gistfile1.diff

diff --git a/src/extended/gff3_out_stream.c b/src/extended/gff3_out_stream.c
index 6dbbc23..f813e8f 100644
--- a/src/extended/gff3_out_stream.c
+++ b/src/extended/gff3_out_stream.c
@@ -18,6 +18,7 @@
 #include "extended/gff3_out_stream.h"
 #include "extended/gff3_visitor.h"
 #include "extended/node_stream_rep.h"
+#include "core/cstr_table.h"
 
 struct GtGFF3OutStream {
   const GtNodeStream parent_instance;
@@ -76,3 +77,11 @@ void gt_gff3_out_stream_set_fasta_width(GtNodeStream *gs,
   gt_assert(gff3_out_stream);
   gt_gff3_visitor_set_fasta_width(gff3_out_stream->gff3_visitor, fasta_width);
 }
+
+void gt_gff3_out_stream_retain_id_attributes(GtNodeStream *gs)
+{
+  GtGFF3OutStream *gff3_out_stream = gff3_out_stream_cast(gs);
+  gt_assert(gff3_out_stream);
+  gt_gff3_visitor_retain_id_attributes(gff3_out_stream->gff3_visitor);
+
+}
diff --git a/src/extended/gff3_out_stream.h b/src/extended/gff3_out_stream.h
index adc3eba..e004c5d 100644
--- a/src/extended/gff3_out_stream.h
+++ b/src/extended/gff3_out_stream.h
@@ -28,5 +28,7 @@ const GtNodeStreamClass* gt_gff3_out_stream_class(void);
 GtNodeStream*            gt_gff3_out_stream_new(GtNodeStream*, GtGenFile*);
 void                     gt_gff3_out_stream_set_fasta_width(GtNodeStream*,
                                                             unsigned long);
+void                     gt_gff3_out_stream_retain_id_attributes(
+                                                      GtNodeStream *);
 
 #endif
diff --git a/src/extended/gff3_visitor.c b/src/extended/gff3_visitor.c
index 8246680..1a26b72 100644
--- a/src/extended/gff3_visitor.c
+++ b/src/extended/gff3_visitor.c
@@ -23,21 +23,27 @@
 #include "core/ma.h"
 #include "core/unused_api.h"
 #include "core/string_distri.h"
+#include "core/cstr_table.h"
+#include "core/str_api.h"
+#include "core/warning_api.h"
 #include "extended/genome_node.h"
 #include "extended/gff3_output.h"
 #include "extended/gff3_parser.h"
 #include "extended/gff3_visitor.h"
 #include "extended/node_visitor_rep.h"
+#include <stdbool.h>
 
 struct GtGFF3Visitor {
   const GtNodeVisitor parent_instance;
   bool version_string_shown,
+       retain_ids,
        fasta_directive_shown;
   GtStringDistri *id_counter;
   GtHashmap *gt_feature_node_to_id_array,
             *gt_feature_node_to_unique_id_str;
   unsigned long fasta_width;
   GtGenFile *outfp;
+  GtCstrTable *gt_used_ids;
 };
 
 typedef struct {
@@ -71,6 +77,7 @@ static void gff3_visitor_free(GtNodeVisitor *gv)
   gt_string_distri_delete(gff3_visitor->id_counter);
   gt_hashmap_delete(gff3_visitor->gt_feature_node_to_id_array);
   gt_hashmap_delete(gff3_visitor->gt_feature_node_to_unique_id_str);
+  gt_cstr_table_delete(gff3_visitor->gt_used_ids);
 }
 
 static int gff3_visitor_comment_node(GtNodeVisitor *gv, GtCommentNode *cn,
@@ -190,18 +197,66 @@ static GtStr* create_unique_id(GtGFF3Visitor *gff3_visitor, GtFeatureNode *gf)
   return id;
 }
 
+static void make_unique_id_string(GtStr *current_id, unsigned long counter)
+{
+  /* name => name.1 */
+  gt_str_append_char(current_id, '.');
+  gt_str_append_ulong(current_id, counter);
+}
+
+static bool id_string_is_unique(GtStr *id, GtStr *buf, GtCstrTable *tab,
+                                unsigned long i)
+{
+  gt_str_reset(buf);
+  gt_str_append_str(buf, id);
+  make_unique_id_string(buf, i);
+  return (gt_cstr_table_get(tab, gt_str_get(buf)) == NULL);
+}
+static void make_id_unique(GtGFF3Visitor *gff3_visitor, GtStr *id)
+{
+  unsigned long i = 1;
+
+  if (gt_cstr_table_get(gff3_visitor->gt_used_ids, gt_str_get(id)))
+  {
+    GtStr *buf = gt_str_new();
+    while (!id_string_is_unique(id, buf, gff3_visitor->gt_used_ids, i))
+    {
+      i++;
+    }
+    gt_warning("feature ID \"%s\" not unique: changing to %s", gt_str_get(id),
+                                                   gt_str_get(buf));
+    /* update table with the new id */
+    gt_str_set(id, gt_str_get(buf));
+    gt_str_delete(buf);
+  }
+  gt_cstr_table_add(gff3_visitor->gt_used_ids, gt_str_get(id));
+}
+
 static int store_ids(GtGenomeNode *gn, void *data, GtError *err)
 {
   GtGFF3Visitor *gff3_visitor = (GtGFF3Visitor*) data;
   GtFeatureNode *gf = (GtFeatureNode*) gn;
   AddIDInfo add_id_info;
   int had_err = 0;
-  GtStr *id;
+  bool has_id = false;
+  const char *id_string = gt_feature_node_get_attribute(gf, "ID");
+  GtStr *id; /* = gt_str_new_cstr(id_string); */
+  bool retain_ids = gff3_visitor->retain_ids;
 
   gt_error_check(err);
   gt_assert(gn && gf && gff3_visitor);
 
-  if (gt_genome_node_has_children(gn) || gt_feature_node_is_multi(gf)) {
+  if (retain_ids && id_string) {
+    id = gt_str_new_cstr(id_string);
+    if (!gt_feature_node_is_multi(gf) ||
+           (gt_feature_node_is_multi(gf)
+            && gt_feature_node_get_multi_representative(gf) == gf)) {
+        make_id_unique(gff3_visitor, id);
+    }
+    gt_hashmap_add(gff3_visitor->gt_feature_node_to_unique_id_str, gf, id);
+    has_id = true;
+  }
+  else if (gt_genome_node_has_children(gn) || gt_feature_node_is_multi(gf)) {
     if (gt_feature_node_is_multi(gf)) {
       id = gt_hashmap_get(gff3_visitor->gt_feature_node_to_unique_id_str,
                           gt_feature_node_get_multi_representative(gf));
@@ -209,14 +264,18 @@ static int store_ids(GtGenomeNode *gn, void *data, GtError *err)
         id = create_unique_id(gff3_visitor,
                               gt_feature_node_get_multi_representative(gf));
       }
+
       if (gt_feature_node_get_multi_representative(gf) != gf) {
         gt_hashmap_add(gff3_visitor->gt_feature_node_to_unique_id_str, gf,
                        gt_str_ref(id));
       }
     }
-    else
+    else {
       id = create_unique_id(gff3_visitor, gf);
-
+    }
+    has_id = true;
+  }
+  if (has_id) {
     /* for each child -> store the parent feature in the hash map */
     add_id_info.gt_feature_node_to_id_array =
       gff3_visitor->gt_feature_node_to_id_array,
@@ -224,6 +283,7 @@ static int store_ids(GtGenomeNode *gn, void *data, GtError *err)
     had_err = gt_genome_node_traverse_direct_children(gn, &add_id_info, add_id,
                                                       err);
   }
+  /* gt_str_delete(id); */
   return had_err;
 }
 
@@ -329,9 +389,18 @@ GtNodeVisitor* gt_gff3_visitor_new(GtGenFile *outfp)
     HASH_DIRECT, NULL, (GtFree) gt_str_delete);
   gff3_visitor->fasta_width = 0;
   gff3_visitor->outfp = outfp;
+  /* if retain_ids is set to true, hen gt_used_ids is .. used. */
+  gff3_visitor->gt_used_ids = gt_cstr_table_new();
+  gff3_visitor->retain_ids = false;
   return gv;
 }
 
+void gt_gff3_visitor_retain_id_attributes(GtNodeVisitor *gv)
+{
+    GtGFF3Visitor *gff3_visitor = gff3_visitor_cast(gv);
+    gff3_visitor->retain_ids = true;
+}
+
 void gt_gff3_visitor_set_fasta_width(GtNodeVisitor *gv,
                                      unsigned long fasta_width)
 {
diff --git a/src/extended/gff3_visitor.h b/src/extended/gff3_visitor.h
index 1939565..530371c 100644
--- a/src/extended/gff3_visitor.h
+++ b/src/extended/gff3_visitor.h
@@ -27,5 +27,6 @@ const GtNodeVisitorClass* gt_gff3_visitor_class(void);
 GtNodeVisitor*            gt_gff3_visitor_new(GtGenFile*);
 void                      gt_gff3_visitor_set_fasta_width(GtNodeVisitor*,
                                                           unsigned long);
+void                      gt_gff3_visitor_retain_id_attributes(GtNodeVisitor *);
 
 #endif
diff --git a/src/tools/gt_gff3.c b/src/tools/gt_gff3.c
index 02e4e35..cc8f7ff 100644
--- a/src/tools/gt_gff3.c
+++ b/src/tools/gt_gff3.c
@@ -35,6 +35,7 @@
 typedef struct {
   bool sort,
        checkids,
+       retainids,
        mergefeat,
        addintrons,
        verbose,
@@ -91,6 +92,14 @@ static GtOptionParser* gt_gff3_option_parser_new(void *tool_arguments)
                            "parsing", &arguments->tidy, false);
   gt_option_parser_add_option(op, option);
 
+  /* -retainids */
+  option = gt_option_new_bool("retainids",
+                           "when available, use the original IDs provided"
+                           "in the source file\n"
+                           "(memory consumption is O(file_size))",
+                           &arguments->retainids, false);
+  gt_option_parser_add_option(op, option);
+
   /* -checkids */
   option = gt_option_new_bool("checkids",
                            "make sure the ID attributes are unique "
@@ -187,6 +196,7 @@ static int gt_gff3_runner(int argc, const char **argv, int parsed_args,
     gt_gff3_in_stream_show_progress_bar((GtGFF3InStream*) gff3_in_stream);
   if (arguments->checkids)
     gt_gff3_in_stream_check_id_attributes((GtGFF3InStream*) gff3_in_stream);
+
   last_stream = gff3_in_stream;
 
   /* set different type checker if necessary */
@@ -243,6 +253,9 @@ static int gt_gff3_runner(int argc, const char **argv, int parsed_args,
     gt_gff3_out_stream_set_fasta_width(gff3_out_stream, arguments->width);
   }
 
+  if (!had_err && arguments->retainids)
+    gt_gff3_out_stream_retain_id_attributes(gff3_out_stream);
+
   /* pull the features through the stream and free them afterwards */
   if (!had_err)
     had_err = gt_node_stream_pull(gff3_out_stream, err);
diff --git a/testdata/multi_feature_simple_retain.gff3 b/testdata/multi_feature_simple_retain.gff3
new file mode 100644
index 0000000..292dfff
--- /dev/null
+++ b/testdata/multi_feature_simple_retain.gff3
@@ -0,0 +1,6 @@
+##gff-version   3
+##sequence-region   ctg123 1 1497228
+ctg123	.	gene	1000	9000	.	+	.	ID=gene1
+ctg123	.	CDS	1201	1500	.	+	0	ID=CDS1;Parent=gene1
+ctg123	.	CDS	3000	3902	.	+	0	ID=CDS1;Parent=gene1
+###
diff --git a/testdata/png_test_2.gff3 b/testdata/png_test_2.gff3
index 5217d18..2aae61d 100644
--- a/testdata/png_test_2.gff3
+++ b/testdata/png_test_2.gff3
@@ -10,3 +10,4 @@ ctg123	.	exon	1050	1500	.	.	.	Parent=mRNA00001,mRNA00002
 ctg123	.	exon	3000	3902	.	.	.	Parent=mRNA00001,mRNA00003
 ctg123	.	exon	5000	5500	.	.	.	Parent=mRNA00001,mRNA00002,mRNA00003
 ctg123	.	exon	7000	9000	.	.	.	Parent=mRNA00001,mRNA00002,mRNA00003
+###
diff --git a/testdata/png_test_2_out.gff3 b/testdata/png_test_2_out.gff3
new file mode 100644
index 0000000..0b20a6b
--- /dev/null
+++ b/testdata/png_test_2_out.gff3
@@ -0,0 +1,13 @@
+##gff-version   3
+##sequence-region   ctg123 1 10000
+ctg123	.	gene	1000	9000	.	.	.	ID=gene00001
+ctg123	.	TF_binding_site	1000	1012	.	.	.	Parent=gene00001
+ctg123	.	mRNA	1050	9000	.	.	.	ID=mRNA00001;Parent=gene00001
+ctg123	.	mRNA	1050	9000	.	.	.	ID=mRNA00002;Parent=gene00001
+ctg123	.	mRNA	1300	9000	.	.	.	ID=mRNA00003;Parent=gene00001
+ctg123	.	exon	1050	1500	.	.	.	Parent=mRNA00001,mRNA00002
+ctg123	.	exon	3000	3902	.	.	.	Parent=mRNA00001,mRNA00003
+ctg123	.	exon	5000	5500	.	.	.	Parent=mRNA00001,mRNA00002,mRNA00003
+ctg123	.	exon	7000	9000	.	.	.	Parent=mRNA00001,mRNA00002,mRNA00003
+ctg123	.	exon	1300	1500	.	.	.	Parent=mRNA00003
+###
diff --git a/testsuite/gt_gff3_include.rb b/testsuite/gt_gff3_include.rb
index 3525dc2..9faa602 100644
--- a/testsuite/gt_gff3_include.rb
+++ b/testsuite/gt_gff3_include.rb
@@ -612,6 +612,20 @@ Test do
   run "diff #{$last_stdout} #{$testdata}two_fasta_seqs.gff3"
 end
 
+Name "gt gff3 (-retainids)"
+Keywords "gt_gff3 retainids"
+Test do
+  run_test "#{$bin}gt gff3 -retainids #{$testdata}png_test_2.gff3"
+  run "diff #{$last_stdout} #{$testdata}png_test_2_out.gff3"
+end
+
+Name "gt gff3 multi-feature (-retainids)"
+Keywords "gt_gff3 multi-feature retainids"
+Test do
+  run_test "#{$bin}gt gff3 -retainids #{$testdata}multi_feature_simple.gff3"
+  run "diff #{$last_stdout} #{$testdata}multi_feature_simple_retain.gff3"
+end
+
 Name "gt gff3 simple multi-feature (round-trip)"
 Keywords "gt_gff3 multi-feature"
 Test do