parsing numbers with spaces

[julochrobak]

Here is an example of input with three different values which contain only digits and spaces:

[ a.num + 0 | a <- [ {num: "123 456"}, {num: "123456"}, {num: " 123456"} ]]

The result of this execution is:

0
123456
0

If there is a tab-delimited file which contains this kind of strings there is no way for users to work with them as with numbers. The suggested solutions are:

• implement function which can replace characters. It would allow users explicitly get rid of the spaces and therefore get the strings converted into numbers. The disadvantage is that a user would have to always call this function when working with any numeric attributes, for example:

  [ replace(a.num, " ", "") + 0 | a <- ... ]

• try to apply this replace function always before converting strings into numbers. The problem is, that if a value looks like "123 456" this can either mean it's a number 123456 or two numbers 123 and 456 separated by a space (e.g. coordinates). Applying the replace automatically would convert into a value which has different meaning.

[ostap]

I've added a replace function which takes three arguments (as in the example above) and replaces all occurrences. I don't see any disadvantages with this approach since the values with spaces are not necessarily numeric, e.g. a file might have dates in the following format "17 08 2013".