ii offline: fix a wrong index constructed bug

It's caused only when vector column is a source and WITH_SECTION isn't specified.
groonga · Feb 18, 2018 · cf5e255 · cf5e255
1 parent b49f803
commit cf5e255
Show file tree

Hide file tree

Showing 5 changed files with 128 additions and 1 deletion.
diff --git a/lib/ii.c b/lib/ii.c
@@ -12317,7 +12317,8 @@ grn_ii_builder_append_obj(grn_ctx *ctx, grn_ii_builder *builder,
         if (sec->length == 0) {
           continue;
         }
-        if (builder->tokenizer) {
+        if ((builder->ii->header->flags & GRN_OBJ_WITH_SECTION) &&
+            builder->tokenizer) {
           sid = i + 1;
         }
         rc = grn_ii_builder_append_value(ctx, builder, rid, sid, sec->weight,

diff --git a/test/command/suite/load/index/offline/vector_text_without_section.expected b/test/command/suite/load/index/offline/vector_text_without_section.expected
@@ -0,0 +1,46 @@
+plugin_register functions/index_column
+[[0,0.0,0.0],true]
+table_create Docs TABLE_NO_KEY
+[[0,0.0,0.0],true]
+column_create Docs sentences COLUMN_VECTOR Text
+[[0,0.0,0.0],true]
+load --table Docs
+[
+{"sentences": ["-", "-", "-"]}
+]
+[[0,0.0,0.0],1]
+table_create Words TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram
+[[0,0.0,0.0],true]
+column_create Words docs_sentences COLUMN_INDEX|WITH_POSITION Docs sentences
+[[0,0.0,0.0],true]
+select Words   --limit -1   --sort_keys _key   --output_columns '_key, index_column_source_records("docs_sentences")'
+[
+  [
+    0,
+    0.0,
+    0.0
+  ],
+  [
+    [
+      [
+        1
+      ],
+      [
+        [
+          "_key",
+          "ShortText"
+        ],
+        [
+          "index_column_source_records",
+          null
+        ]
+      ],
+      [
+        "-",
+        [
+          1
+        ]
+      ]
+    ]
+  ]
+]
diff --git a/test/command/suite/load/index/offline/vector_text_without_section.test b/test/command/suite/load/index/offline/vector_text_without_section.test
@@ -0,0 +1,17 @@
+plugin_register functions/index_column
+
+table_create Docs TABLE_NO_KEY
+column_create Docs sentences COLUMN_VECTOR Text
+
+load --table Docs
+[
+{"sentences": ["-", "-", "-"]}
+]
+
+table_create Words TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram
+column_create Words docs_sentences COLUMN_INDEX|WITH_POSITION Docs sentences
+
+select Words \
+  --limit -1 \
+  --sort_keys _key \
+  --output_columns '_key, index_column_source_records("docs_sentences")'
diff --git a/test/command/suite/load/index/online/vector/text_without_section.expected b/test/command/suite/load/index/online/vector/text_without_section.expected
@@ -0,0 +1,46 @@
+plugin_register functions/index_column
+[[0,0.0,0.0],true]
+table_create Docs TABLE_NO_KEY
+[[0,0.0,0.0],true]
+column_create Docs sentences COLUMN_VECTOR Text
+[[0,0.0,0.0],true]
+table_create Words TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram
+[[0,0.0,0.0],true]
+column_create Words docs_sentences COLUMN_INDEX|WITH_POSITION Docs sentences
+[[0,0.0,0.0],true]
+load --table Docs
+[
+{"sentences": ["-", "-", "-"]}
+]
+[[0,0.0,0.0],1]
+select Words   --limit -1   --sort_keys _key   --output_columns '_key, index_column_source_records("docs_sentences")'
+[
+  [
+    0,
+    0.0,
+    0.0
+  ],
+  [
+    [
+      [
+        1
+      ],
+      [
+        [
+          "_key",
+          "ShortText"
+        ],
+        [
+          "index_column_source_records",
+          null
+        ]
+      ],
+      [
+        "-",
+        [
+          1
+        ]
+      ]
+    ]
+  ]
+]
diff --git a/test/command/suite/load/index/online/vector/text_without_section.test b/test/command/suite/load/index/online/vector/text_without_section.test
@@ -0,0 +1,17 @@
+plugin_register functions/index_column
+
+table_create Docs TABLE_NO_KEY
+column_create Docs sentences COLUMN_VECTOR Text
+
+table_create Words TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram
+column_create Words docs_sentences COLUMN_INDEX|WITH_POSITION Docs sentences
+
+load --table Docs
+[
+{"sentences": ["-", "-", "-"]}
+]
+
+select Words \
+  --limit -1 \
+  --sort_keys _key \
+  --output_columns '_key, index_column_source_records("docs_sentences")'