@x12i/memorix-writer 1.3.0 → 1.3.1

package/README.md

@@ -70,7 +70,7 @@ Override defaults only when you need to:
 
 ## Collection resolution
 
-Writes target **`memorix-entities`** or **`memorix-events`** based on the write descriptor (`target.kind`, defaulting from `identity.idField`). Database names follow [Memorix Database Conventions](./docs/MEMORIX-DATABASE-CONVENTIONS.md) (`MEMORIX_ENTITIES_DB`, `MEMORIX_EVENTS_DB`, …).
+Writes target **`memorix-entities`**, **`memorix-events`**, or **`memorix-knowledge`** based on the write descriptor (`target.kind`, defaulting from `identity.idField`). Database names follow [Memorix Database Conventions](./docs/MEMORIX-DATABASE-CONVENTIONS.md) (`MEMORIX_ENTITIES_DB`, `MEMORIX_EVENTS_DB`, `MEMORIX_KNOWLEDGE_DB`, …).
 
 Collection name resolution order:
 
@@ -105,6 +105,8 @@ npm run catalox:seed:write-descriptors:validate
 
 Seed file: [`catalox-seeds/memorix-write-descriptors.manifest.json`](catalox-seeds/memorix-write-descriptors.manifest.json)
 
+Shipped write descriptors include `asset-analysis-write` (entity) and corpus helpers `content-documents-snapshot-write` / `content-documents-chunk-write` (`knowledge` target). Record builders: [`@x12i/memorix-corpus`](../memorix-corpus/README.md).
+
 ## Acceptance criteria
 
 | ID | Status |

package/catalox-seeds/memorix-write-descriptors.manifest.json

@@ -158,6 +158,181 @@
           "defaultMode": "summary"
         }
       }
+    },
+    {
+      "catalogId": "memorix-write-descriptors",
+      "data": {
+        "id": "content-documents-snapshot-write",
+        "entity": "content-documents",
+        "title": "Write Content Document Snapshot",
+        "target": {
+          "kind": "knowledge",
+          "contentType": "snapshots"
+        },
+        "identity": {
+          "idField": "knowledgeId",
+          "requireExistingCanonical": false
+        },
+        "operations": {
+          "allowed": ["add", "upsert"],
+          "default": "upsert"
+        },
+        "fields": {
+          "title": {
+            "label": "Title",
+            "inputPath": "title",
+            "target": { "path": "data.title" },
+            "writable": true,
+            "valueType": "string"
+          },
+          "summary": {
+            "label": "Summary",
+            "inputPath": "summary",
+            "target": { "path": "data.summary" },
+            "writable": true,
+            "valueType": "string"
+          },
+          "language": {
+            "label": "Language",
+            "inputPath": "language",
+            "target": { "path": "data.language" },
+            "writable": true,
+            "valueType": "string"
+          },
+          "sourceKind": {
+            "label": "Source kind",
+            "inputPath": "sourceKind",
+            "target": { "path": "source.kind" },
+            "writable": true,
+            "valueType": "string"
+          },
+          "sourceUri": {
+            "label": "Source URI",
+            "inputPath": "sourceUri",
+            "target": { "path": "source.uri" },
+            "writable": true,
+            "valueType": "string"
+          },
+          "canonicalUri": {
+            "label": "Canonical URI",
+            "inputPath": "canonicalUri",
+            "target": { "path": "source.canonicalUri" },
+            "writable": true,
+            "valueType": "string"
+          },
+          "contentHash": {
+            "label": "Content hash",
+            "inputPath": "contentHash",
+            "target": { "path": "provenance.contentHash" },
+            "writable": true,
+            "valueType": "string"
+          },
+          "cleanMarkdown": {
+            "label": "Clean Markdown",
+            "inputPath": "cleanMarkdown",
+            "target": { "path": "content.cleanMarkdown" },
+            "writable": true,
+            "valueType": "content",
+            "content": {
+              "enabled": true,
+              "objectType": "memorix-content-object",
+              "acceptBody": true,
+              "acceptPointer": true,
+              "format": "markdown",
+              "maxBytes": 10485760,
+              "contentKeyTemplate": "content-documents/{{knowledgeId}}/{{contentHash}}/clean.md",
+              "generatePreview": true,
+              "previewMaxChars": 500,
+              "computeMetadata": true,
+              "storage": {
+                "provider": "gcs",
+                "bucket": "memorix-content",
+                "prefix": "content-documents"
+              }
+            }
+          }
+        },
+        "content": {
+          "allowed": true,
+          "allowedFields": ["cleanMarkdown"],
+          "generatePreview": true,
+          "previewMaxChars": 500,
+          "computeMetadata": true,
+          "maxBytes": 10485760
+        },
+        "validation": {
+          "rejectUnknownFields": true,
+          "requireAtLeastOneWritableField": true
+        },
+        "return": {
+          "defaultMode": "summary"
+        }
+      }
+    },
+    {
+      "catalogId": "memorix-write-descriptors",
+      "data": {
+        "id": "content-documents-chunk-write",
+        "entity": "content-documents",
+        "title": "Write Content Document Chunk",
+        "target": {
+          "kind": "knowledge",
+          "contentType": "chunks"
+        },
+        "identity": {
+          "idField": "knowledgeId",
+          "requireExistingCanonical": false
+        },
+        "operations": {
+          "allowed": ["add"],
+          "default": "add"
+        },
+        "fields": {
+          "text": {
+            "label": "Chunk text",
+            "inputPath": "text",
+            "target": { "path": "data.text" },
+            "required": true,
+            "writable": true,
+            "valueType": "string"
+          },
+          "sectionTitle": {
+            "label": "Section title",
+            "inputPath": "sectionTitle",
+            "target": { "path": "data.sectionTitle" },
+            "writable": true,
+            "valueType": "string"
+          },
+          "tokenCount": {
+            "label": "Token count",
+            "inputPath": "tokenCount",
+            "target": { "path": "data.tokenCount" },
+            "writable": true,
+            "valueType": "number"
+          },
+          "chunkIndex": {
+            "label": "Chunk index",
+            "inputPath": "chunkIndex",
+            "target": { "path": "provenance.chunkIndex" },
+            "writable": true,
+            "valueType": "number"
+          },
+          "contentHash": {
+            "label": "Parent content hash",
+            "inputPath": "contentHash",
+            "target": { "path": "parent.contentHash" },
+            "writable": true,
+            "valueType": "string"
+          }
+        },
+        "validation": {
+          "rejectUnknownFields": true,
+          "requireAtLeastOneWritableField": true
+        },
+        "return": {
+          "defaultMode": "summary"
+        }
+      }
     }
   ]
 }

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@x12i/memorix-writer",
-  "version": "1.3.0",
+  "version": "1.3.1",
   "description": "Descriptor-driven write layer for Memorix entity/event records and content objects",
   "type": "module",
   "main": "./dist/index.js",
@@ -26,9 +26,9 @@
     "node": ">=18.0.0"
   },
   "dependencies": {
-    "@x12i/catalox": "^5.1.1",
-    "@x12i/memorix-descriptors": "^1.10.0",
-    "@x12i/memorix-retrieval": "^1.13.0",
+    "@x12i/catalox": "^5.9.8",
+    "@x12i/memorix-descriptors": "^1.11.0",
+    "@x12i/memorix-retrieval": "^1.15.0",
     "mongodb": "^6.21.0"
   },
   "devDependencies": {
@@ -44,5 +44,10 @@
   },
   "publishConfig": {
     "access": "public"
+  },
+  "repository": {
+    "type": "git",
+    "url": "git+ssh://git@github.com/x12i/memorix-mono-repo.git",
+    "directory": "memorix-writer"
   }
 }