[GR-20446] Fix argument handling in Kernel.printf implementation.

PullRequest: truffleruby/1258

Author: eregon

CHANGELOG.md

@@ -68,6 +68,7 @@ Bug fixes:
 * Fixed issue with `ENV.each_key`.
 * Fixed `ENV.replace` implementation.
 * Fixed `ENV.udpate` implementation.
+* Fixed argument handling in `Kernel.printf`.
 
 Compatibility:
 

spec/tags/core/kernel/printf_tags.txt

@@ -1,118 +1,17 @@
-fails:Kernel.printf calls write on the first argument when it is not a string
-fails:Kernel.printf formatting io is specified integer formats converts argument into Integer with to_int
-fails:Kernel.printf formatting io is specified integer formats converts argument into Integer with to_i if to_int isn't available
-fails:Kernel.printf formatting io is specified integer formats converts String argument with Kernel#Integer
-fails:Kernel.printf formatting io is specified integer formats b converts argument as a binary number
-fails:Kernel.printf formatting io is specified integer formats b displays negative number as a two's complement prefixed with '..1'
-fails:Kernel.printf formatting io is specified integer formats b collapse negative number representation if it equals 1
-fails:Kernel.printf formatting io is specified integer formats B converts argument as a binary number
-fails:Kernel.printf formatting io is specified integer formats B displays negative number as a two's complement prefixed with '..1'
-fails:Kernel.printf formatting io is specified integer formats B collapse negative number representation if it equals 1
-fails:Kernel.printf formatting io is specified integer formats d converts argument as a decimal number
-fails:Kernel.printf formatting io is specified integer formats d works well with large numbers
-fails:Kernel.printf formatting io is specified integer formats i converts argument as a decimal number
-fails:Kernel.printf formatting io is specified integer formats i works well with large numbers
-fails:Kernel.printf formatting io is specified integer formats u converts argument as a decimal number
-fails:Kernel.printf formatting io is specified integer formats u works well with large numbers
-fails:Kernel.printf formatting io is specified integer formats o converts argument as an octal number
-fails:Kernel.printf formatting io is specified integer formats o displays negative number as a two's complement prefixed with '..7'
-fails:Kernel.printf formatting io is specified integer formats o collapse negative number representation if it equals 7
-fails:Kernel.printf formatting io is specified integer formats x converts argument as a hexadecimal number
-fails:Kernel.printf formatting io is specified integer formats x displays negative number as a two's complement prefixed with '..f'
-fails:Kernel.printf formatting io is specified integer formats x collapse negative number representation if it equals f
-fails:Kernel.printf formatting io is specified integer formats X converts argument as a hexadecimal number with uppercase letters
-fails:Kernel.printf formatting io is specified integer formats X displays negative number as a two's complement prefixed with '..f'
-fails:Kernel.printf formatting io is specified integer formats X collapse negative number representation if it equals F
-fails:Kernel.printf formatting io is specified float formats converts argument into Float
-fails:Kernel.printf formatting io is specified float formats e converts argument into exponential notation [-]d.dddddde[+-]dd
-fails:Kernel.printf formatting io is specified float formats e cuts excessive digits and keeps only 6 ones
-fails:Kernel.printf formatting io is specified float formats e rounds the last significant digit to the closest one
-fails:Kernel.printf formatting io is specified float formats e displays Float::INFINITY as Inf
-fails:Kernel.printf formatting io is specified float formats e displays Float::NAN as NaN
-fails:Kernel.printf formatting io is specified float formats E converts argument into exponential notation [-]d.dddddde[+-]dd
-fails:Kernel.printf formatting io is specified float formats E cuts excessive digits and keeps only 6 ones
-fails:Kernel.printf formatting io is specified float formats E rounds the last significant digit to the closest one
-fails:Kernel.printf formatting io is specified float formats E displays Float::INFINITY as Inf
-fails:Kernel.printf formatting io is specified float formats E displays Float::NAN as NaN
-fails:Kernel.printf formatting io is specified float formats f converts floating point argument as [-]ddd.dddddd
-fails:Kernel.printf formatting io is specified float formats f cuts excessive digits and keeps only 6 ones
-fails:Kernel.printf formatting io is specified float formats f rounds the last significant digit to the closest one
-fails:Kernel.printf formatting io is specified float formats f displays Float::INFINITY as Inf
-fails:Kernel.printf formatting io is specified float formats f displays Float::NAN as NaN
-fails:Kernel.printf formatting io is specified float formats g displays Float::INFINITY as Inf
-fails:Kernel.printf formatting io is specified float formats g displays Float::NAN as NaN
-fails:Kernel.printf formatting io is specified float formats g the exponent is less than -4 converts a floating point number using exponential form
-fails:Kernel.printf formatting io is specified float formats g the exponent is greater than or equal to the precision (6 by default) converts a floating point number using exponential form
-fails:Kernel.printf formatting io is specified float formats g otherwise converts a floating point number in dd.dddd form
-fails:Kernel.printf formatting io is specified float formats g otherwise cuts excessive digits in fractional part and keeps only 4 ones
-fails:Kernel.printf formatting io is specified float formats g otherwise rounds the last significant digit to the closest one in fractional part
-fails:Kernel.printf formatting io is specified float formats g otherwise cuts fraction part to have only 6 digits at all
-fails:Kernel.printf formatting io is specified float formats G displays Float::INFINITY as Inf
-fails:Kernel.printf formatting io is specified float formats G displays Float::NAN as NaN
-fails:Kernel.printf formatting io is specified float formats G the exponent is less than -4 converts a floating point number using exponential form
-fails:Kernel.printf formatting io is specified float formats G the exponent is greater than or equal to the precision (6 by default) converts a floating point number using exponential form
-fails:Kernel.printf formatting io is specified float formats G otherwise converts a floating point number in dd.dddd form
-fails:Kernel.printf formatting io is specified float formats G otherwise cuts excessive digits in fractional part and keeps only 4 ones
-fails:Kernel.printf formatting io is specified float formats G otherwise rounds the last significant digit to the closest one in fractional part
-fails:Kernel.printf formatting io is specified float formats G otherwise cuts fraction part to have only 6 digits at all
-fails:Kernel.printf formatting io is specified float formats a converts floating point argument as [-]0xh.hhhhp[+-]dd
-fails:Kernel.printf formatting io is specified float formats a displays Float::INFINITY as Inf
-fails:Kernel.printf formatting io is specified float formats a displays Float::NAN as NaN
-fails:Kernel.printf formatting io is specified float formats A converts floating point argument as [-]0xh.hhhhp[+-]dd and use uppercase X and P
-fails:Kernel.printf formatting io is specified float formats A displays Float::INFINITY as Inf
-fails:Kernel.printf formatting io is specified float formats A displays Float::NAN as NaN
-fails:Kernel.printf formatting io is specified other formats c displays character if argument is a numeric code of character
-fails:Kernel.printf formatting io is specified other formats c displays character if argument is a single character string
-fails:Kernel.printf formatting io is specified other formats c raises ArgumentError if argument is a string of several characters
 fails:Kernel.printf formatting io is specified other formats c raises ArgumentError if argument is an empty string
 fails:Kernel.printf formatting io is specified other formats c supports Unicode characters
-fails:Kernel.printf formatting io is specified other formats p displays argument.inspect value
-fails:Kernel.printf formatting io is specified other formats s substitute argument passes as a string
-fails:Kernel.printf formatting io is specified other formats s converts argument to string with to_s
 fails:Kernel.printf formatting io is specified other formats s does not try to convert with to_str
 fails:Kernel.printf formatting io is specified other formats % alone displays the percent sign
-fails:Kernel.printf formatting io is specified other formats % is escaped by %
-fails:Kernel.printf formatting io is specified flags space applies to numeric formats bBdiouxXeEfgGaA leaves a space at the start of non-negative numbers
-fails:Kernel.printf formatting io is specified flags space applies to numeric formats bBdiouxXeEfgGaA does not leave a space at the start of negative numbers
-fails:Kernel.printf formatting io is specified flags space applies to numeric formats bBdiouxXeEfgGaA prevents converting negative argument to two's complement form
-fails:Kernel.printf formatting io is specified flags space applies to numeric formats bBdiouxXeEfgGaA treats several white spaces as one
-fails:Kernel.printf formatting io is specified flags (digit)$ specifies the absolute argument number for this field
-fails:Kernel.printf formatting io is specified flags (digit)$ raises exception if argument number is bigger than actual arguments list
-fails:Kernel.printf formatting io is specified flags (digit)$ ignores '-' sign
-fails:Kernel.printf formatting io is specified flags (digit)$ raises ArgumentError exception when absolute and relative argument numbers are mixed
-fails:Kernel.printf formatting io is specified flags # applies to format o increases the precision until the first digit will be `0' if it is not formatted as complements
 fails:Kernel.printf formatting io is specified flags # applies to format o does nothing for negative argument
-fails:Kernel.printf formatting io is specified flags # applies to formats bBxX prefixes the result with 0x, 0X, 0b and 0B respectively for non-zero argument
 fails:Kernel.printf formatting io is specified flags # applies to formats bBxX does nothing for zero argument
 fails:Kernel.printf formatting io is specified flags # applies to formats aAeEfgG forces a decimal point to be added, even if no digits follow
-fails:Kernel.printf formatting io is specified flags # applies to formats aAeEfgG changes format from dd.dddd to exponential form for gG
 fails:Kernel.printf formatting io is specified flags # applies to gG does not remove trailing zeros
-fails:Kernel.printf formatting io is specified flags + applies to numeric formats bBdiouxXaAeEfgG adds a leading plus sign to non-negative numbers
-fails:Kernel.printf formatting io is specified flags + applies to numeric formats bBdiouxXaAeEfgG does not use two's complement form for negative numbers for formats bBoxX
 fails:Kernel.printf formatting io is specified flags - left-justifies the result of conversion if width is specified
-fails:Kernel.printf formatting io is specified flags 0 (zero) applies to numeric formats bBdiouxXaAeEfgG and width is specified pads with zeros, not spaces
 fails:Kernel.printf formatting io is specified flags 0 (zero) applies to numeric formats bBdiouxXaAeEfgG and width is specified uses radix-1 when displays negative argument as a two's complement
-fails:Kernel.printf formatting io is specified flags * uses the previous argument as the field width
 fails:Kernel.printf formatting io is specified flags * left-justifies the result if width is negative
-fails:Kernel.printf formatting io is specified flags * uses the specified argument as the width if * is followed by a number and $
 fails:Kernel.printf formatting io is specified flags * left-justifies the result if specified with $ argument is negative
-fails:Kernel.printf formatting io is specified flags * raises ArgumentError when is mixed with width
-fails:Kernel.printf formatting io is specified width specifies the minimum number of characters that will be written to the result
-fails:Kernel.printf formatting io is specified width is ignored if argument's actual length is greater
-fails:Kernel.printf formatting io is specified precision integer types controls the number of decimal places displayed
-fails:Kernel.printf formatting io is specified precision float types controls the number of decimal places displayed in fraction part
-fails:Kernel.printf formatting io is specified precision float types does not affect G format
 fails:Kernel.printf formatting io is specified precision string formats determines the maximum number of characters to be copied from the string
-fails:Kernel.printf formatting io is specified reference by name %<name>s style uses value passed in a hash argument
-fails:Kernel.printf formatting io is specified reference by name %<name>s style supports flags, width, precision and type
-fails:Kernel.printf formatting io is specified reference by name %<name>s style allows to place name in any position
-fails:Kernel.printf formatting io is specified reference by name %<name>s style cannot be mixed with unnamed style
-fails:Kernel.printf formatting io is specified reference by name %{name} style uses value passed in a hash argument
-fails:Kernel.printf formatting io is specified reference by name %{name} style does not support type style
 fails:Kernel.printf formatting io is specified reference by name %{name} style supports flags, width and precision
-fails:Kernel.printf formatting io is specified reference by name %{name} style cannot be mixed with unnamed style
-fails:Kernel.printf formatting io is specified reference by name %{name} style raises KeyError when there is no matching key
-fails:Kernel.printf formatting io is specified reference by name %{name} style converts value to String with to_s
 fails:Kernel.printf formatting io is not specified other formats c raises ArgumentError if argument is an empty string
 fails:Kernel.printf formatting io is not specified other formats c supports Unicode characters
 fails:Kernel.printf formatting io is not specified other formats s does not try to convert with to_str
@@ -127,7 +26,4 @@ fails:Kernel.printf formatting io is not specified flags * left-justifies the re
 fails:Kernel.printf formatting io is not specified precision string formats determines the maximum number of characters to be copied from the string
 fails:Kernel.printf formatting io is not specified reference by name %{name} style supports flags, width and precision
 fails:Kernel.printf formatting io is specified other formats % alone raises an ArgumentError
-fails:Kernel.printf formatting io is specified faulty key sets the Hash as the receiver of KeyError
-fails:Kernel.printf formatting io is specified faulty key sets the unmatched key as the key of KeyError
 fails:Kernel.printf formatting io is not specified other formats % alone raises an ArgumentError
-fails:Kernel.printf formatting io is specified faulty key raises a KeyError

src/main/ruby/truffleruby/core/kernel.rb

@@ -586,7 +586,14 @@ def __dir__
   module_function :__dir__
 
   def printf(*args)
-    print sprintf(*args)
+    return nil if args.empty?
+    if Truffle::Type.object_kind_of?(args[0], String)
+      print sprintf(*args)
+    else
+      io = args.shift
+      io.write(sprintf(*args))
+    end
+    nil
   end
   module_function :printf