rgw: es: introduce ES types

This commit introduces an enum mapping various ES types. Also encode_json now
has a new member for string type so that this can be changed easily when
upgrading ES versions for eg. wherein new ES engines do not support the string
type anymore.

Signed-off-by: Abhishek Lekshmanan <abhishek@suse.com>

diff --git a/src/rgw/rgw_sync_module_es.cc b/src/rgw/rgw_sync_module_es.cc
index b56741e292141cfb91766e0816c69e51d6bbbdf5..f81dc62102af355c586609e58f82645ac15fa947 100644 (file)
--- a/src/rgw/rgw_sync_module_es.cc
+++ b/src/rgw/rgw_sync_module_es.cc
@@ -167,13 +167,47 @@ struct ElasticConfig {
 
 using ElasticConfigRef = std::shared_ptr<ElasticConfig>;
 
+std::optional<const char*> es_type_to_str(const ESType& t) {
+  switch (t) {
+  case ESType::String: return "string";
+  case ESType::Text: return "text";
+  case ESType::Keyword: return "keyword";
+  case ESType::Long: return "long";
+  case ESType::Integer: return "integer";
+  case ESType::Short: return "short";
+  case ESType::Byte: return "byte";
+  case ESType::Double: return "double";
+  case ESType::Float: return "float";
+  case ESType::Half_Float: return "half_float";
+  case ESType::Scaled_Float: return "scaled_float";
+  case ESType::Date: return "date";
+  case ESType::Boolean: return "boolean";
+  case ESType::Integer_Range: return "integer_range";
+  case ESType::Float_Range: return "float_range";
+  case ESType::Double_Range: return "date_range";
+  case ESType::Date_Range: return "date_range";
+  case ESType::Geo_Point: return "geo_point";
+  case ESType::Ip: return "ip";
+  default:
+    return std::nullopt;
+  }
+
+}
+
 struct es_dump_type {
   const char *type;
   const char *format;
   bool analyzed;
+  std::string default_type;
 
   es_dump_type(const char *t, const char *f = nullptr, bool a = false) : type(t), format(f), analyzed(a) {}
 
+  es_dump_type(ESType t, const char *f = nullptr, bool a = false,
+               std::string default_type="text"): format(f), analyzed(a),
+                                                 default_type(default_type) {
+    type = es_type_to_str(t).value_or(default_type.c_str());
+  }
+
   void dump(Formatter *f) const {
     encode_json("type", type, f);
     if (format) {
@@ -186,37 +220,45 @@ struct es_dump_type {
 };
 
 struct es_index_mappings {
+  ESType string_type {ESType::String};
+
   void dump_custom(Formatter *f, const char *section, const char *type, const char *format) const {
     f->open_object_section(section);
     ::encode_json("type", "nested", f);
     f->open_object_section("properties");
-    encode_json("name", es_dump_type("string"), f);
+    encode_json("name", es_dump_type(string_type), f);
     encode_json("value", es_dump_type(type, format), f);
     f->close_section(); // entry
     f->close_section(); // custom-string
   }
+
+  void dump_custom(Formatter *f, const char* section, ESType t, const char *format) const {
+    const auto s = es_type_to_str(t).value_or("text");
+    dump_custom(f,section,s,format);
+  }
+
   void dump(Formatter *f) const {
     f->open_object_section("object");
     f->open_object_section("properties");
-    encode_json("bucket", es_dump_type("string"), f);
-    encode_json("name", es_dump_type("string"), f);
-    encode_json("instance", es_dump_type("string"), f);
-    encode_json("versioned_epoch", es_dump_type("long"), f);
+    encode_json("bucket", es_dump_type(string_type), f);
+    encode_json("name", es_dump_type(string_type), f);
+    encode_json("instance", es_dump_type(string_type), f);
+    encode_json("versioned_epoch", es_dump_type(ESType::Date), f);
     f->open_object_section("meta");
     f->open_object_section("properties");
-    encode_json("cache_control", es_dump_type("string"), f);
-    encode_json("content_disposition", es_dump_type("string"), f);
-    encode_json("content_encoding", es_dump_type("string"), f);
-    encode_json("content_language", es_dump_type("string"), f);
-    encode_json("content_type", es_dump_type("string"), f);
-    encode_json("etag", es_dump_type("string"), f);
-    encode_json("expires", es_dump_type("string"), f);
+    encode_json("cache_control", es_dump_type(string_type), f);
+    encode_json("content_disposition", es_dump_type(string_type), f);
+    encode_json("content_encoding", es_dump_type(string_type), f);
+    encode_json("content_language", es_dump_type(string_type), f);
+    encode_json("content_type", es_dump_type(string_type), f);
+    encode_json("etag", es_dump_type(string_type), f);
+    encode_json("expires", es_dump_type(string_type), f);
     f->open_object_section("mtime");
     ::encode_json("type", "date", f);
     ::encode_json("format", "strict_date_optional_time||epoch_millis", f);
     f->close_section(); // mtime
-    encode_json("size", es_dump_type("long"), f);
-    dump_custom(f, "custom-string", "string", nullptr);
+    encode_json("size", es_dump_type(ESType::Long), f);
+    dump_custom(f, "custom-string", string_type, nullptr);
     dump_custom(f, "custom-int", "long", nullptr);
     dump_custom(f, "custom-date", "date", "strict_date_optional_time||epoch_millis");
     f->close_section(); // properties

diff --git a/src/rgw/rgw_sync_module_es.h b/src/rgw/rgw_sync_module_es.h
index 0be368e5883b1d59be935526d3a2c67bbe776690..1b5506c35799d1a6a9c6db7f4ba0f40f5b06701c 100644 (file)
--- a/src/rgw/rgw_sync_module_es.h
+++ b/src/rgw/rgw_sync_module_es.h
@@ -6,6 +6,33 @@
 
 #include "rgw_sync_module.h"
 
+enum class ESType {
+  /* string datatypes */
+  String, /* Deprecated Since 5.X+ */
+  Text,
+  Keyword,
+
+  /* Numeric Types */
+  Long, Integer, Short, Byte, Double, Float, Half_Float, Scaled_Float,
+
+  /* Date Type */
+  Date,
+
+  /* Boolean */
+  Boolean,
+
+  /* Binary; Must Be Base64 Encoded */
+  Binary,
+
+  /* Range Types */
+  Integer_Range, Float_Range, Long_Range, Double_Range, Date_Range,
+
+  /* A Few Specialized Types */
+  Geo_Point,
+  Ip
+};
+
+
 class RGWElasticSyncModule : public RGWSyncModule {
 public:
   RGWElasticSyncModule() {}