Make blank node prefixing automatic

Though potentially useful, I don't think the complexity cost of the old
interface (both to the implementation and to the user) is worth it.  A special
tool to transform blank node labels (for example with regular expressions)
would be a better approach to this if it's ever needed in the future.

17 files changed, 30 insertions, 227 deletions

diff --git a/doc/man/serd-pipe.1 b/doc/man/serd-pipe.1
index 056d236f..b1e01990 100644
--- a/doc/man/serd-pipe.1
+++ b/doc/man/serd-pipe.1
@@ -13,9 +13,7 @@
 .Op Fl I Ar syntax
 .Op Fl O Ar syntax
 .Op Fl b Ar bytes
-.Op Fl c Ar prefix
 .Op Fl k Ar bytes
-.Op Fl p Ar prefix
 .Op Fl r Ar root
 .Op Fl s Ar string
 .Op Fl w Ar filename
@@ -162,23 +160,6 @@ I/O block size.
 This is the number of bytes in a file that will be read or written at once.
 The default is 4096, which should perform well in most cases.
 Note that this only applies to files, standard input and output are always processed one byte at a time.
-.It Fl c Ar prefix
-Chop
-.Ar prefix
-from matching blank node IDs.
-This is typically used to revert the effects of
-.Fl p .
-For example, with
-.Ar prefix
-.Dq doc01 ,
-the blank node
-.Li _:doc01b42
-will be emitted as
-.Li _:b42 .
-.It Fl e
-Eat input one character at a time, rather than a page at a time which is the default.
-This is useful when reading from a pipe since output will be generated immediately as input arrives, rather than waiting until an entire page of input has arrived.
-With this option one less page of memory is used, but likely with a performance penalty.
 .It Fl f
 Fast and loose URI mode:
 preserve full URIs (without qualifying or making relative),
@@ -191,20 +172,6 @@ Parsing is performed using a pre-allocated stack for performance and security re
 By default, the stack is 1 MiB, which should be sufficient for most data.
 This can be increased to support unusually structured data and huge literals,
 or decreased to reduce overall memory requirements and reduce startup time.
-.It Fl p Ar prefix
-Add
-.Ar prefix
-to blank node IDs.
-This can be used to avoid clashes between blank node IDs in input documents.
-The effects can be reversed in a later run with
-.Fl c .
-For example, with
-.Ar prefix
-.Dq doc01 ,
-the blank node
-.Li _:b42
-will be emitted as
-.Li _:doc01b42 .
 .It Fl q
 Suppress all output except data.
 .It Fl r Ar root
diff --git a/include/serd/reader.h b/include/serd/reader.h
index 34af6d04..a2bde202 100644
--- a/include/serd/reader.h
+++ b/include/serd/reader.h
@@ -105,19 +105,6 @@ serd_reader_new(SerdWorld* ZIX_NONNULL      world,
                 const SerdSink* ZIX_NONNULL sink);
 
 /**
-   Set a prefix to be added to all blank node identifiers.
-
-   This is useful when multiple files are to be parsed into the same output (a
-   model or a file).  Since Serd preserves blank node IDs, this could cause
-   conflicts where two non-equivalent blank nodes are merged, resulting in
-   corrupt data.  By setting a unique blank node prefix for each parsed file,
-   this can be avoided, while preserving blank node names.
-*/
-SERD_API void
-serd_reader_add_blank_prefix(SerdReader* ZIX_NONNULL  reader,
-                             const char* ZIX_NULLABLE prefix);
-
-/**
    Prepare to read some input.
 
    This sets up the reader to read from the given input, but will not read any
diff --git a/include/serd/writer.h b/include/serd/writer.h
index 812b1851..91baa7e8 100644
--- a/include/serd/writer.h
+++ b/include/serd/writer.h
@@ -103,16 +103,6 @@ SERD_CONST_API const SerdSink* ZIX_NONNULL
 serd_writer_sink(SerdWriter* ZIX_NONNULL writer);
 
 /**
-   Set a prefix to be removed from matching blank node identifiers.
-
-   This is the counterpart to serd_reader_add_blank_prefix() which can be used
-   to "undo" added prefixes.
-*/
-SERD_API void
-serd_writer_chop_blank_prefix(SerdWriter* ZIX_NONNULL  writer,
-                              const char* ZIX_NULLABLE prefix);
-
-/**
    Set the current root URI.
 
    The root URI should be a prefix of the base URI.  The path of the root URI
diff --git a/src/read_ntriples.c b/src/read_ntriples.c
index 57b1e7be..dd5c28fc 100644
--- a/src/read_ntriples.c
+++ b/src/read_ntriples.c
@@ -273,10 +273,8 @@ read_BLANK_NODE_LABEL(SerdReader* const reader,
     return r_err(reader, SERD_BAD_SYNTAX, "expected blank node label");
   }
 
-  if (!(*dest = push_node(reader,
-                          SERD_BLANK,
-                          reader->bprefix ? reader->bprefix : "",
-                          reader->bprefix_len))) {
+  if (!(*dest = push_node(
+          reader, SERD_BLANK, reader->bprefix, reader->bprefix_len))) {
     return SERD_BAD_STACK;
   }
 
diff --git a/src/reader.c b/src/reader.c
index 39e85b47..73509c6f 100644
--- a/src/reader.c
+++ b/src/reader.c
@@ -68,11 +68,10 @@ set_blank_id(SerdReader* const reader,
              SerdNode* const   node,
              const size_t      buf_size)
 {
-  char*       buf    = (char*)(node + 1);
-  const char* prefix = reader->bprefix ? (const char*)reader->bprefix : "";
+  char* const buf = (char*)(node + 1);
 
-  node->length =
-    (size_t)snprintf(buf, buf_size, "%sb%u", prefix, reader->next_id++);
+  node->length = (size_t)snprintf(
+    buf, buf_size, "%sb%u", reader->bprefix, reader->next_id++);
 }
 
 size_t
@@ -201,6 +200,13 @@ serd_reader_read_document(SerdReader* const reader)
     return SERD_BAD_CALL;
   }
 
+  if (!(reader->flags & SERD_READ_GLOBAL)) {
+    reader->bprefix_len = (size_t)snprintf(reader->bprefix,
+                                           sizeof(reader->bprefix),
+                                           "f%u",
+                                           ++reader->world->next_document_id);
+  }
+
   if (reader->syntax != SERD_SYNTAX_EMPTY && !reader->source->prepared) {
     SerdStatus st = serd_reader_prepare(reader);
     if (st) {
@@ -271,6 +277,12 @@ serd_reader_new(SerdWorld* const      world,
   assert(me->rdf_rest);
   assert(me->rdf_nil);
 
+  if (!(flags & SERD_READ_GLOBAL)) {
+    me->bprefix[0]  = 'f';
+    me->bprefix[1]  = '0';
+    me->bprefix_len = 2;
+  }
+
   return me;
 }
 
@@ -286,28 +298,9 @@ serd_reader_free(SerdReader* const reader)
   }
 
   serd_aaligned_free(reader->world->allocator, reader->stack.buf);
-  serd_wfree(reader->world, reader->bprefix);
   serd_wfree(reader->world, reader);
 }
 
-void
-serd_reader_add_blank_prefix(SerdReader* const reader, const char* const prefix)
-{
-  assert(reader);
-
-  serd_wfree(reader->world, reader->bprefix);
-  reader->bprefix_len = 0;
-  reader->bprefix     = NULL;
-
-  const size_t prefix_len = prefix ? strlen(prefix) : 0;
-  if (prefix_len) {
-    reader->bprefix_len = prefix_len;
-    reader->bprefix =
-      (char*)serd_wmalloc(reader->world, reader->bprefix_len + 1);
-    memcpy(reader->bprefix, prefix, reader->bprefix_len + 1);
-  }
-}
-
 static SerdStatus
 skip_bom(SerdReader* const me)
 {
diff --git a/src/reader.h b/src/reader.h
index a98d5ef8..0bb3454b 100644
--- a/src/reader.h
+++ b/src/reader.h
@@ -46,7 +46,7 @@ struct SerdReaderImpl {
   SerdSyntax      syntax;
   SerdReaderFlags flags;
   unsigned        next_id;
-  char*           bprefix;
+  char            bprefix[24];
   size_t          bprefix_len;
   bool            strict; ///< True iff strict parsing
   bool            seen_genid;
diff --git a/src/world.h b/src/world.h
index 2499b761..f615868a 100644
--- a/src/world.h
+++ b/src/world.h
@@ -14,10 +14,11 @@
 #include <stdint.h>
 
 struct SerdWorldImpl {
-  SerdLimits     limits;
   SerdAllocator* allocator;
   SerdLog        log;
+  SerdLimits     limits;
   uint32_t       next_blank_id;
+  uint32_t       next_document_id;
   SerdNode*      blank_node;
 
   bool stderr_color;
diff --git a/src/writer.c b/src/writer.c
index 1de5e055..7afeb1c4 100644
--- a/src/writer.c
+++ b/src/writer.c
@@ -140,8 +140,6 @@ struct SerdWriterImpl {
   size_t          anon_stack_size;
   SerdBlockDumper output;
   WriteContext    context;
-  char*           bprefix;
-  size_t          bprefix_len;
   Sep             last_sep;
   int             indent;
 };
@@ -901,17 +899,7 @@ write_blank(SerdWriter* const        writer,
   }
 
   TRY(st, esink("_:", 2, writer));
-  if (writer->bprefix &&
-      !strncmp(node_str, writer->bprefix, writer->bprefix_len)) {
-    TRY(st,
-        esink(node_str + writer->bprefix_len,
-              node->length - writer->bprefix_len,
-              writer));
-  } else {
-    TRY(st, esink(node_str, node->length, writer));
-  }
-
-  return st;
+  return esink(node_str, node->length, writer);
 }
 
 SERD_NODISCARD static SerdStatus
@@ -1382,25 +1370,7 @@ serd_writer_new(SerdWorld*        world,
   return writer;
 }
 
-void
-serd_writer_chop_blank_prefix(SerdWriter* writer, const char* prefix)
-{
-  assert(writer);
-
-  serd_wfree(writer->world, writer->bprefix);
-  writer->bprefix_len = 0;
-  writer->bprefix     = NULL;
-
-  const size_t prefix_len = prefix ? strlen(prefix) : 0;
-  if (prefix_len) {
-    writer->bprefix_len = prefix_len;
-    writer->bprefix =
-      (char*)serd_wmalloc(writer->world, writer->bprefix_len + 1);
-    memcpy(writer->bprefix, prefix, writer->bprefix_len + 1);
-  }
-}
-
-SERD_NODISCARD static SerdStatus
+static SerdStatus
 serd_writer_set_base_uri(SerdWriter* writer, const SerdNode* uri)
 {
   assert(writer);
@@ -1482,7 +1452,6 @@ serd_writer_free(SerdWriter* writer)
   free_anon_stack(writer);
   serd_block_dumper_close(&writer->output);
   serd_wfree(writer->world, writer->anon_stack);
-  serd_wfree(writer->world, writer->bprefix);
   serd_node_free(writer->world->allocator, writer->root_node);
   serd_wfree(writer->world, writer);
 }
diff --git a/test/extra/prefix/README.md b/test/extra/prefix/README.md
deleted file mode 100644
index 56a07244..00000000
--- a/test/extra/prefix/README.md
+++ /dev/null
@@ -1,4 +0,0 @@
-Prefix Test Suite
-=================
-
-This small suite tests adding/chopping prefixes to/from blank node labels.
diff --git a/test/extra/prefix/manifest.ttl b/test/extra/prefix/manifest.ttl
deleted file mode 100644
index 12a0f9ca..00000000
--- a/test/extra/prefix/manifest.ttl
+++ /dev/null
@@ -1,20 +0,0 @@
-@prefix mf: <http://www.w3.org/2001/sw/DataAccess/tests/test-manifest#> .
-@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
-@prefix rdft: <http://www.w3.org/ns/rdftest#> .
-@prefix serd: <http://drobilla.net/ns/serd#> .
-
-<>
-	a mf:Manifest ;
-	rdfs:comment "Serd blank node prefixing test suite" ;
-	mf:entries (
-		<#remove-prefixes>
-	) .
-
-<#remove-prefixes>
-	a rdft:TestTurtleEval ;
-	mf:action <remove-prefixes.ttl> ;
-	mf:name "remove-prefixes" ;
-	mf:result <remove-prefixes.nt> .
-
-rdft:Test
-	rdfs:subClassOf mf:ManifestEntry .
diff --git a/test/extra/prefix/remove-prefixes.nt b/test/extra/prefix/remove-prefixes.nt
deleted file mode 100644
index 26ac46ee..00000000
--- a/test/extra/prefix/remove-prefixes.nt
+++ /dev/null
@@ -1,4 +0,0 @@
-_:s1 <http://example.org/p1> _:o1 .
-_:s1 <http://example.org/p2> _:o3 .
-_:s1 <http://example.org/p2> _:o4 .
-_:b0 <http://example.org/p3> _:b1 .
diff --git a/test/extra/prefix/remove-prefixes.ttl b/test/extra/prefix/remove-prefixes.ttl
deleted file mode 100644
index f3002a59..00000000
--- a/test/extra/prefix/remove-prefixes.ttl
+++ /dev/null
@@ -1,7 +0,0 @@
-_:tests1
-	<http://example.org/p1> _:testo1 ;
-	<http://example.org/p2> _:testo3 ,
-		_:testo4 .
-
-_:testb0
-	<http://example.org/p3> _:testb1 .
diff --git a/test/meson.build b/test/meson.build
index 2de2348f..d7dcfa05 100644
--- a/test/meson.build
+++ b/test/meson.build
@@ -35,7 +35,6 @@ ttl_metadata_file_paths = [
   'extra/lax/manifest.ttl',
   'extra/pattern/manifest.ttl',
   'extra/perfect/manifest.ttl',
-  'extra/prefix/manifest.ttl',
   'extra/pretty/manifest.ttl',
   'extra/qualify/manifest.ttl',
   'extra/root/manifest.ttl',
@@ -189,12 +188,10 @@ simple_command_tests = {
       ['-b', '1024junk'],
       ['-b', '9223372036854775807'],
       ['-b'],
-      ['-c'],
       ['-k', '-1'],
       ['-k', '1024junk'],
       ['-k', '9223372036854775807'],
       ['-k'],
-      ['-p'],
       ['-qi'],
       ['-r'],
       ['-s', '<foo> a <Bar> .'],
@@ -500,20 +497,6 @@ test_suites = {
     files('extra/perfect/manifest.ttl'),
     ns_serdtest + 'perfect/',
   ],
-  'prefix_add': [
-    '--reverse',
-    files('extra/prefix/manifest.ttl'),
-    ns_serdtest + 'prefix/',
-    '--',
-    ['-I', 'generated'],
-    ['-p', 'test'],
-  ],
-  'prefix_remove': [
-    files('extra/prefix/manifest.ttl'),
-    ns_serdtest + 'prefix/',
-    '--',
-    ['-c', 'test'],
-  ],
   'pretty': [
     files('extra/pretty/manifest.ttl'),
     ns_serdtest + 'pretty/',
diff --git a/test/multifile/output.nq b/test/multifile/output.nq
index dd35dc4d..cf3ea93e 100644
--- a/test/multifile/output.nq
+++ b/test/multifile/output.nq
@@ -1,3 +1,3 @@
-_:f0b1 <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://example.org/Type> .
 _:f1b1 <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://example.org/Type> .
-_:f1b2 <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://example.org/OtherType> <http://example.org/graph> .
+_:f2b1 <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://example.org/Type> .
+_:f2b2 <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://example.org/OtherType> <http://example.org/graph> .
diff --git a/test/test_overflow.c b/test/test_overflow.c
index 8b47cd2b..62a77022 100644
--- a/test/test_overflow.c
+++ b/test/test_overflow.c
@@ -9,7 +9,7 @@
 #include <stdio.h>
 
 static const size_t min_stack_size = 4U * sizeof(size_t) + 238U;
-static const size_t max_stack_size = 1024U;
+static const size_t max_stack_size = 2048U;
 
 static SerdStatus
 test_size(SerdWorld* const      world,
diff --git a/test/test_reader_writer.c b/test/test_reader_writer.c
index aea0976d..9a4c988f 100644
--- a/test/test_reader_writer.c
+++ b/test/test_reader_writer.c
@@ -151,9 +151,6 @@ test_writer(const char* const path)
 
   assert(writer);
 
-  serd_writer_chop_blank_prefix(writer, "tmp");
-  serd_writer_chop_blank_prefix(writer, NULL);
-
   SerdNode* lit = serd_node_new(NULL, serd_a_string("hello"));
 
   const SerdSink* const iface = serd_writer_sink(writer);
@@ -273,17 +270,6 @@ test_reader(const char* path)
   assert(serd_reader_read_chunk(reader) == SERD_BAD_CALL);
   assert(serd_reader_read_document(reader) == SERD_BAD_CALL);
 
-  serd_reader_add_blank_prefix(reader, "tmp");
-
-#if defined(__GNUC__)
-#  pragma GCC diagnostic push
-#  pragma GCC diagnostic ignored "-Wnonnull"
-#endif
-  serd_reader_add_blank_prefix(reader, NULL);
-#if defined(__GNUC__)
-#  pragma GCC diagnostic pop
-#endif
-
   SerdInputStream in = serd_open_input_file(path);
   assert(!serd_reader_start(reader, &in, NULL, 4096));
   assert(!serd_reader_read_document(reader));
diff --git a/tools/serd-pipe.c b/tools/serd-pipe.c
index 16abbd2c..66600006 100644
--- a/tools/serd-pipe.c
+++ b/tools/serd-pipe.c
@@ -44,10 +44,8 @@ print_usage(const char* const name, const bool error)
     "  -O SYNTAX    Output syntax (empty/turtle/ntriples/nquads),\n"
     "               or flag (ascii/expanded/verbatim/terse/lax).\n"
     "  -b BYTES     I/O block size.\n"
-    "  -c PREFIX    Chop PREFIX from matching blank node IDs.\n"
     "  -h           Display this help and exit.\n"
     "  -k BYTES     Parser stack size.\n"
-    "  -p PREFIX    Add PREFIX to blank node IDs.\n"
     "  -q           Suppress all output except data.\n"
     "  -r ROOT_URI  Keep relative URIs within ROOT_URI.\n"
     "  -s STRING    Parse STRING as input.\n"
@@ -76,7 +74,6 @@ read_file(SerdWorld* const      world,
           const SerdSink* const sink,
           const size_t          stack_size,
           const char* const     filename,
-          const char* const     add_prefix,
           const size_t          block_size)
 {
   SerdInputStream in = serd_open_tool_input(filename);
@@ -91,9 +88,7 @@ read_file(SerdWorld* const      world,
   serd_world_set_limits(world, limits);
 
   SerdReader* reader = serd_reader_new(world, syntax, flags, env, sink);
-  serd_reader_add_blank_prefix(reader, add_prefix);
-
-  SerdStatus st = serd_reader_start(reader, &in, NULL, block_size);
+  SerdStatus  st     = serd_reader_start(reader, &in, NULL, block_size);
 
   st = st ? st : serd_reader_read_document(reader);
 
@@ -119,8 +114,6 @@ main(int argc, char** argv)
   size_t          block_size    = 4096U;
   size_t          stack_size    = 1048576U;
   const char*     input_string  = NULL;
-  const char*     add_prefix    = "";
-  const char*     chop_prefix   = NULL;
   const char*     root_uri      = NULL;
   const char*     out_filename  = NULL;
   int             a             = 1;
@@ -192,13 +185,6 @@ main(int argc, char** argv)
         }
         block_size = (size_t)size;
         break;
-      } else if (opt == 'c') {
-        if (argv[a][o + 1] || ++a == argc) {
-          return missing_arg(prog, 'c');
-        }
-
-        chop_prefix = argv[a];
-        break;
       } else if (opt == 'k') {
         if (argv[a][o + 1] || ++a == argc) {
           return missing_arg(prog, 'k');
@@ -212,13 +198,6 @@ main(int argc, char** argv)
         }
         stack_size = (size_t)size;
         break;
-      } else if (opt == 'p') {
-        if (argv[a][o + 1] || ++a == argc) {
-          return missing_arg(prog, 'p');
-        }
-
-        add_prefix = argv[a];
-        break;
       } else if (opt == 'r') {
         if (argv[a][o + 1] || ++a == argc) {
           return missing_arg(prog, 'r');
@@ -313,10 +292,7 @@ main(int argc, char** argv)
     serd_writer_set_root_uri(writer, serd_string(root_uri));
   }
 
-  serd_writer_chop_blank_prefix(writer, chop_prefix);
-
-  SerdStatus st         = SERD_SUCCESS;
-  SerdNode*  input_name = NULL;
+  SerdStatus st = SERD_SUCCESS;
   if (input_string) {
     const char*     position  = input_string;
     SerdInputStream string_in = serd_open_input_string(&position);
@@ -324,8 +300,6 @@ main(int argc, char** argv)
     SerdReader* const reader = serd_reader_new(
       world, input_syntax ? input_syntax : SERD_TRIG, reader_flags, env, sink);
 
-    serd_reader_add_blank_prefix(reader, add_prefix);
-
     if (!(st = serd_reader_start(reader, &string_in, NULL, 1U))) {
       st = serd_reader_read_document(reader);
     }
@@ -334,11 +308,8 @@ main(int argc, char** argv)
     serd_close_input(&string_in);
   }
 
-  size_t prefix_len = 0;
-  char*  prefix     = NULL;
-  if (n_inputs > 1) {
-    prefix_len = 8 + strlen(add_prefix);
-    prefix     = (char*)calloc(1, prefix_len);
+  if (n_inputs == 1) {
+    reader_flags |= SERD_READ_GLOBAL;
   }
 
   for (int i = 0; !st && i < n_inputs; ++i) {
@@ -349,10 +320,6 @@ main(int argc, char** argv)
       }
     }
 
-    if (n_inputs > 1) {
-      snprintf(prefix, prefix_len, "f%d%s", i, add_prefix);
-    }
-
     if ((st = read_file(world,
                         serd_choose_syntax(world, input_syntax, inputs[i]),
                         reader_flags,
@@ -360,16 +327,13 @@ main(int argc, char** argv)
                         sink,
                         stack_size,
                         inputs[i],
-                        n_inputs > 1 ? prefix : add_prefix,
                         block_size))) {
       break;
     }
   }
-  free(prefix);
 
   serd_sink_free(canon);
   serd_writer_free(writer);
-  serd_node_free(NULL, input_name);
   serd_env_free(env);
   serd_node_free(NULL, base);
   serd_world_free(world);