2016-03-14 05:03:52 +00:00
|
|
|
//: For convenience, some instructions will take literal arrays of characters
|
|
|
|
//: (text or strings).
|
2015-03-31 06:15:03 +00:00
|
|
|
//:
|
|
|
|
//: Instead of quotes, we'll use [] to delimit strings. That'll reduce the
|
2015-03-31 17:17:19 +00:00
|
|
|
//: need for escaping since we can support nested brackets. And we can also
|
2016-10-22 23:56:07 +00:00
|
|
|
//: imagine that 'recipe' might one day itself be defined in Mu, doing its own
|
2015-03-31 17:17:19 +00:00
|
|
|
//: parsing.
|
2015-03-31 06:15:03 +00:00
|
|
|
|
2015-04-24 07:28:24 +00:00
|
|
|
:(scenarios load)
|
2015-04-24 17:19:03 +00:00
|
|
|
:(scenario string_literal)
|
2016-03-08 09:30:14 +00:00
|
|
|
def main [
|
2016-08-31 16:53:11 +00:00
|
|
|
1:address:array:character <- copy [abc def]
|
2015-03-31 04:22:29 +00:00
|
|
|
]
|
2016-03-21 09:25:52 +00:00
|
|
|
+parse: ingredient: {"abc def": "literal-string"}
|
2015-03-31 04:22:29 +00:00
|
|
|
|
2015-04-24 17:19:03 +00:00
|
|
|
:(scenario string_literal_with_colons)
|
2016-03-08 09:30:14 +00:00
|
|
|
def main [
|
2015-04-29 05:42:54 +00:00
|
|
|
1:address:array:character <- copy [abc:def/ghi]
|
2015-04-15 17:27:16 +00:00
|
|
|
]
|
2016-03-21 09:25:52 +00:00
|
|
|
+parse: ingredient: {"abc:def/ghi": "literal-string"}
|
2015-04-15 17:27:16 +00:00
|
|
|
|
2015-03-31 04:22:29 +00:00
|
|
|
:(before "End Mu Types Initialization")
|
2015-11-06 19:06:58 +00:00
|
|
|
put(Type_ordinal, "literal-string", 0);
|
2015-03-31 04:22:29 +00:00
|
|
|
|
2015-10-27 18:31:05 +00:00
|
|
|
:(before "End next_word Special-cases")
|
2015-11-17 06:39:14 +00:00
|
|
|
if (in.peek() == '[') {
|
|
|
|
string result = slurp_quoted(in);
|
2015-12-03 01:11:58 +00:00
|
|
|
skip_whitespace_and_comments_but_not_newline(in);
|
2015-11-17 06:39:14 +00:00
|
|
|
return result;
|
|
|
|
}
|
2015-03-31 04:22:29 +00:00
|
|
|
|
|
|
|
:(code)
|
|
|
|
string slurp_quoted(istream& in) {
|
|
|
|
ostringstream out;
|
2015-11-17 09:21:00 +00:00
|
|
|
assert(has_data(in)); assert(in.peek() == '['); out << static_cast<char>(in.get()); // slurp the '['
|
2015-09-15 04:38:58 +00:00
|
|
|
if (is_code_string(in, out))
|
2015-06-14 23:11:47 +00:00
|
|
|
slurp_quoted_comment_aware(in, out);
|
|
|
|
else
|
|
|
|
slurp_quoted_comment_oblivious(in, out);
|
|
|
|
return out.str();
|
|
|
|
}
|
|
|
|
|
2016-08-14 04:21:08 +00:00
|
|
|
// A string is a code string (ignores comments when scanning for matching
|
|
|
|
// brackets) if it contains a newline at the start before any non-whitespace.
|
2015-10-27 21:49:36 +00:00
|
|
|
bool is_code_string(istream& in, ostream& out) {
|
2015-11-17 09:21:00 +00:00
|
|
|
while (has_data(in)) {
|
2015-06-14 23:11:47 +00:00
|
|
|
char c = in.get();
|
|
|
|
if (!isspace(c)) {
|
|
|
|
in.putback(c);
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
out << c;
|
|
|
|
if (c == '\n') {
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
|
|
|
// Read a regular string. Regular strings can only contain other regular
|
|
|
|
// strings.
|
2015-10-27 21:49:36 +00:00
|
|
|
void slurp_quoted_comment_oblivious(istream& in, ostream& out) {
|
2015-06-14 23:11:47 +00:00
|
|
|
int brace_depth = 1;
|
2015-11-17 09:21:00 +00:00
|
|
|
while (has_data(in)) {
|
2015-03-31 04:22:29 +00:00
|
|
|
char c = in.get();
|
2015-05-28 18:28:15 +00:00
|
|
|
if (c == '\\') {
|
2016-08-21 02:29:45 +00:00
|
|
|
slurp_one_past_backslashes(in, out);
|
2015-05-28 18:28:15 +00:00
|
|
|
continue;
|
|
|
|
}
|
2015-03-31 04:22:29 +00:00
|
|
|
out << c;
|
2015-05-28 20:31:20 +00:00
|
|
|
if (c == '[') ++brace_depth;
|
|
|
|
if (c == ']') --brace_depth;
|
|
|
|
if (brace_depth == 0) break;
|
2015-03-31 04:22:29 +00:00
|
|
|
}
|
2015-11-17 09:21:00 +00:00
|
|
|
if (!has_data(in) && brace_depth > 0) {
|
2016-02-26 21:04:55 +00:00
|
|
|
raise << "unbalanced '['\n" << end();
|
2015-06-14 23:11:47 +00:00
|
|
|
out.clear();
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
// Read a code string. Code strings can contain either code or regular strings.
|
2015-10-27 21:49:36 +00:00
|
|
|
void slurp_quoted_comment_aware(istream& in, ostream& out) {
|
2015-06-14 23:11:47 +00:00
|
|
|
char c;
|
|
|
|
while (in >> c) {
|
2015-06-19 20:37:11 +00:00
|
|
|
if (c == '\\') {
|
2016-08-21 02:29:45 +00:00
|
|
|
slurp_one_past_backslashes(in, out);
|
2015-06-19 20:37:11 +00:00
|
|
|
continue;
|
|
|
|
}
|
2015-06-14 23:11:47 +00:00
|
|
|
if (c == '#') {
|
|
|
|
out << c;
|
2015-11-17 09:21:00 +00:00
|
|
|
while (has_data(in) && in.peek() != '\n') out << static_cast<char>(in.get());
|
2015-06-14 23:11:47 +00:00
|
|
|
continue;
|
|
|
|
}
|
|
|
|
if (c == '[') {
|
|
|
|
in.putback(c);
|
|
|
|
// recurse
|
|
|
|
out << slurp_quoted(in);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
out << c;
|
2015-06-19 20:37:11 +00:00
|
|
|
if (c == ']') return;
|
2015-05-30 19:34:40 +00:00
|
|
|
}
|
2016-02-26 21:04:55 +00:00
|
|
|
raise << "unbalanced '['\n" << end();
|
2015-06-19 20:37:11 +00:00
|
|
|
out.clear();
|
2015-03-31 04:22:29 +00:00
|
|
|
}
|
2015-03-31 06:15:03 +00:00
|
|
|
|
2015-07-28 23:38:37 +00:00
|
|
|
:(after "Parsing reagent(string s)")
|
2016-09-12 01:07:29 +00:00
|
|
|
if (starts_with(s, "[")) {
|
2016-06-30 17:45:26 +00:00
|
|
|
if (*s.rbegin() != ']') return; // unbalanced bracket; handled elsewhere
|
2015-07-28 23:38:37 +00:00
|
|
|
name = s;
|
2016-05-26 01:56:37 +00:00
|
|
|
// delete [] delimiters
|
|
|
|
name.erase(0, 1);
|
|
|
|
strip_last(name);
|
2016-02-20 16:54:42 +00:00
|
|
|
type = new type_tree("literal-string", 0);
|
2015-07-28 23:38:37 +00:00
|
|
|
return;
|
|
|
|
}
|
2015-04-15 17:27:16 +00:00
|
|
|
|
2015-10-26 08:19:27 +00:00
|
|
|
//: Unlike other reagents, escape newlines in literal strings to make them
|
|
|
|
//: more friendly to trace().
|
2015-06-14 18:30:32 +00:00
|
|
|
|
2016-02-19 21:42:45 +00:00
|
|
|
:(after "string to_string(const reagent& r)")
|
2016-09-17 06:52:15 +00:00
|
|
|
if (is_literal_text(r))
|
2016-02-19 21:42:45 +00:00
|
|
|
return emit_literal_string(r.name);
|
2015-06-14 06:17:13 +00:00
|
|
|
|
|
|
|
:(code)
|
2016-09-17 06:52:15 +00:00
|
|
|
bool is_literal_text(const reagent& x) {
|
2016-02-21 04:05:52 +00:00
|
|
|
return x.type && x.type->name == "literal-string";
|
2015-08-02 05:16:09 +00:00
|
|
|
}
|
|
|
|
|
2015-06-14 06:17:13 +00:00
|
|
|
string emit_literal_string(string name) {
|
|
|
|
size_t pos = 0;
|
|
|
|
while (pos != string::npos)
|
|
|
|
pos = replace(name, "\n", "\\n", pos);
|
2016-03-21 09:25:52 +00:00
|
|
|
return "{\""+name+"\": \"literal-string\"}";
|
2015-06-14 06:17:13 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
size_t replace(string& str, const string& from, const string& to, size_t n) {
|
|
|
|
size_t result = str.find(from, n);
|
|
|
|
if (result != string::npos)
|
|
|
|
str.replace(result, from.length(), to);
|
|
|
|
return result;
|
|
|
|
}
|
|
|
|
|
2015-10-25 18:55:35 +00:00
|
|
|
void strip_last(string& s) {
|
|
|
|
if (!s.empty()) s.erase(SIZE(s)-1);
|
|
|
|
}
|
|
|
|
|
2016-08-21 02:29:45 +00:00
|
|
|
void slurp_one_past_backslashes(istream& in, ostream& out) {
|
|
|
|
// When you encounter a backslash, strip it out and pass through any
|
|
|
|
// following run of backslashes. If we 'escaped' a single following
|
|
|
|
// character, then the character '\' would be:
|
|
|
|
// '\\' escaped once
|
|
|
|
// '\\\\' escaped twice
|
|
|
|
// '\\\\\\\\' escaped thrice (8 backslashes)
|
|
|
|
// ..and so on. With our approach it'll be:
|
|
|
|
// '\\' escaped once
|
|
|
|
// '\\\' escaped twice
|
|
|
|
// '\\\\' escaped thrice
|
|
|
|
// This only works as long as backslashes aren't also overloaded to create
|
|
|
|
// special characters. So Mu doesn't follow C's approach of overloading
|
|
|
|
// backslashes both to escape quote characters and also as a notation for
|
|
|
|
// unprintable characters like '\n'.
|
|
|
|
while (has_data(in)) {
|
|
|
|
char c = in.get();
|
|
|
|
out << c;
|
|
|
|
if (c != '\\') break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2015-04-24 17:19:03 +00:00
|
|
|
:(scenario string_literal_nested)
|
2016-03-08 09:30:14 +00:00
|
|
|
def main [
|
2015-04-29 05:42:54 +00:00
|
|
|
1:address:array:character <- copy [abc [def]]
|
2015-03-31 17:17:19 +00:00
|
|
|
]
|
2016-03-21 09:25:52 +00:00
|
|
|
+parse: ingredient: {"abc [def]": "literal-string"}
|
2015-04-29 05:45:38 +00:00
|
|
|
|
2015-05-28 18:28:15 +00:00
|
|
|
:(scenario string_literal_escaped)
|
2016-03-08 09:30:14 +00:00
|
|
|
def main [
|
2015-05-28 18:28:15 +00:00
|
|
|
1:address:array:character <- copy [abc \[def]
|
|
|
|
]
|
2016-03-21 09:25:52 +00:00
|
|
|
+parse: ingredient: {"abc [def": "literal-string"}
|
2015-05-28 18:28:15 +00:00
|
|
|
|
2016-08-21 02:29:45 +00:00
|
|
|
:(scenario string_literal_escaped_twice)
|
2016-03-08 09:30:14 +00:00
|
|
|
def main [
|
2015-06-19 20:37:11 +00:00
|
|
|
1:address:array:character <- copy [
|
2016-08-21 02:29:45 +00:00
|
|
|
abc \\[def]
|
2015-06-19 20:37:11 +00:00
|
|
|
]
|
2016-03-21 09:25:52 +00:00
|
|
|
+parse: ingredient: {"\nabc \[def": "literal-string"}
|
2015-06-19 20:37:11 +00:00
|
|
|
|
2015-04-29 05:45:38 +00:00
|
|
|
:(scenario string_literal_and_comment)
|
2016-03-08 09:30:14 +00:00
|
|
|
def main [
|
2015-04-29 05:45:38 +00:00
|
|
|
1:address:array:character <- copy [abc] # comment
|
|
|
|
]
|
2015-10-29 19:15:03 +00:00
|
|
|
+parse: --- defining main
|
2015-04-30 04:49:09 +00:00
|
|
|
+parse: instruction: copy
|
2015-11-01 04:56:17 +00:00
|
|
|
+parse: number of ingredients: 1
|
2016-03-21 09:25:52 +00:00
|
|
|
+parse: ingredient: {"abc": "literal-string"}
|
|
|
|
+parse: product: {1: ("address" "array" "character")}
|
2015-06-14 17:07:00 +00:00
|
|
|
|
|
|
|
:(scenario string_literal_escapes_newlines_in_trace)
|
2016-03-08 09:30:14 +00:00
|
|
|
def main [
|
2015-06-14 17:07:00 +00:00
|
|
|
copy [abc
|
|
|
|
def]
|
|
|
|
]
|
2016-03-21 09:25:52 +00:00
|
|
|
+parse: ingredient: {"abc\ndef": "literal-string"}
|
2015-06-14 23:11:47 +00:00
|
|
|
|
|
|
|
:(scenario string_literal_can_skip_past_comments)
|
2016-03-08 09:30:14 +00:00
|
|
|
def main [
|
2015-06-14 23:11:47 +00:00
|
|
|
copy [
|
|
|
|
# ']' inside comment
|
|
|
|
bar
|
|
|
|
]
|
|
|
|
]
|
2016-03-21 09:25:52 +00:00
|
|
|
+parse: ingredient: {"\n # ']' inside comment\n bar\n ": "literal-string"}
|
2015-06-14 23:11:47 +00:00
|
|
|
|
|
|
|
:(scenario string_literal_empty)
|
2016-03-08 09:30:14 +00:00
|
|
|
def main [
|
2015-06-14 23:11:47 +00:00
|
|
|
copy []
|
|
|
|
]
|
2016-03-21 09:25:52 +00:00
|
|
|
+parse: ingredient: {"": "literal-string"}
|
2016-06-30 17:45:26 +00:00
|
|
|
|
|
|
|
:(scenario multiple_unfinished_recipes)
|
|
|
|
% Hide_errors = true;
|
2016-09-15 20:23:00 +00:00
|
|
|
def f1 [
|
|
|
|
def f2 [
|
2016-06-30 17:45:26 +00:00
|
|
|
+error: unbalanced '['
|