轉http://www.cnblogs.com/cjingzm/archive/2012/03/04/2378900.html
近期研究了一下GIS開源庫shapeLib讀寫ArcGIS資料的API函式，先整理一下，將各個API的用法介紹一下。
分為兩個模組，shape API和DBF API，前者的讀取.shp檔案的空間幾何資訊，後者讀取.dbf檔案的屬性資訊。
Shape API：
Shape Types （shape型別）
shape檔案的型別定義如下：
#define SHPT_NULL 0 2D Shape Types (pre ArcView 3.x): #define SHPT_POINT 1 Points #define SHPT_ARC 3 Arcs (Polylines, possible in parts) #define SHPT_POLYGON 5 Polygons (possible in parts) #define SHPT_MULTIPOINT 8 MultiPoint (related points) 3D Shape Types (may include "measure" values for vertices): #define SHPT_POINTZ 11 #define SHPT_ARCZ 13 #define SHPT_POLYGONZ 15 #define SHPT_MULTIPOINTZ 18 2D + Measure Types: #define SHPT_POINTM 21 #define SHPT_ARCM 23 #define SHPT_POLYGONM 25 #define SHPT_MULTIPOINTM 28 Complex (TIN-like) with Z, and Measure: #define SHPT_MULTIPATCH 31
SHPObject （shape檔案中包含的要素物件）
typedef struct
{
int nSHPType; //Shape Type (SHPT_* - see list above) 要素型別
int nShapeId; //Shape Number (-1 is unknown/unassigned) ID
int nParts; //# of Parts (0 implies single part with no info) 要素有幾個部分組成
int *panPartStart; //Start Vertex of part 要素部分的起始點
int *panPartType; //Part Type (SHPP_RING if not SHPT_MULTIPATCH) 各個部分的型別
int nVertices; //Vertex list
double *padfX;
double *padfY;
double *padfZ; //(all zero if not provided)
double *padfM; //(all zero if not provided)
double dfXMin; //Bounds in X, Y, Z and M dimensions
double dfYMin;
double dfZMin;
double dfMMin;
double dfXMax;
double dfYMax;
double dfZMax;
double dfMMax;
} SHPObject;
SHPOpen()
SHPHandle SHPOpen( const char * pszShapeFile, const char * pszAccess ); pszShapeFile: //shape檔案的全路徑 pszAccess: //開啟方式 "rb" (read-only binary) and "rb+" (read/write binary)
//開啟shape檔案，操作完必須呼叫後面的關閉。
SHPGetInfo() //得到shape物件的資訊
void SHPGetInfo( SHPHandle hSHP, int * pnEntities, int * pnShapeType,
double * padfMinBound, double * padfMaxBound );
hSHP: // shape檔案的控制程式碼 The handle previously returned by SHPOpen()
or SHPCreate().
pnEntities: A pointer to an integer into which the number of
entities/structures should be placed. May be NULL.
pnShapetype: A pointer to an integer into which the shapetype
of this file should be placed. Shapefiles may contain
either SHPT_POINT, SHPT_ARC, SHPT_POLYGON or
SHPT_MULTIPOINT entities. This may be NULL.
padfMinBound: The X, Y, Z and M minimum values will be placed into
this four entry array. This may be NULL.
padfMaxBound: The X, Y, Z and M maximum values will be placed into
this four entry array. This may be NULL.
SHPReadObject() //讀取每條記錄中的資訊
SHPObject *SHPReadObject( SHPHandle hSHP, int iShape ); hSHP: The handle previously returned by SHPOpen() or SHPCreate(). iShape: The entity number of the shape to read. Entity numbers are between 0 and nEntities-1 (as returned by SHPGetInfo()).
SHPClose()
void SHPClose( SHPHandle hSHP );
SHPCreate() //建立一個shape檔案
SHPHandle SHPCreate( const char * pszShapeFile, int nShapeType ); pszShapeFile: //建立檔案的名稱 nShapeType: //型別，參加上面定義的型別
SHPCreateSimpleObject() //建立簡單要素物件，一般呼叫此方法
SHPObject *
SHPCreateSimpleObject( int nSHPType, int nVertices,
double *padfX, double * padfY, double *padfZ, );
nSHPType: The SHPT_ type of the object to be created, such
as SHPT_POINT, or SHPT_POLYGON.
nVertices: The number of vertices being passed in padfX,
padfY, and padfZ.
padfX: An array of nVertices X coordinates of the vertices
for this object.
padfY: An array of nVertices Y coordinates of the vertices
for this object.
padfZ: An array of nVertices Z coordinates of the vertices
for this object. This may be NULL in which case
they are all assumed to be zero.
SHPCreateObject() //建立複雜物件，引數中包含所有資訊
SHPObject *
SHPCreateObject( int nSHPType, int iShape,
int nParts, int * panPartStart, int * panPartType,
int nVertices, double *padfX, double * padfY,
double *padfZ, double *padfM );
nSHPType: The SHPT_ type of the object to be created, such
as SHPT_POINT, or SHPT_POLYGON.
iShape: The shapeid to be recorded with this shape.
nParts: The number of parts for this object. If this is
zero for ARC, or POLYGON type objects, a single
zero valued part will be created internally.
panPartStart: The list of zero based start vertices for the rings
(parts) in this object. The first should always be
zero. This may be NULL if nParts is 0.
panPartType: The type of each of the parts. This is only meaningful
for MULTIPATCH files. For all other cases this may
be NULL, and will be assumed to be SHPP_RING.
nVertices: The number of vertices being passed in padfX,
padfY, and padfZ.
padfX: An array of nVertices X coordinates of the vertices
for this object.
padfY: An array of nVertices Y coordinates of the vertices
for this object.
padfZ: An array of nVertices Z coordinates of the vertices
for this object. This may be NULL in which case
they are all assumed to be zero.
padfM: An array of nVertices M (measure values) of the
vertices for this object. This may be NULL in which
case they are all assumed to be zero.
SHPComputeExtents() //重新計算記錄的取值範圍
void SHPComputeExtents( SHPObject * psObject ); psObject: An existing shape object to be updated in place.
SHPWriteObject() //向shape檔案中新增一條記錄，只有先新增了記錄，才能新增屬性資訊
int SHPWriteObject( SHPHandle hSHP, int iShape, SHPObject *psObject );
hSHP: The handle previously returned by SHPOpen("r+")
or SHPCreate().
iShape: The entity number of the shape to write. A value of
-1 should be used for new shapes.
psObject: The shape to write to the file. This should have
been created with SHPCreateObject(), or
SHPCreateSimpleObject().
SHPDestroyObject() //刪除物件
void SHPDestroyObject( SHPObject *psObject ); psObject: The object to deallocate.
SHPRewindObject()
int SHPRewindObject( SHPHandle hSHP, SHPObject *psObject ); hSHP: The shapefile (not used at this time). psObject: The object to deallocate.
This function will reverse any rings necessary in order to enforce the shapefile restrictions on the required order of inner and outer rings in the Shapefile specification. It returns TRUE if a change is made and FALSE if no change is made. Only polygon objects will be affected though any object may be passed.
前面介紹了shape API，接下來介紹 dBF API
DBFOpen()
DBFHandle DBFOpen( const char * pszDBFFile, const char * pszAccess ); pszDBFFile: The name of the xBase (.dbf) file to access. pszAccess: The fopen() style access string. At this time only "rb" (read-only binary) and "rb+" (read/write binary) should be used.
DBFCreate()
DBFHandle DBFCreate( const char * pszDBFFile ); pszDBFFile: The name of the xBase (.dbf) file to create.
//如果要建立一個完整的shape檔案，首先要建立.shap檔案，還要建立.dbf檔案，兩者缺一不可。
DBFGetFieldCount() //得到欄位的個數
int DBFGetFieldCount( DBFHandle hDBF );
hDBF: The access handle for the file to be queried, as returned
by DBFOpen(), or DBFCreate().
DBFGetRecordCount()
//得到記錄的個數，應該和shape檔案中的object的個數對應
int DBFGetRecordCount( DBFHandle hDBF ); hDBF: The access handle for the file to be queried, as returned by DBFOpen(), or DBFCreate().
DBFGetFieldIndex() //通過欄位名稱得到欄位的編號
int DBFGetFieldIndex( DBFHandle hDBF, const char *pszFieldName ); hDBF: The access handle for the file to be queried, as returned by DBFOpen(), or DBFCreate().
DBFGetFieldInfo() //得到欄位的資訊，包括型別、長度等
DBFFieldType DBFGetFieldInfo( DBFHandle hDBF, int iField, char * pszFieldName,
int * pnWidth, int * pnDecimals );
hDBF: The access handle for the file to be queried, as returned by
DBFOpen(), or DBFCreate().
iField: The field to be queried. This should be a number between
0 and n-1, where n is the number fields on the file, as
returned by DBFGetFieldCount().
pszFieldName: If this pointer is not NULL the name of the requested field
will be written to this location. The pszFieldName buffer
should be at least 12 character is size in order to hold
the longest possible field name of 11 characters plus a
terminating zero character.
pnWidth: If this pointer is not NULL, the width of the requested field
will be returned in the int pointed to by pnWidth. This is
the width in characters.
pnDecimals: If this pointer is not NULL, the number of decimal places
precision defined for the field will be returned. This is
zero for integer fields, or non-numeric fields.
DBFAddField() //向DBF表中新增一個新的屬性欄位
int DBFAddField( DBFHandle hDBF, const char * pszFieldName,
DBFFieldType eType, int nWidth, int nDecimals );
hDBF: The access handle for the file to be updated, as returned by
DBFOpen(), or DBFCreate().
pszFieldName: The name of the new field. At most 11 character will be used.
In order to use the xBase file in some packages it may be
necessary to avoid some special characters in the field names
such as spaces, or arithmetic operators.
eType: One of FTString, FTInteger or FTDouble in order to establish
the type of the new field. Note that some valid xBase field
types cannot be created such as date fields.
nWidth: The width of the field to be created. For FTString fields this
establishes the maximum length of string that can be stored.
For FTInteger this establishes the number of digits of the
largest number that can
be represented. For FTDouble fields this in combination
with the nDecimals value establish the size, and precision
of the created field.
nDecimals: The number of decimal places to reserve for FTDouble fields.
For all other field types this should be zero. For instance
with nWidth=7, and nDecimals=3 numbers would be formatted
similarly to `123.456'.
DBFReadIntegerAttribute()
int DBFReadIntegerAttribute( DBFHandle hDBF, int iShape, int iField );
hDBF: The access handle for the file to be queried, as returned by
DBFOpen(), or DBFCreate().
iShape: The record number (shape number) from which the field value
should be read.
iField: The field within the selected record that should be read.
DBFReadDoubleAttribute()
double DBFReadDoubleAttribute( DBFHandle hDBF, int iShape, int iField );
hDBF: The access handle for the file to be queried, as returned by
DBFOpen(), or DBFCreate().
iShape: The record number (shape number) from which the field value
should be read.
iField: The field within the selected record that should be read.
DBFReadStringAttribute()
const char *DBFReadStringAttribute( DBFHandle hDBF, int iShape, int iField );
hDBF: The access handle for the file to be queried, as returned by
DBFOpen(), or DBFCreate().
iShape: The record number (shape number) from which the field value
should be read.
iField: The field within the selected record that should be read.
DBFIsAttributeNULL()
int DBFIsAttributeNULL( DBFHandle hDBF, int iShape, int iField );
hDBF: The access handle for the file to be queried, as returned by
DBFOpen(), or DBFCreate().
iShape: The record number (shape number) from which the field value
should be read.
iField: The field within the selected record that should be read.
DBFWriteIntegerAttribute
int DBFWriteIntegerAttribute( DBFHandle hDBF, int iShape, int iField,
int nFieldValue );
hDBF: The access handle for the file to be written, as returned by
DBFOpen(), or DBFCreate().
iShape: The record number (shape number) to which the field value
should be written.
iField: The field within the selected record that should be written.
nFieldValue: The integer value that should be written.
DBFWriteDoubleAttribute()
int DBFWriteDoubleAttribute( DBFHandle hDBF, int iShape, int iField,
double dFieldValue );
hDBF: The access handle for the file to be written, as returned by
DBFOpen(), or DBFCreate().
iShape: The record number (shape number) to which the field value
should be written.
iField: The field within the selected record that should be written.
dFieldValue: The floating point value that should be written.
The DBFWriteDoubleAttribute() function is used to write a value to a numeric field (FTInteger, or FTDouble). If the write succeeds the value TRUE will be returned, otherwise FALSE will be returned. If the value is too large to fit in the field, it will be truncated and FALSE returned.
DBFWriteStringAttribute()
int DBFWriteStringAttribute( DBFHandle hDBF, int iShape, int iField,
const char * pszFieldValue );
hDBF: The access handle for the file to be written, as returned by
DBFOpen(), or DBFCreate().
iShape: The record number (shape number) to which the field value
should be written.
iField: The field within the selected record that should be written.
pszFieldValue: The string to be written to the field.
The DBFWriteStringAttribute() function is used to write a value to a string field (FString). If the write succeeds the value TRUE willbe returned, otherwise FALSE will be returned. If the value is too large to fit in the field, it will be truncated and FALSE returned.
DBFWriteNULLAttribute()
int DBFWriteNULLAttribute( DBFHandle hDBF, int iShape, int iField );
hDBF: The access handle for the file to be written, as returned by
DBFOpen(), or DBFCreate().
iShape: The record number (shape number) to which the field value
should be written.
iField: The field within the selected record that should be written.
The DBFWriteNULLAttribute() function is used to clear the indicated field to a NULL value. In the .dbf file this is represented by setting the entire field to spaces. If the write succeeds the value TRUE willbe returned, otherwise FALSE will be returned.
DBFClose()
void DBFClose( DBFHandle hDBF ); hDBF: The access handle for the file to be closed.
The DBFClose() function will close the indicated xBase file (opened with DBFOpen(), or DBFCreate()), flushing out all information to the file on disk, and recovering any resources associated with having the file open. The file handle (hDBF) should not be used again with the DBF API after calling DBFClose().
DBFGetNativeFieldType()
char DBFGetNativeFieldType( DBFHandle hDBF, int iField ); hDBF: The access handle for the file. iField: The field index to query.
This function returns the DBF type code of the indicated field. It will be one of:
- 'C' (String)
- 'D' (Date)
- 'F' (Float)
- 'N' (Numeric, with or without decimal)
- 'L' (Logical)
- 'M' (Memo: 10 digits .DBT block ptr)
- ' ' (field out of range)