Oracle retrieve SQL%ROWCOUNT from external script "@"?

I'm trying to build a query for me to use as a template for running data extracts & give me runtime information.

The "@"&RFLD.&RFLE..&REXT.";" section is intended to call any simple SQL query.

So far it works as expected, but I would like it to add the number of rows that was retrieved by the query in @"&RFLD.&RFLE..&REXT.";. For example, @"Path\AddressBook.sql" would retrieve 20 rows, @"Path\ItemMaster.sql" would retrieve 150, @"Path\ItemBranch.sql" would retrieve 1000, & so on.

The issue I'm at is that I cannot put "@" within a BEGIN - END section, & I wasn't successful retrieving SQL%ROWCOUNT outside of the same section.

I will greatly appreciate any ideas. Here my full code:

SET SERVEROUTPUT ON SIZE UNLIMITED;
SET ECHO OFF;
SET VERIFY OFF;
SET FEEDBACK OFF;
SET PAGES 0;
SET LINES 32767;
SET TRIMOUT ON;
SET TRIMSPOOL ON;
SET SQLFORMAT CSV;
--SET MARKUP CSV ON DELIMITER ',' QUOTE ON;

    VARIABLE VDATESTR VARCHAR2(30);
    
    SET TERMOUT ON;
    DECLARE
        VDATESTR VARCHAR(30);
        VDATE VARCHAR(20);
    BEGIN
        SELECT TO_CHAR(SYSTIMESTAMP, 'YYYY-MM-DD HH24:MI:SS') INTO VDATE FROM DUAL;
            :VDATESTR := VDATE;
        DBMS_OUTPUT.PUT_LINE('Start Date-Time: ' || TO_CHAR(SYSTIMESTAMP, 'YYYY-MM-DD HH24:MI:SS TZH'));
    END;
    /

COLUMN RUN_FOLD NEW_VALUE RFLD HEADING Run_Folder;
COLUMN RUN_FILE NEW_VALUE RFLE HEADING Run_File;
COLUMN RUN_EXTN NEW_VALUE REXT HEADING Run_Ext;

    BEGIN
        DBMS_OUTPUT.PUT_LINE('.-.-.-.');
        DBMS_OUTPUT.PUT_LINE('SQL Query File...');
    END;
    /
        ACCEPT PRUN_FOLD CHAR PROMPT '..... Folder Name? ';
        ACCEPT PRUN_FILE CHAR PROMPT '..... File Name (NO extension)? ';    
    SET TERMOUT OFF;
    SELECT RUN_FOLD || DECODE(SUBSTR(RUN_FOLD, -1, 1), '\', NULL, '\') RUN_FOLD, SUBSTR(RUN_FILE, 1, NVL(NULLIF(INSTR(RUN_FILE, '.'), 0), LENGTH(RUN_FILE) + 1) - 1) RUN_FILE, 'SQL' RUN_EXTN 
        FROM (SELECT TRIM('&PRUN_FOLD.') RUN_FOLD, '&PRUN_FILE.' RUN_FILE FROM DUAL);

SET TERMOUT ON;
    BEGIN
        DBMS_OUTPUT.PUT_LINE('.-.-.-.');
        DBMS_OUTPUT.PUT_LINE('Spool Output File...');
    END;
    /

COLUMN SPL_FOLD NEW_VALUE SFLD HEADING Spool_Folder;
COLUMN SPL_FILE NEW_VALUE SFLE HEADING Spool_File;
COLUMN SPL_EXTN NEW_VALUE SEXT HEADING Spool_Ext;

        ACCEPT SSPL_FOLD CHAR PROMPT '..... Folder Name? ';
        ACCEPT SSPL_FILE CHAR PROMPT '..... File Name (NO extension - CSV only)? ';
    SET TERMOUT OFF;
    SELECT SPL_FOLD || DECODE(SUBSTR(SPL_FOLD, -1, 1), '\', NULL, '\') SPL_FOLD,
        SUBSTR(SPL_FILE, 1, NVL(NULLIF(INSTR(SPL_FILE, '.'), 0), LENGTH(SPL_FILE) + 1) - 1) || TO_CHAR(SYSTIMESTAMP, '_YYYYMMDD_HH24MITZH') SPL_FILE, 'CSV' SPL_EXTN
            FROM (SELECT TRIM('&SSPL_FOLD.') SPL_FOLD, '&SSPL_FILE.' SPL_FILE FROM DUAL);

SPOOL "&SFLD.&SFLE..&SEXT.";

@"&RFLD.&RFLE..&REXT.";

SPOOL OFF;

    SET TERMOUT ON;
    DECLARE
        VDATESSTR VARCHAR(30);
        VDATES TIMESTAMP;
        VDATEE TIMESTAMP;
        VDATE_DIF INTERVAL DAY TO SECOND;
    BEGIN 
        SELECT TO_TIMESTAMP(:VDATESTR, 'YYYY-MM-DD HH24:MI:SS'), SYSTIMESTAMP INTO VDATES, VDATEE FROM DUAL;
        DBMS_OUTPUT.PUT_LINE('.-.-.-.');
        DBMS_OUTPUT.PUT_LINE('File Name: &SFLE.');
        DBMS_OUTPUT.PUT_LINE('End Date-Time: ' || TO_CHAR(SYSTIMESTAMP, 'YYYY-MM-DD HH24:MI:SS TZH'));
        
        VDATE_DIF := VDATEE - VDATES;
        DBMS_OUTPUT.PUT_LINE('Cycle Time: ' || VDATE_DIF);
    END;
    /

SET PAGES 70;
--SET MARKUP CSV OFF;

Solution

SQL*Plus is not really a programming environment, only a basic scripting utility, so by choosing to use it you may find some things difficult to achieve. As for reporting row count from a query:

1. Unless you set feedback off, SQL*Plus automatically reports the row count at the end of query output:

   60 rows selected

That will be in your spool file, which you can grep out if you'd like.

You could also use wc -l [file] in shell after cleaning up the spool file to get the # of lines. I generally find it best to use Unix utilities in shell scripting to post-process raw data that I've obtained via SQL*Plus.
You cannot, unfortunately, use SQL%ROWCOUNT because that's a PL/SQL feature and you didn't execute the query within PL/SQL. You could I suppose read the contents of the query file into a variable (or load it with utl_file, then in PL/SQL use dynamic SQL (EXECUTE IMMEDIATE) making it into an insert to run it within PL/SQL, dumping the results to a global temporary table, report stats, then outside the PL/SQL block query the temporary table to spool the results. Or you can output with dbms_output a delimited string per row, or assemble a column and line delimited CLOB of all rows and assign it to a SQL*Plus variable and print it afterwards.

It would hardly be worth all the plumbing to do any of these, however, just for a row count much more easily obtained outside SQL*Plus entirely. Since you've chosen something in the operating system to initiate the job and receive results, that something in the operating system is best suited to reporting stats about what it got.
Just for academic purposes, not at all an advised real solution, you could add a bogus hint that makes the SQL utterly new/unique from any other SQL ever run, then after spooling the results query v$sql looking for that unique identifier in the sql_text and get the rows_processed from the shared pool. But DBAs would frown on multiplying cursors just to get a row count, and rightly so.