VFP 愛用者社區

[Ruey]

來源:飛狐堂 boe
http://www.vfp.cn/blog/


用SAX實現xmltocursor

SAX:Simple API FOR XML 是和一個基於事件流的 XML 文檔解析方式，比 DOM 方式節省資源，速度快很多。
VFP 的 XMLTOCURSOR 採用 DOM 方式，對於稍大的文檔基本無法處理，筆者的一個 XML 文檔有 43 MB大小，內涵50000多條記錄，17個字段，用 VFP 8 sp1 的 xmltocursor 無法處理(等待20分鐘無果、且系統無法相應)。
利用 SAX 處理，可以做到4分多鐘解析出 cursor，(且可以支持進度條)，系統資源佔用較小，沒有死機的感覺。雖說時間依然蠻長，但鑒於可以支持進度條，而且測試文件較大，所以這種基於 SAX的 XMLTOCURSOR 是可行的。

週末我會成文，與大家分享。

DOM 把 XML當作一棵樹看待，遇到大的 XML 要把這個樹生成，就需要很多資源、時間，慢是肯定的；當 XML 的規模較小的時候， DOM 的使用方便地優勢就很明顯。


employees.rar
－－－非緩衝模式
DEFINE CLASS XmlContentHandler AS custom
IMPLEMENTS IVBSAXContentHandler IN "c:\windows\system32\msxml4.dll"

DIMENSION aRecord[17]
iCol=-1
isData=.F.
isAfterStartElement=.F.
PROCEDURE Init()
CREATE CURSOR XMLTEMP(employeeid int NOT NULL,lastname c(20) NOT null,firstname c(10) NOT null,Title c(30) null,titleofcourtesy c(25) null,birthdate datetime null,hiredate datetime null,address c(60) null,city c(15) null,region c(15) null,postalcode c(10) null,country c(15) null,homephone c(24) null,extension c(4) null,note m null,reportsto i null,photopath m null)
SELECT XMLTEMP
SCATTER TO this.aRecord BLANK
ENDPROC

PROCEDURE IVBSAXContentHandler_documentLocator() AS VARIANT;
HELPSTRING "Receive an object for locating the origin of SAX document events."
* add user code here
ENDPROC

PROCEDURE IVBSAXContentHandler_put_documentLocator( ISAXLocator) AS VARIANT;
HELPSTRING "Receive an object for locating the origin of SAX document events."
* add user code here
ENDPROC
PROCEDURE IVBSAXContentHandler_startDocument() AS VOID;
HELPSTRING "Receive notification of the beginning of a document."
* add user code here
ENDPROC

PROCEDURE IVBSAXContentHandler_endDocument() AS VOID;
HELPSTRING "Receive notification of the end of a document."
* add user code here
ENDPROC

PROCEDURE IVBSAXContentHandler_startPrefixMapping(strPrefix AS STRING @, strURI AS STRING @) AS VOID;
HELPSTRING "Begin the scope of a prefix-URI Namespace mapping."
* add user code here
ENDPROC

PROCEDURE IVBSAXContentHandler_endPrefixMapping(strPrefix AS STRING @) AS VOID;
HELPSTRING "End the scope of a prefix-URI mapping."
* add user code here
ENDPROC

PROCEDURE IVBSAXContentHandler_startElement(strNamespaceURI AS STRING @, strLocalName AS STRING @, strQName AS STRING @, oAttributes AS VARIANT) AS VOID;
HELPSTRING "Receive notification of the beginning of an element."
IF ALLTRIM(strLocalName)=='xml'
this.isData=.T.
this.iCol=-1
ENDIF
IF this.isData
this.iCol = this.iCol + 1
this.isAfterStartElement=.T.
ENDIF

ENDPROC

PROCEDURE IVBSAXContentHandler_endElement(strNamespaceURI AS STRING @, strLocalName AS STRING @, strQName AS STRING @) AS VOID;
HELPSTRING "Receive notification of the end of an element."
IF ALLTRIM(strLocalName)=='xml'
this.iCol=-1
SELECT XMLTEMP
APPEND FROM ARRAY this.aRecord
ENDIF
this.isAfterStartElement=.F.
ENDPROC

PROCEDURE IVBSAXContentHandler_characters(strChars AS STRING @) AS VOID;
HELPSTRING "Receive notification of character data."
IF this.iCol>0 AND this.isAfterStartElement
s=ALLTRIM(strChars)
DO CASE
CASE INLIST(this.iCol,1,16)
this.aRecord[this.iCol]=VAL(s)
CASE INLIST(this.iCol,6,7)
this.aRecord[this.iCol]=CTOT(s)
OTHERWISE
this.aRecord[this.iCol]=s
ENDCASE
ENDIF

ENDPROC

PROCEDURE IVBSAXContentHandler_ignorableWhitespace(strChars AS STRING @) AS VOID;
HELPSTRING "Receive notification of ignorable whitespace in element content."
* add user code here
ENDPROC

PROCEDURE IVBSAXContentHandler_processingInstruction(strTarget AS STRING @, strData AS STRING @) AS VOID;
HELPSTRING "Receive notification of a processing instruction."
* add user code here
ENDPROC

PROCEDURE IVBSAXContentHandler_skippedEntity(strName AS STRING @) AS VOID;
HELPSTRING "Receive notification of a skipped entity."
* add user code here
ENDPROC

ENDDEFINE
－－－－
－－－－50條記錄緩衝
DEFINE CLASS XmlContentHandler AS custom
IMPLEMENTS IVBSAXContentHandler IN "c:\windows\system32\msxml4.dll"

DIMENSION aRecord[50,17]
iCol=-1
isData=.F.
isAfterStartElement=.F.
iRow=0
PROCEDURE Init()
CREATE CURSOR XMLTEMP(employeeid int NOT NULL,lastname c(20) NOT null,firstname c(10) NOT null,Title c(30) null,titleofcourtesy c(25) null,birthdate datetime null,hiredate datetime null,address c(60) null,city c(15) null,region c(15) null,postalcode c(10) null,country c(15) null,homephone c(24) null,extension c(4) null,note m null,reportsto i null,photopath m null)
SELECT XMLTEMP
SCATTER TO this.aRecord BLANK
ENDPROC

PROCEDURE IVBSAXContentHandler_documentLocator() AS VARIANT;
HELPSTRING "Receive an object for locating the origin of SAX document events."
* add user code here
ENDPROC

PROCEDURE IVBSAXContentHandler_put_documentLocator( ISAXLocator) AS VARIANT;
HELPSTRING "Receive an object for locating the origin of SAX document events."
* add user code here
ENDPROC
PROCEDURE IVBSAXContentHandler_startDocument() AS VOID;
HELPSTRING "Receive notification of the beginning of a document."
* add user code here
ENDPROC

PROCEDURE IVBSAXContentHandler_endDocument() AS VOID;
HELPSTRING "Receive notification of the end of a document."
IF this.iRow>0
DIMENSION myRecord[this.iRow,17]
ACOPY(this.aRecord,myRecord,1,AELEMENT(this.aRecord,this.iRow,17))
SELECT XMLTEMP
APPEND FROM ARRAY myRecord
ENDIF
ENDPROC

PROCEDURE IVBSAXContentHandler_startPrefixMapping(strPrefix AS STRING @, strURI AS STRING @) AS VOID;
HELPSTRING "Begin the scope of a prefix-URI Namespace mapping."
* add user code here
ENDPROC

PROCEDURE IVBSAXContentHandler_endPrefixMapping(strPrefix AS STRING @) AS VOID;
HELPSTRING "End the scope of a prefix-URI mapping."
* add user code here
ENDPROC

PROCEDURE IVBSAXContentHandler_startElement(strNamespaceURI AS STRING @, strLocalName AS STRING @, strQName AS STRING @, oAttributes AS VARIANT) AS VOID;
HELPSTRING "Receive notification of the beginning of an element."
IF ALLTRIM(strLocalName)=='xml'
this.isData=.T.
this.iCol=-1
this.iRow = this.iRow + 1
ENDIF
IF this.isData
this.iCol = this.iCol + 1
this.isAfterStartElement=.T.
ENDIF

ENDPROC

PROCEDURE IVBSAXContentHandler_endElement(strNamespaceURI AS STRING @, strLocalName AS STRING @, strQName AS STRING @) AS VOID;
HELPSTRING "Receive notification of the end of an element."
IF ALLTRIM(strLocalName)=='xml'
this.iCol=-1
IF this.iRow=50
SELECT XMLTEMP
APPEND FROM ARRAY this.aRecord
this.iRow=0
ENDIF
ENDIF
this.isAfterStartElement=.F.
ENDPROC

PROCEDURE IVBSAXContentHandler_characters(strChars AS STRING @) AS VOID;
HELPSTRING "Receive notification of character data."
IF this.iCol>0 AND this.isAfterStartElement
s=ALLTRIM(strChars)
DO CASE
CASE INLIST(this.iCol,1,16)
this.aRecord[this.iRow,this.iCol]=VAL(s)
CASE INLIST(this.iCol,6,7)
this.aRecord[this.iRow,this.iCol]=CTOT(s)
OTHERWISE
this.aRecord[this.iRow,this.iCol]=s
ENDCASE
ENDIF

ENDPROC

PROCEDURE IVBSAXContentHandler_ignorableWhitespace(strChars AS STRING @) AS VOID;
HELPSTRING "Receive notification of ignorable whitespace in element content."
* add user code here
ENDPROC

PROCEDURE IVBSAXContentHandler_processingInstruction(strTarget AS STRING @, strData AS STRING @) AS VOID;
HELPSTRING "Receive notification of a processing instruction."
* add user code here
ENDPROC

PROCEDURE IVBSAXContentHandler_skippedEntity(strName AS STRING @) AS VOID;
HELPSTRING "Receive notification of a skipped entity."
* add user code here
ENDPROC

ENDDEFINE

[elleryq]

真是厲害啊~~

[syntech]