X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fanalysis%2FCrossRef.java;h=9b90ca578660434ca954e1eba5ace4a82c54dfed;hb=57738a1f3c19b1c3a00bd3ac5108f8cd0af32f99;hp=e6bae9b1a0c31c5f8057109d188bf69df5072763;hpb=f4766a7bbcfae845fc95923b01fa14ff83d589ff;p=jalview.git

diff --git a/src/jalview/analysis/CrossRef.java b/src/jalview/analysis/CrossRef.java
index e6bae9b..9b90ca5 100644
--- a/src/jalview/analysis/CrossRef.java
+++ b/src/jalview/analysis/CrossRef.java
@@ -99,7 +99,7 @@ public class CrossRef
    */
   public List<String> findXrefSourcesForSequences(boolean dna)
   {
-    List<String> sources = new ArrayList<String>();
+    List<String> sources = new ArrayList<>();
     for (SequenceI seq : fromSeqs)
     {
       if (seq != null)
@@ -143,15 +143,17 @@ public class CrossRef
     /*
      * first find seq's xrefs (dna-to-peptide or peptide-to-dna)
      */
-    DBRefEntry[] rfs = DBRefUtils.selectDbRefs(!fromDna, seq.getDBRefs());
+    List<DBRefEntry> rfs = DBRefUtils.selectDbRefs(!fromDna,
+            seq.getDBRefs());
     addXrefsToSources(rfs, sources);
     if (dataset != null)
     {
       /*
        * find sequence's direct (dna-to-dna, peptide-to-peptide) xrefs
        */
-      DBRefEntry[] lrfs = DBRefUtils.selectDbRefs(fromDna, seq.getDBRefs());
-      List<SequenceI> foundSeqs = new ArrayList<SequenceI>();
+      List<DBRefEntry> lrfs = DBRefUtils.selectDbRefs(fromDna,
+              seq.getDBRefs());
+      List<SequenceI> foundSeqs = new ArrayList<>();
 
       /*
        * find sequences in the alignment which xref one of these DBRefs
@@ -164,7 +166,7 @@ public class CrossRef
        */
       for (SequenceI rs : foundSeqs)
       {
-        DBRefEntry[] xrs = DBRefUtils.selectDbRefs(!fromDna,
+        List<DBRefEntry> xrs = DBRefUtils.selectDbRefs(!fromDna,
                 rs.getDBRefs());
         addXrefsToSources(xrs, sources);
       }
@@ -178,7 +180,7 @@ public class CrossRef
    * @param xrefs
    * @param sources
    */
-  void addXrefsToSources(DBRefEntry[] xrefs, List<String> sources)
+  void addXrefsToSources(List<DBRefEntry> xrefs, List<String> sources)
   {
     if (xrefs != null)
     {
@@ -218,7 +220,7 @@ public class CrossRef
   public Alignment findXrefSequences(String source, boolean fromDna)
   {
 
-    rseqs = new ArrayList<SequenceI>();
+    rseqs = new ArrayList<>();
     AlignedCodonFrame cf = new AlignedCodonFrame();
     matcher = new SequenceIdMatcher(dataset.getSequences());
 
@@ -230,18 +232,18 @@ public class CrossRef
         dss = dss.getDatasetSequence();
       }
       boolean found = false;
-      DBRefEntry[] xrfs = DBRefUtils.selectDbRefs(!fromDna,
+      List<DBRefEntry> xrfs = DBRefUtils.selectDbRefs(!fromDna,
               dss.getDBRefs());
       // ENST & ENSP comes in to both Protein and nucleotide, so we need to
       // filter them
       // out later.
-      if ((xrfs == null || xrfs.length == 0) && dataset != null)
+      if ((xrfs == null || xrfs.size() == 0) && dataset != null)
       {
         /*
          * found no suitable dbrefs on sequence - look for sequences in the
          * alignment which share a dbref with this one
          */
-        DBRefEntry[] lrfs = DBRefUtils.selectDbRefs(fromDna,
+        List<DBRefEntry> lrfs = DBRefUtils.selectDbRefs(fromDna,
                 seq.getDBRefs());
 
         /*
@@ -370,7 +372,8 @@ public class CrossRef
         {
           // do a bit more work - search for sequences with references matching
           // xrefs on this sequence.
-          found = searchDataset(fromDna, dss, xref, rseqs, cf, false);
+          found = searchDataset(fromDna, dss, xref, rseqs, cf, false,
+                  DBRefUtils.SEARCH_MODE_FULL);
         }
         if (found)
         {
@@ -400,7 +403,7 @@ public class CrossRef
   }
 
   private void retrieveCrossRef(List<DBRefEntry> sourceRefs, SequenceI seq,
-          DBRefEntry[] xrfs, boolean fromDna, AlignedCodonFrame cf)
+          List<DBRefEntry> xrfs, boolean fromDna, AlignedCodonFrame cf)
   {
     ASequenceFetcher sftch = SequenceFetcherFactory.getSequenceFetcher();
     SequenceI[] retrieved = null;
@@ -430,8 +433,8 @@ public class CrossRef
     if (retrieved != null)
     {
       boolean addedXref = false;
-      List<SequenceI> newDsSeqs = new ArrayList<SequenceI>(),
-              doNotAdd = new ArrayList<SequenceI>();
+      List<SequenceI> newDsSeqs = new ArrayList<>(),
+              doNotAdd = new ArrayList<>();
 
       for (SequenceI retrievedSequence : retrieved)
       {
@@ -443,6 +446,11 @@ public class CrossRef
         addedXref |= importCrossRefSeq(cf, newDsSeqs, doNotAdd, dss,
                 retrievedDss);
       }
+      // JBPNote: What assumptions are made for dbref structures on
+      // retrieved sequences ?
+      // addedXref will be true means importCrossRefSeq found
+      // sequences with dbrefs with mappings to sequences congruent with dss
+
       if (!addedXref)
       {
         // try again, after looking for matching IDs
@@ -483,19 +491,25 @@ public class CrossRef
   private void removeAlreadyRetrievedSeqs(List<DBRefEntry> sourceRefs,
           boolean fromDna)
   {
-    DBRefEntry[] dbrSourceSet = sourceRefs.toArray(new DBRefEntry[0]);
-    for (SequenceI sq : dataset.getSequences())
+    List<DBRefEntry> dbrSourceSet = new ArrayList<>(sourceRefs);
+    List<SequenceI> dsSeqs = dataset.getSequences();
+    for (int ids = 0, nds = dsSeqs.size(); ids < nds; ids++)
     {
+      SequenceI sq = dsSeqs.get(ids);
       boolean dupeFound = false;
       // !fromDna means we are looking only for nucleotide sequences, not
       // protein
       if (sq.isProtein() == fromDna)
       {
-        for (DBRefEntry dbr : sq.getPrimaryDBRefs())
+        List<DBRefEntry> sqdbrefs = sq.getPrimaryDBRefs();
+        for (int idb = 0, ndb = sqdbrefs.size(); idb < ndb; idb++)
         {
-          for (DBRefEntry found : DBRefUtils.searchRefs(dbrSourceSet, dbr))
+          DBRefEntry dbr = sqdbrefs.get(idb);
+          List<DBRefEntry> searchrefs = DBRefUtils.searchRefs(dbrSourceSet,
+                  dbr, DBRefUtils.SEARCH_MODE_FULL);
+          for (int isr = 0, nsr = searchrefs.size(); isr < nsr; isr++)
           {
-            sourceRefs.remove(found);
+            sourceRefs.remove(searchrefs.get(isr));
             dupeFound = true;
           }
         }
@@ -503,14 +517,17 @@ public class CrossRef
       if (dupeFound)
       {
         // rebuild the search array from the filtered sourceRefs list
-        dbrSourceSet = sourceRefs.toArray(new DBRefEntry[0]);
+        dbrSourceSet.clear();
+        dbrSourceSet.addAll(sourceRefs);
       }
     }
   }
 
   /**
    * process sequence retrieved via a dbref on source sequence to resolve and
-   * transfer data
+   * transfer data JBPNote: as of 2022-02-03 - this assumes retrievedSequence
+   * has dbRefs with Mapping references to a sequence congruent with
+   * sourceSequence
    * 
    * @param cf
    * @param sourceSequence
@@ -526,11 +543,14 @@ public class CrossRef
      * sourceSequence
      */
     boolean imported = false;
-    DBRefEntry[] dbr = retrievedSequence.getDBRefs();
+    List<DBRefEntry> dbr = retrievedSequence.getDBRefs();
     if (dbr != null)
     {
-      for (DBRefEntry dbref : dbr)
+      for (int ib = 0, nb = dbr.size(); ib < nb; ib++)
       {
+
+        DBRefEntry dbref = dbr.get(ib);
+        // matched will return null if the dbref has no map
         SequenceI matched = findInDataset(dbref);
         if (matched == sourceSequence)
         {
@@ -542,9 +562,10 @@ public class CrossRef
         Mapping map = dbref.getMap();
         if (map != null)
         {
-          if (map.getTo() != null && map.getMap() != null)
+          SequenceI ms = map.getTo();
+          if (ms != null && map.getMap() != null)
           {
-            if (map.getTo() == sourceSequence)
+            if (ms == sourceSequence)
             {
               // already called to import once, and most likely this sequence
               // already imported !
@@ -555,7 +576,7 @@ public class CrossRef
               /*
                * sequence is new to dataset, so save a reference so it can be added. 
                */
-              newDsSeqs.add(map.getTo());
+              newDsSeqs.add(ms);
               continue;
             }
 
@@ -567,7 +588,6 @@ public class CrossRef
             {
               // compare ms with dss and replace with dss in mapping
               // if map is congruent
-              SequenceI ms = map.getTo();
               // TODO findInDataset requires exact sequence match but
               // 'congruent' test is only for the mapped part
               // maybe not a problem in practice since only ENA provide a
@@ -589,7 +609,7 @@ public class CrossRef
                         + matched.getName();
                 System.out.println(msg);
 
-                DBRefEntry[] toRefs = map.getTo().getDBRefs();
+                List<DBRefEntry> toRefs = map.getTo().getDBRefs();
                 if (toRefs != null)
                 {
                   /*
@@ -682,7 +702,7 @@ public class CrossRef
     {
       return;
     }
-    DBRefEntry[] dbrefs = mapTo.getDBRefs();
+    List<DBRefEntry> dbrefs = mapTo.getDBRefs();
     if (dbrefs == null)
     {
       return;
@@ -707,7 +727,8 @@ public class CrossRef
   /**
    * Returns null or the first sequence in the dataset which is identical to
    * xref.mapTo, and has a) a primary dbref matching xref, or if none found, the
-   * first one with an ID source|xrefacc
+   * first one with an ID source|xrefacc JBPNote: Could refactor this to
+   * AlignmentI/DatasetI
    * 
    * @param xref
    *          with map and mapped-to sequence
@@ -738,8 +759,8 @@ public class CrossRef
     for (SequenceI seq : dataset.getSequences())
     {
       // first check primary refs.
-      List<DBRefEntry> match = DBRefUtils.searchRefs(
-              seq.getPrimaryDBRefs().toArray(new DBRefEntry[0]), template);
+      List<DBRefEntry> match = DBRefUtils.searchRefs(seq.getPrimaryDBRefs(),
+              template, DBRefUtils.SEARCH_MODE_FULL);
       if (match != null && match.size() == 1 && sameSequence(seq, dss))
       {
         return seq;
@@ -805,14 +826,15 @@ public class CrossRef
   /**
    * Updates any empty mappings in the cross-references with one to a compatible
    * retrieved sequence if found, and adds any new mappings to the
-   * AlignedCodonFrame
+   * AlignedCodonFrame JBPNote: TODO: this relies on sequence IDs like
+   * UNIPROT|ACCESSION - which do not always happen.
    * 
    * @param mapFrom
    * @param xrefs
    * @param retrieved
    * @param acf
    */
-  void updateDbrefMappings(SequenceI mapFrom, DBRefEntry[] xrefs,
+  void updateDbrefMappings(SequenceI mapFrom, List<DBRefEntry> xrefs,
           SequenceI[] retrieved, AlignedCodonFrame acf, boolean fromDna)
   {
     SequenceIdMatcher idMatcher = new SequenceIdMatcher(retrieved);
@@ -921,7 +943,7 @@ public class CrossRef
 
     if (fromDna)
     {
-      AlignmentUtils.computeProteinFeatures(mapFrom, mapTo, mapping);
+      // AlignmentUtils.computeProteinFeatures(mapFrom, mapTo, mapping);
       mappings.addMap(mapFrom, mapTo, mapping);
     }
     else
@@ -946,7 +968,7 @@ public class CrossRef
    * @return true if matches were found.
    */
   private boolean searchDatasetXrefs(boolean fromDna, SequenceI sequenceI,
-          DBRefEntry[] lrfs, List<SequenceI> foundSeqs,
+          List<DBRefEntry> lrfs, List<SequenceI> foundSeqs,
           AlignedCodonFrame cf)
   {
     boolean found = false;
@@ -954,14 +976,14 @@ public class CrossRef
     {
       return false;
     }
-    for (int i = 0; i < lrfs.length; i++)
+    for (int i = 0, n = lrfs.size(); i < n; i++)
     {
-      DBRefEntry xref = new DBRefEntry(lrfs[i]);
-      // add in wildcards
-      xref.setVersion(null);
-      xref.setMap(null);
-      found |= searchDataset(fromDna, sequenceI, xref, foundSeqs, cf,
-              false);
+      // DBRefEntry xref = new DBRefEntry(lrfs.get(i));
+      // // add in wildcards
+      // xref.setVersion(null);
+      // xref.setMap(null);
+      found |= searchDataset(fromDna, sequenceI, lrfs.get(i), foundSeqs, cf,
+              false, DBRefUtils.SEARCH_MODE_NO_MAP_NO_VERSION);
     }
     return found;
   }
@@ -992,11 +1014,13 @@ public class CrossRef
    *          sequenceI or all the returned sequences (eg a genomic reference
    *          associated with a locus and one or more transcripts)</li>
    *          </ul>
+   * @param mode
+   *          SEARCH_MODE_FULL for all; SEARCH_MODE_NO_MAP_NO_VERSION optional
    * @return true if relationship found and sequence added.
    */
   boolean searchDataset(boolean fromDna, SequenceI fromSeq, DBRefEntry xrf,
           List<SequenceI> foundSeqs, AlignedCodonFrame mappings,
-          boolean direct)
+          boolean direct, int mode)
   {
     boolean found = false;
     if (dataset == null)
@@ -1008,8 +1032,8 @@ public class CrossRef
       System.err.println("Empty dataset sequence set - NO VECTOR");
       return false;
     }
-    List<SequenceI> ds;
-    synchronized (ds = dataset.getSequences())
+    List<SequenceI> ds = dataset.getSequences();
+    synchronized (ds)
     {
       for (SequenceI nxt : ds)
       {
@@ -1041,13 +1065,13 @@ public class CrossRef
           }
 
           // look for direct or indirect references in common
-          DBRefEntry[] poss = nxt.getDBRefs();
+          List<DBRefEntry> poss = nxt.getDBRefs();
           List<DBRefEntry> cands = null;
 
           // todo: indirect specifies we select either direct references to nxt
           // that match xrf which is indirect to sequenceI, or indirect
           // references to nxt that match xrf which is direct to sequenceI
-          cands = DBRefUtils.searchRefs(poss, xrf);
+          cands = DBRefUtils.searchRefs(poss, xrf, mode);
           // else
           // {
           // poss = DBRefUtils.selectDbRefs(nxt.isProtein()!fromDna, poss);