(QDataSet bds,
                                        java.lang.String name)

allow inline dataset to be retrieved.

Parameters:

bds - a BundleDescriptor, from BUNDLE_1. This must have been created by this code.

name - the name of the inline dataset

Returns:

the dataset, or null if the dataset is not found.

Throws:

java.lang.IllegalArgumentException - if the dataset not a BundleDescriptor.

getInlineDataSetNames

public static java.lang.String[] getInlineDataSetNames(QDataSet bds)

return the list of inline dataset names. This was probably used during development.

Parameters:

bds - bundle dataset descriptor, though only BundleDescriptor is supported.

Returns:

the inline dataset names.