tobscore / sysprog Goto Github PK

Sometimes the EOF of a file cannot be read correctly, but this bug only occurs sometimes. An example build is here.

In der Datei testProgramm.txt befindet sich in Zeile 5 eine Leerzeile, also eine Zeile ohne Inhalt.
Der Scanner erkennt diese Leerzeile nicht, der Zeilenzähler wird also nicht hochgezählt. Somit ist die Position der nachfolgenden Tokens nicht mehr korrekt.

There needs to be better error handling. If the program tries to read some unknown character, an error token with the unknown character needs to created.

It's not possible to read in the bible or other large files. Somehow the program ends up in an infinite loop.

The reason for this is unknown. This case will be tested against, so this needs to be fixed.

Check out Commit 59506b0 to see a test case, that would fail.

      Expected: "Kommentar"
To be equal to: static_cast<IdentifierToken*>(token)->getLexem()
      Which is: "11111aaaaaKommentar"

The Dot-Character "." is not recognized as an erroneous character.

Our scanner still recognizes large integers and tries to convert the string representation of those integers to actual integers (or in our case long long int).

An error should be thrown, when trying to read in numbers which are to big to be cast to an integer.

One of the test cases checks how our program handles really big identifiers. In order to enhance the memory usage of the program and avoid possible bugs, the buffer size needs to be allocated dynamically, meaning that its size needs to be calculated in accordance to the identifier in the token.

Currently the google test framework has been copied in the lib folder. In order to have the latest and greatest Google Test Version, this could be optimized by using git submodules.