multiconverter

Converter tool from LLM output to QTI itslearning test.

Installation

with pipx (recommended):

pipx install git+https://github.com/Hananja/multiconverter.git

Run GUI app

multiconverter-gui

Run commandline app

python -m multiconverter <output_file>.zip <input_file1>.xml [<input_file2>.xml ...]

(or with pipx just multiconverter ...)

Without any parameter, a help screen is displayed:

$ multiconverter
Usage: multiconverter <output_file>.zip <input_file1>.xml [<input_file2>.xml ...]
Process one or more AI generated XML files

Message to LLM:  ----------------------------------
Die Ausgabe soll ausschließlich im XML Format erfolgen und der folgenden XSD genügen: <?xml version='1.0' encoding='utf-8'?>
<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema" xmlns="https://github.com/Hananja/multiconverter" elementFormDefault="qualified" targetNamespace="https://github.com/Hananja/multiconverter"><!--The comments are also used to inform AI LLM. Note that the less than and greater than symbols must be represented with the entities &lt; and &gt;.--><xs:element name="questions"><xs:complexType><xs:sequence><xs:choice minOccurs="0" maxOccurs="unbounded"><xs:element ref="multiple-choice-question"/><xs:element ref="fill-in-question"/><xs:element ref="map-question"/></xs:choice></xs:sequence></xs:complexType></xs:element><xs:complexType name="option"><xs:simpleContent><xs:extension base="xs:string"><xs:attribute name="correct" use="required"><xs:simpleType><xs:restriction base="xs:string"><xs:enumeration value="true"/><xs:enumeration value="false"/></xs:restriction></xs:simpleType></xs:attribute></xs:extension></xs:simpleContent></xs:complexType><xs:element name="multiple-choice-question"><!--This question type presents a task in the <text> element. This can be one or more correct answers by selecting the correct ones from the <options>.--><xs:complexType><xs:sequence><xs:element name="text" type="xs:string"/><xs:element name="options"><xs:complexType><xs:sequence><xs:element name="option" maxOccurs="unbounded" type="option"/></xs:sequence></xs:complexType></xs:element></xs:sequence></xs:complexType></xs:element><xs:element name="fill-in-question"><!--This question type presents an optional description in the <text> element. The <fill-in-text> element contains a task text that may contain gaps. These are marked with <fill> and must be filled in during the test. The correct possible alternatives for the answer are contained in <alt> tags within a <fill> element.--><xs:complexType><xs:sequence><xs:element name="text" type="xs:string" minOccurs="0"/><xs:element name="fill-in-text"><xs:complexType mixed="true"><xs:sequence><xs:element name="fill" minOccurs="0" maxOccurs="unbounded"><xs:complexType><xs:sequence><xs:element name="alt" type="xs:string" maxOccurs="unbounded"/></xs:sequence></xs:complexType></xs:element></xs:sequence></xs:complexType></xs:element></xs:sequence></xs:complexType></xs:element><!--This question type first has an explanatory text contained in the <text> element. This is followed by a list of <mapping> elements in the <mappings> element, each of which contains related texts in the <left> and <right> elements. Example: <map-question><text>Match the pairs.</text><mappings><mapping><left>1+1</left><right>=2</right></mapping><mapping><left>2+2</left><right>=4</right></mapping></mappings></map-question>--><xs:element name="map-question"><xs:complexType><xs:sequence><!--Explanatory text--><xs:element name="text" type="xs:string"/><!--Container for one or more mappings--><xs:element name="mappings"><xs:complexType><xs:sequence><!--Each mapping has exactly one left and one right--><xs:element name="mapping" minOccurs="2" maxOccurs="unbounded"><xs:complexType><xs:sequence><xs:element name="left" type="xs:string"/><xs:element name="right" type="xs:string"/></xs:sequence></xs:complexType></xs:element></xs:sequence></xs:complexType></xs:element></xs:sequence></xs:complexType></xs:element></xs:schema>

Erstelle ein Beispieldokument mit je zwei Fragen pro Fragetyp

It is recommended to process the entire generated XML file first, because if the AI encounters any errors, an error message will be issued prompting the LLM to correct the output:

$ multiconverter output.zip test_raw.xml 
Ungültig: test_raw.xml
The document test_raw.xml is not correct.

Message to LLM:  ----------------------------------
Überleg noch einmal: Das Dokument entspricht nicht der XSD. Der Parser liefert den Fehler "XML-Syntax-Fehler: StartTag: invalid element name, line 141, column 37 (test_raw.xml, line 141)".

If communication with the LLM has worked without errors after several iterations, the XML files can be edited and adjusted (e.g., by deleting questions).

Supported question types

single and multiple choice

<multiple-choice-question xmlns="https://github.com/Hananja/multiconverter">
    <text>Was bedeutet WWW?</text>
    <options>
        <option correct="false">Willy will's wissen</option>
        <option correct="false">Wow, wie wichtig</option>
        <option correct="true">World Wide Web</option>
        <option correct="false">Wo? Wer? Wann?</option>
    </options>
</multiple-choice-question>

fill in text

<fill-in-question xmlns="https://github.com/Hananja/multiconverter">
    <text>Für jedes Ereignis, das weder sicher noch unmöglich ist, gilt: </text>
    <fill-in-text><fill><alt>0</alt></fill> &lt; P(E) &lt; <fill><alt>1</alt></fill> für
        die Wahrscheinlichkeit des Ereignisses</fill-in-text>
</fill-in-question>

map items

<map-question xmlns="https://github.com/Hananja/multiconverter">
    <text>Match the pairs.</text>
    <mappings>
        <mapping>
            <left>1 + 1</left>
            <right>2</right>
        </mapping>
        <mapping>
            <left>2 + 2</left>
            <right>4</right>
        </mapping>
        <mapping>
            <left>3 + 5</left>
            <right>8</right>
        </mapping>
    </mappings>
</map-question>