optionally, shorten monomer names also for mmCIF files

because old programs may not support 5-character residue names

gemmi convert: added option --shorten-tlc (TLC=three-letter-code, which
actually may not be three-, and may have digits)

shortened and original names are stored in
_chem_comp.id and _chem_comp.three_letter_code.

When reading a file, original names are automatically restored.

Author: wojdyr

docs/convert-help.txt

@@ -37,6 +37,7 @@ Any output options:
   --shorten               Shorten chain names to 1 (if # < 63) or 2 characters.
   --rename-chain=OLD:NEW  Rename chain OLD to NEW (--rename-chain=:A adds
                           missing chain IDs).
+  --shorten-tlc           Change 5-character monomer names to 3-char. aliases.
   --monomer=OLD:NEW       Change monomer name (CCD code) OLD to NEW.
   -s FILE                 Use sequence(s) from FILE in PIR or FASTA format. Each
                           chain is assigned the best matching sequence, if any.

docs/mol.rst

@@ -1992,11 +1992,16 @@ with two functions:
 
 * ``restore_full_ccd_codes()`` restores the original names.
 
-When reading a PDB file with the tilde-hetnam extension,
-the long names are restored automatically. Apart from this,
-switching between long and short names requires function calls.
+When reading a file with monomer names shortened in a gemmi-compatible way:
 
-Internally, the mapping between old and new names is stored in
+* the tilde-hetnam extension in PDB
+* shortened and original names in ``_chem_comp.id`` and
+  ``_chem_comp.three_letter_code`` in mmCIF,
+
+the long names are automatically restored. Apart from this,
+switching between the long and short names requires function calls.
+
+Internally, the mapping between names is stored in
 ``Structure::shortened_ccd_codes``.
 
 .. doctest::

prog/convert.cpp

@@ -47,7 +47,7 @@ enum OptionIndex {
   ExpandNcs, AsAssembly,
   RemoveH, RemoveWaters, RemoveLigWat, TrimAla, Select, Remove, ApplySymop,
   Reframe, ShortTer, Linkr, CopyRemarks, Minimal, ShortenCN, RenameChain,
-  ChangeCcdCode, SetSeq,
+  ShortenTLC, ChangeCcdCode, SetSeq,
   SiftsNum, Biso, Anisou, SetCis, SegmentAsChain, OldPdb, ForceLabel
 };
 
@@ -103,6 +103,8 @@ const option::Descriptor Usage[] = {
   { RenameChain, 0, "", "rename-chain", Arg::ColonPair,
     "  --rename-chain=OLD:NEW  \tRename chain OLD to NEW "
     "(--rename-chain=:A adds missing chain IDs)." },
+  { ShortenTLC, 0, "", "shorten-tlc", Arg::None,
+    "  --shorten-tlc  \tChange 5-character monomer names to 3-char. aliases." },
   { ChangeCcdCode, 0, "", "monomer", Arg::ColonPair,
     "  --monomer=OLD:NEW  \tChange monomer name (CCD code) OLD to NEW." },
   { SetSeq, 0, "s", "", Arg::Required,
@@ -359,6 +361,9 @@ void convert(gemmi::Structure& st,
     for (gemmi::Model& model : st.models)
       split_chains_by_segments(model, gemmi::HowToNameCopiedChain::Dup);
 
+  if (options[ShortenTLC] || output_type == CoorFormat::Pdb)
+    shorten_ccd_codes(st);
+
   gemmi::Ofstream os(output, &std::cout);
 
   if (output_type == CoorFormat::Mmcif || output_type == CoorFormat::Mmjson) {
@@ -383,7 +388,6 @@ void convert(gemmi::Structure& st,
       writer.write_json(doc);
     }
   } else if (output_type == CoorFormat::Pdb) {
-    shorten_ccd_codes(st);
     gemmi::PdbWriteOptions opt;
     if (options[Minimal])
       opt = gemmi::PdbWriteOptions::minimal();

src/mmcif.cpp

@@ -5,6 +5,7 @@
 #include <gemmi/mmcif_impl.hpp> // for set_cell_from_mmcif
 #include <gemmi/atox.hpp>    // for string_to_int
 #include <gemmi/enumstr.hpp> // for entity_type_from_string, polymer_type_from_string
+#include <gemmi/polyheur.hpp>  // for restore_full_ccd_codes
 
 namespace gemmi {
 
@@ -895,6 +896,17 @@ Structure make_structure_from_block(const cif::Block& block_) {
   st.assemblies = read_assemblies(block);
   read_sifts_unp(block, st);
 
+  cif::Table chem_comp_table = block.find("_chem_comp.", {"id", "three_letter_code"});
+  if (chem_comp_table.ok()) {
+    for (auto row : chem_comp_table) {
+      std::string alias = row.str(0);
+      std::string long_id = row.str(1);
+      if (!alias.empty() && !long_id.empty() && alias != long_id && alias.back() == '~')
+        st.shortened_ccd_codes.emplace_back(long_id, alias);
+    }
+    restore_full_ccd_codes(st);
+  }
+
   return st;
 }
 

src/to_mmcif.cpp

@@ -583,8 +583,18 @@ void update_mmcif_block(const Structure& st, cif::Block& block, MmcifOutputGroup
       for (const std::string& item : ent.full_sequence)
         resnames.insert(Entity::first_mon(item));
     cif::Loop& chem_comp_loop = block.init_mmcif_loop("_chem_comp.", {"id", "type"});
-    for (const std::string& name : resnames)
-      chem_comp_loop.add_row({cif::quote(name), "."});
+    if (!st.shortened_ccd_codes.empty())
+      chem_comp_loop.tags.push_back("_chem_comp.three_letter_code");
+    for (const std::string& name : resnames) {
+      chem_comp_loop.values.push_back(cif::quote(name));
+      chem_comp_loop.values.push_back(".");
+      if (!st.shortened_ccd_codes.empty()) {
+        chem_comp_loop.values.push_back(cif::quote(name));
+        for (const auto& old_new : st.shortened_ccd_codes)
+          if (old_new.second == name)
+            chem_comp_loop.values.back() = old_new.first;
+      }
+    }
   }
 
   if (groups.exptl) {