Port JS chat code over from chat-stream

.gitignore

@@ -222,6 +222,7 @@ $RECYCLE.BIN/
 
 .vscode
 /samconfig.toml
+/samconfig.yaml
 /env.json
 /env.*.json
 /*.parameters

.husky/pre-commit

@@ -1,4 +1,3 @@
 #!/usr/bin/env sh
 . "$(dirname -- "$0")/_/husky.sh"
 cd node && npm run lint && npm run prettier && cd -
-cd python && ruff check . && cd -

.tool-versions

@@ -1,3 +1,2 @@
-nodejs 16.14.0
+nodejs 18.16.1
 java corretto-19.0.1.10.1
-python 3.9.17

Makefile

@@ -14,45 +14,28 @@ help:
 	echo "make test           | run all tests"
 	echo "make cover          | run all tests with coverage"
 	echo "make env ENV=[env]  | activate env.\$$ENV.json file (default: dev)"
-	echo "make deps-node      | install node dependencies"
-	echo "make deps-python    | install python dependencies"
-	echo "make style-node     | run node code style check"
-	echo "make style-python   | run python code style check"
-	echo "make test-node      | run node tests"
-	echo "make test-python    | run python tests"
-	echo "make cover-node     | run node tests with coverage"
-	echo "make cover-python   | run python tests with coverage"
+	echo "make deps           | install dependencies"
+	echo "make style          | run code style check"
+	echo "make test           | run tests"
+	echo "make cover          | run tests with coverage"
 .aws-sam/build.toml: ./template.yaml node/package-lock.json node/src/package-lock.json python/requirements.txt python/src/requirements.txt
 	sam build --cached --parallel
-deps-node:
+deps:
 	cd node && npm ci
-cover-node:
+cover:
 	cd node && npm run test:coverage
-style-node:
+style:
 	cd node && npm run prettier
-test-node:
+test:
 	cd node && npm run test
-deps-python:
-	cd python && pip install -r requirements.txt
-cover-python:
-	cd python && coverage run --include='src/**/*' -m unittest -v && coverage report
-style-python:
-	cd python && ruff check .
-test-python:
-	cd python && python -m unittest -v
 build: .aws-sam/build.toml
 link: build
-	cd python/src && for src in *.py **/*.py; do for target in $$(find ../../.aws-sam/build -maxdepth 1 -type d); do if [[ -f $$target/$$src ]]; then ln -f $$src $$target/$$src; fi; done; done
 	cd node/src && for src in *.js *.json **/*.js **/*.json; do for target in $$(find ../../.aws-sam/build -maxdepth 1 -type d); do if [[ -f $$target/$$src ]]; then ln -f $$src $$target/$$src; fi; done; done
 serve: link
 	sam local start-api --host 0.0.0.0 --log-file dc-api.log
-deps: deps-node deps-python
-style: style-node style-python
-test: test-node test-python
-cover: cover-node cover-python
 env:
 	ln -fs ./env.${ENV}.json ./env.json
 secrets:
 	ln -s ../tfvars/dc-api/* .
 clean:
-	rm -rf .aws-sam node/node_modules node/src/node_modules python/**/__pycache__ python/.coverage python/.ruff_cache
+	rm -rf .aws-sam node/node_modules node/src/node_modules

node/src/embedded-weaviate-store.js

@@ -0,0 +1,79 @@
+const { WeaviateStore } = require("langchain/vectorstores/weaviate");
+const { Document } = require("langchain/document");
+
+class EmbeddedWeaviateStore extends WeaviateStore {
+  constructor(...args) {
+    super(...args);
+  }
+
+  async similaritySearch(query, k = 4, filter = undefined) {
+    try {
+      let builder = await this.client.graphql
+        .get()
+        .withClassName(this.indexName)
+        .withFields(this.queryAttrs.join(" "))
+        .withNearText({
+          concepts: [query],
+        })
+        .withLimit(k);
+
+      if (filter?.where) {
+        builder = builder.withWhere(filter.where);
+      }
+
+      const result = await builder.do();
+
+      const documents = [];
+      for (const data of result.data.Get[this.indexName]) {
+        const { [this.textKey]: text, ...rest } = data;
+
+        documents.push(
+          new Document({
+            pageContent: text,
+            metadata: rest,
+          })
+        );
+      }
+      return documents;
+    } catch (e) {
+      throw Error(`'Error in similaritySearch' ${e}`);
+    }
+  }
+
+  async similaritySearchWithScore(query, k = 4, filter = undefined) {
+    try {
+      let builder = await this.client.graphql
+        .get()
+        .withClassName(this.indexName)
+        .withFields(`${this.queryAttrs.join(" ")} _additional { distance }`)
+        .withNearText({
+          concepts: [query],
+        })
+        .withLimit(k);
+
+      if (filter?.where) {
+        builder = builder.withWhere(filter.where);
+      }
+
+      const result = await builder.do();
+
+      const documents = [];
+      for (const data of result.data.Get[this.indexName]) {
+        const { [this.textKey]: text, _additional, ...rest } = data;
+
+        documents.push([
+          new Document({
+            pageContent: text,
+            metadata: rest,
+          }),
+          _additional.distance,
+        ]);
+      }
+      return documents;
+    } catch (e) {
+      throw Error(`'Error in similaritySearch' ${e}`);
+    }
+  }
+}
+
+module.exports = { EmbeddedWeaviateStore };

node/src/handlers/chat-endpoint.js

@@ -0,0 +1,23 @@
+const { wrap } = require("./middleware");
+const Honeybadger = require("../honeybadger-setup");
+
+/**
+ * ChatEndpoint - Returns the function URL of the streaming chat handler
+ */
+exports.handler = wrap(async (event) => {
+  try {
+    return {
+      statusCode: 200,
+      body: JSON.stringify({
+        url: process.env.CHAT_ENDPOINT,
+        auth: event.userToken.sign(),
+      }),
+    };
+  } catch (error) {
+    await Honeybadger.notifyAsync(error);
+    return {
+      statusCode: 401,
+      body: "Error: " + error.message,
+    };
+  }
+});

python/src/helpers/prompts.py → node/src/handlers/chat-prompts.js

@@ -1,6 +1,5 @@
-# ruff: noqa: E501
-def prompt_template(): 
-  return """Using all of the provided source documents, create a helpful and thorough answer to the supplied question.
+function promptTemplate() {
+  return `Using all of the provided source documents, create a helpful and thorough answer to the supplied question.
   If you don't know the answer, just say that you don't know. Don't try to make up an answer, but you should use the documents provided in order to ground your response.
   It may be helpful to explain why a provided document does not pertain to the query as well.
   Feel free to reference various aspects of the sources in your explanation, but please don't include the full sources in the answer.
@@ -188,10 +187,19 @@ def prompt_template():
 
     QUESTION: {question}  
     =========
-    HELPFUL ANSWER:"""
+    HELPFUL ANSWER: `;
+}
 
-def document_template(attributes):
-  lines = (["Content: {page_content}", "Metadata:"] +
-            [f"  {attribute}: {{{attribute}}}" for attribute in attributes] +
-            ["Source: {source}"])
-  return "\n".join(lines)
+function documentTemplate(attributes) {
+  let lines = [
+    "Content: {pageContent}",
+    "Metadata:",
+    [...attributes.map((attribute) => `  ${attribute}: {${attribute}}`)].join(
+      "\n"
+    ),
+    "Source: {source}",
+  ];
+  return lines.join("\n");
+}
+
+module.exports = { promptTemplate, documentTemplate };