From e905f9b03f9705388dfea9df961f5dfabb7902a9 Mon Sep 17 00:00:00 2001
From: Mark Olesen <Mark.Olesen@Germany>
Date: Sun, 9 Aug 2009 12:43:45 +0200
Subject: [PATCH] tokenizing Scalar is more stringent
- require a digit somewhere. This traps '.' or '.+E' type of sequences
instead of returning them as 0
- avoid buffer overruns on really long sequences
- should we be considering strtod() as an alternative to atof() and checking
the errno?
The same might be true for using strtoul() or strtol() instead of atol()
---
src/OpenFOAM/db/IOstreams/Sstreams/ISstream.C | 75 ++++++++++++-------
1 file changed, 47 insertions(+), 28 deletions(-)
diff --git a/src/OpenFOAM/db/IOstreams/Sstreams/ISstream.C b/src/OpenFOAM/db/IOstreams/Sstreams/ISstream.C
index 482abb8bf38..a05a5e72c4c 100644
--- a/src/OpenFOAM/db/IOstreams/Sstreams/ISstream.C
+++ b/src/OpenFOAM/db/IOstreams/Sstreams/ISstream.C
@@ -102,7 +102,7 @@ char Foam::ISstream::nextValid()
Foam::Istream& Foam::ISstream::read(token& t)
{
- static char numberBuffer[100];
+ static char charBuffer[128];
// Return the put back token if it exists
if (Istream::getBack(t))
@@ -113,7 +113,7 @@ Foam::Istream& Foam::ISstream::read(token& t)
// Assume that the streams supplied are in working order.
// Lines are counted by '\n'
- // Get next 'valid character': i.e. proceed through any white space
+ // Get next 'valid character': i.e. proceed through any whitespace
// and/or comments until a semantically valid character is hit upon.
char c = nextValid();
@@ -144,7 +144,7 @@ Foam::Istream& Foam::ISstream::read(token& t)
case token::COMMA :
case token::ASSIGN :
case token::ADD :
- // case token::SUBTRACT : // Handled later as the posible start of a number
+ // case token::SUBTRACT : // Handled later as the possible start of a number
case token::MULTIPLY :
case token::DIVIDE :
{
@@ -152,6 +152,7 @@ Foam::Istream& Foam::ISstream::read(token& t)
return *this;
}
+
// Strings: enclosed by double quotes.
case token::BEGIN_STRING :
{
@@ -170,21 +171,24 @@ Foam::Istream& Foam::ISstream::read(token& t)
return *this;
}
+
// Numbers: do not distinguish at this point between Types.
+ //
+ // we ideally wish to match the equivalent of this regular expression
+ //
+ // /^[-+]?([0-9]+\.?[0-9]*|\.[0-9]+)([Ee][-+]?[0-9]+)?$/
+ //
case '-' :
case '.' :
case '0' : case '1' : case '2' : case '3' : case '4' :
case '5' : case '6' : case '7' : case '8' : case '9' :
{
- bool isScalar = false;
-
- if (c == '.')
- {
- isScalar = true;
- }
+ // has a floating point or digit
+ bool isScalar = (c == '.');
+ bool hasDigit = isdigit(c);
- int i=0;
- numberBuffer[i++] = c;
+ unsigned int nChar = 0;
+ charBuffer[nChar++] = c;
while
(
@@ -199,14 +203,23 @@ Foam::Istream& Foam::ISstream::read(token& t)
)
)
{
- numberBuffer[i++] = c;
+ charBuffer[nChar++] = c;
+ if (nChar >= sizeof(charBuffer))
+ {
+ nChar--;
+ break;
+ }
- if (!isdigit(c))
+ if (isdigit(c))
+ {
+ hasDigit = true;
+ }
+ else
{
isScalar = true;
}
}
- numberBuffer[i] = '\0';
+ charBuffer[nChar] = '\0';
setState(is_.rdstate());
@@ -214,26 +227,31 @@ Foam::Istream& Foam::ISstream::read(token& t)
{
is_.putback(c);
- if (i == 1 && numberBuffer[0] == '-')
+ if (hasDigit)
{
- t = token::punctuationToken(token::SUBTRACT);
+ if (!isScalar)
+ {
+ long lt = atol(charBuffer);
+ t = label(lt);
+
+ // return as a scalar if doesn't fit in a label
+ isScalar = (t.labelToken() != lt);
+ }
+
+ if (isScalar)
+ {
+ t = scalar(atof(charBuffer));
+ }
}
- else if (isScalar)
+ else if (nChar == 1 && charBuffer[0] == '-')
{
- t = scalar(atof(numberBuffer));
+ // a single '-' is punctuation
+ t = token::punctuationToken(token::SUBTRACT);
}
else
{
- long lt = atol(numberBuffer);
- t = label(lt);
-
- // If the integer is too large to be represented as a label
- // return it as a scalar
- if (t.labelToken() != lt)
- {
- isScalar = true;
- t = scalar(atof(numberBuffer));
- }
+ // some really bad sequence - eg, ".+E"
+ t.setBad();
}
}
else
@@ -244,6 +262,7 @@ Foam::Istream& Foam::ISstream::read(token& t)
return *this;
}
+
// Should be a word (which can be a single character)
default:
{
--
GitLab