Retrieving yesterday’s date is a frequent task in Linux scripting and automation. This guide presents two reliable methods using the built-in date command and explores the considerations for handling time zones.
Table of Contents
Using the date Command
The simplest approach involves directly using the date command’s built-in “yesterday” functionality:
date -d "yesterday" +%Y-%m-%d
This command:
date: Invokes the date command.-d "yesterday": Specifies that we want yesterday’s date.+%Y-%m-%d: Formats the output as YYYY-MM-DD. You can customize this (e.g.,+%F,+%Y/%m/%d,+%d/%m/%Y).
Example: If today is 2024-10-27, the output will be: 2024-10-26
Handling Time Zones for Greater Accuracy
For scripts requiring precise date calculations across different time zones, a more robust method is needed. While the previous method usually suffices, inconsistencies can arise due to daylight saving time transitions. This refined approach uses epoch time:
date -d "@$(($(date +%s) - 86400))" +%Y-%m-%d
This command:
date +%s: Gets the current time in seconds since the epoch (January 1, 1970, 00:00:00 UTC).$(...): Command substitution, capturing the output of the inner command.- 86400: Subtracts 86400 seconds (one day).@(...): Tellsdateto interpret the preceding value as seconds since the epoch.date -d "..." +%Y-%m-%d: Formats the resulting epoch time as YYYY-MM-DD.
This method ensures accuracy by directly manipulating the epoch time, avoiding potential timezone-related ambiguities.
For most scenarios, the first method using date -d "yesterday" is sufficient. However, for critical applications or cross-timezone scripts, the second approach provides a more reliable solution. Remember to tailor the output format to your specific requirements.