Auto Multiple Choice for Linux. AMC. Python-assisted source.tex creation. Only for tests with separate answer sheets!

This is an automation of AMC redaction. The internal AMC-TXT should be working in a similar way, but I wanted the one that reflects my needs, if you want you can freely use it :)

Place both python_for_AMC.py and wrongCorrect.png files start.sh file in the folder with questions you want to convert. Give it permission to be executed, normally in terminal type chmod +x start.sh. The wrongCorrect.png image will be fetched into the same project directory while executing python_for_AMC.py start.sh since 0.4.0 1.0.0 Version

The main rules:

1. Questions should be saved in separate Qs.txt, Qms.txt or Qhs.txt one any file Qs.txt, file utf-8 encoded depending on the type of questions
2. Simple question should start with 'qqq' (see the example below)
3. Multi correct answer questions should start with triple 'qmN' (see the example below)
4. Horizontal answer questions should start with 'qhq' (5 columns by default) or 'qhN', where N is an integer reflecting the number of columns. I decided not to use the {choiceshoriz} since this option won't align the answers left. If you have 3 and fewer answers (e.g. True or False) you could leave default 'qhq'
5. Correct answer starts with '+++'
6. Wrong answer starts with '---'
7. Please note, that 'Fifty Shades of Grey' will be omitted because they don't have a recognised prefix

So far, most of the goals I set for the project achieved (simple questions, qmN, qhq/qhN)

In future, the code will also process image containing questions (as 'qiq').

The example for processing will look like:

qm2What is the best movie in the world?
+++Interstellar
---Harry Potter
+++Alita
---Saw
Fifty Shades of Grey

qqqWhat is the best animated film?
+++Frozen
---Barbi
---Masha and the Bear

qhqI, Robot novel was written by Isaac Asimov
+++True
---False

qh4What is our genome made of
+++DNA and proteins
---Proteins
---Lipids
---Sugars

The processed file will be in the same folder, called prcssdQs.txt

List of changes:

1.2.4

Here we have a huge update! Now you only need to download start.sh, then in your terminal (where the start.sh) type chmod +x start.sh. This will let the terminal execute the script from the shell. Unfortunately, it doesn't work with mouse rightclick "Run as Program", but we work on that. Funnily, it was working before the last update to new 6th Linux kernel.

So what does the script do? It automatically updates > upgrades Linux > installs AMC (if needed), then creates a specified folder in the MC-Projects (AMC folder where all default projects are located), then downloads all needed files for a proper functioning of AMC within this project. This steps saves your time on: 1. Open AMC, 2. Press "+" to create new project, 3. Type down project name, 4. Choose mode of the project (now it is Empty by default as the script will produce structured folder contents anyway). Then it will ask you to navigate to the file (previously, you needed to name it specifically, now just double-click on the needed any-name-file with the questions), and based on the selection, the python script will be run to produce source LaTeX file.

As for now, negative marking is applied to all multiple correct choice answers, it is -0.25, and the bottom is set to 0. In the next revisions, we will allow you to choose how much you'd like to subtract for choosing incorrect options. Also we will implement "All or nothing" option. For now, it is as it is.

NB! setting up questions with multiple correct choices is on qmN pattern, where N is a number of correct answers for this particular question. This number defines points for choosing the option, e.g. qm3 makes all correct answers weigh 0.333333, and qm4 - 0.25, all incorrect choices will give -0.25

0.4.2

Fixed the problem of improper ending of the string in the line 353

NB! To insert an image you will need to paste "\includegraphics[width=8cm]{Image_name.png}" indicating a desired width of it on the page. Image should be in the same project folder on the step of compilation of the final LaTeX pdf of the exam

0.4.1

A lot of optimisations. Changed the SheBang

Fixed the bug with \insertgroup

0.4.0

Now the wrongCorrect.png image is downloaded in the end of the process

Also added the log file creation. You will need to delete it yourself, with time it is going to take some space on your computer

Reason for adding: in past hundreds of script executions it never failed. However, while it was a Midterm exam, the question FOR LOOP was terminated and never raised an issue. This led to incomplete test (about 20% of questions were omitted), while the rest of the script went well. I couldn't catch the error, and couldn't reproduce it, so HAIL THE LOG. If you have any issues like I mentioned, send the log to me along with your Qs.txt and prcssdQs.txt, my working email is mirakklys.gh@gmail.com

0.3.1

Minor changes

0.3.0

Now the script will process a single file Qs.txt

0.2.1

I jumped 0.2.0 because introduced the open answer questions in 0.1.1 version

So, finally it is one step closer to the final line :)

Minor code changes. Some windows machines when process txt in utf-8 encoding put some additional character at the beginning of the file. Since most people shouldn't use Win with AMC, I changed the code so it is suitable to Linux/Unix users :)

Now all files will have the same version numbering, so that it doesn't look stupidly messy

0.1.1 (9.1.1)

Added feature of column number choice. Now you will need to declare the feature with 'qhq' (default) and 'qh4' (4 columns)

Minor improvements to the code

Multiple correct horizontal answered questions are coming in the next release

Also released the open question script - open_Qs_for_AMC.py. You will need to declare the number of lines and maximum mark for the question, e.g. qoLM, where L is a number of lines and M is the max points received for the all correct answer. Partial point will be calculated depending on the number of lines and max point -

partialPoint = maxPoint / (numberOfLines - 1)

0.1.0 (9.1.0 for fixed 9-digit student ID number)

Major update with the addition of qmq and qhq support

Now you will need to put all multiple correct answer questions into Qms.txt file utf-8 encoded, and all horizontal questions into Qhs.txt file utf-8 encoded

Please note that the number of columns per question (qhq ones) is defaulted to 5. Adjust according to your needs by changing this parameter in final .tex source file of AMC. In next version i will adjust it so that person could regulate it by coding qhq tag, e.g. qh2q or qh4q

Also, now code is much cleaner. Fewer possible weak points

0.0.3

Added the check-ups for user input, which catch wrong type

Added the number of digits in student ID number, now it is forom 2 to 12

Removed some repeated codes, functions rule!

Forked the NU-specific python script (it has 9-digit student ID number by default)

Next update will be a major one, with the support for qmq and qhq!

0.0.2

Added date, number of copy and name managing feature. In the terminal it will ask sequentially: number of copies, the date of the exam, the name of the exam

Please, note that the date is stricted to 2099 Dec 31 max, 2020 Jan 01 min, any other dates will cause the stop and error message

Inside the prcssdQs.txt you will need to change/correct the section {etu}, and \clearpage <=> \AMCcleardoublepage depending on your needs

For now it prints answer sheet on a separate piece of paper, should be easier to handle scanning

There is still no python script to handle qmq. It is comming in the next version

0.0.1

The beginning of the project