Examples with Perl5Matcher org.apache.oro.text.regex.Perl5Matcher used on opensource projects

Example 16 with Perl5Matcher

use of org.apache.oro.text.regex.Perl5Matcher in project Lucee by lucee.

the class Perl5Util method match.

public static Array match(String strPattern, String strInput, int offset, boolean caseSensitive) throws MalformedPatternException {
    Perl5Matcher matcher = new Perl5Matcher();
    PatternMatcherInput input = new PatternMatcherInput(strInput);
    int compileOptions = caseSensitive ? 0 : Perl5Compiler.CASE_INSENSITIVE_MASK;
    compileOptions += Perl5Compiler.MULTILINE_MASK;
    if (offset < 1)
        offset = 1;
    Pattern pattern = getPattern(strPattern, compileOptions);
    Array rtn = new ArrayImpl();
    MatchResult result;
    while (matcher.contains(input, pattern)) {
        result = matcher.getMatch();
        rtn.appendEL(result.toString());
    }
    return rtn;
}

Example 17 with Perl5Matcher

use of org.apache.oro.text.regex.Perl5Matcher in project nutch by apache.

the class OutlinkExtractor method getOutlinks.

/**
 * Extracts <code>Outlink</code> from given plain text and adds anchor to the
 * extracted <code>Outlink</code>s
 *
 * @param plainText
 *          the plain text from wich URLs should be extracted.
 * @param anchor
 *          the anchor of the url
 *
 * @return Array of <code>Outlink</code>s within found in plainText
 */
public static Outlink[] getOutlinks(final String plainText, String anchor, Configuration conf) {
    long start = System.currentTimeMillis();
    final List<Outlink> outlinks = new ArrayList<>();
    try {
        final PatternCompiler cp = new Perl5Compiler();
        final Pattern pattern = cp.compile(URL_PATTERN, Perl5Compiler.CASE_INSENSITIVE_MASK | Perl5Compiler.READ_ONLY_MASK | Perl5Compiler.MULTILINE_MASK);
        final PatternMatcher matcher = new Perl5Matcher();
        final PatternMatcherInput input = new PatternMatcherInput(plainText);
        MatchResult result;
        String url;
        // loop the matches
        while (matcher.contains(input, pattern)) {
            // do not unnecessarily hit this limit.)
            if (System.currentTimeMillis() - start >= 60000L) {
                if (LOG.isWarnEnabled()) {
                    LOG.warn("Time limit exceeded for getOutLinks");
                }
                break;
            }
            result = matcher.getMatch();
            url = result.group(0);
            try {
                outlinks.add(new Outlink(url, anchor));
            } catch (MalformedURLException mue) {
                LOG.warn("Invalid url: '" + url + "', skipping.");
            }
        }
    } catch (Exception ex) {
        // on
        if (LOG.isErrorEnabled()) {
            LOG.error("getOutlinks", ex);
        }
    }
    final Outlink[] retval;
    // create array of the Outlinks
    if (outlinks != null && outlinks.size() > 0) {
        retval = outlinks.toArray(new Outlink[0]);
    } else {
        retval = new Outlink[0];
    }
    return retval;
}

Example 18 with Perl5Matcher

use of org.apache.oro.text.regex.Perl5Matcher in project jspwiki by apache.

the class AbstractReferralPlugin method filterCollection.

/**
 *  Filters a collection according to the include and exclude parameters.
 *
 *  @param c The collection to filter.
 *  @return A filtered collection.
 */
protected Collection filterCollection(Collection c) {
    ArrayList<Object> result = new ArrayList<Object>();
    PatternMatcher pm = new Perl5Matcher();
    for (Iterator i = c.iterator(); i.hasNext(); ) {
        String pageName = null;
        Object objectje = i.next();
        if (objectje instanceof WikiPage) {
            pageName = ((WikiPage) objectje).getName();
        } else {
            pageName = (String) objectje;
        }
        // 
        // If include parameter exists, then by default we include only those
        // pages in it (excluding the ones in the exclude pattern list).
        // 
        // include='*' means the same as no include.
        // 
        boolean includeThis = m_include == null;
        if (m_include != null) {
            for (int j = 0; j < m_include.length; j++) {
                if (pm.matches(pageName, m_include[j])) {
                    includeThis = true;
                    break;
                }
            }
        }
        if (m_exclude != null) {
            for (int j = 0; j < m_exclude.length; j++) {
                if (pm.matches(pageName, m_exclude[j])) {
                    includeThis = false;
                    // The inner loop, continue on the next item
                    break;
                }
            }
        }
        if (includeThis) {
            if (objectje instanceof WikiPage) {
                result.add(objectje);
            } else {
                result.add(pageName);
            }
            // 
            // if we want to show the last modified date of the most recently change page, we keep a "high watermark" here:
            WikiPage page = null;
            if (m_lastModified) {
                page = m_engine.getPage(pageName);
                if (page != null) {
                    Date lastModPage = page.getLastModified();
                    if (log.isDebugEnabled()) {
                        log.debug("lastModified Date of page " + pageName + " : " + m_dateLastModified);
                    }
                    if (lastModPage.after(m_dateLastModified)) {
                        m_dateLastModified = lastModPage;
                    }
                }
            }
        }
    }
    return result;
}

Example 19 with Perl5Matcher

use of org.apache.oro.text.regex.Perl5Matcher in project ofbiz-framework by apache.

the class RegexpCondition method checkCondition.

@Override
public boolean checkCondition(MethodContext methodContext) throws MiniLangException {
    Object fieldVal = fieldFma.get(methodContext.getEnvMap());
    if (fieldVal == null) {
        fieldVal = "";
    } else if (!(fieldVal instanceof String)) {
        try {
            fieldVal = MiniLangUtil.convertType(fieldVal, String.class, methodContext.getLocale(), methodContext.getTimeZone(), null);
        } catch (Exception e) {
            throw new MiniLangRuntimeException(e, this);
        }
    }
    String regExp = exprFse.expandString(methodContext.getEnvMap());
    Pattern pattern = null;
    try {
        pattern = PatternFactory.createOrGetPerl5CompiledPattern(regExp, true);
    } catch (MalformedPatternException e) {
        Debug.logError(e, "Regular Expression [" + regExp + "] is mal-formed: " + e.toString(), module);
        throw new MiniLangRuntimeException(e, this);
    }
    PatternMatcher matcher = new Perl5Matcher();
    if (matcher.matches((String) fieldVal, pattern)) {
        // Debug.logInfo("The string [" + fieldVal + "] matched the pattern expr [" + pattern.getPattern() + "]", module);
        return true;
    } else {
        // Debug.logInfo("The string [" + fieldVal + "] did NOT match the pattern expr [" + pattern.getPattern() + "]", module);
        return false;
    }
}

Example 20 with Perl5Matcher

use of org.apache.oro.text.regex.Perl5Matcher in project ofbiz-framework by apache.

the class CatalogUrlSeoTransform method getNiceName.

/**
 * Get a string lower cased and hyphen connected.
 *
 * @param name a String to be transformed
 * @return String nice name
 */
protected static String getNiceName(String name) {
    Perl5Matcher matcher = new Perl5Matcher();
    String niceName = null;
    if (UtilValidate.isNotEmpty(name)) {
        name = name.trim().replaceAll(" ", URL_HYPHEN);
        if (UtilValidate.isNotEmpty(name) && matcher.matches(name, asciiPattern)) {
            niceName = name;
        }
    }
    return niceName;
}