Fill in missing periods in between periods where missing transactions

With a lof of help from the community I have been able to create a query that automatically adds new rows for project with accumulated number from previous row if there are no transactions. Unfortunately I came across on thing I would like to change. If there are no transactions in periods in between periods, those periods will be missing. I would like to fill those as well.

Example: Project 12345 GL:

project
amount
period

12345
10
202001

12345
10
202002

12345
10
202003

12345
10
202004

12345
10
202005

12345
10
202006

12345
10
202009

My code is the following:

DECLARE @d date = DATEFROMPARTS(YEAR(getdate()), MONTH(getdate()), 1);

;WITH src AS -- mostly your existing query
(
  SELECT project, 
     period,
     rn = ROW_NUMBER() OVER 
          (PARTITION BY project ORDER BY period DESC),
     amount = SUM(SUM(amount)) OVER 
              (PARTITION BY project ORDER BY period
               ROWS UNBOUNDED PRECEDING)
  FROM dbo.GLProject
  GROUP BY project, period
),
recur AS
(
  SELECT project, period = DATEADD(MONTH, 1, period), amount
  FROM src 
  WHERE rn = 1 AND period < @d
  UNION ALL
  SELECT project, DATEADD(MONTH, 1, period), amount
  FROM recur WHERE period < @d 
)
SELECT project, amount, period = CONVERT(char(6), period, 112)
FROM src
UNION ALL
SELECT project, amount, period = CONVERT(char(6), period, 112)
FROM recur
ORDER BY project, period;

I would like it to fill in the missing periods 202007 and 202008 as well. Like this:

project
amount
period

12345
10
202001

12345
20
202002

12345
30
202003

12345
40
202004

12345
50
202005

12345
60
202006

12345
60
202007

12345
60
202008

12345
70
202009

12345
70
202010

12345
70
202011

12345
70
202012

12345
70
202101

12345
70
202102

12345
70
202103

12345
70
202104

12345
70
202105

12345
70
202106

12345
70
202107

12345
70
202108

12345
70
202109

Any ideas on what to add to my query to solve this?

Fiddle for testing here: https://dbfiddle.uk/?rdbms=sqlserver_2019&fiddle=615712cf8911

You probably won't believe me, but adding the requirement of filling in missing gaps actually drove a simpler solution.

DECLARE @d date = DATEFROMPARTS(YEAR(getdate()), MONTH(getdate()), 1);

;WITH total_range AS 
(
  SELECT project, period = MIN(period)
    FROM dbo.GLProject
    GROUP BY project
  UNION ALL 
  SELECT project, DATEADD(MONTH, 1, period)
  FROM total_range 
  WHERE period < @d
)
SELECT 
  tr.project, 
  amount = SUM(COALESCE(p.amount,0)) OVER (
             PARTITION BY tr.project ORDER BY tr.period
             -- ROWS UNBOUNDED PRECEDING -- not strictly necessary here
           ), 
  period = CONVERT(char(6), tr.period, 112)
FROM total_range AS tr
LEFT OUTER JOIN dbo.GLProject AS p 
ON tr.project = p.project
AND tr.period = p.period;

• Example: db<>fiddle

StackExchange Database Administrators Q#299401, answer score: 5