Synchronize independent spreadsheet rows, filled by IMPORTRANGE()

Question

I need to synchronize the contents of 2 spreadsheets that reference each other, keeping their rows in sync if a new row is added in one of the sheets.

I've got 2 spreadsheets in Google Sheets (although if there is a cross spreadsheet solution, both Excel and GS that would be great):

• Spreadsheet1 has data in A:F and party1 (a set of users) writes their data in it.
• Spreadsheet2 is and import range of A:F from spreadsheet1 and then has further details written in G:M, the data is written in by party2.

The way it works is party1 writes in their data in rows A1-F10 then party2 writes their additional data in spreadsheet2 based on what party1 has written in.

For example if Spreadsheet1 A1:F10 was a name, price, est delivery time, qty etc. of an item, Spreadsheet2 G1:M10 might be a bunch of data on order date, delivered (yes / no) etc.

The issue I'm currently having is that when the spreadsheets are setup they read across fine i.e. 1-10 in spreadsheet1 lines up with 1-10 in spreadsheet2, but after a while some new rows get added into spreadsheet1 between the old rows 2-5. This throws out the order in spreadsheet2 (now row 4 in spreadsheet1 doesn't line up with the row 4 in spreadsheet2 and the data becomes out of line). Is there away around this so that even if someone adds additional rows in the middle of existing rows both spreadsheets will update?

Answer 1

This is a classic problem in database design; how to associate information in two tables. The usual solution is to use key data; one or more columns that exist in both tables and provide a unique identifier, or key, to associate rows.

We can adapt that idea to your situation, with a script that will adjust the location of rows in Spreadsheet 2 to synchronize with Spreadsheet 1. To do that, we need to identify a key - say the Name column - which must exist in both spreadsheets.

This entails a small change in spreadsheet 2, where a Name column will now appear in column G, following the imported range in columns A-F.

    A      B             C            D       E         F        G         H           I           J
| Name | Price | est delivery time | qty | etc. of | an item | Name  | order date | delivered | blah blah |
 < - - - - - - - - - - - -  Imported  - - - - - - - - - - - >  *KEY*  < - - - - - -  sheet 2  - - - - - >

Demo

Here's how that would look in action! This example is using two sheets in the same spreadsheet, just for convenience. In the demo, a new "Item" row is added in the middle of sheet 1, which automatically appears on sheet 2 thanks to the =IMPORTRANGE() function. The synchronizing function is running on a 1-minute timed Trigger, and you'll see it move things around about 20 seconds in.

You can grab a copy of the spreadsheet + embedded script here.

Code

/**
 * Call syncTables() with the name of a key column.
 */
function doSyncTables() {
  syncTables( "Name" );
}

/*
 * Sync "Orders" spreadsheet with imported rows from "Items" spreadsheet.
 *
 * From: http://stackoverflow.com/a/33172975/1677912
 *
 * @param {String}  keyName    Column header used as key colum, appears
 *                             at start of "Orders" data, following
 *                             "Items" data.
 */
function syncTables( keyName ) {
  var sheet2 = SpreadsheetApp.openById( sheetId2 ).getSheetByName('Orders');

  // Get data
  var lastCol = sheet2.getLastColumn();
  var lastRow = sheet2.getLastRow();      // Includes all rows, even blank, because of =importRange()
  var headers = sheet2.getRange(1, 1, 1, lastCol).getValues()[0];
  var keyCol = headers.lastIndexOf( keyName ) + 1;
  var itemKeys = sheet2.getSheetValues(1, 1, lastRow, 1).map(function(row) {return row[0]});
  var itemData = sheet2.getSheetValues(1, 1, lastRow, keyCol-1);
  var orderData = sheet2.getSheetValues(1, keyCol, lastRow, lastCol-keyCol+1);

  var ordersByKey = [];  // To keep track of orders by key

  // Scan keys in orderData
  for (var row=1; row<orderData.length; row++) {
    // break loop if we've run out of data.
    var orderKey = orderData[row][0];
    if (orderKey === '') break;

    ordersByKey[ orderKey ] = orderData.slice(row, row+1)[0];

    var orderKey = orderData[row][0];
  }

  var newOrderData = [];  // To store reordered rows

  // Reconcile with Items, fill out array of matching orders
  for (row = 1; row<itemData.length; row++) {
    // break loop if we've run out of data.
    var itemKey = itemData[row][0];
    if (itemKey === '') break;

    // With each item row, match existing order data, or add new
    if (ordersByKey.hasOwnProperty(itemKey)) {
      // There is a matching order row for this item
      newOrderData.push(ordersByKey[itemKey]);
    }
    else {
      // This is a new item, create a new order row with same key
      var newRow = [itemKey];
      // Pad out all columns for the new row
      for (var col=1; col<orderData[0].length; col++) newRow.push('');
      newOrderData.push(newRow);
    }
  }

  // Update spreadsheet with reorganized order data
  sheet2.getRange(2, keyCol, newOrderData.length, newOrderData[0].length).setValues(newOrderData);
}

Answer 2

the current answer by mogsdad is great as always. i just wanted to point out a less complex alternative:

if you can live with preventing spreadsheet1 from allowing insertions or deletion of rows, you will avoid the issue. instead of removing rows you could use a column to mark "deleted" for example (and use filters to remove from view).

to prevent row insertions and deletions in spreadsheet1, simply select an entire unused column to the right, and create a protected range so none of the editors have permission. that prevents modifying at the row level up to the last existing row (but new rows can still be inserted below the range)

it also doesnt prevent users from swapping two row's data. but its still good to know about this simpler alternative.