summaryrefslogtreecommitdiff
path: root/core/string
diff options
context:
space:
mode:
authorRémi Verschelde <rverschelde@gmail.com>2021-10-28 15:43:36 +0200
committerRémi Verschelde <rverschelde@gmail.com>2021-10-28 15:43:36 +0200
commit3a6be64c12a3ba8edb646d5362c34de1a5fcc0bf (patch)
treee43708b20321cf07262918d1bd6fef02c70dabf9 /core/string
parent3b11e33a099daa0978147a7550dd84ba5dd14f35 (diff)
clang-format: Various fixes to comments alignment from `clang-format` 13
All reviewed manually and occasionally rewritten to avoid bad auto formatting.
Diffstat (limited to 'core/string')
-rw-r--r--core/string/ustring.cpp114
1 files changed, 61 insertions, 53 deletions
diff --git a/core/string/ustring.cpp b/core/string/ustring.cpp
index 397743fb6e..8d6da31cf3 100644
--- a/core/string/ustring.cpp
+++ b/core/string/ustring.cpp
@@ -2317,28 +2317,33 @@ bool String::is_numeric() const {
}
template <class C>
-static double built_in_strtod(const C *string, /* A decimal ASCII floating-point number,
- * optionally preceded by white space. Must
- * have form "-I.FE-X", where I is the integer
- * part of the mantissa, F is the fractional
- * part of the mantissa, and X is the
- * exponent. Either of the signs may be "+",
- * "-", or omitted. Either I or F may be
- * omitted, or both. The decimal point isn't
- * necessary unless F is present. The "E" may
- * actually be an "e". E and X may both be
- * omitted (but not just one). */
- C **endPtr = nullptr) /* If non-nullptr, store terminating Cacter's
- * address here. */
-{
- static const int maxExponent = 511; /* Largest possible base 10 exponent. Any
- * exponent larger than this will already
- * produce underflow or overflow, so there's
- * no need to worry about additional digits.
- */
- static const double powersOf10[] = { /* Table giving binary powers of 10. Entry */
- 10., /* is 10^2^i. Used to convert decimal */
- 100., /* exponents into floating-point numbers. */
+static double built_in_strtod(
+ /* A decimal ASCII floating-point number,
+ * optionally preceded by white space. Must
+ * have form "-I.FE-X", where I is the integer
+ * part of the mantissa, F is the fractional
+ * part of the mantissa, and X is the
+ * exponent. Either of the signs may be "+",
+ * "-", or omitted. Either I or F may be
+ * omitted, or both. The decimal point isn't
+ * necessary unless F is present. The "E" may
+ * actually be an "e". E and X may both be
+ * omitted (but not just one). */
+ const C *string,
+ /* If non-nullptr, store terminating Cacter's
+ * address here. */
+ C **endPtr = nullptr) {
+ /* Largest possible base 10 exponent. Any
+ * exponent larger than this will already
+ * produce underflow or overflow, so there's
+ * no need to worry about additional digits. */
+ static const int maxExponent = 511;
+ /* Table giving binary powers of 10. Entry
+ * is 10^2^i. Used to convert decimal
+ * exponents into floating-point numbers. */
+ static const double powersOf10[] = {
+ 10.,
+ 100.,
1.0e4,
1.0e8,
1.0e16,
@@ -2353,25 +2358,28 @@ static double built_in_strtod(const C *string, /* A decimal ASCII floating-point
const double *d;
const C *p;
int c;
- int exp = 0; /* Exponent read from "EX" field. */
- int fracExp = 0; /* Exponent that derives from the fractional
- * part. Under normal circumstances, it is
- * the negative of the number of digits in F.
- * However, if I is very long, the last digits
- * of I get dropped (otherwise a long I with a
- * large negative exponent could cause an
- * unnecessary overflow on I alone). In this
- * case, fracExp is incremented one for each
- * dropped digit. */
- int mantSize; /* Number of digits in mantissa. */
- int decPt; /* Number of mantissa digits BEFORE decimal
- * point. */
- const C *pExp; /* Temporarily holds location of exponent in
- * string. */
+ /* Exponent read from "EX" field. */
+ int exp = 0;
+ /* Exponent that derives from the fractional
+ * part. Under normal circumstances, it is
+ * the negative of the number of digits in F.
+ * However, if I is very long, the last digits
+ * of I get dropped (otherwise a long I with a
+ * large negative exponent could cause an
+ * unnecessary overflow on I alone). In this
+ * case, fracExp is incremented one for each
+ * dropped digit. */
+ int fracExp = 0;
+ /* Number of digits in mantissa. */
+ int mantSize;
+ /* Number of mantissa digits BEFORE decimal point. */
+ int decPt;
+ /* Temporarily holds location of exponent in string. */
+ const C *pExp;
/*
- * Strip off leading blanks and check for a sign.
- */
+ * Strip off leading blanks and check for a sign.
+ */
p = string;
while (*p == ' ' || *p == '\t' || *p == '\n') {
@@ -2388,9 +2396,9 @@ static double built_in_strtod(const C *string, /* A decimal ASCII floating-point
}
/*
- * Count the number of digits in the mantissa (including the decimal
- * point), and also locate the decimal point.
- */
+ * Count the number of digits in the mantissa (including the decimal
+ * point), and also locate the decimal point.
+ */
decPt = -1;
for (mantSize = 0;; mantSize += 1) {
@@ -2405,11 +2413,11 @@ static double built_in_strtod(const C *string, /* A decimal ASCII floating-point
}
/*
- * Now suck up the digits in the mantissa. Use two integers to collect 9
- * digits each (this is faster than using floating-point). If the mantissa
- * has more than 18 digits, ignore the extras, since they can't affect the
- * value anyway.
- */
+ * Now suck up the digits in the mantissa. Use two integers to collect 9
+ * digits each (this is faster than using floating-point). If the mantissa
+ * has more than 18 digits, ignore the extras, since they can't affect the
+ * value anyway.
+ */
pExp = p;
p -= mantSize;
@@ -2455,8 +2463,8 @@ static double built_in_strtod(const C *string, /* A decimal ASCII floating-point
}
/*
- * Skim off the exponent.
- */
+ * Skim off the exponent.
+ */
p = pExp;
if ((*p == 'E') || (*p == 'e')) {
@@ -2486,10 +2494,10 @@ static double built_in_strtod(const C *string, /* A decimal ASCII floating-point
}
/*
- * Generate a floating-point number that represents the exponent. Do this
- * by processing the exponent one bit at a time to combine many powers of
- * 2 of 10. Then combine the exponent with the fraction.
- */
+ * Generate a floating-point number that represents the exponent. Do this
+ * by processing the exponent one bit at a time to combine many powers of
+ * 2 of 10. Then combine the exponent with the fraction.
+ */
if (exp < 0) {
expSign = true;