We Keep Coding

How to document merge a parent record and all of its child records

I'm creating a document merge (mail merge) from Google App Maker to a Google Document template. I can do so successfully when merging one single record, but how do you merge several records into the one document?
I have an purchase_orders parent record which has several purchase_order_line_items child records but I can't seem to get all of these records into a single document merge.
A similar question (Document Merge with Google App Maker) was asked by by Johan W with a comprehensive answer by Markus Malessa and Pavel Shkleinik (thank you!). However, it only caters for cases when you are merging one single record.
I have tried to build on their answer by using a second for loop to get the data of all associated child records. The script runs but only seems to merge the first child record; not all of them.
Here is an example of the server-side code I've tried to use:
function Export(key, key2) {
// Get the parent record by its key, which was passed by the first parameter above
var record = app.models.Purchase_Orders.getRecord(key);
// Get the first child record by its key, which was passed by the second parameter above
var childRecord = app.models.Purchase_Order_Line_Items.getRecord(key2);
// Get the Google Document which will be used as a template for this merge
var templateId = '1Xbt8camqHJYrhBnx0a6G2-RvTvybqU0PclHifcdiLLA';
//Set the filename of the new merge document to be created
var filename = 'Document for Customer ' + new Date();
//Make a copy of the template to use as the merge document
var copyFile = DriveApp.getFileById(templateId).makeCopy(filename);
//Get the Google Docs ID of the newly created merge document
var copyDoc = DocumentApp.openById(copyFile.getId());
var copyBody = copyDoc.getBody();
// Replace the field names in the template with the field data from the parent record
var fields = app.metadata.models.Purchase_Orders.fields;
for (var i in fields) {
console.log(i);
var text = '<<' + fields[i].name + '>>';
var data = record[fields[i].name];
if (data !== null) {
copyBody.replaceText(text, data);
} else {
// do nothing
}
}
// Replace the field names in the template with the field data from the child records
childFields = app.metadata.models.Purchase_Order_Line_Items.fields;
for (i in childFields) {
console.log(i);
var childtext = '<<' + childFields[i].name + '>>';
var childdata = childRecord[childFields[i].name];
if (childdata !== null) {
copyBody.replaceText(childtext, childdata);
} else {
// do nothing
}
}
}
How can I improve my code so that all associated child records are merged into a single document?
How can I set up my Google Document template to cater for any number of child records?

Rather than passing in the child record key via a second parameter, I would suggest just passing in the parent key and then changing your function as follows:
function Export(key) {
// Get the parent record by its key, which was passed by the first parameter above
var record = app.models.Purchase_Orders.getRecord(key);
// Get the first child record by its key, which was passed by the second parameter above
var childRecords = record.Purchase_Order_Line_Items;
// Get the Google Document which will be used as a template for this merge
var templateId = '1Xbt8camqHJYrhBnx0a6G2-RvTvybqU0PclHifcdiLLA';
//Set the filename of the new merge document to be created
var filename = 'Document for Customer ' + new Date();
//Make a copy of the template to use as the merge document
var copyFile = DriveApp.getFileById(templateId).makeCopy(filename);
//Get the Google Docs ID of the newly created merge document
var copyDoc = DocumentApp.openById(copyFile.getId());
var copyBody = copyDoc.getBody();
// Replace the field names in the template with the field data from the parent record
var fields = app.metadata.models.Purchase_Orders.fields;
for (var i in fields) {
console.log(i);
var text = '<<' + fields[i].name + '>>';
var data = record[fields[i].name];
if (data !== null) {
copyBody.replaceText(text, data);
} else {
// do nothing
}
}
// Replace the field names in the template with the field data from the child records
var childFields = app.metadata.models.Purchase_Order_Line_Items.fields;
var table = [];
var tableheader = [];
for (i in childFields) {
console.log(i);
tableheader.push(childFields[i].displayName);
}
table.push(tableheader);
for (i in childRecords) {
var data = [];
for (var j in childFields) {
data.push(childRecords[i][childFields[j].name]);
}
table.push(data);
}
copyBody.appendTable(table);
The table building is based on a 2D array and the documentation is here https://developers.google.com/apps-script/reference/document/table. But you will also need to remove your prebuilt table in favor of just appending a table instead. This way you are not dependent on the quantity of child records being fixed like they currently are in your document template. Also, the variable for childRecords may or may not work, I have not tested this since I am unsure if prefetch works in conjunction with .getRecord(key). This may require some additional testing but hopefully this will provide enough guidance.
Thought I would add this as an alternative. Lets say you keep your table but remove all the rows with exception for the header row then you could still use DocumentApp service to add your rows to the table like so:
var tableheaderfieldnames = ['Quantity_for_PO', 'Inventory_Item.id', 'Unit_Price']; //set a fixed table header with the field names, uncertain if the table header for the related inventory item will work or not
var table = copyBody.getTables()[0];
for (i in childRecords) {
var row = table.appendRow();
for (var j in tableheaderfieldnames) {
row.appendTableCell(childRecords[i][tableheaderfieldnames[j]]);
}
}
Keep in mind that AM does not allow you to use FK references, so for your inventory item that appears to use a fk field you may need to tinker around with setting the proper name reference for when you are trying to fill in the item in your table.

Updating only some of object

I have a table in my database that I would like to be able to change some of the sections and keep the other functions as they were however it is updating the table so that the two are changed but the other 3 become empty. is there any way to change this?
$(function Tuesday(){
// CREATE A REFERENCE TO FIREBASE
var dateTuesdayRef = new Firebase('https://shiftsapp.firebaseio.com/roster');
// REGISTER DOM ELEMENTS
var date2Field = $('#date2Input');
var emp1put2Field = $('#emp1Input2');
var emp2put2Field = $('#emp2Input2');
var emp3put2Field = $('#emp3Input2');
var emp4put2Field = $('#emp4Input2');
var emp5put2Field = $('#emp5Input2');
var enter2Field = $('#enter2');
// LISTEN FOR KEYPRESS EVENT
enter2Field.keypress(function (e) {
if (e.keyCode == 13) {
//FIELD VALUES
var dateTuesday = date2Field.val();
var emp1put2 = emp1put2Field.val();
var emp2put2 = emp2put2Field.val();
var emp3put2 = emp3put2Field.val();
var emp4put2 = emp4put2Field.val();
var emp5put2 = emp5put2Field.val();
var enter2 = enter2Field.val();
//SAVE DATA TO FIREBASE AND EMPTY FIELD
var obj2 = {};
obj2[dateTuesday] = {
emp1:emp1put2,
emp2:emp2put2,
emp3:emp3put2,
emp4:emp4put2,
emp5:emp5put2
}
dateTuesdayRef.child(dateTuesday).set({emp1:emp1put2,
emp2:emp2put2,
emp3:emp3put2,
emp4:emp4put2,
emp5:emp5put2});
enter2Field.val('');
}
});
});

Get the values for the things you want to stay the same from your server, and feed them back when you set the object. You could also use a custom function to autofill undefined values like I have suggested.

From the table at the top of the Firebase guide on saving data:
set( ): Write or replace data to a defined path, like messages/users/
update( ): Update some of the keys for a defined path without replacing all of the data
So if you call update() instead of replace, it will only change the values of the properties you pass in and leave other values unmodified.

Get all items in NotesXSPDocument

In my Notes Database, I perform an audit when the document is saved. Pretty easy in LotusScript. I grab the original document (oDoc) from the server, then in the document I modified (mDoc), I do a Forall loop that gets the names of each item; forall item in mDoc.items. Grab the same item from oDoc, execute a function with the new item as an argument that will run down a case statement that will see if its a field we care about. if so, I update a set of list values in the document with "When", "Who", "What field", and the "New Value".
I'm doing this in a server side script. In trying this, I discovered a couple of interesting things;
currentDocument is the NotesXSPDocument that contains everything that was just changed.
currentDocument.getDocument() contains the pre-change values. It also returns a NotesDocument which has the "items" field that I can run through.
Thing is, I need something similar in the NotesXSPDocument. Is there a way in an iterative loop to grab the names and values of all items from there?
Here's the broken code. (Currently it's walking through the NotesDocument items, but those are the old values. I'd rather walk down the XSP document items)
function FInvoice_beginAudit() {
var original_doc:NotesDocument = currentDocument.getDocument();
var oItem:NotesItem;
var oItems:java.util.Vector = original_doc.getItems();
var iterator = oItems.iterator();
while (iterator.hasNext()) {
var oItem:NotesItem = iterator.next();
item = currentDocument.getItemValue(oItem.getName());
if (oItem == undefined) {
var MasterItem = ScreenAudit(doc,item,True)
if (MasterItem) { return true }
} else {
if (item.getValueString() != oItem.getValueString()) {
var MasterItem = ScreenAudit(doc,Item,True);
if (MasterItem) { return true }
}
}
}
}

You can get both versions of a document after submit - the original and the one with changed/new values:
original: var original_doc:NotesDocument = currentDocument.getDocument();
changed: var changed_doc:NotesDocument = currentDocument.getDocument(true);
This way you can compare the items for changes.
But, there is a pitfall: after assigning "changed_doc" to currentDocument.getDocument(true) the "original_doc" has the changed values too because both variables point to the same document. That's why we have to copy all items from currentDocument.getDocument() to a new temporary document first and only after get the changed values with currentDocument.getDocument(true). As an alternative you could read the original document from server like you do in LotusScript.
This is a code for detecting changed items as a starting point:
var original_doc:NotesDocument = database.createDocument();
currentDocument.getDocument().copyAllItems(original_doc, true);
var changed_doc:NotesDocument = currentDocument.getDocument(true);
var oItems:java.util.Vector = original_doc.getItems();
var iterator = oItems.iterator();
while (iterator.hasNext()) {
var oItem:NotesItem = iterator.next();
var itemName = oItem.getName();
var cItem:NotesItem = changed_doc.getFirstItem(itemName);
if (cItem.getText() !== oItem.getText()) {
print("changed: " + itemName);
}
oItem.recycle();
cItem.recycle();
}
original_doc.remove(true);
original_doc.recycle();

IndexedDB: Retrieve item with max value

Suppose I have an IndexedDB collection with name items. All items have fields:
id
name
revision
revision field is a number field. I need to retrieve an item with max value of revision (or at least just retrive max revision value). What is the best way to do it?

First thing you need to do is create index on the revision field.
Then you need a search function which will use that index and open the index with inverse order of the objects. Then the first object will be the object you are looking for.
var index = objectStore.index('revision');
index.openCursor(null, 'prev');
The null states that you are searching for all values not a specific one, and the second parameter is the direction of the search.
Here is the sample code:
function getMaxNumber (callback) {
var openReq = indexedDB.open(baseName);
openReq.onsuccess = function() {
var db = openReq.result;
var transaction = db.transaction(objectStoreName, 'readonly');
var objectStore = transaction.objectStore(objectStoreName);
var index = objectStore.index('revision');
var openCursorRequest = index.openCursor(null, 'prev');
var maxRevisionObject = null;
openCursorRequest.onsuccess = function (event) {
if (event.target.result) {
maxRevisionObject = event.target.result.value; //the object with max revision
}
};
transaction.oncomplete = function (event) {
db.close();
if(callback) //you'll need a calback function to return to your code
callback(maxRevisionObject);
};
}
}
Since the IndexedDB api is async you would need a callback function to return the value to your code.

Creating hash array in Google Apps Script

I've been trying to work with Trello and the Google Apps Script this week. I am trying to create an array of hashes that I can then use to load the spreadsheet. Google apps script doesn't like the typical javascript code of creating hashes. I've looked up the docs but they don't have anything like hashes...they say to:
var object = [];
var object1 = {};
object.push(object1);
This wont work because I'm essentially trying to do something like:
var hash={name: , label: };
var n= someNumber;
var l= someLabel
var hash.push(name: n, label: l);
Essentially that is the code I have right now. But here is my entire function:
function getData(){
var list={};
//get the list of delivered cards from Trello
var listRequest = authorizeToTrello(); // get authorization
var result = UrlFetchApp.fetch("https://trello.com/1/lists/4fea3a2c3a7038911ebff2d8/cards",
listRequest);//fetch list
var listOfCards = Utilities.jsonParse(result.getContentText());//Google app utility format json
//outer loop to iterate through list of Cards
for(var i=0; i < listOfCards.length; i++){
var cardId = listOfCards[i].id; //get the id of a single card
var l = listOfCards[i]["label"]; //get the label for the our structure
//get a json object for a single card within the list of cards iteration
var cardRequest = authorizeToTrello();
var getCard = UrlFetchApp.fetch("https://trello.com/1/cards/" + cardId + "/actions", cardRequest);
var singleCard = Utilities.jsonParse(getCard.getContentText());
//inner loop to iterate the single cards JSON objects
for(var j=0; j < singleCard.length; j++) {
if(singleCard[j].data != undefined && singleCard[j].data.listAfter != undefined)
{
var str = singleCard[j]["data"]["listAfter"]['name'];
if(str === "Delivered Q3 2012"){
var n = singleCard[j]['memberCreator']['fullName'];
}
}
}
//push the data to list
list.push(n,l);
}
return name, label; //return list for output
}

Reading the question, I understood that the author needs to know how to create an associative array in a GAS. If it is correct then here is a couple of links (here and here) and a sample code is bellow.
function testMap() {
var map = {};
map["name1"] = "value1";
map["name2"] = "value2";
return map;
}
If the author needs really
an array of hashes
then there are a couple of ways depending on which hash algorithm is required.
to use the Utilities.computeDigest method to calculate a hash of a string using one of available algorithms.
if the required hash calculation algorithm is not supported by the Utilities.computeDigest, then is possible to write own implementation as it is done for the BLAKE function.
Here is a sample of how to create an array of hashes using the MD5 hash.
function testHash() {
var array = [];
array.push(Utilities.computeDigest(Utilities.DigestAlgorithm.MD5, "value1"));
array.push(Utilities.computeDigest(Utilities.DigestAlgorithm.MD5, "value2"));
return array;
}
P.S. The return line of the author code return name, label; //return list for output
is not correct - only the label variable value is returned. To return a couple of variables as an array is necessary to write return [name, label];. Or may be the author needs to return the list variable and not name and label.

I know this is an old post / question, but i would like to update my answer since the original anwer (1st answer) is misleading. I was myself looking for how to return associative arrays back to a cell in the spreadsheet, but alas.. "YOU CANNOT". Google spreadsheet MUST want an numerically indexed array or an object. Otherwise it returns "#ERROR".
Here are the steps to replicate the issue.
function testMap() {
var map = {};
map["name1"] = "value1";
map["name2"] = "value2";
return map
Formula in your cell: =testMap()
Value in your cell: Thinking... #ERROR
Solution (rather a workaround)
1: Transfer your objects from your associative array into a numerically indexed array using for-each type loop.
var temp = new Array();
for (var i in map) {
temp.push([i,map[i]])
// optionally use activeSheet.getRange(X:X).setValue([i,map[i]])) function here.
// set values will not work in cell functions. To use it via cell functions, rerun / trigger the functions using an on_edit event.
}
If you used a temp like numerically indexed array, you can return "temp" back to the calling cell.

Summary: For onEdit() purposes, use Cache Service to define associative array data.
Here's a shared Gsheet demonstrating this curious behavior. I tried the following solution in programmatically defining an associative array based on data in a Google sheet.
var assocArr = {
labels: {},
init: function () {
var sheet = SpreadsheetApp.getActiveSpreadsheet().getSheetByName('sheetName');
var values = sheet.getDataRange().getValues();
for(var row in values) {
assocArr.labels[values[row][0]] = values[row][1];
};
for(var key in assocArr.labels) {
Logger.log("key: %s, value: %s",key, assocArr.labels[key]);
};
return(void(0));
},
};
To execute this, you run the init() method in the onOpen() event handler.
function onOpen() {
assocArr.init();
var key = 'test';
SpreadsheetApp.getUi().alert( assocArr.labels[key] );
Logger.log("onOpen: key: %s, value: %s",key, assocArr.labels[key]);
};
The logger message confirms that init() loads the data from the worksheet.
Now if I try to reference this assocArr object in onEdit() it returns undefined for all key values.
function onEdit(event) {
var key = 'test';
SpreadsheetApp.getUi().alert( assocArr.labels[key] );
Logger.log("onEdit: key: %s, value: %s",key, assocArr.labels[key]);
};
I infer that for security reasons, Google limited the simple-trigger onEdit() to not have global variable scope, same as they voided the utility of the event.user property.
Now instead if I simply put the key-value pair in the cache, it works! Here is the complete code that works using the Cache Service.
var cache = CacheService.getPrivateCache();
var assocArr = {
init: function () {
var sheet = SpreadsheetApp.getActiveSpreadsheet().getSheetByName('Account Labels');
var values = sheet.getDataRange().getValues();
for(var row in values) {
cache.put(values[row][0], values[row][1], 3600);
};
return(void(0));
},
};
function onOpen() {
assocArr.init();
var key = 'test';
SpreadsheetApp.getUi().alert( cache.get(key) );
Logger.log("onOpen: key: %s, value: %s",key, cache.get(key));
};
function onEdit(event) {
var key = 'test';
SpreadsheetApp.getUi().alert( cache.get(key) );
Logger.log("onEdit: key: %s, value: %s",key, cache.get(key));
};
Curiously, the onEdit() has the cache variable in its scope.
Here again is the shared Gsheet demonstrating this curious behavior.

I found this really quick way that is not listed
Create a json object (array style)
var myArray = {
1:{"id": "inprogress","title" : "in Progress"},
2:{"id": "notstarted","title" : "Not Started"},
3:{"id": "completed" ,"title" : "Completed"}
};
read the json
// get the lenght of the json object
var jsonSize = Object.keys(myArray).length;
// use this in a loop
for (var i = 1; i < Object.keys(jsonSize).length; i++) {
var title = myArray[i].title;
}
Works like a charm for me