Skip to content

Commit

Permalink
Refactoring: better processing of (un)quoted values for tags in the c…
Browse files Browse the repository at this point in the history 
…onfiguration file (part 2)
  • Loading branch information
@doxygen
doxygen committed Dec 25, 2020
1 parent eb3d1eb commit e0c4e9f
Show file tree
Hide file tree
Showing 4 changed files with 314 additions and 72 deletions.
290 changes: 266 additions & 24 deletions addon/doxywizard/config_doxyw.l
View file
Original file line number Diff line number Diff line change
Expand Up @@ -65,13 +65,13 @@ static QStack<ConfigFileState*> g_includeStack;
static int g_includeDepth;
static QVariant *g_arg;
static Input *g_curOption=0;
static QString g_elemStr;
static QByteArray g_str;
static QTextCodec *g_codec = QTextCodec::codecForName("UTF-8");
static QString g_codecName = QString::fromLatin1("UTF-8");
static int g_lastState;
static QByteArray g_tmpString;
static QString g_cmd;
static int g_part;
static bool g_isEnum;

static const char *stateToString(int state);

Expand Down Expand Up @@ -110,6 +110,193 @@ static void checkEncoding()
}
}

static QByteArray stripComment(const QByteArray &s)
{
// check if there is a comment at the end of the string
bool insideQuote=false;
int l = s.length();
for (int i=0;i<l;i++)
{
char c = s.at(i);
if (c=='\\') // skip over escaped characters
{
i++;
}
else if (c=='"') // toggle inside/outside quotation
{
insideQuote=!insideQuote;
}
else if (!insideQuote && c=='#') // found start of a comment
{
return s.left(i).trimmed();
}
}
return s;
}


static void processString()
{
// strip leading and trailing whitespace
QByteArray s = stripComment(g_str.trimmed());
int l = s.length();

// remove surrounding quotes if present (and not escaped)
bool quotedString=false;
if (l>=2 && s.at(0)=='"' && s.at(l-1)=='"' && // remove quotes
(s.at(l-2)!='\\' || (s.at(l-2)=='\\' && s.at(l-3)=='\\')))
{
s=s.mid(1,s.length()-2);
l=s.length();
quotedString=true;
}

// check for invalid and/or escaped quotes
bool warned=false;
QByteArray result;
for (int i=0;i<l;i++)
{
char c = s.at(i);
if (c=='\\') // escaped character
{
if (i<l-1 && s.at(i+1)=='"') // unescape the quote character
{
result+='"';
}
else // keep other escaped characters in escaped form
{
result+=c;
if (i<l-1)
{
result+=s.at(i+1);
}
}
i++; // skip over the escaped character
}
else if (c=='"') // unescaped quote
{
if (!warned)
{
std::string str = g_str.trimmed().toStdString();
config_warn("Invalid value for '%s' tag at line %d, file %s: Value '%s' is not properly quoted\n",
qPrintable(g_cmd),yylineno,qPrintable(g_yyFileName),str.c_str());
}
warned=true;
}
else // normal character
{
result+=c;
}
}

// recode the string
if (g_isEnum)
{
InputString *cur = dynamic_cast<InputString *>(g_curOption);
*g_arg = cur->checkEnumVal(g_codec->toUnicode(result));
}
else
{
*g_arg = QVariant(g_codec->toUnicode(result));
}

// update encoding
checkEncoding();

//printf("Processed string '%s'\n",g_string->data());
}

static void processList()
{
bool allowCommaAsSeparator = g_cmd!=QString::fromLatin1("PREDEFINED");

const QByteArray s = stripComment(g_str.trimmed());
int l = s.length();

QByteArray elemStr;

// helper to push elemStr to the list and clear it
auto addElem = [&elemStr]()
{
if (!elemStr.isEmpty())
{
//printf("Processed list element '%s'\n",e.data());
*g_arg = QVariant(g_arg->toStringList() << g_codec->toUnicode(elemStr));
elemStr="";
}
};

bool needsSeparator=false;
int insideQuote=false;
bool warned=false;
for (int i=0;i<l;i++)
{
char c = s.at(i);
if (!needsSeparator && c=='\\') // escaped character
{
if (i<l-1 && s.at(i+1)=='"') // unescape the quote character
{
elemStr+='"';
}
else // keep other escaped characters in escaped form
{
elemStr+=c;
if (i<l-1)
{
elemStr+=s.at(i+1);
}
}
i++; // skip over the escaped character
}
else if (!needsSeparator && c=='"') // quote character
{
if (!insideQuote)
{
insideQuote=true;
}
else // this quote ends an element
{
insideQuote=false;
needsSeparator=true;
}
}
else if (!insideQuote && ((c==',' && allowCommaAsSeparator) || isspace(c))) // separator
{
needsSeparator=false;
addElem();
}
else // normal content character
{
if (needsSeparator)
{
if (!warned)
{
std::string str = g_str.trimmed().toStdString();
config_warn("Invalid value for '%s' tag at line %d, file %s: Values in list '%s' are not properly space %sseparated\n",
qPrintable(g_cmd),yylineno,qPrintable(g_yyFileName),str.c_str(),allowCommaAsSeparator?"or comma ":"");
warned=true;
}
needsSeparator=false;
i--; // try the character again as part of a new element
addElem();
}
else
{
elemStr+=c;
}
}
}
// add last part
addElem();
if (insideQuote)
{
std::string str = g_str.trimmed().toStdString();
config_warn("Invalid value for '%s' tag at line %d, file %s: Values in list '%s' are not properly quoted\n",
g_cmd.data(),yylineno,g_yyFileName.data(),str.c_str());
}
}


static FILE *tryPath(const QString &path,const QString &fileName)
{
QString absName=!path.isEmpty() ? path+QString::fromLatin1("/")+fileName : fileName;
Expand Down Expand Up @@ -198,7 +385,6 @@ static void readIncludeFile(const QString &incName)
%option yylineno

%x Start
%x SkipComment
%x SkipInvalid
%x GetString
%x GetEnum
Expand All @@ -209,7 +395,13 @@ static void readIncludeFile(const QString &incName)
%%

<*>\0x0d
<Start,GetString,GetEnum,GetStrList,SkipInvalid>"#" { BEGIN(SkipComment); }

/*-------------- Comments ---------------*/

<Start>"#".*\n { /* Skip comment */ }

/*-------------- TAG start ---------------*/

<Start>[a-z_A-Z][a-z_A-Z0-9]*[ \t]*"=" { g_cmd = g_codec->toUnicode(yytext);
g_part = 0;
g_cmd=g_cmd.left(g_cmd.length()-1).trimmed();
Expand All @@ -222,24 +414,17 @@ static void readIncludeFile(const QString &incName)
}
else // known tag
{
//option->setEncoding(encoding);
g_arg = &g_curOption->value();
g_str = QByteArray();
switch(g_curOption->kind())
{
case Input::StrList:
g_elemStr = QString();
*g_arg = QStringList();
BEGIN(GetStrList);
break;
case Input::String:
if (dynamic_cast<InputString *>(g_curOption)->stringMode() == InputString::StringFixed)
{
BEGIN(GetEnum);
}
else
{
BEGIN(GetString);
}
g_isEnum = dynamic_cast<InputString *>(g_curOption)->stringMode() == InputString::StringFixed;
BEGIN(GetString);
break;
case Input::Int:
BEGIN(GetString);
Expand Down Expand Up @@ -272,7 +457,7 @@ static void readIncludeFile(const QString &incName)
{
case Input::StrList:
g_arg = &g_curOption->value();
g_elemStr=QString();
g_str=QByteArray();
BEGIN(GetStrList);
break;
case Input::String:
Expand All @@ -292,7 +477,10 @@ static void readIncludeFile(const QString &incName)
}
}
}
<Start>"@INCLUDE_PATH"[ \t]*"=" { BEGIN(GetStrList); g_arg=&g_includePathList; *g_arg = QStringList(); g_elemStr=QString(); }

/*-------------- INCLUDE* ---------------*/

<Start>"@INCLUDE_PATH"[ \t]*"=" { BEGIN(GetStrList); g_arg=&g_includePathList; *g_arg = QStringList(); g_str=QByteArray(); }
/* include a config file */
<Start>"@INCLUDE"[ \t]*"=" { BEGIN(Include);}
<Include>([^ \"\t\r\n]+)|("\""[^\n\"]+"\"") {
Expand Down Expand Up @@ -321,22 +509,72 @@ static void readIncludeFile(const QString &incName)
}

<Start>[a-z_A-Z0-9]+ { config_warn("ignoring unknown tag '%s' at line %d, file %s\n",yytext,yylineno,qPrintable(g_yyFileName)); }
<GetString,GetEnum,SkipInvalid>\n { BEGIN(Start); }

/*-------------- GetString ---------------*/

<GetString>\n { // end of string
processString();
BEGIN(Start);
}
<GetString>\\[ \r\t]*\n { // line continuation
g_str+=' ';
}
<GetString>"\\" { // escape character
g_str+=yytext;
}
<GetString>[^\n\\]+ { // string part without escape characters
g_str+=yytext;
}

/*-------------- GetStrList ---------------*/

<GetStrList>\n { // end of list
processList();
BEGIN(Start);
}
<GetStrList>\\[ \r\t]*\n { // line continuation
g_str+=' ';
}
<GetStrList>"\\" { // escape character
g_str+=yytext;
}
<GetStrList>[^\n\\]+ { // string part without escape characters
g_str+=yytext;
}

/*-------------- SkipInvalid ---------------*/

<SkipInvalid>\n { // end of skipped part
BEGIN(Start);
}
<SkipInvalid>\\[ \r\t]*\n { // line continuation
g_str+=' ';
}
<SkipInvalid>"\\" { // escape character
g_str+=yytext;
}
<SkipInvalid>[^\n\\]+ { // string part without escape characters
g_str+=yytext;
}

/*
<GetString,GetEnum,SkipInvalid>\n { BEGIN(Start); }
<GetStrList>\n {
if (!g_elemStr.isEmpty())
if (!g_str.isEmpty())
{
//printf("elemStr1='%s'\n",qPrintable(elemStr));
*g_arg = QVariant(g_arg->toStringList() << g_elemStr);
*g_arg = QVariant(g_arg->toStringList() << g_str);
}
BEGIN(Start);
}
<GetStrList>[ \t]+ {
if (!g_elemStr.isEmpty())
if (!g_str.isEmpty())
{
//printf("elemStr2='%s'\n",qPrintable(elemStr));
*g_arg = QVariant(g_arg->toStringList() << g_elemStr);
*g_arg = QVariant(g_arg->toStringList() << g_str);
}
g_elemStr = QString();
g_str = QString();
}
<GetString>[^ \"\t\r\n]+ {
if (g_part == 1) // multiple unquoted parts, reset to default
Expand Down Expand Up @@ -430,7 +668,7 @@ static void readIncludeFile(const QString &incName)
}
else
{
g_elemStr+=g_codec->toUnicode(g_tmpString);
g_str+=g_codec->toUnicode(g_tmpString);
}
if (*yytext=='\n')
{
Expand All @@ -447,11 +685,15 @@ static void readIncludeFile(const QString &incName)
}
<GetQuotedString>. { g_tmpString+=*yytext; }
<GetStrList>[^ \#\"\t\r\n]+ {
g_elemStr+=g_codec->toUnicode(yytext);
g_str+=g_codec->toUnicode(yytext);
}
<SkipComment>\n { BEGIN(Start); }
<SkipComment>\\[ \r\t]*\n { BEGIN(Start); }
<SkipComment,SkipInvalid>. { }
*/

/*-------------- fall through -------------*/

<*>\\[ \r\t]*\n { }
<*>[ \r\t] { }
<*>\n
Expand Down
Loading

0 comments on commit e0c4e9f

Please sign in to comment.