请求提供移除数据库中重复及重复模式价格记录的算法

Got it, let's tackle this duplicate price record cleanup problem. The goal here is to strip out consecutive entries with identical prices—since they don't add any meaningful data to your price trend tracking, right? We only need to keep the records where the price actually shifts.

Core Approach #

We'll build a cleaned list of records by only retaining entries where the price differs from the last saved record. Consecutive duplicates get discarded because they don't reflect a meaningful change in the price trend.

Solution Code #

Record = Struct.new(:id, :created_at, :price)
records = [
  Record.new(1, Date.parse('2017-01-01'), 150_000),
  Record.new(2, Date.parse('2017-01-02'), 150_000),
  Record.new(3, Date.parse('2017-01-03'), 130_000),
  Record.new(4, Date.parse('2017-01-04'), 140_000),
  Record.new(5, Date.parse('2017-01-05'), 140_000),
  Record.new(6, Date.parse('2017-01-06'), 130_000)
]

# Build cleaned list by keeping only price-change records
cleaned_records = records.inject([]) do |result, record|
  # Add the record if it's the first entry, or if price differs from last saved entry
  if result.empty? || record.price != result.last.price
    result << record
  end
  result
end

# Print the cleaned output
cleaned_records.each do |r|
  puts "ID: #{r.id}, Date: #{r.created_at}, Price: #{r.price}"
end

What This Does #

• We use Ruby's inject method to iterate through the records and build our cleaned array.
• For each record, we check two things: if we're dealing with the first entry, or if the current price is different from the last record we kept.
• If either condition is true, we add the record to our result list—otherwise, we skip it.

Sample Output #

ID: 1, Date: 2017-01-01, Price: 150000
ID: 3, Date: 2017-01-03, Price: 130000
ID: 4, Date: 2017-01-04, Price: 140000
ID: 6, Date: 2017-01-06, Price: 130000

Edge Cases Covered #

• All records have the same price: Only the first record is kept (no redundant duplicates).
• Single record: The record is returned as-is, no changes needed.
• Non-consecutive duplicates: Records with the same price but separated by different prices are both kept—since this reflects a price change back to a previous value, which is useful for trend tracking.

Optional: Keep the Latest Consecutive Duplicate #

If you'd rather retain the last entry in a sequence of identical prices (instead of the earliest), adjust the logic to replace the last entry in the result array until a price change is detected:

cleaned_records = records.inject([]) do |result, record|
  if result.empty? || record.price != result.last.price
    result << record
  else
    # Replace the last entry with the current one to keep the latest date for same price
    result[-1] = record
  end
  result
end

This version will give you the most recent date for each consecutive price sequence.

内容的提问来源于stack exchange，提问作者Niels Kristian