boxtec
/
yacy_search_server
mirror of https://github.com/yacy/yacy_search_server
4
1
Fork 0
Code Issues Projects Releases Wiki Activity

Extended WikiCode template inclusion syntax support.

Browse Source 
Wiki templates are not rendered but syntax support is improved, which
greatly enhance snippets rendering on search results coming from a
MediaWiki dump import.
Tested on various dumps from Wikimedia at
https://dumps.wikimedia.org/backup-index.html
See also Wikipedia transclusion documentation at
https://en.wikipedia.org/wiki/Wikipedia:Transclusion
pull/122/head
luccioman 8 years ago
parent 973d74712f
commit 31fff2c986
2 changed files with 217 additions and 64 deletions
Show Stats Download Patch File Download Diff File

98
source/net/yacy/data/wiki/WikiCode.java
Unescape View File

@ -107,7 +107,9 @@ public class WikiCode extends AbstractWikiParser implements WikiParser {
private static final String WIKI_CLOSE_LINK = "]]";
private static final String WIKI_OPEN_LINK = "[[";
/** Wiki template inclusion closing tag */
private static final String WIKI_CLOSE_METADATA = "}}";
/** Wiki template inclusion opening tag */
private static final String WIKI_OPEN_METADATA = "{{";
private static final String WIKI_CLOSE_EXTERNAL_LINK = "]";
private static final String WIKI_OPEN_EXTERNAL_LINK = "[";
@ -127,6 +129,8 @@ public class WikiCode extends AbstractWikiParser implements WikiParser {
private static final char SIX = '6';
private static final char WIKI_FORMATTED = ' ';
private static final char WIKI_INDENTION = ':';
/** Wiki template parameter separator */
private static final char WIKI_METADATA_PARAMETER_SEPARATOR = '|';
private static final int LEN_WIKI_CLOSE_PRE_ESCAPED = WIKI_CLOSE_PRE_ESCAPED.length();
private static final int LEN_WIKI_OPEN_PRE_ESCAPED = WIKI_OPEN_PRE_ESCAPED.length();
@ -140,6 +144,7 @@ public class WikiCode extends AbstractWikiParser implements WikiParser {
private static final int LEN_WIKI_HR_LINE = WIKI_HR_LINE.length();
private static final int LEN_PIPE_ESCAPED = PIPE_ESCAPED.length();
private static final int LEN_WIKI_OPEN_METADATA = WIKI_OPEN_METADATA.length();
private static final int LEN_WIKI_CLOSE_METADATA = WIKI_CLOSE_METADATA.length();
/** List of properties which can be used in tables. */
private final static String[] TABLE_PROPERTIES = {"rowspan", "colspan", "vspace", "hspace", "cellspacing", "cellpadding", "border"};
@ -1042,16 +1047,29 @@ public class WikiCode extends AbstractWikiParser implements WikiParser {
/**
* Process line with geo coordinate metadata
* @param line of wiki text
* @return line with geo coordinate formatted to be recogizeable by parser
* Process template inclusions in line, eventually with geo coordinate metadata
* @param line line of wiki text
* @return cleaned text with eventual geo coordinates formatted to be recognizable by parser
* @see https://en.wikipedia.org/wiki/Wikipedia:Transclusion
*/
private static String processMetadata(String line) {
int p, q, s = 0;
while ((p = line.indexOf(WIKI_OPEN_METADATA, s)) >= 0 && (q = line.indexOf(WIKI_CLOSE_METADATA, p + 1)) >= 0) {
s = q; // continue with next position
final String a = line.substring(p + LEN_WIKI_OPEN_METADATA, q);
if (a.toLowerCase().startsWith("coordinate")) {
protected static String processMetadata(final String line) {
StringBuilder processedLine = new StringBuilder(line);
int openIndex, closeIndex, fromIndex = 0;
while ((openIndex = processedLine.indexOf(WIKI_OPEN_METADATA, fromIndex)) >= 0) {
closeIndex = processedLine.indexOf(WIKI_CLOSE_METADATA, openIndex + LEN_WIKI_OPEN_METADATA);
/* Closing tag position : handle eventually nested tags */
int nextOpenIndex = processedLine.indexOf(WIKI_OPEN_METADATA, openIndex + LEN_WIKI_OPEN_METADATA);
while(nextOpenIndex >= 0 && nextOpenIndex < closeIndex) {
closeIndex = processedLine.indexOf(WIKI_CLOSE_METADATA, closeIndex + LEN_WIKI_CLOSE_METADATA);
if(closeIndex < 0) {
/* Parent closing mark is missing: likely a multi-line template inclusion */
break;
}
nextOpenIndex = processedLine.indexOf(WIKI_OPEN_METADATA, nextOpenIndex + LEN_WIKI_OPEN_METADATA);
}
if(closeIndex > 0) {
final String content = processedLine.substring(openIndex + LEN_WIKI_OPEN_METADATA, closeIndex);
if (content.toLowerCase().startsWith("coordinate")) {
// parse Geographical Coordinates as described in
// http://en.wikipedia.org/wiki/Wikipedia:Manual_of_Style_%28dates_and_numbers%29#Geographical_coordinates
// looks like:
@ -1062,7 +1080,7 @@ public class WikiCode extends AbstractWikiParser implements WikiParser {
// and if passed through this parser:
// {{Coordinate |NS 45/37/43.0/N |EW. 07/58/41.0/E |type=landmark |region=IT-BI}} ## means: degree/minute/second
// {{Coordinate |NS 51.48994 |EW. 7.33249 |type=landmark |region=DE-NW}}
final String b[] = a.split("\\|");
final String b[] = content.split("\\|");
float lon = Float.NaN, lat = Float.NaN; // degree
float lonm = 0.0f, latm = 0.0f; // minutes (including sec as fraction)
String lono = "E", lato = "N";
@ -1095,17 +1113,71 @@ public class WikiCode extends AbstractWikiParser implements WikiParser {
}
} catch (NumberFormatException nsExcept) {
// catch parseFloat exception (may still happen if wiki code contains expressions)
processedLine.delete(closeIndex, closeIndex + LEN_WIKI_CLOSE_METADATA);
processedLine.delete(openIndex, openIndex + LEN_WIKI_OPEN_METADATA);
fromIndex = openIndex;
continue;
}
if (!Float.isNaN(lon) && !Float.isNaN(lat)) {
// replace this with a format that the html parser can understand
line = line.substring(0, p) + (name.length() > 0 ? (" " + name) : "") + " <nobr> " + lato + " " + lat + "\u00B0 " + latm + "'</nobr><nobr>" + lono + " " + lon + "\u00B0 " + lonm + "'</nobr> " + line.substring(q + WIKI_CLOSE_METADATA.length());
s = p;
final String htmlCoord = (name.length() > 0 ? (" " + name) : "") +
WIKI_FORMATTED +"<nobr> " + lato + " " + lat + "\u00B0 " + latm + "'</nobr><nobr>" + lono + " " + lon + "\u00B0 " + lonm + "'</nobr>" + WIKI_FORMATTED;
processedLine.replace(openIndex, closeIndex + LEN_WIKI_CLOSE_METADATA, htmlCoord);
/* Set next position to openIndex as some parameters can still contain nested template inclusion tags */
fromIndex = openIndex;
continue;
}
fromIndex = closeIndex; // continue with next position
} else {
String processedContent;
/* Any other template inclusion : only remove opening and closing tag and parameter separators */
int nestedOpenTagIndex = content.indexOf(WIKI_OPEN_METADATA);
int lastNestedCloseTagIndex = content.lastIndexOf(WIKI_CLOSE_METADATA);
if(nestedOpenTagIndex >= 0 && lastNestedCloseTagIndex > 0) {
processedContent = WIKI_FORMATTED + content.substring(0, nestedOpenTagIndex).replace(WIKI_METADATA_PARAMETER_SEPARATOR, ' ').replace('=', ' ')
+ content.substring(nestedOpenTagIndex, lastNestedCloseTagIndex)
+ content.substring(lastNestedCloseTagIndex).replace(WIKI_METADATA_PARAMETER_SEPARATOR, ' ').replace('=', ' ') + WIKI_FORMATTED;
fromIndex = openIndex; // continue with next nested position
} else {
/* No nested tag : we can now replace parameter separators with spaces in all remaining content */
processedContent = WIKI_FORMATTED + content.replace(WIKI_METADATA_PARAMETER_SEPARATOR, ' ').replace('=', ' ') + WIKI_FORMATTED;
fromIndex = openIndex + processedContent.length(); // continue with next position
}
processedLine.replace(openIndex, closeIndex + LEN_WIKI_CLOSE_METADATA, processedContent);
}
return line;
} else {
/* Multi-line template inclusion : only remove opening tag and parameter separators until eventually first nested tag */
int nestedOpenTagIndex = processedLine.indexOf(WIKI_OPEN_METADATA, openIndex + LEN_WIKI_OPEN_METADATA);
if(nestedOpenTagIndex >= 0) {
processedLine.replace(openIndex, nestedOpenTagIndex,
WIKI_FORMATTED
+ processedLine.substring(openIndex + LEN_WIKI_OPEN_METADATA, nestedOpenTagIndex)
.replace(WIKI_METADATA_PARAMETER_SEPARATOR, ' ').replace('=', ' '));
fromIndex = openIndex;
} else {
processedLine.replace(openIndex, processedLine.length(), WIKI_FORMATTED
+ processedLine.substring(openIndex + LEN_WIKI_OPEN_METADATA).replace(WIKI_METADATA_PARAMETER_SEPARATOR, ' ').replace('=', ' '));
break;
}
}
}
/* Handle any eventual multi-line template remaining closing tags */
fromIndex = 0;
while ((closeIndex = processedLine.indexOf(WIKI_CLOSE_METADATA, fromIndex)) >= 0) {
processedLine.replace(fromIndex, closeIndex, processedLine.substring(fromIndex, closeIndex).replace(WIKI_METADATA_PARAMETER_SEPARATOR, ' ').replace('=', ' '));
processedLine.delete(closeIndex, closeIndex + LEN_WIKI_CLOSE_METADATA);
fromIndex = closeIndex;
}
/* Handle any eventual multi-line template remaining parameter lines */
String result = processedLine.toString();
if(result.matches("^\\s*\\" + WIKI_METADATA_PARAMETER_SEPARATOR + "\\s*[^\\-\\}\\|].*")) {
result = result.replace(WIKI_METADATA_PARAMETER_SEPARATOR, ' ').replace('=', ' ');
}
return result;
}
private class TableOfContent {

87
test/java/net/yacy/data/wiki/WikiCodeTest.java
Unescape View File

@ -1,6 +1,7 @@
package net.yacy.data.wiki;
import org.junit.Test;
import static org.junit.Assert.*;
@ -10,16 +11,18 @@ public class WikiCodeTest {
* test geo location metadata convert
*/
@Test
public void testProcessMetadata() {
public void testProcessMetadataCoordinates() {
String[] testmeta = new String[]{
"{{coordinate|NS=52.205944|EW=0.117593|region=GB-CAM|type=landmark}}", // decimal N-E location
"{{coordinate|NS=43/50/29/N|EW=73/23/17/W|type=landmark|region=US-NY}}", // N-W location
"{{Coordinate |text=DMS |NS=50/7/49/N |EW=6/8/09/E |type=landmark |region=BE-WLG |name=Monument des trois Frontières}}",
"{{Coordinate |text=DMS |NS= 49.047169|EW=7.899148|region=DE-RP |type=landmark |name=Europadenkmal (Rheinland-Pfalz)}}",
"{{Coordinate |text=DMS |NS= 49.047169|EW=7.899148|region=DE-RP |type=landmark |name={{de}}Europadenkmal (Rheinland-Pfalz)}}",// with nested language template
"{{coordinate|NS=0.00000|EW=0.117593}}", // testing equator coord
"{{coordinate|NS=-10.00000|EW=-10.10000}}" // testing S-E location
"{{coordinate|NS=-10.00000|EW=-10.10000}}", // testing S-E location
"{{coordinate|NS=12a5|EW=-10.10000}}" // testing malformed coordinates value
};
WikiCode wc = new WikiCode();
@ -27,9 +30,87 @@ public class WikiCodeTest {
String result = wc.transform("http://wiki:8080",testmeta[i]);
System.out.println(testmeta[i] + " --> " + result);
// simply check if replacement took place, if no coordinate recognized original string is just html encoded
assertFalse(result.contains("#124;")); // simple check - result not containing char code for "{",
assertFalse(result.contains("#123;")); // simple check - result not containing char code for "{",
assertFalse(result.contains("#125;")); // simple check - result not containing char code for "}"
}
}
/**
* Test multi-line template inclusion processing
*/
@Test
public void testTransformMultilineTemplateInclusion() {
String wikitext = "{{Infobox|Example\n"
+ "<!-- *** Name section *** -->\n"
+ "| name = Example\n"
+ "| category = [[Infobox Examples|Example]]\n"
+ "<!-- *** Website *** -->\n"
+ "| website = {{URL|http://example.com}}\n"
+ "}}";
WikiCode wc = new WikiCode();
String result = wc.transform("http://wiki:8080", wikitext);
System.out.println(wikitext + " --> " + result);
assertFalse(result.contains("#123;")); // simple check - result not containing char code for "{",
assertFalse(result.contains("#125;")); // simple check - result not containing char code for "}"
}
/**
* Test single line template inclusion processing
*/
@Test
public void testProcessMetadataTransclusion() {
final String[] wikitexts = new String[]{
"{{Like}}", // most simple template inclusion
"{{Stochastic processes}}", // page name including space
"{{:Stochastic processes}}", // page inclusion with implicit namespace
"{{WP:Assume good faith}}", // page inclusion from Wikipedia namespace
"{{Pagename|parameter1|parameter2|parameter3}}", // with unnamed parameters
"{{Pagename|parameter1=value1|parameter2=value2|parameter3=value3}}", // with named parameters
"{{Template|This is the title text|This is a custom warning line}}", // with parameters including spaces
"{{Special:Recentchangeslinked/General}}", // subpage inclusion
"{{Template1}} text {{Template2}} {{Template3|parameter value1|param2}}", // multiple templates on the same line
"{{Template|[[Page]]}}", // with link parameter
"{{Template|parameter1={{en}}value1|parameter2}}", // nested template inclusion
"{{Template|parameter1={{en|param1|param2=val2}}value1}}", // nested template with parameters inclusion
"{{Template", // Multi-line template inclusion beginning
"simple text {{Template", // Multi-line template inclusion beginning with text before
"{{Template|parameter1={{en}} value1", // Multi-line template inclusion beginning with nested tag
"{{Template|parameter1={{subTemplate", // Multi-line nested template inclusion
"|parameter", // Multi-line template inclusion unnamed parameter line
"|parameter=value", // Multi-line template inclusion named parameter line
"|parameter={{subTemplate|param1|param2}}value", // Multi-line template inclusion with nested template inclusion
"|[[Page]]", // Multi-line template inclusion with unnamed link parameter
"|parameter=[[Page]]", // Multi-line template inclusion with named link parameter
"}}", // Multi-line template inclusion closing
"|lastParameter}}", // Multi-line template inclusion closing with unnamed parameter
"|lastParameter=value}}", // Multi-line template inclusion closing with named parameter
"|lastParameter={{en}}value}}", // Multi-line template inclusion closing with nested tag
"}}}}" // Multi-line nested template inclusion closing
};
for (String wikitext : wikitexts) {
String result = WikiCode.processMetadata(wikitext);
System.out.println(wikitext + " --> " + result);
// simply check if replacement took place
assertFalse(result.contains("{"));
assertFalse(result.contains("|"));
assertFalse(result.contains("="));
assertFalse(result.contains("}"));
}
final String[] wikitextsNotToModify = new String[]{
"", // empty string
"Simple text",
"<pre>Simple preformatted text</pre>",
"[[Page]]", // link
"{|", // table start
"|-", // new table line
"||", // table cell divider
"|}", // table end
};
for (String wikitext : wikitextsNotToModify) {
assertEquals("Text sould not have been modified", wikitext, WikiCode.processMetadata(wikitext));
}
}
/**

Write Preview
Loading…
Cancel
Save