DEV Community: fiona

Jaspersoft Studio: Using a new font in the report

fiona — Sun, 19 May 2024 21:43:27 +0000

Problem (murmurs)

This is my first time using Jaspersoft Studio. Once my coworkers downloaded a new font, the font would appear in the selection box automatically after restarting Jaspersoft Studio. However that is not my case. Neither restarting the application nor the computer works for me.

Solution

I refer to the documentation from Jaspersoft community.

On the tool bar, select Window > Preferences. Then find Jaspersoft Studio > Fonts.
Select Add From Path on the right and select the folder of the .ttf file of the downloaded font. (Not from the Fonts folder in Windows OS but the actual file path.) Select Finish and after the application analyzes the folder, the font would appear in Font List.
Apply and close the Preferences window. Right click the project you are working on and select Properties. You can see the font in Font List of Workspace Settings. 🙌
The font would now appear in the selection box as expected. 🎉🎉

replace the last comma to period in string

fiona — Sat, 12 Mar 2022 08:37:11 +0000

To slightly modify a string, like replacing its first character to capital or last character from comma to period, there are two ways to achieve.

use `replace()` with regex

There are two things worth notice:

without using regular expression, only the first-matched substring will be replaced.
replace() function returns a new string and left the old string unchanged.

// replace the last comma to period
let names = "anne, ben, carol, dennis, evans,";
names = names.replace(/,$/, '.');
// anne, ben, carol, dennis, evans.

use `slice()` and append

// replace the last comma to period
let names = "anne, ben, carol, dennis, evans,";
names = names.slice(0, -1) + '.';
// anne, ben, carol, dennis, evans.

slice(indexStart, indexEnd) and substring(indexStart, indexEnd) are almost identical with some slight differences:

indexEnd is optional in both functions, if not specified it would extract to the end of the string. Otherwise, it is the first character to exclude.
If value is less than 0 or NaN, substring() would consider 0. slice() would consider NaN 0, and negative values count backwards from the end.
If value is greater than stringName.length, it would consider stringName.length in both functions.
substring() swap the arguements if indexStart is greater than indexEnd, slice() returns an empty string.

set up keyboard shortcut to open terminal on mac

fiona — Sat, 05 Mar 2022 07:03:20 +0000

I am currently using Monterey. These are actually simple steps to set up the shortcut. (NO coding required! yay! 😌)

open Automator

It is under Applications > Other > Automator. Once open a new file, select Quick Action.

For convenience, if the window is blank like below, select View > Show Library.

Use the search box to find Launch Application. Then double-click or drag it to the right-hand side.

Change the application from the default Contact to Terminal. (Select Other... and Terminal can be found in the Utilities folder)

Select Run on the upper right of the window to make sure the terminal is launched as expected. Then save the file. I have my file named open terminal.

open System Preferences

Under System Preferences > Keyboard > Shortcuts > Services, find the file name. By clicking on it, select Add Shortcut and set up the keyboard shortcut you prefer. I use cmd+opt+T as linux command.

And now it is all set up! 😬

problem: Java Date and Time

fiona — Sun, 27 Feb 2022 19:15:16 +0000

Link to the problem

Input would be month day year, and output the day of week.

//sample input
08 05 2015

//sample output
WEDNESDAY

proper solution

DISCLAIMER: the solution referenced from the editorial

public static String findDay(int year, int month, int day) {
    Calendar date = Calendar.getInstance();
    date.set(year, month, day);
    return date.get(Calendar.DAY_OF_WEEK); //error
}

The code would have error because date.get(Calendar.DAY_OF_WEEK) returns integer value of day of week. For example, returns 1 for SUNDAY, 2 for MONDAY, and 7 for SATURDAY.

We need to create an additional array to reference day of week:

// outside main method
static String[] dayOfWeek = {"SUNDAY", "MONDAY", "TUESDAY", "WEDNESDAY", "THURSDAY", "FRIDAY", "SATURDAY"};

public static String findDay(int year, int month, int day) {
    Calendar date = Calendar.getInstance();
    date.set(year, month, day);
    return dayOfWeek[date.get(Calendar.DAY_OF_WEEK)-1];
}

use another class

Instead of the class java.util.Calendar it mentioned, java.time.LocalDate (Java 8 and higher) actually finish all the work nice and neat.

public static String findDay(int year, int month, int day) {
    LocalDate date = LocalDate.of(year, month, day);
    return date.getDayOfWeek().toString(); //getDayOfWeek() returns a DayOfWeek object
}

android to mac file transfer

fiona — Thu, 23 Dec 2021 09:52:37 +0000

I simply want to transfer file from android to mac. Initially I use Android File Transfer, however it stucks at 4GB.

By considering this post, I install android-platform-tools for file transfer.

step 1

on mac, install android-platform-tools.

brew install android-platform-tools

step 2

connect the android device with USB-C to mac.
on the device, go to settings > developer options > enable USB debugging

NOTE: on xiaomi 9, tap MIUI version 7 times to enable developer options

step 3

on mac, run adb devices to confirm that android is connected to mac.

step 4

to copy file to mac, run:

adb pull <device_file_location> <mac_location>

to copy file to device, run:

adb push <mac_file_location> <device_location>

please let me know if there is any other way more straightforward! 😬

Heroku: deploy with an existing git repo

fiona — Tue, 27 Oct 2020 19:55:58 +0000

This is my first time using Heroku. Most tutorials I found on the internet deploy their app while creating a new git repository. However, I already have an existing git repository. After trial and error, I realize the documentation is brief and clear all the time but I wasn't able to understand at first. 😧

With an existing git repository currently working on origin develop branch, in order not to let heroku settings affect develop branch, there is NO need to create a new origin branch then push it there. (<-- it wouldn't affect at all...a silly thing I did at first)

Required packages (pip install)

gunicorn
whitenoise
django-heroku
psycopg2 # not sure if preinstalled from prior

Django settings for Heroku

create Procfile
- do not add extension after file name!
- most tutorials write web: gunicorn mysite.wsgi in Procfile, use web: gunicorn --pythonpath mysite mysite.wsgi instead. (<-- repository location reseaon)
create requirements.txt
- pip freeze > requirements.txt
- in order for heroku to detect python buildpack
modify settings.py
- add import django_heroku on the top of the file
- add django_heroku.settings(locals()) in the file

It's important to notice the working directory. Normally, the django mysite project folder will also be the github repository folder. But I have my github folder outside mysite project folder.

Below is how I locate my required files.

# my local github repository folder
.
├── Procfile
├── README.md
├── data
├── mysite
│   ├── mysite
│   │   ├── __init__.py
│   │   ├── settings.py
│   │   ├── urls.py
│   │   └── wsgi.py
│   ├── prediction
│   │   ├── admin.py
│   │   ├── apps.py
│   │   ├── models.py
│   │   └── views
│   ├── static
└── requirements.txt

1. Create a Heroku account

Install Heroku CLI in order to use heroku command in the command line.

2. Create an app

Under the working directory (for my case, my local repository), run:

heroku create <app_name>

Heroku generates a random name if <app_name> left blank.
Or create an app from the website > dashboard > create an app.
(Most settings can actually be done from the website as well.)

3. Create a Heroku remote

heroku git:remote -a <app_name>

# to check if remote is added
git remote -v

4. Deploy the app

# either one is fine
git push heroku master
git push heroku main

Note that the local branch name can only be master or main. If the local branch is neither, run git push heroku <local_branch_name>:master

5. Visit the app

heroku open

#if required login
heroku login

Git: push to a new remote branch

fiona — Tue, 27 Oct 2020 17:19:13 +0000

It is important to specify <remote> <branch> in every pull and push in the scenario. If not specified, would be default master.

branch

# list all branches, including local and remote
git branch -a

# create new local branch
git branch <new_branch_name>

# switch to another branch
git checkout <another_branch>

# switch no to new local branch (combine the steps above)
git checkout -b <new_branch_name>

# delete a branch, use -D for force delete
# need to checkout to another branch before delete
git branch -d <branch>

clone

# clone a repository in specific remote branch
git clone --branch <branch_name> <remote_repo_url>
git clone -b <branch_name> <remote_repo_url>

# clone a specific remote branch 
# and does not track other remote branch
git clone -b <branch_name> --single-branch <remote_repo_url>

pull (fetch+merge)

git pull origin <branch>

push

# push to a new remote branch not yet created
git push --set-upstream <remote> <branch_name>
git push -u <remote> <branch_name>

# push to remote branch
git push <remote> <branch_name>

# delete a remote branch
git push <remote> --delete <remote_branch_name>

Usually <remote> would be origin unless working with another repository (eg. deploying app to heroku)
origin stands for the remote repository that a project was originally cloned from.

overwrite local repository

git fetch --all
git reset --hard origin/master

replace master with whatever branch originally wanted

git lfs

fiona — Mon, 24 Aug 2020 08:28:24 +0000

I encounter an error when I push my database to Github, and git suggest to checkout git lfs (large files storage) .

Quoted from the article by Atlassian,

Git LFS does this by replacing large files in your repository with tiny pointer files. During normal usage, you'll never see these pointer files as they are handled automatically by Git LFS

It's so easy to apply even with an existing large file, for the case, my database.

# for the very first time, run install
git lfs install

# add the large files you want to track
git lfs track '*.sqlite3'

# without argument, you can see the patterns being tracked
git lfs track

# show the files being tracked
git lfs ls-files

Note that after running git lfs track '*.sqlite3', it will auto create a file .gitattributes record the patterns.

Then commit your large file and .gitattributes to github, done.

When others trying to git pull the remote repo, after running git lfs install (only the very first time), they can git pull as usual.

reference:

UPDATE:
After adding more data to the database, github reaches its limit even with lfs.😓
Let's just keep the database in local file.
(^._.^)ﾉ

Django: set default value for datetime field

fiona — Sat, 15 Aug 2020 16:42:47 +0000

When running makemigrations after adding new fields to the current model, django would ask for a default value in order to save in the datebase when fields created.

However, I do not set default value to fields I know it is not nullable and would have a certain value assigned already before saving in the database.

The way I compromise is to give a default value when django asks me to in the terminal. For numeric fields I go with 0, and an empty string for character fields.

I find it quite confusing when I first encounter the issue of setting a default value for datetime field since I do not often encounter this problem if I set the field in the first place as the model initially created.

The terminal gives a useful tip suggest that you can use timezone.now() as default. But I still kind of prefer a more generic datetime. And datetime.datetime(1900,1,1) works.

PS. I am referring to both date and datetime field when mentioning datetime field.

2020/8/17 edit: It is rather dumb to not set default in field argument, cause later giving default values in terminal would be such a pain when there are many fields.

Django: reset auto-increment id to 1

fiona — Tue, 04 Aug 2020 08:32:13 +0000

The discussion in stackoverflow here gave a clear view of how to handle the problem, which many argue and I do agree that the id should not be reset since it is unique and should not be recreated.

Anyway, for the clarity of my database id, I still tend to do so. Simply erase table data from SQLiteStudio, or run objects.all().delete() does not reset the id. The answer provided in the above stackoverflow for sqlite is to run a sql command

-- reset index

DELETE FROM <table_name>; 
DELETE FROM SQLite_sequence WHERE name='<table_name>';

UPDATE:
It also works when adding data to table with existing data.
In a senario where you import the wrong data, you can remove the data by the following:

-- delete certain rows (only keep rows that are between)

DELETE FROM <table_name>
WHERE id NOT IN (SELECT id FROM <table_name> WHERE id BETWEEN <int> AND <int>)

Then run the previous reset index command to let the index remain consistent before import another data.

Django: switch from sqlite3 to mysql

fiona — Mon, 02 Dec 2019 19:05:43 +0000

assume there is an existing sqlite3 db with data in django app

in command prompt:

python manage.py dumpdata > datadump.json
# to export existing data into a json file

create a new schema in mysql server
(mysql should have been installed already)
for my own information: Difference Between Schema / Database in MySQL

in settings.py:

DATABASES = {
    'default': {
        'ENGINE': 'django.db.backends.mysql',
        'NAME': '<schema_name>',
        'USER': 'root',
        'PASSWORD': '*****',
        'HOST': 'localhost',
        'PORT': '<port_num>',
    }
}