Merge pull request #14 from rdhyee/issue-13-parquet-duckdb

integrating Raymond's experiments so far with quarto

Author: datadavev

.github/workflows/hello-world.yml

@@ -0,0 +1,21 @@
+name: Hello World Workflow
+
+on:
+  workflow_dispatch:
+  push:
+    branches:
+      - '**'
+  pull_request:
+    types: [opened, synchronize, reopened]
+
+jobs:
+  say-hello:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v3
+      - name: Say Hello
+        run: |
+          echo "Hello, World!"
+          echo "Running on branch: ${{ github.ref_name }}"
+          echo "Triggered by: ${{ github.event_name }}"

_quarto.yml

@@ -38,6 +38,13 @@ website:
         - icon: github
           text: Github
           href: "https://github.com/isamplesorg/"
+      - section: "Tutorials"
+        contents:
+          - text: "iSamples Tutorials Overview"
+            href: tutorials/index.qmd
+          - text: "iSamples Parquet Tutorial"
+            href: tutorials/parquet.qmd
+
 
 
 # configure for correct source repository

tutorials/index.qmd

@@ -0,0 +1,67 @@
+---
+title: "Tutorials: Overview"
+---
+
+Here's where we park our various tutorials!
+
+Get the OpenAPI spec.
+
+```{ojs}
+//| echo: true
+
+// Get the OpenAPI specification and display detailed endpoint information
+viewof apiEndpointDetails = {
+  // Show loading indicator
+  const loadingElement = html`<div>Loading API endpoints...</div>`;
+  document.body.appendChild(loadingElement);
+
+  try {
+    const OPENAPI_URL = 'https://central.isample.xyz/isamples_central/openapi.json';
+
+    // Fetch the OpenAPI spec
+    const response = await fetch(OPENAPI_URL);
+    if (!response.ok) throw new Error(`Failed to fetch API spec: ${response.status}`);
+
+    const apiSpec = await response.json();
+
+    // Extract detailed information about each endpoint
+    const endpointDetails = [];
+
+    for (const [path, pathMethods] of Object.entries(apiSpec.paths)) {
+      for (const [method, details] of Object.entries(pathMethods)) {
+        endpointDetails.push({
+          endpoint: path,
+          method: method.toUpperCase(),
+          summary: details.summary || '',
+          operationId: details.operationId || '',
+          tags: (details.tags || []).join(', '),
+          parameters: (details.parameters || [])
+            .map(p => `${p.name} (${p.required ? 'required' : 'optional'})`)
+            .join(', ')
+        });
+      }
+    }
+
+    // Create a table with the detailed endpoint information
+    return Inputs.table(
+      endpointDetails,
+      {
+        label: "iSamples API Endpoints Details",
+        width: {
+          endpoint: 150,
+          method: 80,
+          summary: 200,
+          operationId: 200,
+          tags: 100,
+          parameters: 300
+        }
+      }
+    );
+  } catch (error) {
+    return html`<div style="color: red">Error fetching API endpoints: ${error.message}</div>`;
+  } finally {
+    // Remove loading indicator
+    loadingElement.remove();
+  }
+}
+```

tutorials/parquet.qmd

@@ -0,0 +1,64 @@
+---
+title: "Parquet"
+---
+
+Let's query Eric's parquet file using duckdb+parquet
+
+
+simpler query:
+
+```{ojs}
+//| echo: true
+
+// Import Observable's libraries
+import {DuckDBClient} from "@observablehq/duckdb"
+
+// Create a DuckDB instance
+db = DuckDBClient.of()
+
+// Set the Parquet file path
+parquet_path = 'https://storage.googleapis.com/opencontext-parquet/oc_isamples_pqg.parquet'
+
+// For testing, use a smaller dataset or limit rows
+// Option 1: Use LIMIT to reduce data transferred
+viewof testResults = {
+  // Show loading indicator
+  const loadingElement = html`<div>Running query...</div>`;
+  document.body.appendChild(loadingElement);
+  
+  try {
+    // Test with a small LIMIT to verify connection works
+    const data = await db.query(`
+      SELECT otype, pid
+      FROM read_parquet('${parquet_path}')
+      LIMIT 10
+    `);
+    return Inputs.table(data);
+  } finally {
+    // Remove loading indicator when done (whether success or error)
+    loadingElement.remove();
+  }
+}
+
+```
+
+now the full query
+
+
+```{ojs}
+//| echo: true
+
+
+// Query the Parquet file
+viewof results = Inputs.table(
+  await db.query(`
+    SELECT COUNT(pid) as count, otype 
+    FROM read_parquet('${parquet_path}')
+    GROUP BY otype
+    ORDER BY count DESC
+  `)
+)
+```
+
+
+