*/
package jalview.util;
+import static jalview.util.UrlConstants.DB_ACCESSION;
+import static jalview.util.UrlConstants.DELIM;
+import static jalview.util.UrlConstants.SEP;
+import static jalview.util.UrlConstants.SEQUENCE_ID;
+
+import jalview.datamodel.DBRefEntry;
+import jalview.datamodel.SequenceI;
+
+import java.util.Arrays;
+import java.util.List;
+import java.util.Map;
import java.util.Vector;
public class UrlLink
* Jalview 2.4 extension allows regular expressions to be used to parse ID
* strings and replace the result in the URL. Regex's operate on the whole ID
* string given to the matchURL method, if no regex is supplied, then only
- * text following the first pipe symbol will be susbstituted. Usage
+ * text following the first pipe symbol will be substituted. Usage
* documentation todo.
*/
- private String url_suffix, url_prefix, target, label, regexReplace;
+
+ private static final String EQUALS = "=";
+
+ private static final String SPACE = " ";
+
+ private String urlSuffix;
+
+ private String urlPrefix;
+
+ private String target;
+
+ private String label;
+
+ private String dbname;
+
+ private String regexReplace;
private boolean dynamic = false;
+ private boolean usesDBaccession = false;
+
private String invalidMessage = null;
/**
- * parse the given linkString of the form '<label>|<url>' into parts url may
+ * parse the given linkString of the form '<label>SEP<url>' into parts url may
* contain a string $SEQUENCE_ID<=optional regex=>$ where <=optional regex=>
* must be of the form =/<perl style regex>/=$
*
*/
public UrlLink(String link)
{
- int sep = link.indexOf("|"), psqid = link.indexOf("$SEQUENCE_ID");
+ int sep = link.indexOf(SEP);
+ int psqid = link.indexOf(DELIM + DB_ACCESSION);
+ int nsqid = link.indexOf(DELIM + SEQUENCE_ID);
if (psqid > -1)
{
dynamic = true;
- int p = sep;
- do
- {
- sep = p;
- p = link.indexOf("|", sep + 1);
- } while (p > sep && p < psqid);
- // Assuming that the URL itself does not contain any '|' symbols
- // sep now contains last pipe symbol position prior to any regex symbols
- label = link.substring(0, sep);
- if (label.indexOf("|") > -1)
- {
- // | terminated database name / www target at start of Label
- target = label.substring(0, label.indexOf("|"));
- }
- else if (label.indexOf(" ") > 2)
+ usesDBaccession = true;
+
+ sep = parseLabel(sep, psqid, link);
+
+ int endOfRegex = parseUrl(link, DB_ACCESSION, psqid, sep);
+ parseTarget(link, sep, endOfRegex);
+ }
+ else if (nsqid > -1)
+ {
+ dynamic = true;
+ sep = parseLabel(sep, nsqid, link);
+
+ int endOfRegex = parseUrl(link, SEQUENCE_ID, nsqid, sep);
+
+ parseTarget(link, sep, endOfRegex);
+ }
+ else
+ {
+ label = link.substring(0, sep).trim();
+
+ // if there's a third element in the url link string
+ // it is the target name, otherwise target=label
+ int lastsep = link.lastIndexOf(SEP);
+ if (lastsep != sep)
{
- // space separated Label - matches database name
- target = label.substring(0, label.indexOf(" "));
+ urlPrefix = link.substring(sep + 1, lastsep).trim();
+ target = link.substring(lastsep + 1).trim();
}
else
{
+ urlPrefix = link.substring(sep + 1).trim();
target = label;
}
- // Parse URL : Whole URL string first
- url_prefix = link.substring(sep + 1, psqid);
- if (link.indexOf("$SEQUENCE_ID=/") == psqid
- && (p = link.indexOf("/=$", psqid + 14)) > psqid + 14)
- {
- // Extract Regex and suffix
- url_suffix = link.substring(p + 3);
- regexReplace = link.substring(psqid + 14, p);
- try
- {
- com.stevesoft.pat.Regex rg = com.stevesoft.pat.Regex.perlCode("/"
- + regexReplace + "/");
- if (rg == null)
- {
- invalidMessage = "Invalid Regular Expression : '"
- + regexReplace + "'\n";
- }
- } catch (Exception e)
- {
- invalidMessage = "Invalid Regular Expression : '" + regexReplace
- + "'\n";
- }
- }
- else
- {
- regexReplace = null;
- // verify format is really correct.
- if (link.indexOf("$SEQUENCE_ID$") == psqid)
- {
- url_suffix = link.substring(psqid + 13);
- regexReplace = null;
- }
- else
- {
- invalidMessage = "Warning: invalid regex structure for URL link : "
- + link;
- }
- }
- }
- else
- {
- target = link.substring(0, sep);
- label = link.substring(0, sep = link.lastIndexOf("|"));
- url_prefix = link.substring(sep + 1);
+
regexReplace = null; // implies we trim any prefix if necessary //
- // regexReplace=".*\\|?(.*)";
- url_suffix = null;
+ urlSuffix = null;
}
+
+ label = label.trim();
+ target = target.trim();
+ }
+
+ /**
+ * Alternative constructor for separate name, link and description
+ *
+ * @param name
+ * The string used to match the link to a DB reference id
+ * @param url
+ * The url to link to
+ * @param desc
+ * The description of the associated target DB
+ */
+ public UrlLink(String name, String url, String desc)
+ {
+ this(name + SEP + url + SEP + desc);
}
/**
* @return the url_suffix
*/
- public String getUrl_suffix()
+ public String getUrlSuffix()
{
- return url_suffix;
+ return urlSuffix;
}
/**
* @return the url_prefix
*/
- public String getUrl_prefix()
+ public String getUrlPrefix()
{
- return url_prefix;
+ return urlPrefix;
}
/**
return label;
}
+ public String getUrlWithToken()
+ {
+ String var = (usesDBaccession ? DB_ACCESSION : SEQUENCE_ID);
+
+ return urlPrefix
+ + (dynamic ? (DELIM + var + ((regexReplace != null) ? EQUALS
+ + regexReplace + EQUALS + DELIM : DELIM)) : "")
+ + ((urlSuffix == null) ? "" : urlSuffix);
+ }
+
/**
* @return the regexReplace
*/
}
/**
+ *
+ * @return whether link is dynamic
+ */
+ public boolean isDynamic()
+ {
+ return dynamic;
+ }
+
+ /**
+ *
+ * @return whether link uses DB Accession id
+ */
+ public boolean usesDBAccession()
+ {
+ return usesDBaccession;
+ }
+
+ /**
+ * Set the label
+ *
+ * @param newlabel
+ */
+ public void setLabel(String newlabel)
+ {
+ this.label = newlabel;
+ }
+
+ /**
+ * Set the target
+ *
+ * @param desc
+ */
+ public void setTarget(String desc)
+ {
+ target = desc;
+ }
+
+ /**
* return one or more URL strings by applying regex to the given idstring
*
* @param idstring
{
// take whole regex
return new String[] { rg.stringMatched(),
- url_prefix + rg.stringMatched() + url_suffix };
+ urlPrefix + rg.stringMatched() + urlSuffix };
} /*
* else if (ns==1) { // take only subgroup match return new String[]
* { rg.stringMatched(1), url_prefix+rg.stringMatched(1)+url_suffix
+ rg.stringMatched(s) + "'");
}
// try to collate subgroup matches
- Vector subs = new Vector();
+ Vector<String> subs = new Vector<String>();
// have to loop through submatches, collating them at top level
// match
int s = 0; // 1;
if (mtch.length() > 0)
{
subs.addElement(mtch);
- subs.addElement(url_prefix + mtch + url_suffix);
+ subs.addElement(urlPrefix + mtch + urlSuffix);
}
s = r;
}
if (rg.matchedFrom(s) > -1)
{
subs.addElement(rg.stringMatched(s));
- subs.addElement(url_prefix + rg.stringMatched(s)
- + url_suffix);
+ subs.addElement(urlPrefix + rg.stringMatched(s)
+ + urlSuffix);
}
s++;
}
String[] res = new String[subs.size()];
for (int r = 0, rs = subs.size(); r < rs; r++)
{
- res[r] = (String) subs.elementAt(r);
+ res[r] = subs.elementAt(r);
}
subs.removeAllElements();
return res;
}
}
/* Otherwise - trim off any 'prefix' - pre 2.4 Jalview behaviour */
- if (idstring.indexOf("|") > -1)
+ if (idstring.indexOf(SEP) > -1)
{
- idstring = idstring.substring(idstring.lastIndexOf("|") + 1);
+ idstring = idstring.substring(idstring.lastIndexOf(SEP) + 1);
}
// just return simple url substitution.
- return new String[] { idstring, url_prefix + idstring + url_suffix };
+ return new String[] { idstring, urlPrefix + idstring + urlSuffix };
}
else
{
- return new String[] { "", url_prefix };
+ return new String[] { "", urlPrefix };
}
}
+ @Override
public String toString()
{
- return label
- + "|"
- + url_prefix
- + (dynamic ? ("$SEQUENCE_ID" + ((regexReplace != null) ? "="
- + regexReplace + "=$" : "$")) : "")
- + ((url_suffix == null) ? "" : url_suffix);
+ return label + SEP + getUrlWithToken();
+ }
+ /**
+ * @return delimited string containing label, url and target
+ */
+ public String toStringWithTarget()
+ {
+ return label + SEP + getUrlWithToken() + SEP + target;
}
- private static void testUrls(UrlLink ul, String idstring, String[] urls)
+ /**
+ * Parse the label from the link string
+ *
+ * @param firstSep
+ * Location of first occurrence of separator in link string
+ * @param psqid
+ * Position of sequence id or name in link string
+ * @param link
+ * Link string containing database name and url
+ * @return Position of last separator symbol prior to any regex symbols
+ */
+ protected int parseLabel(int firstSep, int psqid, String link)
+ {
+ int p = firstSep;
+ int sep = firstSep;
+ do
+ {
+ sep = p;
+ p = link.indexOf(SEP, sep + 1);
+ } while (p > sep && p < psqid);
+ // Assuming that the URL itself does not contain any SEP symbols
+ // sep now contains last pipe symbol position prior to any regex symbols
+ label = link.substring(0, sep);
+
+ return sep;
+ }
+
+ /**
+ * Parse the target from the link string
+ *
+ * @param link
+ * Link string containing database name and url
+ * @param sep
+ * Location of first separator symbol
+ * @param endOfRegex
+ * Location of end of any regular expression in link string
+ */
+ protected void parseTarget(String link, int sep, int endOfRegex)
{
+ int lastsep = link.lastIndexOf(SEP);
- if (urls == null)
+ if ((lastsep != sep) && (lastsep > endOfRegex))
{
- System.out.println("Created NO urls.");
+ // final element in link string is the target
+ target = link.substring(lastsep + 1).trim();
}
else
{
- System.out.println("Created " + (urls.length / 2) + " Urls.");
- for (int uls = 0; uls < urls.length; uls += 2)
- {
- System.out.println("URL Replacement text : " + urls[uls]
- + " : URL : " + urls[uls + 1]);
- }
+ target = label;
+ }
+
+ if (target.indexOf(SEP) > -1)
+ {
+ // SEP terminated database name / www target at start of Label
+ target = target.substring(0, target.indexOf(SEP));
+ }
+ else if (target.indexOf(SPACE) > 2)
+ {
+ // space separated label - first word matches database name
+ target = target.substring(0, target.indexOf(SPACE));
}
}
- public static void main(String argv[])
+ /**
+ * Parse the URL part of the link string
+ *
+ * @param link
+ * Link string containing database name and url
+ * @param varName
+ * Name of variable in url string (e.g. SEQUENCE_ID, SEQUENCE_NAME)
+ * @param sqidPos
+ * Position of id or name in link string
+ * @param sep
+ * Position of separator in link string
+ * @return Location of end of any regex in link string
+ */
+ protected int parseUrl(String link, String varName, int sqidPos, int sep)
{
- String[] links = new String[] {
- /*
- * "AlinkT|Target|http://foo.foo.soo/",
- * "myUrl1|http://$SEQUENCE_ID=/[0-9]+/=$.someserver.org/foo",
- * "myUrl2|http://$SEQUENCE_ID=/(([0-9]+).+([A-Za-z]+))/=$.someserver.org/foo"
- * ,
- * "myUrl3|http://$SEQUENCE_ID=/([0-9]+).+([A-Za-z]+)/=$.someserver.org/foo"
- * , "myUrl4|target|http://$SEQUENCE_ID$.someserver.org/foo|too",
- * "PF1|http://us.expasy.org/cgi-bin/niceprot.pl?$SEQUENCE_ID=/(?:PFAM:)?(.+)/=$"
- * ,
- * "PF2|http://us.expasy.org/cgi-bin/niceprot.pl?$SEQUENCE_ID=/(PFAM:)?(.+)/=$"
- * ,
- * "PF3|http://us.expasy.org/cgi-bin/niceprot.pl?$SEQUENCE_ID=/PFAM:(.+)/=$"
- * , "NOTFER|http://notfer.org/$SEQUENCE_ID=/(?<!\\s)(.+)/=$",
- */
- "NESTED|http://nested/$SEQUENCE_ID=/^(?:Label:)?(?:(?:gi\\|(\\d+))|([^:]+))/=$/nested" };
- String[] idstrings = new String[] {
- /*
- * //"LGUL_human", //"QWIQW_123123", "uniprot|why_do+_12313_foo",
- * //"123123312", "123123 ABCDE foo", "PFAM:PF23943",
- */
- "Label:gi|9234|pdb|102L|A" };
- // TODO: test the setLabel method.
- for (int i = 0; i < links.length; i++)
+ urlPrefix = link.substring(sep + 1, sqidPos).trim();
+
+ // delimiter at start of regex: e.g. $SEQUENCE_ID=/
+ String startDelimiter = DELIM + varName + "=/";
+
+ // delimiter at end of regex: /=$
+ String endDelimiter = "/=" + DELIM;
+
+ int startLength = startDelimiter.length();
+
+ // Parse URL : Whole URL string first
+ int p = link.indexOf(endDelimiter, sqidPos + startLength);
+
+ if (link.indexOf(startDelimiter) == sqidPos
+ && (p > sqidPos + startLength))
{
- UrlLink ul = new UrlLink(links[i]);
- if (ul.isValid())
+ // Extract Regex and suffix
+ urlSuffix = link.substring(p + endDelimiter.length());
+ regexReplace = link.substring(sqidPos + startLength, p);
+ try
{
- System.out.println("\n\n\n");
- System.out.println("Link " + i + " " + links[i] + " : "
- + ul.toString());
- System.out.println(" pref : "
- + ul.getUrl_prefix()
- + "\n suf : "
- + ul.getUrl_suffix()
- + "\n : "
- + ((ul.getRegexReplace() != null) ? ul.getRegexReplace()
- : ""));
- for (int ids = 0; ids < idstrings.length; ids++)
+ com.stevesoft.pat.Regex rg = com.stevesoft.pat.Regex.perlCode("/"
+ + regexReplace + "/");
+ if (rg == null)
{
- System.out.println("ID String : " + idstrings[ids]
- + "\nWithout onlyIfMatches:");
- String[] urls = ul.makeUrls(idstrings[ids], false);
- testUrls(ul, idstrings[ids], urls);
- System.out.println("With onlyIfMatches set.");
- urls = ul.makeUrls(idstrings[ids], true);
- testUrls(ul, idstrings[ids], urls);
+ invalidMessage = "Invalid Regular Expression : '" + regexReplace
+ + "'\n";
}
+ } catch (Exception e)
+ {
+ invalidMessage = "Invalid Regular Expression : '" + regexReplace
+ + "'\n";
+ }
+ }
+ else
+ {
+ // no regex
+ regexReplace = null;
+ // verify format is really correct.
+ if (link.indexOf(DELIM + varName + DELIM) == sqidPos)
+ {
+ int lastsep = link.lastIndexOf(SEP);
+ if (lastsep < sqidPos + startLength - 1)
+ {
+ // the last SEP character was before the regex, ignore
+ lastsep = link.length();
+ }
+ urlSuffix = link.substring(sqidPos + startLength - 1, lastsep)
+ .trim();
+ regexReplace = null;
}
else
{
- System.err.println("Invalid URLLink : " + links[i] + " : "
- + ul.getInvalidMessage());
+ invalidMessage = "Warning: invalid regex structure for URL link : "
+ + link;
}
}
+
+ return p;
}
- public boolean isDynamic()
+ /**
+ * Create a set of URL links for a sequence
+ *
+ * @param seq
+ * The sequence to create links for
+ * @param linkset
+ * Map of links: key = id + SEP + link, value = [target, label, id,
+ * link]
+ */
+ public void createLinksFromSeq(final SequenceI seq,
+ Map<String, List<String>> linkset)
{
- // TODO Auto-generated method stub
- return dynamic;
+ if (seq != null && dynamic)
+ {
+ createDynamicLinks(seq, linkset);
+ }
+ else
+ {
+ createStaticLink(linkset);
+ }
}
- public void setLabel(String newlabel)
+ /**
+ * Create a static URL link
+ *
+ * @param linkset
+ * Map of links: key = id + SEP + link, value = [target, label, id,
+ * link]
+ */
+ protected void createStaticLink(Map<String, List<String>> linkset)
{
- this.label = newlabel;
+ if (!linkset.containsKey(label + SEP + getUrlPrefix()))
+ {
+ // Add a non-dynamic link
+ linkset.put(label + SEP + getUrlPrefix(),
+ Arrays.asList(target, label, null, getUrlPrefix()));
+ }
+ }
+
+ /**
+ * Create dynamic URL links
+ *
+ * @param seq
+ * The sequence to create links for
+ * @param linkset
+ * Map of links: key = id + SEP + link, value = [target, label, id,
+ * link]
+ */
+ protected void createDynamicLinks(final SequenceI seq,
+ Map<String, List<String>> linkset)
+ {
+ // collect id string too
+ String id = seq.getName();
+ String descr = seq.getDescription();
+ if (descr != null && descr.length() < 1)
+ {
+ descr = null;
+ }
+
+ if (usesDBAccession()) // link is ID
+ {
+ // collect matching db-refs
+ DBRefEntry[] dbr = DBRefUtils.selectRefs(seq.getDBRefs(),
+ new String[] { target });
+
+ // if there are any dbrefs which match up with the link
+ if (dbr != null)
+ {
+ for (int r = 0; r < dbr.length; r++)
+ {
+ // create Bare ID link for this URL
+ createBareURLLink(dbr[r].getAccessionId(), true, linkset);
+ }
+ }
+ }
+ else if (!usesDBAccession() && id != null) // link is name
+ {
+ // create Bare ID link for this URL
+ createBareURLLink(id, false, linkset);
+ }
+
+ // Create urls from description but only for URL links which are regex
+ // links
+ if (descr != null && getRegexReplace() != null)
+ {
+ // create link for this URL from description where regex matches
+ createBareURLLink(descr, false, linkset);
+ }
+ }
+
+ /*
+ * Create a bare URL Link
+ * Returns map where key = id + SEP + link, and value = [target, label, id, link]
+ */
+ protected void createBareURLLink(String id, Boolean combineLabel,
+ Map<String, List<String>> linkset)
+ {
+ String[] urls = makeUrls(id, true);
+ if (urls != null)
+ {
+ for (int u = 0; u < urls.length; u += 2)
+ {
+ if (!linkset.containsKey(urls[u] + SEP + urls[u + 1]))
+ {
+ String thisLabel = label;
+ if (combineLabel)
+ {
+ // incorporate label with idstring
+ thisLabel = label + SEP + urls[u];
+ }
+
+ linkset.put(urls[u] + SEP + urls[u + 1],
+ Arrays.asList(target, thisLabel, urls[u], urls[u + 1]));
+ }
+ }
+ }
}
}
git://source.jalview.org / jalview.git / blobdiff