Enjoy every bit: May 2021

Monday, May 31, 2021

Salesforce Permission : Manual Share Record using Apex Bulk API 分享單筆資料代碼

There are different way of sharing a "private" object record in sfdc.

Include using inherit role hierarchy ,sharing rule , or pressing custom button "Sharing"

on record detail page in user interface.

Codes below try to use apex API to do the same thing as the "Sharing" button.

#Salesforce分享單筆資料代碼

    // Create new sharing object for the custom object named MyCustomObject.
      MyCustomObject__Share myCustomObjectShr  = new MyCustomObject__Share();
   
      // Set the ID of record being shared.
      myCustomObjectShr.ParentId = recordId;
        
      // Set the ID of user or group being granted access.
      myCustomObjectShr.UserOrGroupId = userOrGroupId;
        
      // Set the access level.
      myCustomObjectShr.AccessLevel = 'Read';
        
      // Set rowCause to 'manual' for manual sharing.
      // You may change to other row cause if you have created other sharing reason.
      myCustomObjectShr.RowCause = Schema.myCustomObject__Share.RowCause.Manual;
        
      // Insert the sharing record and capture the save result. 
      Database.SaveResult sr = Database.insert(myCustomObjectShr,false);
 
      // Process the save results.
      if(sr.isSuccess()){
         // Indicates success
         System.debug(LoggingLevel.INFO, 'Done');
      }
	else{
          // read first save result error.
         Database.Error err = sr.getErrors()[0];
         System.debug(LoggingLevel.ERROR, 'Error code :' + err.getStatusCode() );
      }

Thursday, May 13, 2021

Batch script commands to split big text file in windows with efficiency comparison 命令提示字元分割打開超大型檔案

When we try to open the big file (e.g. : server log), we may meet an error when we open with text editor.

Notepad : File is too large for Notepad .

Notepad++ : File is too big to be opened .

At this point, the only way is try to split the file using program. Surely it is always good to use those "close to metal" language like C++ . However , you may not want to install compiler , SDK etc.

Is there any convenient way ? Yes ,using batch script is always a solution .
But always remember , it may takes you more than half hour to split 1GB file into 100 small files.

Code in Batch .bat script (Method 1 Faster)

@echo off
setLocal EnableDelayedExpansion
set limit=50000 #Rows per file
REM can be in any extension (e.g. csv ) , as long as it is a text file 
set file=YourFileName.txt
set lineCounter=1
set filenameCounter=1

set name=
set extension=
for %%a in (%file%) do (
    set "name=%%~na"
    set "extension=%%~xa"
)

for /f "tokens=*" %%a in (%file%) do (
  
    if !lineCounter! gtr !limit! (
        set /a filenameCounter=!filenameCounter! + 1
        set lineCounter=1
        echo Created !splitFile!.
    )
    REM Output filename pattern YourFileName-part1.csv , YourFileName-part2.csv
    set splitFile=!name!-part!filenameCounter!!extension! 
    echo %%a>> !splitFile!
  
    set /a lineCounter=!lineCounter! + 1
)

Code in Batch .bat script (Method 2 Slower)

@echo off
setlocal enableextensions disabledelayedexpansion
set STARTTIME=%TIME%
set "nLines=50000" #Rows in each file
set "line=0"
REM can be in any extension (e.g. csv ) , as long as it is a text file 

for /f "usebackq delims=" %%a in ("InputFileName.txt") do (
    set /a "file=line/%nLines%", "line+=1"
    setlocal enabledelayedexpansion
    for %%b in (!file!) do (
        endlocal
         >>"OutputName_%%b.txt" echo %%a 
         REM Ouput filename pattern : OutputName_1.txt , OutputName_2.txt

         REM Filename prefix will NOT follow the input file in this way.
    )
)

Efficiency Comparison

The way of writing the script could lead to double process time.

Test with 100K rows , >18 MB file.

For example : Split a file 18.7MB into 3 files (each file with max 31670 rows).

Using method 1 one takes 23s , while using method 2 takes 46s.

Test with 5000K rows , >1 GB file

With method 1 , if a 1.17GB (1230188 KB , around 5000Krows inside) has to be split into 100 small files (50K rows @file) , it takes 30m42s .

With method 2, believer me, you don't want to try.

PC configuration reference

16GB RAM , with 8 x QuaCores , i5 CPU.

Conclusion

- Better use C++ to split file over 1GB . ;)