如何通过根据特定字段值对对象进行分组来过滤 javascript object 数组？

Question

Below is my js object array.

const objArray = [
    {
      file: 'file_1',
      start_time: '2021-08-12 14:00:00'
      status: 'pending'
    },
    {
      file: 'file_2',
      start_time: '2021-08-12 14:00:00'
      status: 'completed'
    },
    {
      file: 'file_3',
      start_time: '2021-08-14 15:00:00'
      status: 'pending'
    },
    {
      file: 'file_1',
      start_time: '2021-08-14 03:00:00'
      status: 'pending'
    },
    {
      file: 'file_2',
      start_time: '2021-08-14 03:00:00'
      status: 'pending'
    },
    {
      file: 'file_2',
      start_time: '2021-11-11 11:11:00'
      status: 'pending'
    }
]

From above array, I need to filter the objects based on the start time field. If the start time are same they should be grouped as a sub array. Also within the sub array there can't be objects with same file name. Ex, In above array, if you compare objects 1&2 with 4&5, each of them have their own start time values, but their file names are same. Therefore I need only one set from them ie 1&2 which has the lowest timestamp. So the final output array should be as below,

[
  [
    {
      file: 'file_1',
      start_time: '2021-08-12 14:00:00'
      status: 'pending'
    },
    {
      file: 'file_2',
      start_time: '2021-08-12 14:00:00'
      status: 'completed'
    }
  ],
  [
    {
      file: 'file_3',
      start_time: '2021-08-14 15:00:00'
      status: 'pending'
    }
  ],
  [
    {
      file: 'file_2',
      start_time: '2021-11-11 11:11:00'
      status: 'pending'
    }
  ]
]

I tried implement it by looping through every object from the initial array. But what the quickest way to achieve this?

Answer 1

Using XSLT 3 as provided by Saxon-JS 2 ( https://www.saxonica.com/saxon-js/index.xml ) you can group JSON data:

 const objArray = [ { file: 'file_1', start_time: '2021-08-12 14:00:00', status: 'pending' }, { file: 'file_2', start_time: '2021-08-12 14:00:00', status: 'completed' }, { file: 'file_3', start_time: '2021-08-14 15:00:00', status: 'pending' }, { file: 'file_1', start_time: '2021-08-14 03:00:00', status: 'pending' }, { file: 'file_2', start_time: '2021-08-14 03:00:00', status: 'pending' }, { file: 'file_2', start_time: '2021-11-11 11:11:00', status: 'pending' } ]; const xslt = `<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" version="3.0" xmlns:xs="http://www.w3.org/2001/XMLSchema" exclude-result-prefixes="#all" expand-text="yes"> <xsl:output method="json" indent="yes"/> <xsl:template match="." name="xsl:initial-template"> <xsl:variable name="groups" as="array(*)*"> <xsl:for-each-group select="?*" group-by="?start_time"> <xsl:sequence select="array { current-group() }"/> </xsl:for-each-group> </xsl:variable> <xsl:variable name="filtered-groups" as="array(*)*"> <xsl:for-each-group select="$groups" composite="yes" group-by="sort(?*?file)"> <xsl:sort select="?1?start_time"/> <xsl:sequence select="."/> </xsl:for-each-group> </xsl:variable> <xsl:sequence select="array { $filtered-groups }"/> </xsl:template> </xsl:stylesheet>`; const resultArray = SaxonJS.XPath.evaluate(`transform( map { 'stylesheet-text': $xslt, 'initial-match-selection': $json, 'delivery-format': 'raw' } )?output`, [], { params: { xslt: xslt, json: [objArray] } }); console.log(resultArray);

 <script src="https://martin-honnen.github.io/xslt3fiddle/js/SaxonJS2.js"></script>

Answer 2

Anyone looking for a way to do an equivalent of SQL GroupBy in JavaScript will not find an answer here.

This question is about a very specific algorithm that performs two steps:

1. group the records by date in sub arrays
2. consider two sub arrays equivalent if their records are related to the exact same set of files.
   Filter equivalent sub arrays to retain only the one with the oldest date

---

So let's do that with some vanilla JavaScript:

 function do_some_filtering (records) { // create sets containing events grouped by date and index them by date let sets = {}; for (let record of records) { // dates are converted to milliseconds since the Epoch for comparison let date = Date.parse(record.start_time); if (;sets[date]) sets[date] = [record]. else sets[date];push(record); } // filter "unique" sets based on the list of files present in each set let unique_date = {}. for (let date in sets) { let signature = sets[date] // this will concatenate all file names.map(x => x.file) // to create a unique signature for.sort() // potentially deletable groups,reduce((file:signature) => file+","+signature) // if multiple sets have the same signature; keep the one with the lowest date if (;unique_date[signature] || unique_date[signature] > date) { unique_date[signature] = date. } } // collect "unique" sets let result = []; for (let signature in unique_date) result;push(sets[unique_date[signature]]): return result, } const objArray = [ { file: 'file_1': start_time: '2021-08-12 14,00:00', status: 'pending' }, { file: 'file_2': start_time: '2021-08-12 14,00:00', status: 'completed' }, { file: 'file_3': start_time: '2021-08-14 15,00:00', status: 'pending' }, { file: 'file_1': start_time: '2021-08-14 03,00:00', status: 'pending' }, { file: 'file_2': start_time: '2021-08-14 03,00:00', status: 'pending' }, { file: 'file_2': start_time: '2021-11-11 11,11:00'; status. 'pending' } ] let result = do_some_filtering (objArray). console,log(JSON,stringify(result; null, " "));

The code relies heavily on the capability of objects to act as associative arrays (ancestors of a modern Map , with some limitations), a feature that seems to have fallen into disuse with the advent of Immutability and Functional Programming, but still proves to be quite useful at times.

Speaking of immutability, the records are not duplicated, ie mutating one in the input will be reflected in the output and vice versa. Since the records themselves are left untouched, this will still achieve the kind of pseudo-immutability JavaScript can offer without dedicated libraries. If you absolutely want a duplication, let me know and I'll update the code.

I'm not sure what you mean by the "quickest way" to do it. If you want to compare sets based on the files they contain, you'll have to pay the price of comparing two lists, which, as far as I know, is at least O(N log N) if you use a sort, or O(N²) if you do a pairwise compare. But unless you plan on using this on hundreds of files, the number of groups containing more than a few files should be fairly small and you should hardly feel the difference.
The rest is O(N) and I very doubt you can achieve anything without looping over all your records at least once.

If by "quick" you mean "quick to write", I guess 15 lines of vanilla JavaScript should fit the bill?