static struct phyloTree *parseSubTree(char **ptrPtr)
/* the recursive workhorse function, parses a tree from ptr */
{
struct phyloTree *node = NULL;
char *ptr = *ptrPtr;
/* trees are terminated by one of these three chars */
if ((*ptr == ';') || (*ptr == ',') || (*ptr == ')') )
return NULL;
AllocVar(node);
if (*ptr == '(')
{
struct phyloTree *edge;
ptr++;
do
{
struct phyloTree *child = parseSubTree(&ptr);
if (!child)
errAbort("missing child/subTree at (%s)",ptr-1);
edge = newEdge(node,child);
edge->parent = node;
} while (*ptr++ == ',');
--ptr;
if (*ptr++ != ')')
errAbort("unbalanced parenthesis at (%s)",ptr-1);
node->ident = parseIdent(&ptr);
}
else
if ((*ptr == ':') || (isalpha(*ptr))|| (isdigit(*ptr))
|| (*ptr == '\'') || (*ptr == '.'))
node->ident = parseIdent(&ptr);
else
errAbort("illegal char '%c' in phyloString",*ptr);
if (*ptr == '[')
{
if (startsWith("[&&NHX:D=Y]",ptr))
node->isDup = TRUE;
while(*ptr != ']')
ptr++;
ptr++;
}
*ptrPtr = ptr;
return node;
}
/*
* Parse an external acl helper line.
*
* Squid can be configured to pass various formats, we assume something
* similar to the normal redirector format:
*
* %URI %SRC %LOGIN
*
* For example:
* external_acl_type foo ttl=60 children=1 %URI %SRC %LOGIN /path/to/sg
*/
int parseAuthzLine(char *line, struct SquidInfo *s)
{
char *field = NULL;
sgLogDebug("got authz line %s", line);
resetSquidInfo(s);
/* get the URL and parse */
if ((field = strtok(line, "\t ")) == NULL)
return 0;
HTUnEscape(field);
if (!parseUrl(field, s))
return 0;
/* get the source address and parse */
if ((field = strtok(NULL, " \t\n")) == NULL)
return 0;
HTUnEscape(field); /* just in case, IPs should not need escaping */
strcpy(s->src, field);
/* get the login and parse */
if ((field = strtok(NULL, " \t\n")) == NULL)
return 0;
if (!parseIdent(field,s))
return 0;
sgLogDebug("parsed authz line: furl='%s' domain='%s' surl='%s' src=%s ident='%s'",
s->furl, s->domain, s->surl, s->src, s->ident);
return 1;
}
bool TScanner::nextToken()
{
skipSpace();
tokenCol=current-lineStart;
tokenLine=line;
if(current>=condition.length()){
token=TToken::FILE_END;
tokenText="";
return false;
}
QChar l_ch=condition.at(current);
if(l_ch=='"'||l_ch=='\''){
return parseString(l_ch);
} else if((l_ch>='a' && l_ch <='z') ||(l_ch>='A' && l_ch<='Z') || l_ch=='_'){
return parseIdent();
}
token=TToken::CHAR;
tokenText=l_ch;
current++;
if(l_ch=='='){
token=TToken::COND_EQUAL;
} else if(l_ch=='>'){
token=TToken::COND_BIGGER;
if(current<condition.length()){
if(condition.at(current)=='='){
token=TToken::COND_BE;
tokenText=">=";
current++;
}
}
} else if(l_ch=='<'){
token=TToken::COND_SMALLER;
if(current<condition.length()){
if(condition.at(current)=='='){
token=TToken::COND_SE;
tokenText="<=";
current++;
}
}
} else if(l_ch=='!'){
token=TToken::COND_NOT;
if(current<condition.length()){
if(condition.at(current)=='='){
token=TToken::COND_NE;
tokenText="!=";
current++;
}
}
} else if(l_ch=='('){
token=TToken::HOOK_L;
} else if(l_ch==')'){
token=TToken::HOOK_R;
}
return false;
}
/*
* Parse a redirector input line, format is:
*
* URL ip-address/fqdn ident method
*
* for example
* http://www.example.com/page1.html 192.168.2.3/- andi GET
*/
int parseLine(char *line, struct SquidInfo *s)
{
char *field = NULL;
char *p = NULL;
sgLogDebug("got redirector line %s", line);
resetSquidInfo(s);
/* get the URL and parse */
if ((field = strtok(line, "\t ")) == NULL)
return 0;
HTUnEscape(field);
if (!parseUrl(field, s))
return 0;
/* get the source address and parse */
if ((field = strtok(NULL, " \t\n")) == NULL)
return 0;
if ((p = strchr(field, '/')) != NULL) {
*p = 0;
strcpy(s->src, field);
strcpy(s->srcDomain, p + 1);
if (s->srcDomain[0] == '-' && s->srcDomain[1] == 0)
s->srcDomain[0] = 0;
} else {
strcpy(s->src, field);
}
/* get the identity */
if ((field = strtok(NULL, " \t\n")) == NULL)
return 0;
if (!parseIdent(field,s))
return 0;
/* get the method */
if ((field = strtok(NULL, " \t\n")) == NULL)
return 0;
strcpy(s->method, field);
if (s->method[0] == '\0')
return 0;
sgLogDebug("parsed redirector line: furl='%s' domain='%s' surl='%s' src=%s ident='%s'",
s->furl, s->domain, s->surl, s->src, s->ident);
return 1;
}
void parseExpr(B& in, S& solver, vec<Lit>& out_ps, vec<Int>& out_Cs, vec<char>& tmp)
// NOTE! only uses "getVar()" method of solver; doesn't add anything.
// 'tmp' is a tempory, passed to avoid frequent memory reallocation.
{
bool empty = true;
for(;;){
skipWhitespace(in);
if ((*in < '0' || *in > '9') && *in != '+' && *in != '-') break;
out_Cs.push(parseInt(in));
skipWhitespace(in);
if (*in != '*') throw xstrdup("Missing '*' after coefficient.");
++in;
out_ps.push(Lit(solver.getVar(parseIdent(in, tmp))));
empty = false;
}
if (empty) throw xstrdup("Empty expression.");
}
FlowToken FlowLexer::nextToken()
{
bool expectsValue = token() == FlowToken::Ident || FlowTokenTraits::isOperator(token());
lastPos_ = currPos_;
if (consumeSpace())
return token_ = FlowToken::Eof;
// printf("FlowLexer.nextToken: currentChar %s curr[%zu:%zu.%zu] next[%zu:%zu.%zu]\n",
// escape(currentChar_).c_str(),
// currPos_.line, currPos_.column, currPos_.offset,
// nextPos_.line, nextPos_.column, nextPos_.offset);
content_.clear();
content_ += static_cast<char>(currentChar_);
lastLocation_ = currLocation_;
currLocation_.begin = currPos_;
switch (currentChar_) {
case EOF: // (-1)
return token_ = FlowToken::Eof;
case '=':
switch (nextChar()) {
case '=':
nextChar();
return token_ = FlowToken::Equal;
case '^':
nextChar();
return token_ = FlowToken::PrefixMatch;
case '$':
nextChar();
return token_ = FlowToken::SuffixMatch;
case '~':
nextChar();
return token_ = FlowToken::RegexMatch;
case '>':
nextChar();
return token_ = FlowToken::HashRocket;
default:
return token_ = FlowToken::Assign;
}
case '<':
switch (nextChar()) {
case '<':
nextChar();
return token_ = FlowToken::Shl;
case '=':
nextChar();
return token_ = FlowToken::LessOrEqual;
default:
return token_ = FlowToken::Less;
}
case '>':
switch (nextChar()) {
case '>':
nextChar();
return token_ = FlowToken::Shr;
case '=':
nextChar();
return token_ = FlowToken::GreaterOrEqual;
default:
return token_ = FlowToken::Greater;
}
case '|':
switch (nextChar()) {
case '|':
nextChar();
return token_ = FlowToken::Or;
case '=':
nextChar();
return token_ = FlowToken::OrAssign;
default:
return token_ = FlowToken::BitOr;
}
case '&':
switch (nextChar()) {
case '&':
nextChar();
return token_ = FlowToken::And;
case '=':
nextChar();
return token_ = FlowToken::AndAssign;
default:
return token_ = FlowToken::BitAnd;
}
case '.':
if (nextChar() == '.') {
if (nextChar() == '.') {
nextChar();
return token_ = FlowToken::Ellipsis;
}
return token_ = FlowToken::DblPeriod;
}
return token_ = FlowToken::Period;
case ':':
if (peekChar() == ':') {
stringValue_.clear();
return continueParseIPv6(false);
} else {
nextChar();
return token_ = FlowToken::Colon;
}
case ';':
nextChar();
return token_ = FlowToken::Semicolon;
case ',':
nextChar();
return token_ = FlowToken::Comma;
case '{':
nextChar();
return token_ = FlowToken::Begin;
case '}':
if (interpolationDepth_) {
return token_ = parseInterpolationFragment(false);
} else {
nextChar();
return token_ = FlowToken::End;
}
case '(':
nextChar();
return token_ = FlowToken::RndOpen;
case ')':
nextChar();
return token_ = FlowToken::RndClose;
case '[':
nextChar();
return token_ = FlowToken::BrOpen;
case ']':
nextChar();
return token_ = FlowToken::BrClose;
case '+':
nextChar();
return token_ = FlowToken::Plus;
case '-':
nextChar();
return token_ = FlowToken::Minus;
case '*':
switch (nextChar()) {
case '*':
nextToken();
return token_ = FlowToken::Pow;
default:
return token_ = FlowToken::Mul;
}
case '/':
if (expectsValue)
return token_ = parseString('/', FlowToken::RegExp);
nextChar();
return token_ = FlowToken::Div;
case '%':
nextChar();
return token_ = FlowToken::Mod;
case '!':
switch (nextChar()) {
case '=':
nextChar();
return token_ = FlowToken::UnEqual;
default:
return token_ = FlowToken::Not;
}
case '\'':
return token_ = parseString(true);
case '"':
++interpolationDepth_;
return token_ = parseInterpolationFragment(true);
case '0':
case '1':
case '2':
case '3':
case '4':
case '5':
case '6':
case '7':
case '8':
case '9':
return parseNumber();
default:
if (std::isalpha(currentChar()) || currentChar() == '_')
return token_ = parseIdent();
if (std::isprint(currentChar()))
printf("lexer: unknown char %c (0x%02X)\n", currentChar(), currentChar());
else
printf("lexer: unknown char %u (0x%02X)\n", currentChar() & 0xFF, currentChar() & 0xFF);
nextChar();
return token_ = FlowToken::Unknown;
}
}
/**
Implements the parsing of expressions that are part of the core language
The core language supports:
- Functions (but not closures)
- Integer literals (decimal only)
- String literals
- Tuple literals [a,b]
- If-else expressions
- Sequencing blocks with {a;b}
- Function calls fun(a,b)
*/
Tuple coreParseExpr(Input* input)
{
// Consume whitespace
input->eatWS();
// Numerical constant
if (isdigit(input->peekCh()))
{
return parseInt(input);
}
// String literal
if (input->matchCh('\''))
{
return parseStringLit(input, '\'');
}
if (input->matchCh('\"'))
{
return parseStringLit(input, '\"');
}
// Tuple expression
if (input->matchCh('['))
{
return Tuple{
Tuple("tuple"),
parseExprList(input, ']')
};
}
// Block/sequence expression (i.e { a; b; c }
if (input->matchCh('{'))
{
return parseBlockExpr(input, '}');
}
// Keywords
if (isalnum(input->peekCh()))
{
// If expression
if (input->matchStr("if"))
return parseIfExpr(input);
// Let expression
if (input->matchStr("let"))
return parseLetExpr(input);
// Function definition expression
if (input->matchStr("fun"))
return parseFunExpr(input);
// Function call expression
if (input->matchStr("call"))
return parseCallExpr(input);
}
// Host/magic constants
if (input->matchCh('$'))
{
return parseHostRef(input);
}
// Variable references
if (input->peekCh() == '_' || isalpha(input->peekCh()))
{
return parseIdent(input);
}
// Parsing failed
throw ParseError(input, "failed to parse expression");
}
