SCC is C++ snippets evaluator (REPL without L), with some AWK features. Implemented as GCC wrapper.
For all examples to work use commit tagged as v0.1; gcc-4.6 or newer and Boost . This page describes v0.1. If used as simple REPL, older gcc (tested with gcc-4.4) will work too and Boost is not needed.
Tip of master branch (do not use) is about x100 faster for AWK ops, but only gcc-4.7 is supported and Boost is not used.
C++ expression evaluated and printed:
scc 2+2 4
If snippet have Bash special characters, it must be in single or double quotes. If expresion starts with minus, end-of-options indicator "--" must be used:
scc '"hello world"' hello world scc 'sin(0.5)' 0.479426 scc -- -1-INT_MIN 2147483647
A multi-statement with explicit printing.
Note trailing semicolon - meaning do not send last expression
to std::cout
scc 'double x=0.5; cout << sin(x) << endl;' 0.479426
SCC includes simple.h which has a lot of shortcuts (à la Google CodeJam).
Same as previous example, but with SCC’s shortcuts (
bar-print and pre-declared vars)
scc 'x=0.5; __ sin(x);' 0.479426
All STL containers can be sent to std::cout. Typedef vint is std::vector<int>:
scc 'vint v{1,2,3}; v'
{1, 2, 3}
scc 'tuple<const char*,int> v{"aaa",2}; v'
⟨aaa, 2⟩
scc 'tuple<const char*,int,tuple<int,float>> v{"aaa",2,make_tuple(3,4.5f)}; v'
⟨aaa, 2, ⟨3, 4.5⟩⟩
scc 'map<int,int> v{{1,1},{2,2}}; v'
{⟨1,1⟩, ⟨2,2⟩}
scc 'set<int> v{2,2}; v'
{2}
scc 'vector<vint> v{{1,1}, {2,2}}; v'
{{1, 1}, {2, 2}}
scc 'int v[3]{1,2,3}; v'
{1, 2, 3}
scc 'vector<map<str,int>> v{{{"aa",1}, {"bb",2}}}; v'
{{⟨aa,1⟩, ⟨bb,2⟩}}
Square an array. C++ 11 is on by default. gcc-4.6.0 or newer is needed for this example
scc 'vint V{1,2,3}; for(auto v:V) v*=v; V'
{1, 4, 9}
Words frequencies (typedef str is std::string):
echo aa bb aa | scc 'str w; map<str,int> M; while(cin>>w) M[w]++; M'
{⟨aa,2⟩, ⟨bb,1⟩}
Sum-up DF(1) used-disk-space column:
df | scc 'WRL n+=F2; n' 31399199
Explanation:
Shortcut WRL expands to while(read_line()).
Function read_line(), reads input line, splits it into fields. One input line is an array of fields F.
To access 3nd field use F(2) or F2.
Class F derived from std::deque<field>.
Type field (derived from std::string) can be directly converted to/from numeric types.
The F member function operator() is safe version of operator[].
Variables i,j,k,n are
pre-declared as long; variables x,y,z are pre-declared as double;
and s - pre-declared as std::string. Numeric types initialized to 0.
You don’t need to remember these, you can re-declare/re-define these freely.
Prepend line number to every line.
Option -p is similar to PERL’s. It does while(read_line()), evaluates snippet and prints F.
Global variable NR - number of record:
echo -e 'aaa\nbbb' | scc -p 'F.push_front(NR);' 0 aaa 1 bbb
Another way for doing the same:
echo -e 'aaa\nbbb' | scc -p NR 0 aaa 1 bbb
Make comma separated fields out of colon separated. Option -o sets OFS
(output field separator), -i - set IFS Snippet is empty in this example.
echo 1:2:3 | scc -i: -o, -p 1,2,3
Read CSV file with quoted fields (spaces, commas and escaped quotes in field are allowed):
Option -I, like -i, also sets IFS, but now you can use regex; CSV - predefined regex constant.
How to make your own regexes for IFS - look in scc.h
echo '"aa", "bb\"-, bb", "cc"' | scc -I CSV -o '"\n"' -p aa bb\"-, bb cc
Replace "-", "none" and empty fields with "n/a" in 2nd column using boost::regex.
Boost use in SCC is optional, it will be used if /usr/include/boost exist.
RR is shortcut for boost::regex_replace; shortcut R is boost::regex
echo \
'1 -
2
3 none
4 abc' | scc -p 'F1=RR(F1,R("^(none|-)?$"),"n/a");'
1 n/a
2 n/a
3 n/a
4 abc