|
|
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN" "http://www.w3.org/TR/html4/strict.dtd">
<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
<title>Mu - 013literal_string.cc</title>
<meta name="Generator" content="Vim/7.4">
<meta name="plugin-version" content="vim7.4_v1">
<meta name="syntax" content="cpp">
<meta name="settings" content="use_css,pre_wrap,no_foldcolumn,expand_tabs,prevent_copy=">
<meta name="colorscheme" content="minimal">
<style type="text/css">
<!--
pre { white-space: pre-wrap; font-family: monospace; color: #eeeeee; background-color: #080808; }
body { font-family: monospace; color: #eeeeee; background-color: #080808; }
* { font-size: 1.05em; }
.Identifier { color: #804000; }
.cSpecial { color: #008000; }
.Comment { color: #9090ff; }
.Delimiter { color: #a04060; }
.Special { color: #ff6060; }
.CommentedCode { color: #6c6c6c; }
.Constant { color: #00a0a0; }
.traceContains { color: #008000; }
-->
</style>
<script type='text/javascript'>
<!--
-->
</script>
</head>
<body>
<pre id='vimCodeElement'>
<span class="Comment">//: For convenience, some instructions will take literal arrays of characters (strings).</span>
<span class="Comment">//:</span>
<span class="Comment">//: Instead of quotes, we'll use [] to delimit strings. That'll reduce the</span>
<span class="Comment">//: need for escaping since we can support nested brackets. And we can also</span>
<span class="Comment">//: imagine that 'recipe' might one day itself be defined in mu, doing its own</span>
<span class="Comment">//: parsing.</span>
<span class="Delimiter">:(scenarios load)</span>
<span class="Delimiter">:(scenario string_literal)</span>
recipe main [
<span class="Constant">1</span>:address:array:character<span class="Special"> <- </span>copy [abc def] <span class="Comment"># copy can't really take a string</span>
]
<span class="traceContains">+parse: ingredient: {name: "abc def", properties: [_: "literal-string"]}</span>
<span class="Delimiter">:(scenario string_literal_with_colons)</span>
recipe main [
<span class="Constant">1</span>:address:array:character<span class="Special"> <- </span>copy [abc:def/ghi]
]
<span class="traceContains">+parse: ingredient: {name: "abc:def/ghi", properties: [_: "literal-string"]}</span>
<span class="Delimiter">:(before "End Mu Types Initialization")</span>
Type_ordinal[<span class="Constant">"literal-string"</span>] = <span class="Constant">0</span><span class="Delimiter">;</span>
<span class="Delimiter">:(after "string next_word(istream& in)")</span>
if <span class="Delimiter">(</span>in<span class="Delimiter">.</span>peek<span class="Delimiter">()</span> == <span class="Constant">'['</span><span class="Delimiter">)</span> <span class="Delimiter">{</span>
string result = slurp_quoted<span class="Delimiter">(</span>in<span class="Delimiter">);</span>
skip_whitespace<span class="Delimiter">(</span>in<span class="Delimiter">);</span>
skip_comment<span class="Delimiter">(</span>in<span class="Delimiter">);</span>
<span class="CommentedCode">//? cerr << '^' << result << "$\n"; //? 1</span>
<span class="Identifier">return</span> result<span class="Delimiter">;</span>
<span class="Delimiter">}</span>
<span class="Delimiter">:(code)</span>
string slurp_quoted<span class="Delimiter">(</span>istream& in<span class="Delimiter">)</span> <span class="Delimiter">{</span>
ostringstream out<span class="Delimiter">;</span>
assert<span class="Delimiter">(</span>!in<span class="Delimiter">.</span>eof<span class="Delimiter">());</span> assert<span class="Delimiter">(</span>in<span class="Delimiter">.</span>peek<span class="Delimiter">()</span> == <span class="Constant">'['</span><span class="Delimiter">);</span> out << static_cast<char><span class="Delimiter">(</span>in<span class="Delimiter">.</span>get<span class="Delimiter">());</span> <span class="Comment">// slurp the '['</span>
if <span class="Delimiter">(</span>code_string<span class="Delimiter">(</span>in<span class="Delimiter">,</span> out<span class="Delimiter">))</span>
slurp_quoted_comment_aware<span class="Delimiter">(</span>in<span class="Delimiter">,</span> out<span class="Delimiter">);</span>
else
slurp_quoted_comment_oblivious<span class="Delimiter">(</span>in<span class="Delimiter">,</span> out<span class="Delimiter">);</span>
<span class="Identifier">return</span> out<span class="Delimiter">.</span>str<span class="Delimiter">();</span>
<span class="Delimiter">}</span>
<span class="Comment">// A string is a code string if it contains a newline before any non-whitespace</span>
<span class="Comment">// todo: support comments before the newline. But that gets messy.</span>
bool code_string<span class="Delimiter">(</span>istream& in<span class="Delimiter">,</span> ostringstream& out<span class="Delimiter">)</span> <span class="Delimiter">{</span>
while <span class="Delimiter">(</span>!in<span class="Delimiter">.</span>eof<span class="Delimiter">())</span> <span class="Delimiter">{</span>
char c = in<span class="Delimiter">.</span>get<span class="Delimiter">();</span>
if <span class="Delimiter">(</span>!isspace<span class="Delimiter">(</span>c<span class="Delimiter">))</span> <span class="Delimiter">{</span>
in<span class="Delimiter">.</span>putback<span class="Delimiter">(</span>c<span class="Delimiter">);</span>
<span class="CommentedCode">//? cerr << "code_string: " << out.str() << '\n'; //? 1</span>
<span class="Identifier">return</span> <span class="Constant">false</span><span class="Delimiter">;</span>
<span class="Delimiter">}</span>
out << c<span class="Delimiter">;</span>
if <span class="Delimiter">(</span>c == <span class="cSpecial">'\n'</span><span class="Delimiter">)</span> <span class="Delimiter">{</span>
<span class="CommentedCode">//? cerr << "code_string: " << out.str() << '\n'; //? 1</span>
<span class="Identifier">return</span> <span class="Constant">true</span><span class="Delimiter">;</span>
<span class="Delimiter">}</span>
<span class="Delimiter">}</span>
<span class="Identifier">return</span> <span class="Constant">false</span><span class="Delimiter">;</span>
<span class="Delimiter">}</span>
<span class="Comment">// Read a regular string. Regular strings can only contain other regular</span>
<span class="Comment">// strings.</span>
void slurp_quoted_comment_oblivious<span class="Delimiter">(</span>istream& in<span class="Delimiter">,</span> ostringstream& out<span class="Delimiter">)</span> <span class="Delimiter">{</span>
<span class="CommentedCode">//? cerr << "comment oblivious\n"; //? 1</span>
int brace_depth = <span class="Constant">1</span><span class="Delimiter">;</span>
while <span class="Delimiter">(</span>!in<span class="Delimiter">.</span>eof<span class="Delimiter">())</span> <span class="Delimiter">{</span>
char c = in<span class="Delimiter">.</span>get<span class="Delimiter">();</span>
<span class="CommentedCode">//? cerr << '%' << (int)c << ' ' << brace_depth << ": " << out.str() << "%$\n"; //? 1</span>
<span class="CommentedCode">//? cout << (int)c << ": " << brace_depth << '\n'; //? 2</span>
if <span class="Delimiter">(</span>c == <span class="cSpecial">'\\'</span><span class="Delimiter">)</span> <span class="Delimiter">{</span>
out << static_cast<char><span class="Delimiter">(</span>in<span class="Delimiter">.</span>get<span class="Delimiter">());</span>
<span class="Identifier">continue</span><span class="Delimiter">;</span>
<span class="Delimiter">}</span>
out << c<span class="Delimiter">;</span>
<span class="CommentedCode">//? cout << out.str() << "$\n"; //? 1</span>
if <span class="Delimiter">(</span>c == <span class="Constant">'['</span><span class="Delimiter">)</span> ++brace_depth<span class="Delimiter">;</span>
if <span class="Delimiter">(</span>c == <span class="Constant">']'</span><span class="Delimiter">)</span> --brace_depth<span class="Delimiter">;</span>
if <span class="Delimiter">(</span>brace_depth == <span class="Constant">0</span><span class="Delimiter">)</span> <span class="Identifier">break</span><span class="Delimiter">;</span>
<span class="Delimiter">}</span>
if <span class="Delimiter">(</span>in<span class="Delimiter">.</span>eof<span class="Delimiter">()</span> && brace_depth > <span class="Constant">0</span><span class="Delimiter">)</span> <span class="Delimiter">{</span>
raise << <span class="Constant">"unbalanced '['</span><span class="cSpecial">\n</span><span class="Constant">"</span> << end<span class="Delimiter">();</span>
out<span class="Delimiter">.</span>clear<span class="Delimiter">();</span>
<span class="Delimiter">}</span>
<span class="Delimiter">}</span>
<span class="Comment">// Read a code string. Code strings can contain either code or regular strings.</span>
void slurp_quoted_comment_aware<span class="Delimiter">(</span>istream& in<span class="Delimiter">,</span> ostringstream& out<span class="Delimiter">)</span> <span class="Delimiter">{</span>
<span class="CommentedCode">//? cerr << "comment aware\n"; //? 1</span>
char c<span class="Delimiter">;</span>
while <span class="Delimiter">(</span>in >> c<span class="Delimiter">)</span> <span class="Delimiter">{</span>
<span class="CommentedCode">//? cerr << '^' << (int)c << ": " << out.str() << "$\n"; //? 1</span>
if <span class="Delimiter">(</span>c == <span class="cSpecial">'\\'</span><span class="Delimiter">)</span> <span class="Delimiter">{</span>
out << static_cast<char><span class="Delimiter">(</span>in<span class="Delimiter">.</span>get<span class="Delimiter">());</span>
<span class="Identifier">continue</span><span class="Delimiter">;</span>
<span class="Delimiter">}</span>
if <span class="Delimiter">(</span>c == <span class="Constant">'#'</span><span class="Delimiter">)</span> <span class="Delimiter">{</span>
out << c<span class="Delimiter">;</span>
while <span class="Delimiter">(</span>!in<span class="Delimiter">.</span>eof<span class="Delimiter">()</span> && in<span class="Delimiter">.</span>peek<span class="Delimiter">()</span> != <span class="cSpecial">'\n'</span><span class="Delimiter">)</span> out << static_cast<char><span class="Delimiter">(</span>in<span class="Delimiter">.</span>get<span class="Delimiter">());</span>
<span class="Identifier">continue</span><span class="Delimiter">;</span>
<span class="Delimiter">}</span>
if <span class="Delimiter">(</span>c == <span class="Constant">'['</span><span class="Delimiter">)</span> <span class="Delimiter">{</span>
in<span class="Delimiter">.</span>putback<span class="Delimiter">(</span>c<span class="Delimiter">);</span>
<span class="Comment">// recurse</span>
out << slurp_quoted<span class="Delimiter">(</span>in<span class="Delimiter">);</span>
<span class="Identifier">continue</span><span class="Delimiter">;</span>
<span class="Delimiter">}</span>
out << c<span class="Delimiter">;</span>
if <span class="Delimiter">(</span>c == <span class="Constant">']'</span><span class="Delimiter">)</span> <span class="Identifier">return</span><span class="Delimiter">;</span>
<span class="Delimiter">}</span>
raise << <span class="Constant">"unbalanced '['</span><span class="cSpecial">\n</span><span class="Constant">"</span> << end<span class="Delimiter">();</span>
out<span class="Delimiter">.</span>clear<span class="Delimiter">();</span>
<span class="Delimiter">}</span>
<span class="Delimiter">:(after "Parsing reagent(string s)")</span>
if <span class="Delimiter">(</span>s<span class="Delimiter">.</span>at<span class="Delimiter">(</span><span class="Constant">0</span><span class="Delimiter">)</span> == <span class="Constant">'['</span><span class="Delimiter">)</span> <span class="Delimiter">{</span>
assert<span class="Delimiter">(</span>*s<span class="Delimiter">.</span>rbegin<span class="Delimiter">()</span> == <span class="Constant">']'</span><span class="Delimiter">);</span>
<span class="Comment">// delete [] delimiters</span>
s<span class="Delimiter">.</span>erase<span class="Delimiter">(</span><span class="Constant">0</span><span class="Delimiter">,</span> <span class="Constant">1</span><span class="Delimiter">);</span>
s<span class="Delimiter">.</span>erase<span class="Delimiter">(</span>SIZE<span class="Delimiter">(</span>s<span class="Delimiter">)</span>-<span class="Constant">1</span><span class="Delimiter">);</span>
name = s<span class="Delimiter">;</span>
types<span class="Delimiter">.</span>push_back<span class="Delimiter">(</span><span class="Constant">0</span><span class="Delimiter">);</span>
properties<span class="Delimiter">.</span>push_back<span class="Delimiter">(</span>pair<string<span class="Delimiter">,</span> vector<string> ><span class="Delimiter">(</span>name<span class="Delimiter">,</span> vector<string><span class="Delimiter">()));</span>
properties<span class="Delimiter">.</span>back<span class="Delimiter">().</span>second<span class="Delimiter">.</span>push_back<span class="Delimiter">(</span><span class="Constant">"literal-string"</span><span class="Delimiter">);</span>
<span class="Identifier">return</span><span class="Delimiter">;</span>
<span class="Delimiter">}</span>
<span class="Comment">//: Two tweaks to printing literal strings compared to other reagents:</span>
<span class="Comment">//: a) Don't print the string twice in the representation, just put '_' in</span>
<span class="Comment">//: the property list.</span>
<span class="Comment">//: b) Escape newlines in the string to make it more friendly to trace().</span>
<span class="Delimiter">:(after "string reagent::to_string()")</span>
if <span class="Delimiter">(</span>is_literal_string<span class="Delimiter">(</span>*this<span class="Delimiter">))</span>
<span class="Identifier">return</span> emit_literal_string<span class="Delimiter">(</span>name<span class="Delimiter">);</span>
<span class="Delimiter">:(code)</span>
bool is_literal_string<span class="Delimiter">(</span>const reagent& x<span class="Delimiter">)</span> <span class="Delimiter">{</span>
<span class="Identifier">return</span> !x<span class="Delimiter">.</span>properties<span class="Delimiter">.</span>at<span class="Delimiter">(</span><span class="Constant">0</span><span class="Delimiter">).</span>second<span class="Delimiter">.</span>empty<span class="Delimiter">()</span> && x<span class="Delimiter">.</span>properties<span class="Delimiter">.</span>at<span class="Delimiter">(</span><span class="Constant">0</span><span class="Delimiter">).</span>second<span class="Delimiter">.</span>at<span class="Delimiter">(</span><span class="Constant">0</span><span class="Delimiter">)</span> == <span class="Constant">"literal-string"</span><span class="Delimiter">;</span>
<span class="Delimiter">}</span>
string emit_literal_string<span class="Delimiter">(</span>string name<span class="Delimiter">)</span> <span class="Delimiter">{</span>
size_t pos = <span class="Constant">0</span><span class="Delimiter">;</span>
while <span class="Delimiter">(</span>pos != string::npos<span class="Delimiter">)</span>
pos = replace<span class="Delimiter">(</span>name<span class="Delimiter">,</span> <span class="Constant">"</span><span class="cSpecial">\n</span><span class="Constant">"</span><span class="Delimiter">,</span> <span class="Constant">"</span><span class="cSpecial">\\</span><span class="Constant">n"</span><span class="Delimiter">,</span> pos<span class="Delimiter">);</span>
<span class="Identifier">return</span> <span class="Constant">"{name: </span><span class="cSpecial">\"</span><span class="Constant">"</span>+name+<span class="Constant">"</span><span class="cSpecial">\"</span><span class="Constant">, properties: [_: </span><span class="cSpecial">\"</span><span class="Constant">literal-string</span><span class="cSpecial">\"</span><span class="Constant">]}"</span><span class="Delimiter">;</span>
<span class="Delimiter">}</span>
size_t replace<span class="Delimiter">(</span>string& str<span class="Delimiter">,</span> const string& from<span class="Delimiter">,</span> const string& to<span class="Delimiter">,</span> size_t n<span class="Delimiter">)</span> <span class="Delimiter">{</span>
size_t result = str<span class="Delimiter">.</span>find<span class="Delimiter">(</span>from<span class="Delimiter">,</span> n<span class="Delimiter">);</span>
if <span class="Delimiter">(</span>result != string::npos<span class="Delimiter">)</span>
str<span class="Delimiter">.</span>replace<span class="Delimiter">(</span>result<span class="Delimiter">,</span> from<span class="Delimiter">.</span>length<span class="Delimiter">(),</span> to<span class="Delimiter">);</span>
<span class="Identifier">return</span> result<span class="Delimiter">;</span>
<span class="Delimiter">}</span>
<span class="Delimiter">:(scenario string_literal_nested)</span>
recipe main [
<span class="Constant">1</span>:address:array:character<span class="Special"> <- </span>copy [abc [def]]
]
<span class="traceContains">+parse: ingredient: {name: "abc [def]", properties: [_: "literal-string"]}</span>
<span class="Delimiter">:(scenario string_literal_escaped)</span>
recipe main [
<span class="Constant">1</span>:address:array:character<span class="Special"> <- </span>copy [abc \[def]
]
<span class="traceContains">+parse: ingredient: {name: "abc [def", properties: [_: "literal-string"]}</span>
<span class="Delimiter">:(scenario string_literal_escaped_comment_aware)</span>
recipe main [
<span class="Constant">1</span>:address:array:character<span class="Special"> <- </span>copy [
abc \\\[def]
]
<span class="traceContains">+parse: ingredient: {name: "\nabc \[def", properties: [_: "literal-string"]}</span>
<span class="Delimiter">:(scenario string_literal_and_comment)</span>
recipe main [
<span class="Constant">1</span>:address:array:character<span class="Special"> <- </span>copy [abc] <span class="Comment"># comment</span>
]
<span class="traceContains">+parse: instruction: copy</span>
<span class="traceContains">+parse: ingredient: {name: "abc", properties: [_: "literal-string"]}</span>
<span class="traceContains">+parse: product: {name: "1", properties: ["1": "address":"array":"character"]}</span>
<span class="Comment"># no other ingredients</span>
$parse: <span class="Constant">3</span>
<span class="Delimiter">:(scenario string_literal_escapes_newlines_in_trace)</span>
recipe main [
copy [abc
def]
]
<span class="traceContains">+parse: ingredient: {name: "abc\ndef", properties: [_: "literal-string"]}</span>
<span class="Delimiter">:(scenario string_literal_can_skip_past_comments)</span>
recipe main [
copy [
<span class="Comment"># ']' inside comment</span>
bar
]
]
<span class="traceContains">+parse: ingredient: {name: "\n # ']' inside comment\n bar\n ", properties: [_: "literal-string"]}</span>
<span class="Delimiter">:(scenario string_literal_empty)</span>
recipe main [
copy []
]
<span class="traceContains">+parse: ingredient: {name: "", properties: [_: "literal-string"]}</span>
</pre>
</body>
</html>
<!-- vim: set foldmethod=manual : -->
|