loci.poi.hssf.record

Class SSTRecord

java.lang.Object

loci.poi.hssf.record.Record

loci.poi.hssf.record.SSTRecord

public class SSTRecord
extends Record

Title: Static String Table Record

Description: This holds all the strings for LabelSSTRecords.

REFERENCE: PG 389 Microsoft Excel 97 Developer's Kit (ISBN: 1-57231-498-2)

Author:

Andrew C. Oliver (acoliver at apache dot org), Marc Johnson (mjohnson at apache dot org), Glen Stampoultzis (glens at apache.org)

See Also:

LabelSSTRecord, ContinueRecord

Field Summary

Fields

Modifier and Type	Field and Description
(package private) int[]	bucketAbsoluteOffsets Offsets from the beginning of the SST record (even across continuations)
(package private) int[]	bucketRelativeOffsets Offsets relative the start of the current SST or continue record
private SSTDeserializer	deserializer
private static UnicodeString	EMPTY_STRING
private int	field_1_num_strings union of strings in the SST and EXTSST
private int	field_2_num_unique_strings according to docs ONLY SST
private IntMapper	field_3_strings
(package private) static int	MAX_DATA_SPACE how much data can we stuff into an SST record? That would be _max minus the standard SST record overhead
(package private) static int	MAX_RECORD_SIZE how big can an SST record be? As big as any record can be: 8228 bytes
static short	sid
(package private) static int	SST_RECORD_OVERHEAD SST overhead: the standard record overhead, plus the number of strings and the number of unique strings -- two ints
(package private) static int	STD_RECORD_OVERHEAD standard record overhead: two shorts (record id plus data space size)
(package private) static int	STRING_MINIMAL_OVERHEAD overhead for each string includes the string's character count (a short) and the flag describing its characteristics (a byte)

Constructor Summary

Constructors

Constructor and Description
SSTRecord() default constructor
SSTRecord(RecordInputStream in) Constructs an SST record and sets its fields appropriately.

Method Summary

Methods

Modifier and Type	Method and Description
int	addString(UnicodeString string) Add a string.
int	calcExtSSTRecordSize() Calculates the size in bytes of the EXTSST record as it would be if the record was serialized.
(package private) int	countStrings()
ExtSSTRecord	createExtSSTRecord(int sstOffset) Creates an extended string record based on the current contents of the current SST record.
boolean	equals(Object o)
protected void	fillFields(RecordInputStream in) Fill the fields from the data
(package private) SSTDeserializer	getDeserializer()
int	getNumStrings()
int	getNumUniqueStrings()
int	getRecordSize() gives the current serialized size of the record.
short	getSid() return the non static version of the id for this record.
UnicodeString	getString(int id) Get a particular string by its index
(package private) Iterator	getStrings()
int	hashCode()
boolean	isString16bit(int id)
int	serialize(int offset, byte[] data) called by the class that is responsible for writing this sucker.
void	setNumStrings(int count) USE THIS METHOD AT YOUR OWN PERIL: THE addString METHODS MANIPULATE THE NUMBER OF STRINGS AS A SIDE EFFECT; YOUR ATTEMPTS AT MANIPULATING THE STRING COUNT IS LIKELY TO BE VERY WRONG AND WILL RESULT IN BAD BEHAVIOR WHEN THIS RECORD IS WRITTEN OUT AND ANOTHER PROCESS ATTEMPTS TO READ THE RECORD
void	setNumUniqueStrings(int count) USE THIS METHOD AT YOUR OWN PERIL: THE addString METHODS MANIPULATE THE NUMBER OF UNIQUE STRINGS AS A SIDE EFFECT; YOUR ATTEMPTS AT MANIPULATING THE UNIQUE STRING COUNT IS LIKELY TO BE VERY WRONG AND WILL RESULT IN BAD BEHAVIOR WHEN THIS RECORD IS WRITTEN OUT AND ANOTHER PROCESS ATTEMPTS TO READ THE RECORD
String	toString() Return a debugging string representation
protected void	validateSid(short id) validate SID

Methods inherited from class loci.poi.hssf.record.Record

clone, isInValueSection, isValue, serialize

Methods inherited from class java.lang.Object

finalize, getClass, notify, notifyAll, wait, wait, wait

Field Detail

EMPTY_STRING

private static UnicodeString EMPTY_STRING

MAX_RECORD_SIZE

static final int MAX_RECORD_SIZE

how big can an SST record be? As big as any record can be: 8228 bytes

See Also:

Constant Field Values

STD_RECORD_OVERHEAD

static final int STD_RECORD_OVERHEAD

standard record overhead: two shorts (record id plus data space size)

See Also:

Constant Field Values

SST_RECORD_OVERHEAD

static final int SST_RECORD_OVERHEAD

SST overhead: the standard record overhead, plus the number of strings and the number of unique strings -- two ints

See Also:

Constant Field Values

MAX_DATA_SPACE

static final int MAX_DATA_SPACE

how much data can we stuff into an SST record? That would be _max minus the standard SST record overhead

See Also:

Constant Field Values

STRING_MINIMAL_OVERHEAD

static final int STRING_MINIMAL_OVERHEAD

overhead for each string includes the string's character count (a short) and the flag describing its characteristics (a byte)

See Also:

Constant Field Values

sid

public static final short sid

See Also:

Constant Field Values

field_1_num_strings

private int field_1_num_strings

union of strings in the SST and EXTSST

field_2_num_unique_strings

private int field_2_num_unique_strings

according to docs ONLY SST

field_3_strings

private IntMapper field_3_strings

deserializer

private SSTDeserializer deserializer

bucketAbsoluteOffsets

int[] bucketAbsoluteOffsets

Offsets from the beginning of the SST record (even across continuations)

bucketRelativeOffsets

int[] bucketRelativeOffsets

Offsets relative the start of the current SST or continue record

Constructor Detail

SSTRecord

public SSTRecord()

default constructor

SSTRecord

public SSTRecord(RecordInputStream in)

Constructs an SST record and sets its fields appropriately.

Parameters:

in - the RecordInputstream to read the record from

Method Detail

addString

public int addString(UnicodeString string)

Add a string.

Parameters:

string - string to be added

Returns:

the index of that string in the table

getNumStrings

public int getNumStrings()

Returns:

number of strings

getNumUniqueStrings

public int getNumUniqueStrings()

Returns:

number of unique strings

setNumStrings

public void setNumStrings(int count)

USE THIS METHOD AT YOUR OWN PERIL: THE addString METHODS MANIPULATE THE NUMBER OF STRINGS AS A SIDE EFFECT; YOUR ATTEMPTS AT MANIPULATING THE STRING COUNT IS LIKELY TO BE VERY WRONG AND WILL RESULT IN BAD BEHAVIOR WHEN THIS RECORD IS WRITTEN OUT AND ANOTHER PROCESS ATTEMPTS TO READ THE RECORD

Parameters:

count - number of strings

setNumUniqueStrings

public void setNumUniqueStrings(int count)

USE THIS METHOD AT YOUR OWN PERIL: THE addString METHODS MANIPULATE THE NUMBER OF UNIQUE STRINGS AS A SIDE EFFECT; YOUR ATTEMPTS AT MANIPULATING THE UNIQUE STRING COUNT IS LIKELY TO BE VERY WRONG AND WILL RESULT IN BAD BEHAVIOR WHEN THIS RECORD IS WRITTEN OUT AND ANOTHER PROCESS ATTEMPTS TO READ THE RECORD

Parameters:

count - number of strings

getString

public UnicodeString getString(int id)

Get a particular string by its index

Parameters:

id - index into the array of strings

Returns:

the desired string

isString16bit

public boolean isString16bit(int id)

toString

public String toString()

Return a debugging string representation

Overrides:

toString in class Record

Returns:

string representation

getSid

public short getSid()

Description copied from class: Record

return the non static version of the id for this record.

Specified by:

getSid in class Record

Returns:

sid

hashCode

public int hashCode()

Overrides:

hashCode in class Object

Returns:

hashcode

equals

public boolean equals(Object o)

Overrides:

equals in class Object

validateSid

protected void validateSid(short id)
                    throws RecordFormatException

validate SID

Specified by:

validateSid in class Record

Parameters:

id - the alleged SID

Throws:

RecordFormatException - if validation fails

fillFields

protected void fillFields(RecordInputStream in)

Fill the fields from the data

The data consists of sets of string data. This string data is arranged as follows:

 short  string_length;   // length of string data
 byte   string_flag;     // flag specifying special string
                         // handling
 short  run_count;       // optional count of formatting runs
 int    extend_length;   // optional extension length
 char[] string_data;     // string data, can be byte[] or
                         // short[] (length of array is
                         // string_length)
 int[]  formatting_runs; // optional formatting runs (length of
                         // array is run_count)
 byte[] extension;       // optional extension (length of array
                         // is extend_length)
 

The string_flag is bit mapped as follows:

Bit number	Meaning if 0	Meaning if 1
0	string_data is byte[]	string_data is short[]
1	Should always be 0	string_flag is defective
2	extension is not included	extension is included
3	formatting run data is not included	formatting run data is included
4	Should always be 0	string_flag is defective
5	Should always be 0	string_flag is defective
6	Should always be 0	string_flag is defective
7	Should always be 0	string_flag is defective

We can handle eating the overhead associated with bits 2 or 3 (or both) being set, but we have no idea what to do with the associated data. The UnicodeString class can handle the byte[] vs short[] nature of the actual string data

Specified by:

fillFields in class Record

Parameters:

in - the RecordInputstream to read the record from

getStrings

Iterator getStrings()

Returns:

an iterator of the strings we hold. All instances are UnicodeStrings

countStrings

int countStrings()

Returns:

count of the strings we hold.

serialize

public int serialize(int offset,
            byte[] data)

called by the class that is responsible for writing this sucker. Subclasses should implement this so that their data is passed back in a byte array.

Specified by:

serialize in class Record

Parameters:

offset - to begin writing at

data - byte array containing instance data

Returns:

size

getRecordSize

public int getRecordSize()

Description copied from class: Record

gives the current serialized size of the record. Should include the sid and reclength (4 bytes).

Overrides:

getRecordSize in class Record

getDeserializer

SSTDeserializer getDeserializer()

createExtSSTRecord

public ExtSSTRecord createExtSSTRecord(int sstOffset)

Creates an extended string record based on the current contents of the current SST record. The offset within the stream to the SST record is required because the extended string record points directly to the strings in the SST record.

NOTE: THIS FUNCTION MUST ONLY BE CALLED AFTER THE SST RECORD HAS BEEN SERIALIZED.

Parameters:

sstOffset - The offset in the stream to the start of the SST record.

Returns:

The new SST record.

calcExtSSTRecordSize

public int calcExtSSTRecordSize()

Calculates the size in bytes of the EXTSST record as it would be if the record was serialized.

Returns:

The size of the ExtSST record in bytes.