Skip to content

BUGS.md

Latest commit

 

History

History
103 lines (76 loc) · 2.27 KB

BUGS.md

File metadata and controls

103 lines (76 loc) · 2.27 KB

The Robotics Language BUGS

RoL parser

Strings are not HTML safe

The code wii return an error:

print('x > 1')

Issue with parsing real numbers in scientific notation. Conflicting with negation operator:

Parsing: -1.3e-44 results in:

<negate p="32">
      <real p="32">1.3e-44</real>
</negate>

Parsing: - 1.3 e - 44 results in:

<negate p="29">
  <real p="29">1.3</real>
</negate>
<minus p="36">
  <variable p="31">e</variable>
  <integer p="36">44</integer>
</minus>

In the command line if two files are repeated a message is displayed

rol helloworld.rol helloworld.rol
WARNING: the following files are disregarded:
/Users/glopes/Projects/RoboticsLanguage/Examples/helloworld.rol

if first element of the custom vector is '' it does not work

In the custom language operator definition the first parameter cannot be ''. This works:

'functionDefinition': {
    'input': {
        'RoL': {
            'custom': ['define', ':','->',',','->', '']
        }
    },
 ...

An so a function can be defined in RoL by:

define f:R -> R,
         x -> x+1

This does not work:

'functionDefinition': {
    'input': {
        'RoL': {
            'custom': ['', ':','->',',','->', '']
        }
    },
 ...
f:R -> R,
  x -> x+1

Performance issues

caching

The function cache can store static elements in a file saved at ~/.ros/cache. If the cache exists it loads the data from a file. Otherwise, it uses the function functionThatCreatesData to generate the data online:

data =  cache('name', functionThatCreatesData )

RoL parser

the RoL parser currently works by generating the composition of strings of text (see functions xml and xmlInfix) that are eventually parsed into XML. This makes the implementation simple but my incur some performance issues when mini languages are used. Mini languages return XML objects that are converted to text to integrate into the language. When this parser returns the result is converts everything into an XML object, which means that the mini language code is converted to xml text and then back to XML objects, thus waisting CPU time. If performance become critical in the future this should be addressed.