NAME
XML::Table2XML - Generic conversion of tabular data to XML by reverting Excel's flattener methodology.
SYNOPSIS
useXML::Table2XML;my$outXML="";# first parse column path headers for attribute names, id columns and special common sibling mark ("//")parseHeaderForXML("rootNodeName", ['/@id','/@name2','/a']);# then walk through the whole data to build the actual XML string into $outXMLmy@datarows= ([1,"testName","testA"],[1,"testName","testB"],[1,"testName","testC"]);formy$lineData(@datarows) {$outXML.=addXMLLine($lineData);}#finally finish the XML and reset the static vars$outXML.=addXMLLine(undef);$outXML;# yields:# <?xml version="1.0"?># <rootNodeName id="1" name2="testName"><a>testA</a><a>testB</a><a>testC</a></rootNodeName>
DESCRIPTION
table2xml is an algorithm having two functions that allow the conversion of tabular data to XML without using XSLT. This is achieved by reverting the "Flattener" methodology used by Microsoft Excel to convert the XML tree format to a two-dimensional table (see Opening XML Files in Excel and INFO: Microsoft Excel 2002 and XML).
This reversion is achieved by:
1. (possibly) modifying the flattened table a bit to enable a simpler processing of the data,
2. sequentially processing the data column- and row wise.
The whole algorithm is done without the aid of any XML library, so it lends itself to easy translation into other environments and languages.
Producing the XML:
1. invoke parseHeaderForXML, using a line with the rootnode and path information.
2. After parsing the header info, the table data can be processed row by row by calling addXMLLine. The current data row is provided in the single argument lineData, the built XML is string returned and can be collected/written.
3. A final call to addXMLLine with lineData == undef restores the static variables and finalizes the XML string (closes any still open tags).
Public Functions
parseHeaderForXML ($rootNodeName,\@header,$LINEBREAKS,$XMLDIRECTIVE,$ENCODING)
rootNodeName is the name of the common root node. Any /@rootAttributes and /#text will be placed under respectively after this root node.
header is a list of paths denoting the "place" of the data in the targeted XML. Following special cases are allowed:
Plain elements
are denoted by /node/subnode/subsubnode/etc.../elementName
Attributes
are denoted by /node/subnode/subsubnode/etc.../@attributeName
"ID" nodes
are denoted by /node/subnode/subsubnode/etc.../#id (they are not being ouptut)
special common sibling nodes
are denoted by a leading double slash (//) special common sibling nodes are used for nested common sibling nodes (e.g.,
<root><a><b>test</b></a><otherData>...<root>or<root><a><b>test1</b><z>test2</z></a><otherData>...<root>) must be located at the beginning of the last node within the nested sibling.a root text element
is denoted by /#text
root attributes
are given as /@rootNodeAttribute
LINEBREAKS specifies whether '\n' should be added after each datarow for easier readablity, default is no linebreaks
XMLDIRECTIVE specifies any header being inserted before the root element, default is '<?xml version="1.0"?'>.
ENCODING denotes the Unicode Codification used to encode the string(s) returned by addXMLLine(), default is 'iso-8859-1'
$returnedXML = addXMLLine(\@lineData)
lineData is a list of data elements that are converted to XML following the parsed header information.
The produced XML is returned as a function value which can be concatenated or written to a file. Bear in mind that the returned XML is just a part of a larger structure, so only after the last line has been processed and addXMLLine(undef) has been called, the XML structure is finished.
Prerequisites for column order and data layout
The layout of the columns (header = "data paths" and respective column data below) has to follow a certain layout:
child nodes always have to follow their parent nodes (i.e. /a/b/c is after /a or /a/b).
"#id" columns and attributes belong to the same element node (e.g. /a/b/#id, /a/b/@att1 and /a/b/@att2) and therefore have to be given consecutively and with the "#id" column first (attributes and element node order is not important).
related subnodes have to be grouped together (i.e. /a/b, /a/c, /a/x, /a/x/@att, ...), other subnodes have to follow.
The layout of the data below has to be as follows (recursively similar within the blocks for any sub-blocks):
Block1PathHeaders Block2PathHeaders Block3PathHeadersBlock1Data EMPTY EMPTY... EMPTY EMPTYBlock1Data EMPTY EMPTYEMPTY Block2Data EMPTYEMPTY ... EMPTYEMPTY Block2Data EMPTYEMPTY EMPTY Block3DataEMPTY EMPTY ...EMPTY EMPTY Block3Datawhere the corresponding XML would then look like:
<root><Block1><Block1subnode>...</Block1subnode><Block1subnode>...</Block1subnode>..</Block1><Block2><Block2subnode>...</Block2subnode><Block2subnode>...</Block2subnode>..</Block2><Block3><Block3subnode>...</Block3subnode><Block3subnode>...</Block3subnode>..</Block3></root>Sibling nodes that are "common" to a whole subnode (e.g.
<subnode><commonSibling>value1</commonSibling><otherNodes>...</otherNodes></subnode>) have to be first in the subnode and need to "span" the data there.Example for subnode <a>:
<?xml version="1.0"?><root><a><z>TestB</z><c>TestA1</c><c>TestA2</c><c>TestA3</c><c>TestA4</c></a></root>/root/a/z/@x/a/cTestB TestA1TestB TestA2TestB TestA3TestB TestA4In case you happen to own MS Excel, the easiest way to get that layout is to follow the steps below:
- 1. Open Target XML File in Excel (don't forget the XML directive there: "<?xml version="1.0"?>" !!!!)
- 2. remove any "#agg" columns (used to differentiate between numerical common siblings and "real" data)
- 3. move the common root (or the common subnode) siblings leftmost of the root (or resp. Subnode)
-
Examples:
<?xml version="1.0"?><root><x z="testAttX">testX</x><a><b><c>TestA1</c><c>TestA2</c><c>TestA3</c><c>TestA4</c></b></a></root>/root /root/a/b/c /x /x/@z/x /x/@z/a/b/cTestA1 testX testAttX testX testAttX TestA1TestA2 testX testAttX modify to-> testX testAttX TestA2TestA3 testX testAttX testX testAttX TestA3TestA4 testX testAttX testX testAttX TestA4<?xml version="1.0"?><root><a><z x="TestB"></z><b><c>TestA1</c><c>TestA2</c><c>TestA3</c><c>TestA4</c></b></a></root>/root /root/a/b/c /a/z/@x/a/z/@x/a/b/cTestA1 TestB TestB TestA1TestA2 TestB modify to-> TestB TestA2TestA3 TestB TestB TestA3TestA4 TestB TestB TestA4 - 4. For nested common sibling nodes (e.g.,
<root><a><b>test</b></a><otherData>...<root>or<root><a><b>test1</b><c>test2</c></a><otherData>...<root>), write a double slash at the beginning of the last node within the nested sibling. -
Example (also includes column moving as in the examples above):
<?xml version="1.0"?><root><a n=""CW""><l c=""oalp""><p v=""A1""></p></l><f c=""oalvl""><p v=""W""></p></f><p n=""target""></p></a><a n=""CD""><l c=""oalp""><p v=""A1""></p></l><f c=""oalvl""><p v=""D""></p></f></a><r><pr v=""TEST""></pr><ar r=""test2""></ar><ar r=""test4""></ar></r></root>/root/a/@n/a/f/@c/a/f/p/@v/a/l/@c/a/l/p/@v/a/p/@n/r/ar/@r/r/pr/@vCW oalvl W oalp A1 targetCD oalvl D oalp A1test2 TESTtest4 TESTmodify to -->/root/a/@n/a/l/@c//a/l/p/@v/a/f/@c//a/f/p/@v/a/p/@n/r/pr/@v/r/ar/@rCW oalp A1 oalvl W targetCD oalp A1 oalvl DTEST test2TEST test4 - 5. For a first column of a subnode list that is not being a "primary key" column (i.e., having empty cells or continuous equal values), introduce an artificial #id column.
-
Examples:
<?xml version="1.0"?><root><a x="test1">testA</a><a x="test2"></a></root>/root /root/a /a/@xmodify to-> /a/#id /a /a/@xtestA test1 1 testA test1test2 2 test2<?xml version="1.0"?><root><co><f><a>Numeric</a></f></co><co><f><a>VarChar</a></f></co><co><f><a>VarChar</a></f></co><co><f><a>VarChar</a></f></co><co><f><a>VarChar</a></f></co><co><f><a>DBTimeStamp</a></f></co><co><f><a>VarChar</a><fk>JOB_ID</fk><fl>JOB_TITLE</fl></f></co><co><f><a>Numeric</a><fk>TESTID</fk><fl>TESTn</fl></f></co><co><f><a>Numeric</a></f></co><co><f><a>Numeric</a><fk>EMPLOYEE_ID</fk><fl>FIRST_n</fl></f></co><co><f><a>Numeric</a><fk>DEPARTMENT_ID</fk><fl>DEPARTMENT_n</fl></f></co></root>/root modify to-> /root/co/f/a /co/f/fk /co/f/fl /co/#id /co/f/a /co/f/fk /co/f/flNumeric 1 NumericVarChar 2 VarCharVarChar 3 VarCharVarChar 4 VarCharVarChar 5 VarCharDBTimeStamp 6 DBTimeStampVarChar JOB_ID JOB_TITLE 7 VarChar JOB_ID JOB_TITLENumeric TESTID TESTn 8 Numeric TESTID TESTnNumeric 9 NumericNumeric EMPLOYEE_ID FIRST_n 10 Numeric EMPLOYEE_ID FIRST_nNumeric DEPARTMENT_ID DEPARTMENT_n 11 Numeric DEPARTMENT_ID DEPARTMENT_n - 6. Use the header row and rootnodeName for your data layout.
LIMITATIONS
Generally, pay close attention to the ordering of columns and constraints on the data as described above, since the algorithm in writeLine doesn't check for validity, thus producing invalid XML in case of failing to follow preparation steps correctly.
In mixed content nodes, the only way to correctly (re)produce the XML is for ONE content being right after the node name. There's currently no way to produce mixed content nodes with more than one text node (e.g., <node>text1<subnode>Test</subnode>text2</node> and the like).
Same sequential parent nodes are "factored" out by the flattener, so the unflattening algorithm treats them as being factored out, which means there is no way to exactly reproduce (<a><b>test1</b></a><a><b>test2</b></a>, this would be processed as <a><b>test1</b><b>test2</b></a>, which is semantically equal, but not the same...).
REFERENCE
for a detailed discussion of the flattening algorithmm in Excel see https://web.archive.org/web/20041124175116/http://support.microsoft.com/kb/282161/EN-US/ and https://web.archive.org/web/20050210015617/http://support.microsoft.com/kb/288215/EN-US/
AUTHOR
Roland Kapl, rkapl@cpan.org
COPYRIGHT AND LICENSE
Copyright (C) 2006 by Roland Kapl
This library is free software; you can redistribute it and/or modify it under the same terms as Perl itself, either Perl version 5.8.8 or, at your option, any later version of Perl 5 you may have available.