Save paths for alts using plain variant ID (when possible)

src/constructor.cpp

@@ -696,7 +696,7 @@ namespace vg {
 
                     // Name the variant and place it in the order that we'll
                     // actually construct nodes in (see utility.hpp)
-                    string variant_name = make_variant_id(*variant);
+                    string variant_name = sha1_variant_name ? make_variant_id(*variant) : get_or_make_variant_id(*variant);
                     if (variants_by_name.count(variant_name)) {
                         // Some VCFs may include multiple variants at the same
                         // position with the same ref and alt. We will only take the

src/constructor.hpp

@@ -80,6 +80,7 @@ class Constructor : public Progressive, public NameMapper {
     // _alt_6079b4a76d0ddd6b4b44aeb14d738509e266961c_0 and
     // _alt_6079b4a76d0ddd6b4b44aeb14d738509e266961c_1?
     bool alt_paths = false;
+    bool sha1_variant_name = true;
 
     // Should we handle structural variants in the VCF file,
     // or at least the ones we know how to?

src/subcommand/construct_main.cpp

@@ -25,7 +25,9 @@ void help_construct(char** argv) {
          << "    -r, --reference FILE   input FASTA reference (may repeat)" << endl
          << "    -v, --vcf FILE         input VCF (may repeat)" << endl
          << "    -n, --rename V=F       match contig V in the VCFs to contig F in the FASTAs (may repeat)" << endl
-         << "    -a, --alt-paths        save paths for alts of variants by variant ID" << endl
+         << "    -a, --alt-paths        save paths for alts of variants by SHA1 hash" << endl
+         << "    -A, --alt-paths-plain  save paths for alts of variants by variant ID if possible, otherwise SHA1" << endl
+         << "                           (IDs must be unique across all input VCFs)" << endl
          << "    -R, --region REGION    specify a VCF contig name or 1-based inclusive region (may repeat, if on different contigs)" << endl
          << "    -C, --region-is-chrom  don't attempt to parse the regions (use when the reference" << endl
          << "                           sequence name could be inadvertently parsed as a region)" << endl
@@ -87,6 +89,7 @@ int main_construct(int argc, char** argv) {
                 {"drop-msa-paths", no_argument, 0, 'd'},
                 {"rename", required_argument, 0, 'n'},
                 {"alt-paths", no_argument, 0, 'a'},
+                {"alt-paths-plain", no_argument, 0, 'A'},
                 {"handle-sv", no_argument, 0, 'S'},
                 {"insertions", required_argument, 0, 'I'},
                 {"progress",  no_argument, 0, 'p'},
@@ -103,7 +106,7 @@ int main_construct(int argc, char** argv) {
             };
 
         int option_index = 0;
-        c = getopt_long (argc, argv, "v:r:n:ph?z:t:R:m:aCfl:SI:M:dF:iN",
+        c = getopt_long (argc, argv, "v:r:n:ph?z:t:R:m:aACfl:SI:M:dF:iN",
                          long_options, &option_index);
 
         /* Detect the end of the options. */
@@ -170,6 +173,11 @@ int main_construct(int argc, char** argv) {
             constructor.alt_paths = true;
             break;
 
+        case 'A':
+            constructor.alt_paths = true;
+            constructor.sha1_variant_name = false;
+            break;
+
         case 'p':
             show_progress = true;
             break;