Skip to main content

Advanced Timestamp Patterns

Once you have mastered the basic "detect, parse, convert, enrich" pattern, you can handle more complex timestamp scenarios.

Pattern 1: Handling Custom String Formats​

Not all systems use standard formats. You can use the .parse_timestamp() function with a layout string to handle any custom format. The layout string is based on Go's time formatting rules, using 2006 for the year, 01 for the month, 02 for the day, 15 for the hour, etc.

Input:

{"timestamp": "10-20-2025_18:23:45"}

Processor:

- mapping: |
    root.normalized_timestamp = this.timestamp.parse_timestamp("01-02-2006_15:04:05")

Pattern 2: Business Logic (Holidays & Weekends)​

You can use the extracted time components to add valuable business-specific flags to your data.

Use Case: Flagging events that occurred on a weekend or a US federal holiday.

Add Business Logic Flags
- mapping: |
    root = this
    let ts = this.timestamp_utc.parse_timestamp()

    let month = ts.ts_month()
    let day = ts.ts_day()
    let day_of_week = ts.ts_weekday() # Sunday = 0, Saturday = 6

    # Weekend Detection
    root.time_metadata.is_weekend = day_of_week == 0 || day_of_week == 6

    # Holiday Detection (simplified example)
    root.time_metadata.is_holiday = (month == 1 && day == 1) ||      # New Year's
                                    (month == 12 && day == 25)        # Christmas

    # Business Day Flag
    root.time_metadata.is_business_day = !root.time_metadata.is_weekend && !root.time_metadata.is_holiday

Pattern 3: Fiscal Calendars​

For financial data, you can add fiscal calendar information, such as the fiscal quarter.

Use Case: A company whose fiscal year starts in October.

Fiscal Quarter Calculation
- mapping: |
    root = this
    let month = this.time_metadata.month
    let year = this.time_metadata.year

    # Fiscal Year (October 1st - September 30th)
    root.time_metadata.fiscal_year = if month >= 10 { year + 1 } else { year }

    root.time_metadata.fiscal_quarter = if month >= 10 { "FQ1" }
                                       else if month >= 7 { "FQ4" }
                                       else if month >= 4 { "FQ3" }
                                       else { "FQ2" }

These advanced patterns allow you to create highly valuable, analytics-ready time series data directly in your processing pipeline, saving significant time and cost in downstream systems.