Estonia Events Coda

Workflow

0. Trigger Google Calendar Sync in Coda (Run This First!)

The Coda table syncs from Google Calendar, but it does not auto-refresh. You must manually trigger a sync before working with the data.

1. Open the Coda page in the Playwright browser:

   browser_navigate to: https://coda.io/d/Estonia-IT-events_dzkj730WT5a/
   

2. Take a snapshot and find the Refresh button (near the table header, next to the search icon).
3. Click the Refresh button.
4. The sync can take up to a couple of minutes. Take periodic snapshots to check progress — look for the status message (e.g. "Getting updates from Google Calendar...") to disappear and confirm new events have appeared in the table.
5. Once the sync is complete, proceed to the next steps.

Note: There is no API-based way to trigger this sync — the browser click is the only option.

1. Auto-Archive Past Events

Automatically mark past events as archived to keep the working set small:

# Get all non-archived events
QUERY=$(printf '"Archived":false' | jq -sRr @uri)

curl -s -H "Authorization: Bearer $CODA_API_TOKEN" \
  "https://coda.io/apis/v1/docs/$CODA_DOC_ID/tables/$CODA_TABLE_ID/rows?useColumnNames=true&query=$QUERY" \
  | jq -r '
    .items[]
    | select(.values.End < "'$(date -I)'")
    | {
        id: .id,
        name: .values.Name,
        end: .values.End
      }
  ' | jq -s '. | length as $count | if $count > 0 then ("Found " + ($count | tostring) + " past events to archive") else "No past events to archive" end'

For each past event, set Archived=true:

EVENT_ID="i-xxxxx..."

curl -s -X PUT \
  -H "Authorization: Bearer $CODA_API_TOKEN" \
  -H "Content-Type: application/json" \
  "https://coda.io/apis/v1/docs/$CODA_DOC_ID/tables/$CODA_TABLE_ID/rows/$EVENT_ID" \
  -d '{
    "row": {
      "cells": [
        {"column": "Archived", "value": true}
      ]
    }
  }' | jq '.'

Note: This keeps the API response small by filtering archived events server-side.

2. Fetch Events Missing Labels or Links

Get non-archived events that are missing either Labels OR Links:

# Server-side filter: only non-archived events
QUERY=$(printf '"Archived":false' | jq -sRr @uri)

curl -s -H "Authorization: Bearer $CODA_API_TOKEN" \
  "https://coda.io/apis/v1/docs/$CODA_DOC_ID/tables/$CODA_TABLE_ID/rows?useColumnNames=true&query=$QUERY" \
  | jq -r '
    .items[]
    | select(
        ((.values.Labels == "" or .values.Labels == null) or
         (.values.Link == "" or .values.Link == null))
      )
    | .values.Description as $desc
    | ($desc | split("\n")[0]) as $firstLine
    | {
        id: .id,
        name: .values.Name,
        start: .values.Start,
        end: .values.End,
        labels: (.values.Labels // "❌ MISSING"),
        link: (.values.Link // "❌ MISSING"),
        url_in_description: (if ($firstLine | test("^https?://")) then $firstLine else "No URL" end),
        location: .values.Location,
        description: $desc
      }
  ' | jq -s '.'

Server-side filtering: Uses "Archived":false query to fetch only non-archived events, keeping the payload small.

Pro tip: Most events have the URL as the first line of description - extract and populate the Link field from there!

3. Fetch Available Labels

Get the current list of valid labels from Coda:

curl -s -H "Authorization: Bearer $CODA_API_TOKEN" \
  "https://coda.io/apis/v1/docs/$CODA_DOC_ID/tables/$CODA_TABLE_ID/columns" \
  | jq -r '.items[] | select(.name == "Labels") | .format.options[] | .name' | sort

4. Remove "🔥 New" Label (Always Run This!)

Remove "🔥 New" from events that were previously labeled (this cleans up old labels):

QUERY=$(printf '"Archived":false' | jq -sRr @uri)

# Find events with "New" in labels
curl -s -H "Authorization: Bearer $CODA_API_TOKEN" \
  "https://coda.io/apis/v1/docs/$CODA_DOC_ID/tables/$CODA_TABLE_ID/rows?useColumnNames=true&query=$QUERY" \
  | jq -r '
    .items[]
    | select(.values.Labels | contains("New"))
    | {
        id: .id,
        name: .values.Name,
        labels: .values.Labels
      }
  ' | jq -s '.'

For each event found, remove "🔥 New" from the labels (keep other labels):

EVENT_ID="i-xxxxx..."
# Example: if labels are "🔥 New,GameDev,Hackathon", set to "GameDev,Hackathon"
NEW_LABELS="GameDev,Hackathon"

curl -s -X PUT \
  -H "Authorization: Bearer $CODA_API_TOKEN" \
  -H "Content-Type: application/json" \
  "https://coda.io/apis/v1/docs/$CODA_DOC_ID/tables/$CODA_TABLE_ID/rows/$EVENT_ID" \
  -d "{
    \"row\": {
      \"cells\": [
        {\"column\": \"Labels\", \"value\": \"$NEW_LABELS\"}
      ]
    }
  }"

Note: This step runs EVERY time to clean up "🔥 New" labels from previous runs. Only newly labeled events in step 4 get "🔥 New" added back.

5. Manual Labeling Process

⚠️ Important:

• DO NOT use automated keyword matching - it produces too many false positives
• Actually read each event description carefully
• Use judgment - assign labels based on actual content, not keywords
• ALWAYS add "🔥 New" to the labels when labeling events that were missing labels/links

6. Extract URLs from Descriptions

Check the url_in_description field from step 1. If present, use it to populate the Link field.

7. Update Event Row

Update both Labels and Link fields:

# Example variables
EVENT_ID="i-xxxxx..."
LABELS="🔥 New,Startups,Hackathon"  # ALWAYS add "🔥 New" for newly labeled events
LINK="https://example.com/event"

curl -X PUT \
  -H "Authorization: Bearer $CODA_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d "{
    \"row\": {
      \"cells\": [
        {\"column\": \"Labels\", \"value\": \"$LABELS\"},
        {\"column\": \"Link\", \"value\": \"$LINK\"}
      ]
    }
  }" \
  "https://coda.io/apis/v1/docs/$CODA_DOC_ID/tables/$CODA_TABLE_ID/rows/$EVENT_ID"

Label format:

• Multiple labels are comma-separated with NO spaces
• Example: "🔥 New,Dev,Students,Hackathon"

Rate limiting:

• Wait 2 seconds between API calls to avoid Coda rate limiting (1 second is not enough when doing bulk updates)

Quality Checklist

Before updating events:

• ✅ Triggered Google Calendar sync in Coda (browser Refresh) and waited for it to complete
• ✅ Environment variables sourced
• ✅ "Archived" column exists in Coda table
• ✅ Ran auto-archive step to mark past events
• ✅ Fetched available labels from Coda
• ✅ Removed "🔥 New" label from existing events (if present)
• ✅ Read event descriptions carefully
• ✅ Assigned labels based on actual content (not keywords)
• ✅ Added "🔥 New" to newly labeled events
• ✅ Extracted URLs from description first lines
• ✅ Using correct comma-separated format (no spaces)

Common Pitfalls

❌ Not triggering Google Calendar sync before starting (data will be stale) ❌ Forgetting to run auto-archive step first ❌ Missing "Archived" column in Coda table ❌ Using keyword matching instead of reading descriptions ❌ Forgetting to add "🔥 New" to newly labeled events ❌ Wrong label format (spaces after commas) ❌ Not extracting URLs from description first lines ❌ API rate limiting (going too fast) ❌ Hardcoding IDs instead of using env vars