Merge pull request #48 from harp-lab/string_spaces

String spaces

backend/src/ast.h

@@ -27,13 +27,16 @@ enum class slogc_token_type { none,
                               right_paren,
                               symbol,
                               boolean,
-                              number };
+                              number,
+                              string };
 enum class slogc_char_type { left_paren,
                              right_paren,
                              space,
+                             double_quote,
                              other };
 enum class parse_state { init,
-                         symbol };
+                         symbol,
+                         string };
 enum class slogc_relation_version { DELTA,
                                     FULL };
 

backend/src/slog.cpp

@@ -43,6 +43,8 @@ slogc_char_type get_char_type(char ch) {
         return slogc_char_type::left_paren;
     case ')':
         return slogc_char_type::right_paren;
+    case '"':
+        return slogc_char_type::double_quote;
     }
     if (isspace(static_cast<unsigned char>(ch)))
         return slogc_char_type::space;
@@ -79,6 +81,16 @@ bool get_token(std::istream &in, slogc_token &tok) {
                 break;
             }
             str += ch;
+        } else if (state == parse_state::string) {
+            if (ctype == slogc_char_type::double_quote) {
+                str += ch;
+                break;
+            }
+            str += ch;
+        } else if (ctype == slogc_char_type::double_quote) {
+            state = parse_state::string;
+            type = slogc_token_type::string;
+            str = ch;
         } else if (ctype == slogc_char_type::other) {
             state = parse_state::symbol;
             type = slogc_token_type::symbol;
@@ -96,6 +108,8 @@ bool get_token(std::istream &in, slogc_token &tok) {
     if (type == slogc_token_type::symbol) {
         if (!parse_number(str, tok))
             tok.data = str;
+    } else if (type == slogc_token_type::string){
+        tok.data = str;
     }
     return true;
 }

compiler/src/interpreter.rkt

@@ -537,7 +537,7 @@
       (match key
         [`(interned string ,n)
          (set! x (max x n))
-         (displayln (format "~a\t~a" n (hash-ref intern-map key)) out)]
+         (displayln (format "~a\t\"~a\"" n (hash-ref intern-map key)) out)]
         [_ (void)]))
     (close-output-port out)
     x)

doc/repl_and_runslog.md

@@ -61,4 +61,3 @@ Options:
 (print-welcome 1)
 ```
 - Here, This is similar to a printf-style debugging, and can be useful in undestanding what each rule is doing in a program.
-- Note, Here there are no spaces in the strings, because the current parser does not support spaces in strings.

slog/common/client.py

@@ -324,7 +324,7 @@ def _update_intern_strings(self, db_id):
                     if s_line.strip() == '':
                         continue
                     sv = s_line.split('\t')[1]
-                    sw_without_quote = f'"{sv.strip()}"'
+                    sw_without_quote = f'{sv.strip()}'
                     if string_hash(sw_without_quote) in self.intern_string_dict.keys():
                         if sv.strip() != self.intern_string_dict[string_hash(sw_without_quote)]:
                             print(f"Hash collision {sv.strip()} {self.intern_string_dict[string_hash(sw_without_quote)]}")

slog/tests/testcase/stringtest/ground_truth

@@ -1 +1,2 @@
 foobar,2,2
+just_str,1,5

slog/tests/testcase/stringtest/input/strs.csv

@@ -0,0 +1,2 @@
+6	"fact file string"
+5	String with spaces from fact file without quotes

slog/tests/testcase/stringtest/stringtest.slog

@@ -6,3 +6,11 @@
 (foo 2 "bar")
 
 [(foobar x y) <-- (foo x y) (bar y)]
+
+; ------------------------------------
+; Strings with spaces
+(strs 1 "well string")
+(strs 2 "well another string)")
+(strs 3 "yet another ! ? string")
+
+[(strs _ str_val) --> (just_str str_val)]