GeoTools学习笔记-----ShapeFile文件说明----头文件的说明DbaseFileHeader类

 

DBF的文件格式：

    

文件头DbaseFileHeader类

记录1Record

记录2

记录3

…

…

…

记录n

BDFHeader的详细格式：

 

在文件中的位置	内容	说明
0	1个字节	表示当前的版本信息
1－3	3个字节	表示最近的更新日期，按照YYMMDD格式。
4－7	1个32位数	文件中的记录条数。
8－9	1个16位数	文件头中的字节数。
10－11	1个16位数	一条记录中的字节长度。
12－13	2个字节	保留字节，用于以后添加新的说明性信息时使用，这里用0来填写。
14	1个字节	表示未完成的操作。
15	1个字节	dBASE IV编密码标记。
16－27	12个字节	保留字节，用于多用户处理时使用。
28	1个字节	DBF文件的MDX标识。在创建一个DBF 表时 ，如果使用了MDX 格式的索引文件，那么 DBF 表的表头中的这个字节就自动被设置了一个标志，当你下次试图重新打开这个DBF表的时候，数据引擎会自动识别这个标志，如果此标志为真，则数据引擎将试图打开相应的MDX 文件。
29	1个字节	Language driver ID.
30－31	2个字节	保留字节，用于以后添加新的说明性信息时使用，这里用0来填写。
32－X	（n*32）个字节	记录项信息描述数组。n表示记录项的个数。这个数组的结构在表2.8中有详细的解释。
X＋1	1个字节	作为记录项终止标识。

DBF文件头中Field的详细格式：

类文件中的内部类DbaseField类

以及fields的数组

 

位置	内容	说明
0－10	11个字节	记录项名称，是ASCII码值。
11	1个字节	记录项的数据类型，是ASCII码值。（B、C、D、G、L、M和N，具体的解释见表2.9）。
12－15	4个字节	保留字节，用于以后添加新的说明性信息时使用，这里用0来填写。
16	1个字节	记录项长度，二进制型。
17	1个字节	记录项的精度，二进制型。
18－19	2个字节	保留字节，用于以后添加新的说明性信息时使用，这里用0来填写。
20	1个字节	工作区ID。
21－30	10个字节	保留字节，用于以后添加新的说明性信息时使用，这里用0来填写。
31	1个字节	MDX标识。如果存在一个MDX 格式的索引文件，那么这个记录项为真，否则为空。

  DBF文件中的数据类型：

FieldType

代码	数据类型	允许输入的数据
B	二进制型	各种字符。
C	字符型	各种字符。
D	日期型	用于区分年、月、日的数字和一个字符，内部存储按照YYYYMMDD格式。
G	(General or OLE)	各种字符。
N	数值型(Numeric)	- . 0 1 2 3 4 5 6 7 8 9
L	逻辑型（Logical）	? Y y N n T t F f (? 表示没有初始化)。
M	(Memo)	各种字符。

 

举例说明：
假如文件中有10条记录，每个记录有4个字段，4个字段的长度分别为：12、14、16、18，那么文件的详细格式如下：
文件头占32+32*4+2个字节：
前32个字节是文件头中的基本信息，32*4个字节是记录项（即字段的定义信息），最后两个字节分别是16进制的0D 和20，0D代表上文表格中说明的记录项终止标识。20代表一个空格。
数据信息占(12+14+16+18)*10+1个字节：
12+14+16+18个字节代表一条记录，共10条记录。最后一个字节是数据的终止表示通常是16进制表示的1A
补充说明：
由于上述文件格式的定义，决定了字段名称不能超过11个字节（或者5个中文字符），字符类型的字段，数据最大不能超过255个字节。在从文本、Excel、大型数据库导出数据到DBF格式文件时，一定要考虑到这些长度限制。 

 

以下是类文件的详细描述

/* * GeoTools - The Open Source Java GIS Toolkit * http://geotools.org * * (C) 2002-2008, Open Source Geospatial Foundation (OSGeo) * * This library is free software; you can redistribute it and/or * modify it under the terms of the GNU Lesser General Public * License as published by the Free Software Foundation; * version 2.1 of the License. * * This library is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU * Lesser General Public License for more details. * * This file is based on an origional contained in the GISToolkit project: * http://gistoolkit.sourceforge.net/ * -- * 张恒才程序声明 * * GPL声明 */ package org.geotools.data.shapefile.dbf; import java.io.EOFException; import java.io.IOException; import java.nio.ByteBuffer; import java.nio.ByteOrder; import java.nio.channels.ReadableByteChannel; import java.nio.channels.WritableByteChannel; import java.util.ArrayList; import java.util.Calendar; import java.util.Date; import java.util.List; import java.util.logging.Level; import java.util.logging.Logger; import org.geotools.resources.NIOUtilities; /** * * Class to represent the header of a Dbase III file. Creation date: (5/15/2001 * 5:15:30 PM) * * @source $URL: * http://svn.geotools.org/geotools/trunk/gt/modules/plugin/shapefile/src/main/java/org/geotools/data/shapefile/dbf/DbaseFileHeader.java $ *----------------------------- * * 该类代表ShapeFile的DBF的header文件 * 采用类的观点来进行表达 * 该类中Fields数组具有重要的作用 * * -- * 需要详细区分几个概念 * （1）什么是Table * (2)Field DbaseField类代表什么 * (3)Record * (4)Column row * (5)Header */ public class DbaseFileHeader { // Constant for the size of a record //该记录的常量表达头文件的大小 private static final int FILE_DESCRIPTOR_SIZE = 32; // type of the file, must be 03h //唯一的表示该文件的魔数 private static final byte MAGIC = 0x03; private static final int MINIMUM_HEADER = 33; // Date the file was last updated. //文件的更新日期 private Date date = new Date(); //记录的个数 private int recordCnt = 0; //字段的个数 private int fieldCnt = 0; // set this to a default length of 1, which is enough for one "space" // character which signifies an empty record //记录 的长度 private int recordLength = 1; // set this to a flagged value so if no fields are added before the write, // we know to adjust the headerLength to MINIMUM_HEADER private int headerLength = -1; private int largestFieldSize = 0; //日志记录 private Logger logger = org.geotools.util.logging.Logging .getLogger("org.geotools.data.shapefile"); /** * Class for holding the information assicated with a record. * 一条记录的Field的描述 */ class DbaseField { // Field Name //字段名字 String fieldName; // Field Type (C N L D or M) //字段的类型 char fieldType; // Field Data Address offset from the start of the record. //一条记录开始的偏移地址 int fieldDataAddress; // Length of the data in bytes //字段的长度 int fieldLength; // Field decimal count in Binary, indicating where the decimal is int decimalCount; } // collection of header records. // lets start out with a zero-length array, just in case //field 数组表示 //-- //重要的数组变量 private DbaseField[] fields = new DbaseField[0]; //读写文件 private void read(ByteBuffer buffer, ReadableByteChannel channel) throws IOException { while (buffer.remaining() > 0) { if (channel.read(buffer) == -1) { throw new EOFException("Premature end of file"); } } } /** * Determine the most appropriate Java Class for representing the data in * the field. * * <PRE> * All packages are java.lang unless otherwise specified. * C (Character) -> String * N (Numeric) -> Integer or Double (depends on field's decimal count) * F (Floating) -> Double * L (Logical) -> Boolean * D (Date) -> java.util.Date * Unknown -> String * </PRE> * * @param i * The index of the field, from 0 to * <CODE>getNumFields() - 1</CODE> . * @return A Class which closely represents the dbase field type. */ public Class getFieldClass(int i) { Class typeClass = null; switch (fields[i].fieldType) { case 'C': typeClass = String.class; break; case 'N': if (fields[i].decimalCount == 0) { if (fields[i].fieldLength < 10) { typeClass = Integer.class; } else { typeClass = Long.class; } } else { typeClass = Double.class; } break; case 'F': typeClass = Double.class; break; case 'L': typeClass = Boolean.class; break; case 'D': typeClass = Date.class; break; default: typeClass = String.class; break; } return typeClass; } /** * Add a column to this DbaseFileHeader. The type is one of (C N L or D) * character, number, logical(true/false), or date. The Field length is the * total length in bytes reserved for this column. The decimal count only * applies to numbers(N), and floating point values (F), and refers to the * number of characters to reserve after the decimal point. <B>Don't expect * miracles from this...</B> * * <PRE> * Field Type MaxLength * ---------- --------- * C 254 * D 8 * F 20 * N 18 * </PRE> * * @param inFieldName * The name of the new field, must be less than 10 characters * or it gets truncated. * @param inFieldType * A character representing the dBase field, ( see above ). * Case insensitive. * @param inFieldLength * The length of the field, in bytes ( see above ) * @param inDecimalCount * For numeric fields, the number of decimal places to track. * @throws DbaseFileException * If the type is not recognized. *---------------------- *增加一个字段（也就是增加一列） */ public void addColumn(String inFieldName, char inFieldType, int inFieldLength, int inDecimalCount) throws DbaseFileException { //容错检验 if (inFieldLength <= 0) { throw new DbaseFileException("field length <= 0"); } //第一个字段 if (fields == null) { fields = new DbaseField[0]; } //记录偏移地址 int tempLength = 1; // the length is used for the offset, and there is a // * for deleted as the first byte //临时变量 DbaseField[] tempFieldDescriptors = new DbaseField[fields.length + 1]; for (int i = 0; i < fields.length; i++) { fields[i].fieldDataAddress = tempLength; tempLength = tempLength + fields[i].fieldLength; tempFieldDescriptors[i] = fields[i]; } //给新的字段赋值 tempFieldDescriptors[fields.length] = new DbaseField(); tempFieldDescriptors[fields.length].fieldLength = inFieldLength; tempFieldDescriptors[fields.length].decimalCount = inDecimalCount; tempFieldDescriptors[fields.length].fieldDataAddress = tempLength; // set the field name String tempFieldName = inFieldName; if (tempFieldName == null) { tempFieldName = "NoName"; } // Fix for GEOT-42, ArcExplorer will not handle field names > 10 chars // Sorry folks. //文件名不能超过10个字符 // 超过后去前10个字符并且给出警告 if (tempFieldName.length() > 10) { tempFieldName = tempFieldName.substring(0, 10); if (logger.isLoggable(Level.WARNING)) { logger.warning("FieldName " + inFieldName + " is longer than 10 characters, truncating to " + tempFieldName); } } tempFieldDescriptors[fields.length].fieldName = tempFieldName; // the field type if ((inFieldType == 'C') || (inFieldType == 'c')) { tempFieldDescriptors[fields.length].fieldType = 'C'; if (inFieldLength > 254) { if (logger.isLoggable(Level.FINE)) { logger .fine("Field Length for " + inFieldName + " set to " + inFieldLength + " Which is longer than 254, not consistent with dbase III"); } } } else if ((inFieldType == 'S') || (inFieldType == 's')) { tempFieldDescriptors[fields.length].fieldType = 'C'; if (logger.isLoggable(Level.WARNING)) { logger .warning("Field type for " + inFieldName + " set to S which is flat out wrong people!, I am setting this to C, in the hopes you meant character."); } if (inFieldLength > 254) { if (logger.isLoggable(Level.FINE)) { logger .fine("Field Length for " + inFieldName + " set to " + inFieldLength + " Which is longer than 254, not consistent with dbase III"); } } tempFieldDescriptors[fields.length].fieldLength = 8; } else if ((inFieldType == 'D') || (inFieldType == 'd')) { tempFieldDescriptors[fields.length].fieldType = 'D'; if (inFieldLength != 8) { if (logger.isLoggable(Level.FINE)) { logger.fine("Field Length for " + inFieldName + " set to " + inFieldLength + " Setting to 8 digets YYYYMMDD"); } } tempFieldDescriptors[fields.length].fieldLength = 8; } else if ((inFieldType == 'F') || (inFieldType == 'f')) { tempFieldDescriptors[fields.length].fieldType = 'F'; if (inFieldLength > 20) { if (logger.isLoggable(Level.FINE)) { logger .fine("Field Length for " + inFieldName + " set to " + inFieldLength + " Preserving length, but should be set to Max of 20 not valid for dbase IV, and UP specification, not present in dbaseIII."); } } } else if ((inFieldType == 'N') || (inFieldType == 'n')) { tempFieldDescriptors[fields.length].fieldType = 'N'; if (inFieldLength > 18) { if (logger.isLoggable(Level.FINE)) { logger .fine("Field Length for " + inFieldName + " set to " + inFieldLength + " Preserving length, but should be set to Max of 18 for dbase III specification."); } } if (inDecimalCount < 0) { if (logger.isLoggable(Level.FINE)) { logger.fine("Field Decimal Position for " + inFieldName + " set to " + inDecimalCount + " Setting to 0 no decimal data will be saved."); } tempFieldDescriptors[fields.length].decimalCount = 0; } if (inDecimalCount > inFieldLength - 1) { if (logger.isLoggable(Level.WARNING)) { logger.warning("Field Decimal Position for " + inFieldName + " set to " + inDecimalCount + " Setting to " + (inFieldLength - 1) + " no non decimal data will be saved."); } tempFieldDescriptors[fields.length].decimalCount = inFieldLength - 1; } } else if ((inFieldType == 'L') || (inFieldType == 'l')) { tempFieldDescriptors[fields.length].fieldType = 'L'; if (inFieldLength != 1) { if (logger.isLoggable(Level.FINE)) { logger.fine("Field Length for " + inFieldName + " set to " + inFieldLength + " Setting to length of 1 for logical fields."); } } tempFieldDescriptors[fields.length].fieldLength = 1; } else { throw new DbaseFileException("Undefined field type " + inFieldType + " For column " + inFieldName); } // the length of a record tempLength = tempLength + tempFieldDescriptors[fields.length].fieldLength; // set the new fields. //更新新的值 fields = tempFieldDescriptors; fieldCnt = fields.length; headerLength = MINIMUM_HEADER + 32 * fields.length; recordLength = tempLength;//字段的个数 } /** * Remove a column from this DbaseFileHeader. * 移除 * * @todo This is really ugly, don't know who wrote it, but it needs fixin... * @param inFieldName * The name of the field, will ignore case and trim. * @return index of the removed column, -1 if no found * * ---- * 根据列的名字移除一列 * 移除一个字段 */ public int removeColumn(String inFieldName) { int retCol = -1; int tempLength = 1; DbaseField[] tempFieldDescriptors = new DbaseField[fields.length - 1]; for (int i = 0, j = 0; i < fields.length; i++) { if (!inFieldName.equalsIgnoreCase(fields[i].fieldName.trim())) { // if this is the last field and we still haven't found the // named field if (i == j && i == fields.length - 1) { System.err.println("Could not find a field named '" + inFieldName + "' for removal"); return retCol; } tempFieldDescriptors[j] = fields[i]; tempFieldDescriptors[j].fieldDataAddress = tempLength; tempLength += tempFieldDescriptors[j].fieldLength; // only increment j on non-matching fields j++; } else { retCol = i; } } // set the new fields. fields = tempFieldDescriptors; headerLength = 33 + 32 * fields.length; recordLength = tempLength; return retCol; } // Retrieve the length of the field at the given index /** * Returns the field length in bytes. * * @param inIndex * The field index. * @return The length in bytes. * --------- * 返回字段的长度 */ public int getFieldLength(int inIndex) { return fields[inIndex].fieldLength; } // Retrieve the location of the decimal point within the field. /** * Get the decimal count of this field. * * @param inIndex * The field index. * @return The decimal count. * ------ * 返回字段的位数 */ public int getFieldDecimalCount(int inIndex) { return fields[inIndex].decimalCount; } // Retrieve the Name of the field at the given index /** * Get the field name. * * @param inIndex * The field index. * @return The name of the field. * * * ------------ * 返回字段的名字 */ public String getFieldName(int inIndex) { return fields[inIndex].fieldName; } // Retrieve the type of field at the given index /** * Get the character class of the field. * * @param inIndex * The field index. * @return The dbase character representing this field. * ——------- * 返回字段的类型 */ public char getFieldType(int inIndex) { return fields[inIndex].fieldType; } /** * Get the date this file was last updated. * * @return The Date last modified. * -------- * 返回字段的更新日期 */ public Date getLastUpdateDate() { return date; } /** * Return the number of fields in the records. * * @return The number of fields in this table. * * -------- * 返回字段的个数 */ public int getNumFields() { return fields.length; } /** * Return the number of records in the file * * @return The number of records in this table. * * * ---返回记录的个数 */ public int getNumRecords() { return recordCnt; } /** * Get the length of the records in bytes. * * @return The number of bytes per record. * ------- * 返回记录的长度 */ public int getRecordLength() { return recordLength; } /** * Get the length of the header * * @return The length of the header in bytes. * ------- * 返回头文件的长度 */ public int getHeaderLength() { return headerLength; } /** * Read the header data from the DBF file. * * @param channel * A readable byte channel. If you have an InputStream you * need to use, you can call * java.nio.Channels.getChannel(InputStream in). * @throws IOException * If errors occur while reading. * ------------------- * * 从dbf文件读取文件头文件的说明 */ public void readHeader(ReadableByteChannel channel) throws IOException { // we'll read in chunks of 1K //1024字节 ByteBuffer in = ByteBuffer.allocateDirect(1024); // do this or GO CRAZY // ByteBuffers come preset to BIG_ENDIAN ! in.order(ByteOrder.LITTLE_ENDIAN); // only want to read first 10 bytes... in.limit(10); read(in, channel); in.position(0); // type of file. byte magic = in.get(); if (magic != MAGIC) { throw new IOException("Unsupported DBF file Type " + Integer.toHexString(magic)); } // parse the update date information. int tempUpdateYear = in.get(); int tempUpdateMonth = in.get(); int tempUpdateDay = in.get(); // ouch Y2K uncompliant if (tempUpdateYear > 90) { tempUpdateYear = tempUpdateYear + 1900; } else { tempUpdateYear = tempUpdateYear + 2000; } Calendar c = Calendar.getInstance(); c.set(Calendar.YEAR, tempUpdateYear); c.set(Calendar.MONTH, tempUpdateMonth - 1); c.set(Calendar.DATE, tempUpdateDay); date = c.getTime(); // read the number of records. recordCnt = in.getInt(); // read the length of the header structure. // ahhh.. unsigned little-endian shorts // mask out the byte and or it with shifted 2nd byte headerLength = (in.get() & 0xff) | ((in.get() & 0xff) << 8); // if the header is bigger than our 1K, reallocate if (headerLength > in.capacity()) { NIOUtilities.clean(in); in = ByteBuffer.allocateDirect(headerLength - 10); } in.limit(headerLength - 10); in.position(0); read(in, channel); in.position(0); // read the length of a record // ahhh.. unsigned little-endian shorts recordLength = (in.get() & 0xff) | ((in.get() & 0xff) << 8); // skip / skip thesreserved bytes in the header. in.position(in.position() + 20); // calculate the number of Fields in the header fieldCnt = (headerLength - FILE_DESCRIPTOR_SIZE - 1) / FILE_DESCRIPTOR_SIZE; // read all of the header records List lfields = new ArrayList(); for (int i = 0; i < fieldCnt; i++) { DbaseField field = new DbaseField(); // read the field name byte[] buffer = new byte[11]; in.get(buffer); String name = new String(buffer); int nullPoint = name.indexOf(0); if (nullPoint != -1) { name = name.substring(0, nullPoint); } field.fieldName = name.trim(); // read the field type field.fieldType = (char) in.get(); // read the field data address, offset from the start of the record. field.fieldDataAddress = in.getInt(); // read the field length in bytes int length = (int) in.get(); if (length < 0) { length = length + 256; } field.fieldLength = length; if (length > largestFieldSize) { largestFieldSize = length; } // read the field decimal count in bytes field.decimalCount = (int) in.get(); // rreservedvededved bytes. // in.skipBytes(14); in.position(in.position() + 14); // some broken shapefiles have 0-length attributes. The reference // implementation // (ArcExplorer 2.0, built with MapObjects) just ignores them. if (field.fieldLength > 0) { lfields.add(field); } } // Last byte is a marker for the end of the field definitions. // in.skipBytes(1); in.position(in.position() + 1); NIOUtilities.clean(in); fields = new DbaseField[lfields.size()]; fields = (DbaseField[]) lfields.toArray(fields); } /** * Get the largest field size of this table. * * @return The largt field size iiin bytes. * --------------- * 返回该表中最大的字段的大小 */ public int getLargestFieldSize() { return largestFieldSize; } /** * Set the number of records in the file * * @param inNumRecords * The number of records. * ----------- * 设置记录的个数 */ public void setNumRecords(int inNumRecords) { recordCnt = inNumRecords; } /** * Write the header data to the DBF file. * * @param out * A channel to write to. If you have an OutputStream you can * obtain the correct channel by using * java.nio.Channels.newChannel(OutputStream out). * @throws IOException * If errors occur. * ----------------- * 写一个头文件 */ public void writeHeader(WritableByteChannel out) throws IOException { // take care of the annoying case where no records have been added... if (headerLength == -1) { headerLength = MINIMUM_HEADER; } ByteBuffer buffer = ByteBuffer.allocateDirect(headerLength); buffer.order(ByteOrder.LITTLE_ENDIAN); // write the output file type. buffer.put((byte) MAGIC); // write the date stuff Calendar c = Calendar.getInstance(); c.setTime(new Date()); buffer.put((byte) (c.get(Calendar.YEAR) % 100)); buffer.put((byte) (c.get(Calendar.MONTH) + 1)); buffer.put((byte) (c.get(Calendar.DAY_OF_MONTH))); // write the number of records in the datafile. buffer.putInt(recordCnt); // write the length of the header structure. buffer.putShort((short) headerLength); // write the length of a record buffer.putShort((short) recordLength); // // write the reserved bytes in the header // for (int i=0; i<20; i++) out.writeByteLE(0); buffer.position(buffer.position() + 20); // write all of the header records int tempOffset = 0; for (int i = 0; i < fields.length; i++) { // write the field name for (int j = 0; j < 11; j++) { if (fields[i].fieldName.length() > j) { buffer.put((byte) fields[i].fieldName.charAt(j)); } else { buffer.put((byte) 0); } } // write the field type buffer.put((byte) fields[i].fieldType); // // write the field data address, offset from the start of the // record. buffer.putInt(tempOffset); tempOffset += fields[i].fieldLength; // write the length of the field. buffer.put((byte) fields[i].fieldLength); // write the decimal count. buffer.put((byte) fields[i].decimalCount); // write the reserved bytes. // for (in j=0; jj<14; j++) out.writeByteLE(0); buffer.position(buffer.position() + 14); } // write the end of the field definitions marker buffer.put((byte) 0x0D); buffer.position(0); int r = buffer.remaining(); while ((r -= out.write(buffer)) > 0) { ; // do nothing } NIOUtilities.clean(buffer); } /** * Get a simple representation of this header. * * @return A String representing the state of the header. * -- * 返回该头文件包含的每个字段的描述 * Header包含字段的描述 * *程序的输出： * （1）日期： * （2）记录的个数：RecordsCNt * （3）字段的个数：FieldCnt * （4）迭代输出每个字段的描述信息 */ public String toString() { StringBuffer fs = new StringBuffer(); for (int i = 0, ii = fields.length; i < ii; i++) { DbaseField f = fields[i]; fs.append(f.fieldName + " " + f.fieldType + " " + f.fieldLength + " " + f.decimalCount + " " + f.fieldDataAddress + "/n"); } return "DB3 Header/n" + "Date : " + date + "/n" + "Records : " + recordCnt + "/n" + "Fields : " + fieldCnt + "/n" + fs; } }