Workaround broken signif call

Author: ScottPJones

src/fmtcore.jl

@@ -219,6 +219,36 @@ function _pfmt_floate(out::IO, sch::Char, zs::Integer, u::Real, prec::Int, e::In
 end
 
 
+# Pull in definition of signif from v0.6.2 base, since it is currently broken for
+# BigFloat and DecFP (at least) on master
+
+@static if VERSION >= v"0.7.0-DEV.4804"
+# adapted from Matlab File Exchange roundsd: http://www.mathworks.com/matlabcentral/fileexchange/26212
+# for round, og is the power of 10 relative to the decimal point
+# for signif, og is the absolute power of 10
+# digits and base must be integers, x must be convertable to float
+
+function signif(x::Real, digits::Integer, base::Integer=10)
+    digits < 1 && throw(DomainError())
+
+    x = float(x)
+    (x == 0 || !isfinite(x)) && return x
+    if base == 10
+        e = floor(log10(abs(x)) - digits + 1.)
+        og = oftype(x, exp10(abs(e)))
+    elseif base == 2
+        e = exponent(abs(x)) - digits + 1.
+        og = oftype(x, exp2(abs(e)))
+    else
+        e = floor(log(base, abs(x)) - digits + 1.)
+        og = oftype(x, float(base) ^ abs(e))
+    end
+    # for numeric stability
+    r = e >= 0 ? round(x/og)*og : round(x*og)/og
+    isfinite(r) ? r : x
+end
+end
+
 function _pfmt_e(out::IO, fs::FormatSpec, x::AbstractFloat)
     # extract sign, significand, and exponent
     ax = abs(x)
@@ -227,8 +257,7 @@ function _pfmt_e(out::IO, fs::FormatSpec, x::AbstractFloat)
         e = 0
         u = zero(x)
     else
-        rax = (@static VERSION < v"0.7.0-DEV.4804" ? signif(ax, fs.prec + 1) :
-               round(ax; sigdigits = fs.prec + 1))
+        rax = signif(ax, fs.prec + 1) # round(ax; sigdigits = fs.prec + 1)
         e = floor(Integer, log10(rax))  # exponent
         u = rax * exp10(-e)  # significand
     end