/*
- * Jalview - A Sequence Alignment Editor and Viewer (Version 2.9.0b2)
- * Copyright (C) 2015 The Jalview Authors
+ * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
+ * Copyright (C) $$Year-Rel$$ The Jalview Authors
*
* This file is part of Jalview.
*
*/
package jalview.util;
+import static jalview.util.UrlConstants.DB_ACCESSION;
+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
private boolean dynamic = false;
+ private boolean uses_db_accession = false;
+
private String invalidMessage = null;
/**
*/
public UrlLink(String link)
{
- int sep = link.indexOf("|"), psqid = link.indexOf("$SEQUENCE_ID");
+ int sep = link.indexOf("|");
+ int psqid = link.indexOf("$" + DB_ACCESSION);
+ int nsqid = link.indexOf("$" + SEQUENCE_ID);
if (psqid > -1)
{
dynamic = true;
- int p = sep;
- do
- {
- sep = p;
- p = link.indexOf("|", sep + 1);
- } while (p > sep && p < psqid);
- // Assuming that the URL itself does not contain any '|' symbols
- // sep now contains last pipe symbol position prior to any regex symbols
- label = link.substring(0, sep);
- if (label.indexOf("|") > -1)
- {
- // | terminated database name / www target at start of Label
- target = label.substring(0, label.indexOf("|"));
- }
- else if (label.indexOf(" ") > 2)
- {
- // space separated Label - matches database name
- target = label.substring(0, label.indexOf(" "));
- }
- else
- {
- target = label;
- }
- // Parse URL : Whole URL string first
- url_prefix = link.substring(sep + 1, psqid);
- if (link.indexOf("$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$") == psqid)
- {
- url_suffix = link.substring(psqid + 13);
- regexReplace = null;
- }
- else
- {
- invalidMessage = "Warning: invalid regex structure for URL link : "
- + link;
- }
- }
+ uses_db_accession = true;
+
+ sep = parseTargetAndLabel(sep, psqid, link);
+
+ parseUrl(link, DB_ACCESSION, psqid, sep);
+ }
+ else if (nsqid > -1)
+ {
+ dynamic = true;
+ sep = parseTargetAndLabel(sep, nsqid, link);
+
+ parseUrl(link, SEQUENCE_ID, nsqid, sep);
}
else
{
target = link.substring(0, sep);
- label = link.substring(0, sep = link.lastIndexOf("|"));
- url_prefix = link.substring(sep + 1);
+ sep = link.lastIndexOf("|");
+ label = link.substring(0, sep);
+ url_prefix = link.substring(sep + 1).trim();
regexReplace = null; // implies we trim any prefix if necessary //
// regexReplace=".*\\|?(.*)";
url_suffix = null;
}
+
+ label = label.trim();
+ target = target.trim();
+ target = target.toUpperCase(); // DBRefEntry uppercases DB names
+ // NB getCanonicalName might be better but does not currently change case
}
/**
}
}
+ @Override
public String toString()
{
+ String var = (uses_db_accession ? DB_ACCESSION : SEQUENCE_ID);
+
return label
+ "|"
+ url_prefix
- + (dynamic ? ("$SEQUENCE_ID" + ((regexReplace != null) ? "="
+ + (dynamic ? ("$" + var + ((regexReplace != null) ? "="
+ regexReplace + "=$" : "$")) : "")
+ ((url_suffix == null) ? "" : url_suffix);
+ }
+
+ /**
+ *
+ * @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 parseTargetAndLabel(int firstSep, int psqid, String link)
+ {
+ int p = firstSep;
+ int sep = firstSep;
+ 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)
+ {
+ // space separated Label - matches database name
+ target = label.substring(0, label.indexOf(" "));
+ }
+ else
+ {
+ target = label;
+ }
+ return sep;
+ }
+
+ /**
+ * 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
+ */
+ protected void parseUrl(String link, String varName, int sqidPos, int sep)
+ {
+ url_prefix = link.substring(sep + 1, sqidPos).trim();
+
+ // delimiter at start of regex: e.g. $SEQUENCE_ID=/
+ String startDelimiter = "$" + varName + "=/";
+
+ // delimiter at end of regex: /=$
+ String endDelimiter = "/=$";
+
+ int startLength = startDelimiter.length();
+
+ // Parse URL : Whole URL string first
+ int p = link.indexOf(endDelimiter, sqidPos + startLength);
+
+ if (link.indexOf(startDelimiter) == sqidPos
+ && (p > sqidPos + startLength))
+ {
+ // Extract Regex and suffix
+ url_suffix = link.substring(p + endDelimiter.length());
+ regexReplace = link.substring(sqidPos + startLength, p);
+ try
+ {
+ com.stevesoft.pat.Regex rg = com.stevesoft.pat.Regex.perlCode("/"
+ + regexReplace + "/");
+ if (rg == null)
+ {
+ invalidMessage = "Invalid Regular Expression : '" + regexReplace
+ + "'\n";
+ }
+ } catch (Exception e)
+ {
+ invalidMessage = "Invalid Regular Expression : '" + regexReplace
+ + "'\n";
+ }
+ }
+ else
+ {
+ // no regex
+ regexReplace = null;
+ // verify format is really correct.
+ if (link.indexOf("$" + varName + "$") == sqidPos)
+ {
+ url_suffix = link.substring(sqidPos + startLength - 1);
+ regexReplace = null;
+ }
+ else
+ {
+ invalidMessage = "Warning: invalid regex structure for URL link : "
+ + link;
+ }
+ }
+ }
+
+ /**
+ * Create a set of URL links for a sequence
+ *
+ * @param seq
+ * The sequence to create links for
+ * @param linkset
+ * Map of links: key = id | link, value = [target, label, id, link]
+ */
+ public void createLinksFromSeq(final SequenceI seq,
+ Map<String, List<String>> linkset)
+ {
+ if (seq != null && dynamic)
+ {
+ createDynamicLinks(seq, linkset);
+ }
+ else
+ {
+ createStaticLink(linkset);
+ }
+ }
+
+ /**
+ * Create a static URL link
+ *
+ * @param linkset
+ * Map of links: key = id | link, value = [target, label, id, link]
+ */
+ protected void createStaticLink(Map<String, List<String>> linkset)
+ {
+ if (!linkset.containsKey(label + "|" + getUrl_prefix()))
+ {
+ // Add a non-dynamic link
+ linkset.put(label + "|" + getUrl_prefix(),
+ Arrays.asList(target, label, null, getUrl_prefix()));
+ }
+ }
+ /**
+ * Create dynamic URL links
+ *
+ * @param seq
+ * The sequence to create links for
+ * @param linkset
+ * Map of links: key = id | 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 | 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] + "|" + urls[u + 1]))
+ {
+ String thisLabel = label;
+ if (combineLabel)
+ {
+ // incorporate label with idstring
+ thisLabel = label + "|" + urls[u];
+ }
+
+ linkset.put(urls[u] + "|" + urls[u + 1],
+ Arrays.asList(target, thisLabel, urls[u], urls[u + 1]));
+ }
+ }
+ }
}
private static void testUrls(UrlLink ul, String idstring, String[] urls)
* "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+))|([^:]+))/=$/nested" };
+ "NESTED|http://nested/$" + DB_ACCESSION
+ + "=/^(?:Label:)?(?:(?:gi\\|(\\d+))|([^:]+))/=$/nested" };
String[] idstrings = new String[] {
/*
* //"LGUL_human", //"QWIQW_123123", "uniprot|why_do+_12313_foo",
public boolean isDynamic()
{
- // TODO Auto-generated method stub
return dynamic;
}
+ public boolean usesDBAccession()
+ {
+ return uses_db_accession;
+ }
+
public void setLabel(String newlabel)
{
this.label = newlabel;
git://source.jalview.org / jalview.git / blobdiff