Mercurial > hg > mpdl-group
diff software/eXist/webapp/mpdl/_stuff/oxygen-projects/monte-project/archimedes-xml-fragment.xsl @ 7:5589d865af7a
Erstellung XQL/XSL Applikation
| author | Josef Willenborg <jwillenborg@mpiwg-berlin.mpg.de> |
|---|---|
| date | Tue, 08 Feb 2011 15:16:46 +0100 |
| parents | |
| children |
line wrap: on
line diff
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/software/eXist/webapp/mpdl/_stuff/oxygen-projects/monte-project/archimedes-xml-fragment.xsl Tue Feb 08 15:16:46 2011 +0100 @@ -0,0 +1,311 @@ +<xsl:stylesheet version="2.0" + xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + xmlns:xlink="http://www.w3.org/1999/xlink" + xmlns:xs="http://www.w3.org/2001/XMLSchema" + xmlns:functx="http://www.functx.com" + xmlns:text="http://www.mpiwg-berlin.mpg.de/ns/mpdl/text"> + +<xsl:output method="html"/> + +<!-- siehe unter http://www.xsltfunctions.com --> +<xsl:function name="functx:contains-any-of" as="xs:boolean"> + <xsl:param name="arg" as="xs:string?"/> + <xsl:param name="searchStrings" as="xs:string*"/> + <xsl:sequence select="some $searchString in $searchStrings satisfies contains($arg,$searchString)"/> +</xsl:function> + +<!-- Highlight all term occurrences for the fulltext query. It recognizes the first occorrence and then does +this function recursive with the first substring before the first term cut off. +Result is a sequence of text and highlight nodes: +Example: +LE<lb></lb> +<span ...>MECHANICHE</span> +...<lb></lb> +...<lb></lb> +<span ...>MECHANICHE</span> +...<lb></lb> +...<lb></lb> +... + --> +<xsl:function name="text:highlight"> + <xsl:param name="inputString" /> + <xsl:param name="luceneQuery" as="xs:string" /> + <xsl:param name="mode" as="xs:string" /> + <!-- Translation from Lucene fulltext query to text query --> + <xsl:variable name="textQuery"> + <xsl:choose> + <xsl:when test="$mode = 'phrase'"> + <xsl:value-of select="substring-before(substring-after($luceneQuery, '"'), '"')"/> + </xsl:when> + <xsl:when test="$mode = 'fakeWord'"> + <xsl:value-of select="$luceneQuery"/> + </xsl:when> + <xsl:when test="$mode = 'word'"> + <xsl:value-of select="$luceneQuery"/> + </xsl:when> + <xsl:otherwise><xsl:value-of select="''"/></xsl:otherwise> + </xsl:choose> + </xsl:variable> + <!-- Word delimiter: not tested yet --> + <xsl:variable name="wordDelimRegExpr" select="'[\s\(\)\[\]\.\\\{\}\$\^\+\?\!\* ยง%:,;=/]+'"/> <!-- TODO: bol, eol, ", &, <, > --> + <!-- Recognizes the beginning of the line with ^ and the substring up to the query term --> + <xsl:variable name="queryRegExprSubstringBefore"> + <xsl:choose> + <xsl:when test="$mode = 'phrase'"> + <xsl:value-of select="concat('^.*?', $textQuery)"/> + </xsl:when> + <xsl:when test="$mode = 'fakeWord'"> + <xsl:value-of select="concat('^.*?', $textQuery)"/> + </xsl:when> + <xsl:when test="$mode = 'word'"> + <xsl:value-of select="concat('^.*?', $wordDelimRegExpr, $textQuery, '(', $wordDelimRegExpr, ')')"/> + </xsl:when> + <xsl:otherwise><xsl:value-of select="''"/></xsl:otherwise> + </xsl:choose> + </xsl:variable> + <!-- Recognizes the substring after the query term --> + <xsl:variable name="queryRegExprSubstringAfter"> + <xsl:choose> + <xsl:when test="$mode = 'phrase'"> + <xsl:value-of select="concat($textQuery, '.*')"/> + </xsl:when> + <xsl:when test="$mode = 'fakeWord'"> + <xsl:value-of select="concat($textQuery, '.*')"/> + </xsl:when> + <xsl:when test="$mode = 'word'"> + <xsl:value-of select="concat('(', $wordDelimRegExpr, ')', $textQuery, $wordDelimRegExpr, '.*')"/> + </xsl:when> + <xsl:otherwise><xsl:value-of select="''"/></xsl:otherwise> + </xsl:choose> + </xsl:variable> + <!-- Deletes the substring up to the query term --> + <xsl:variable name="substringBefore" select="replace($inputString, $queryRegExprSubstringAfter, '$1', 'i')"/> + <!-- Deletes the substring after the query term --> + <xsl:variable name="substringAfter" select="replace($inputString, $queryRegExprSubstringBefore, '$1', 'i')"/> + <xsl:choose> + <xsl:when test="matches($inputString, $queryRegExprSubstringBefore, 'i')"> + <!-- Prints the original part of the substring up to the first occurrence of the query term --> + <xsl:value-of select="$substringBefore"/> + <!-- Highlight the query term --> + <xsl:variable name="matchQueryTermRegExpr"> + <xsl:choose> + <xsl:when test="$mode = 'phrase'"> + <xsl:value-of select="concat('^.*?', '(', $textQuery, ')', '.*')"/> + </xsl:when> + <xsl:when test="$mode = 'fakeWord'"> + <xsl:value-of select="concat('^.*?', '(', $textQuery, ')', '.*')"/> + </xsl:when> + <xsl:when test="$mode = 'word'"> + <xsl:value-of select="concat('^.*?', $wordDelimRegExpr, '(', $textQuery, ')', $wordDelimRegExpr, '.*')"/> + </xsl:when> + <xsl:otherwise><xsl:value-of select="''"/></xsl:otherwise> + </xsl:choose> + </xsl:variable> + <xsl:variable name="matchQueryTerm" select="replace($inputString, $matchQueryTermRegExpr, '$1', 'i')"/> + <span style="background-color: yellow;"> + <xsl:value-of select="$matchQueryTerm"/> + </span> + <!-- Recursive call of this function with the substring after the first occurrence of the term: further occurrences of the query + term --> + <xsl:sequence select="text:highlight($substringAfter, $luceneQuery, $mode)"/> + </xsl:when> + <!-- if no occurrence of the query term could be found the whole string is printed --> + <xsl:otherwise> + <xsl:value-of select="$inputString"/> + </xsl:otherwise> + </xsl:choose> +</xsl:function> + +<!-- +<xsl:function name="text:limitHighligthedString"> + <xsl:param name="inputString" /> + <xsl:param name="countChar" as="xs:integer"/> + <xsl:variable name="substringBeforeHighlight" select="substring-before($inputString, '<span')"/> + <xsl:variable name="substringAfterHighlight" select="substring-after($inputString, '<span>>')"/> + <xsl:variable name="result"> + ... <xsl:value-of select="substring($substringBeforeHighlight, string-length($substringBeforeHighlight) - $countChar)"/><span style="background-color: yellow;"><xsl:value-of select="' BlaBla '"/></span><xsl:value-of select="substring($substringAfterHighlight, 0, $countChar)"/> ... + </xsl:variable> + <xsl:sequence select="$result"/> +</xsl:function> +--> + +<xsl:function name="text:limitHighligthedString"> + <xsl:param name="inputSequence" as="node()*"/> + <xsl:param name="countChar" as="xs:integer"/> + <xsl:variable name="before" select="$inputSequence[1]"/> + <xsl:variable name="beforeLimited" select="substring($before, string-length($before) - $countChar)"/> + <xsl:variable name="firstSpan" select="$inputSequence[2]"/> + <xsl:variable name="after" select="$inputSequence[position() > 2]"/> + <xsl:variable name="afterLimited" select="substring($after, 0, $countChar)"/> + <xsl:variable name="result"> + ... <xsl:sequence select="$before"/><xsl:sequence select="$firstSpan"/><xsl:sequence select="$after"/> ... + </xsl:variable> + <xsl:sequence select="$result"/> +</xsl:function> + +<xsl:variable name="ftQueryName" select="/result/ft-query/name"/> +<xsl:variable name="ftQueryStr" select="concat('&', 'ft-query=', $ftQueryName)"/> +<xsl:variable name="ftQueryMode"> + <xsl:choose> + <xsl:when test="matches($ftQueryName, concat('"', '.*', '"'))"> + <xsl:value-of select="'phrase'"/> + </xsl:when> + <xsl:when test="string-length($ftQueryName) = 0"> + <xsl:value-of select="'false'"/> + </xsl:when> + <!-- Because there are problems in recognizing words (in mode word) this fake mode is used (same as phrase mode) --> + <xsl:otherwise><xsl:value-of select="'fakeWord'"/></xsl:otherwise> + </xsl:choose> +</xsl:variable> +<xsl:variable name="ftQueryValue"> + <xsl:choose> + <xsl:when test="/result/ft-query[node()]"><xsl:value-of select="$ftQueryStr"/></xsl:when> + <xsl:otherwise><xsl:value-of select="''"/></xsl:otherwise> + </xsl:choose> +</xsl:variable> + +<xsl:template match="result"> + <html><body> + <xsl:variable name="ft-query-name" select="/result/ft-query/name"/> + <xsl:for-each select="document-description"> + <h2> + <xsl:for-each select="info"> + <xsl:value-of select="author"/>. + <xsl:value-of select="title"/>. + <xsl:value-of select="place"/>, + <xsl:value-of select="date"/> + </xsl:for-each> + </h2> + </xsl:for-each> + <form action="" method="get"></form> + <xsl:for-each select="page"> + <xsl:variable name="documentName" select="/result/document-description/document-name"/> + <xsl:variable name="countPages" select="/result/document-description/count-pages"/> + <xsl:variable name="number" select="number(number)"/> + <xsl:variable name="documentValue" select="concat('document=', $documentName)"/> + <xsl:variable name="pnValue" select="concat('pn=', $number)"/> + <xsl:variable name="modeImageValue" select="concat('mode=', 'image')"/> + <xsl:variable name="modeXmlValue" select="concat('mode=', 'xml')"/> + <xsl:variable name="modeTextValue" select="concat('mode=', 'text')"/> + <xsl:variable name="imageLink" select="concat($documentValue, '&', $pnValue, '&', $modeImageValue, $ftQueryValue)"/> + <xsl:variable name="textLink" select="concat($documentValue, '&', $pnValue, '&', $modeTextValue, $ftQueryValue)"/> + <xsl:variable name="xmlLink" select="concat($documentValue, '&', $pnValue, '&', $modeXmlValue, $ftQueryValue)"/> + <table width="100%"> + <colgroup> + <col width="30%"/> + <col width="70%"/> + </colgroup> + <tr> + <td align="left" nowrap="true"> + [<a href="?{$textLink}">Text</a>] + [<a href="?{$imageLink}">Image</a>] + [XML] + </td> + <td align="right" nowrap="true"> + <td> + <xsl:choose> + <xsl:when test="$number > 1"> + <a href="?{$documentValue}&pn={$number - 1}&{$modeXmlValue}{$ftQueryValue}"><img src="images/left.gif" alt="page-down" border="0"/></a> + </xsl:when> + <xsl:otherwise> + <a href="?{$xmlLink}"><img src="images/left.gif" alt="page-down" border="0"/></a> + </xsl:otherwise> + </xsl:choose> + </td> + <td nowrap="true"> + <xsl:value-of select="$number"/> / <xsl:value-of select="$countPages"/> + </td> + <td> + <xsl:choose> + <xsl:when test="$number < $countPages"> + <a href="?{$documentValue}&pn={$number + 1}&{$modeXmlValue}{$ftQueryValue}"><img src="images/right.gif" alt="page-up" border="0"/></a> + </xsl:when> + <xsl:otherwise> + <a href="?{$xmlLink}"><img src="images/right.gif" alt="page-down" border="0"/></a> + </xsl:otherwise> + </xsl:choose> + </td> + <td nowrap="true"> + <input type="hidden" name="document" value="{$documentName}"/> + Page: <input type="text" size="3" name="pn" value="{$number}"/> + <input type="hidden" name="mode" value="xml"/> + <xsl:if test="/result/ft-query[node()]"> + <input type="hidden" name="ft-query" value="{$ftQueryName}"/> + </xsl:if> + </td> + </td> + </tr> + </table> + <hr/> + <table align="middle" width="100%"> + <colgroup> + <col width="70%"/> + <col width="30%"/> + </colgroup> + <tr> + <td align="left" valign="top"> + <xsl:for-each select="content"> + <xsl:apply-templates/> + </xsl:for-each> + </td> + <xsl:if test="$ftQueryValue!=''"> + <td align="left" valign="top"> + <b><xsl:value-of select="/result/ft-query/result/size"/> Hits: "<xsl:value-of select="$ftQueryName"/>"</b> + <ol> + <xsl:for-each select="/result/ft-query/result/hits/hit"> + <li> + <xsl:variable name="hitPN" select="pn"/> + <xsl:variable name="hitPosOfS" select="pos-of-s"/> + <a href="?{$documentValue}&pn={$hitPN}&{$modeXmlValue}{$ftQueryValue}">Page <xsl:value-of select="$hitPN"/>, Sentence <xsl:value-of select="$hitPosOfS"/></a>:<br></br> + <!-- Highlight the query terms in each hit --> + <xsl:variable name="highlightedSentenceSeq" select="text:highlight(s, $ftQueryName, $ftQueryMode)"/> + <xsl:sequence select="text:limitHighligthedString($highlightedSentenceSeq, 100)"/> + </li> + </xsl:for-each> + </ol> + </td> + </xsl:if> + </tr> + </table> + <hr/> + Elapsed time: <xsl:value-of select="/result/document-description/performance"/> ms, Back to <a href="query.xql">query page</a>, see the <a href="page-query-result.xql?_source=yes">XQuery source</a> and <a href="/exist/rest/db/xsl/archimedes-xml-fragment.xsl?_source=yes">XSL source</a> of this page + <br/> + [<a href="page-query-result.xql?{$xmlLink}">fast mechanism (file system search)</a>] [<a href="page-query-result-old.xql?{$xmlLink}">slower mechanism (XQL search)</a>] + </xsl:for-each> + </body></html> +</xsl:template> + + +<xsl:template match="attribute()|element()|text()|comment()|processing-instruction()"> + <xsl:copy> + <xsl:apply-templates select="attribute()|element()|text()|comment()|processing-instruction()"/> + </xsl:copy> +</xsl:template> + +<xsl:template match="element()|text()|comment()|processing-instruction()"> + <xsl:variable name="elementName" select="name()"/> + <span style="color: brown;"><<xsl:value-of select="$elementName"/><xsl:apply-templates select="attribute()"/>></span> + <xsl:apply-templates select="element()|text()|comment()|processing-instruction()"/> + <span style="color: brown;"></<xsl:value-of select="$elementName"/>></span> + <br/> +</xsl:template> + +<xsl:template match="attribute()"> + <xsl:variable name="attributeName" select="name()"/> + <xsl:text> </xsl:text><xsl:value-of select="$attributeName"/>=<xsl:value-of select="."/> + <xsl:apply-templates select="attribute()"/> +</xsl:template> + +<!-- If ft-query is set then highlight all term occurrences for the fulltext query else show the text --> +<xsl:template match="text()"> + <xsl:choose> + <xsl:when test="$ftQueryMode != 'false'"> + <xsl:sequence select="text:highlight(., $ftQueryName, $ftQueryMode)"/> + </xsl:when> + <xsl:otherwise> + <xsl:value-of select="."/> + </xsl:otherwise> + </xsl:choose> +</xsl:template> + +</xsl:stylesheet>