Support thousands separator in fmtnum

docs/src/manpage.md

@@ -2416,9 +2416,14 @@ This is simply a copy of what you should see on running `man mlr` at a command p
        $* = fmtifnum($*, "%.6f") formats numeric fields in the current record, leaving non-numeric ones alone
 
    1mfmtnum0m
-        (class=conversion #args=2) Convert int/float/bool to string using printf-style format string (https://pkg.go.dev/fmt), e.g. '$s = fmtnum($n, "%08d")' or '$t = fmtnum($n, "%.6e")'. This function recurses on array and map values.
-       Example:
-       $x = fmtnum($x, "%.6f")
+        (class=conversion #args=2) Convert int/float/bool to string using printf-style format string (https://pkg.go.dev/fmt), e.g. '$s = fmtnum($n, "%08d")' or '$t = fmtnum($n, "%.6e")'. Miller-specific extension: "%_d" and "%_f" for comma-separated thousands. This function recurses on array and map values.
+       Examples:
+       $y = fmtnum($x, "%.6f")
+       $o = fmtnum($n, "%d")
+       $o = fmtnum($n, "%12d")
+       $y = fmtnum($x, "%.6_f")
+       $o = fmtnum($n, "%_d")
+       $o = fmtnum($n, "%12_d")
 
    1mfold0m
         (class=higher-order-functions #args=3) Given a map or array as first argument and a function as second argument, accumulates entries into a final output -- for example, sum or product. For arrays, the function should take two arguments, for accumulated value and array element. For maps, it should take four arguments, for accumulated key and value, and map-element key and value; it should return the updated accumulator as a new key-value pair (i.e. a single-entry map). The start value for the accumulator is taken from the third argument.

docs/src/manpage.txt

@@ -2395,9 +2395,14 @@
        $* = fmtifnum($*, "%.6f") formats numeric fields in the current record, leaving non-numeric ones alone
 
    1mfmtnum0m
-        (class=conversion #args=2) Convert int/float/bool to string using printf-style format string (https://pkg.go.dev/fmt), e.g. '$s = fmtnum($n, "%08d")' or '$t = fmtnum($n, "%.6e")'. This function recurses on array and map values.
-       Example:
-       $x = fmtnum($x, "%.6f")
+        (class=conversion #args=2) Convert int/float/bool to string using printf-style format string (https://pkg.go.dev/fmt), e.g. '$s = fmtnum($n, "%08d")' or '$t = fmtnum($n, "%.6e")'. Miller-specific extension: "%_d" and "%_f" for comma-separated thousands. This function recurses on array and map values.
+       Examples:
+       $y = fmtnum($x, "%.6f")
+       $o = fmtnum($n, "%d")
+       $o = fmtnum($n, "%12d")
+       $y = fmtnum($x, "%.6_f")
+       $o = fmtnum($n, "%_d")
+       $o = fmtnum($n, "%12_d")
 
    1mfold0m
         (class=higher-order-functions #args=3) Given a map or array as first argument and a function as second argument, accumulates entries into a final output -- for example, sum or product. For arrays, the function should take two arguments, for accumulated value and array element. For maps, it should take four arguments, for accumulated key and value, and map-element key and value; it should return the updated accumulator as a new key-value pair (i.e. a single-entry map). The start value for the accumulator is taken from the third argument.

docs/src/reference-dsl-builtin-functions.md

@@ -534,9 +534,14 @@ $* = fmtifnum($*, "%.6f") formats numeric fields in the current record, leaving
 
 ### fmtnum
 <pre class="pre-non-highlight-non-pair">
-fmtnum  (class=conversion #args=2) Convert int/float/bool to string using printf-style format string (https://pkg.go.dev/fmt), e.g. '$s = fmtnum($n, "%08d")' or '$t = fmtnum($n, "%.6e")'. This function recurses on array and map values.
-Example:
-$x = fmtnum($x, "%.6f")
+fmtnum  (class=conversion #args=2) Convert int/float/bool to string using printf-style format string (https://pkg.go.dev/fmt), e.g. '$s = fmtnum($n, "%08d")' or '$t = fmtnum($n, "%.6e")'. Miller-specific extension: "%_d" and "%_f" for comma-separated thousands. This function recurses on array and map values.
+Examples:
+$y = fmtnum($x, "%.6f")
+$o = fmtnum($n, "%d")
+$o = fmtnum($n, "%12d")
+$y = fmtnum($x, "%.6_f")
+$o = fmtnum($n, "%_d")
+$o = fmtnum($n, "%12_d")
 </pre>
 
 

man/manpage.txt

@@ -2395,9 +2395,14 @@
        $* = fmtifnum($*, "%.6f") formats numeric fields in the current record, leaving non-numeric ones alone
 
    1mfmtnum0m
-        (class=conversion #args=2) Convert int/float/bool to string using printf-style format string (https://pkg.go.dev/fmt), e.g. '$s = fmtnum($n, "%08d")' or '$t = fmtnum($n, "%.6e")'. This function recurses on array and map values.
-       Example:
-       $x = fmtnum($x, "%.6f")
+        (class=conversion #args=2) Convert int/float/bool to string using printf-style format string (https://pkg.go.dev/fmt), e.g. '$s = fmtnum($n, "%08d")' or '$t = fmtnum($n, "%.6e")'. Miller-specific extension: "%_d" and "%_f" for comma-separated thousands. This function recurses on array and map values.
+       Examples:
+       $y = fmtnum($x, "%.6f")
+       $o = fmtnum($n, "%d")
+       $o = fmtnum($n, "%12d")
+       $y = fmtnum($x, "%.6_f")
+       $o = fmtnum($n, "%_d")
+       $o = fmtnum($n, "%12_d")
 
    1mfold0m
         (class=higher-order-functions #args=3) Given a map or array as first argument and a function as second argument, accumulates entries into a final output -- for example, sum or product. For arrays, the function should take two arguments, for accumulated value and array element. For maps, it should take four arguments, for accumulated key and value, and map-element key and value; it should return the updated accumulator as a new key-value pair (i.e. a single-entry map). The start value for the accumulator is taken from the third argument.

man/mlr.1

@@ -3338,9 +3338,14 @@ $* = fmtifnum($*, "%.6f") formats numeric fields in the current record, leaving
 .RS 0
 .\}
 .nf
- (class=conversion #args=2) Convert int/float/bool to string using printf-style format string (https://pkg.go.dev/fmt), e.g. '$s = fmtnum($n, "%08d")' or '$t = fmtnum($n, "%.6e")'. This function recurses on array and map values.
-Example:
-$x = fmtnum($x, "%.6f")
+ (class=conversion #args=2) Convert int/float/bool to string using printf-style format string (https://pkg.go.dev/fmt), e.g. '$s = fmtnum($n, "%08d")' or '$t = fmtnum($n, "%.6e")'. Miller-specific extension: "%_d" and "%_f" for comma-separated thousands. This function recurses on array and map values.
+Examples:
+$y = fmtnum($x, "%.6f")
+$o = fmtnum($n, "%d")
+$o = fmtnum($n, "%12d")
+$y = fmtnum($x, "%.6_f")
+$o = fmtnum($n, "%_d")
+$o = fmtnum($n, "%12_d")
 .fi
 .if n \{\
 .RE

pkg/dsl/cst/builtin_function_manager.go

@@ -2000,10 +2000,15 @@ Note that NaN has the property that NaN != NaN, so you need 'is_nan(x)' rather t
 			name:  "fmtnum",
 			class: FUNC_CLASS_CONVERSION,
 			help: `Convert int/float/bool to string using printf-style format string (https://pkg.go.dev/fmt), e.g.
-'$s = fmtnum($n, "%08d")' or '$t = fmtnum($n, "%.6e")'. This function recurses on array and map values.`,
+'$s = fmtnum($n, "%08d")' or '$t = fmtnum($n, "%.6e")'. Miller-specific extension: "%_d" and "%_f" for comma-separated thousands. This function recurses on array and map values.`,
 			binaryFunc: bifs.BIF_fmtnum,
 			examples: []string{
-				`$x = fmtnum($x, "%.6f")`,
+				`$y = fmtnum($x, "%.6f")`,
+				`$o = fmtnum($n, "%d")`,
+				`$o = fmtnum($n, "%12d")`,
+				`$y = fmtnum($x, "%.6_f")`,
+				`$o = fmtnum($n, "%_d")`,
+				`$o = fmtnum($n, "%12_d")`,
 			},
 		},
 

pkg/mlrval/mlrval_format.go

@@ -2,8 +2,12 @@ package mlrval
 
 import (
 	"fmt"
+	"os"
 	"strconv"
 	"strings"
+
+	"golang.org/x/text/language"
+	"golang.org/x/text/message"
 )
 
 //----------------------------------------------------------------
@@ -103,16 +107,26 @@ func newFormatter(
 	goFormatString = strings.ReplaceAll(goFormatString, "le", "e")
 	goFormatString = strings.ReplaceAll(goFormatString, "lg", "g")
 
-	// MIller 5 and below required C format strings compatible with 64-bit ints
+	// Miller 5 and below required C format strings compatible with 64-bit ints
 	// and double-precision floats: e.g. "%08lld" and "%9.6lf". For Miller 6,
-	// We must still accept these for backward compatibility.
+	// we must still accept these for backward compatibility.
+	if strings.HasSuffix(goFormatString, "_d") {
+		// Special sub-case of "d"; must be checked first
+		n := len(goFormatString)
+		return newFormatterToSeparatedInt(goFormatString[:n-2] + "d"), nil
+	}
 	if strings.HasSuffix(goFormatString, "d") {
 		return newFormatterToInt(goFormatString), nil
 	}
 	if strings.HasSuffix(goFormatString, "x") {
 		return newFormatterToInt(goFormatString), nil
 	}
 
+	if strings.HasSuffix(goFormatString, "_f") {
+		// Special sub-case of "f"; must be checked first
+		n := len(goFormatString)
+		return newFormatterToSeparatedFloat(goFormatString[:n-2] + "f"), nil
+	}
 	if strings.HasSuffix(goFormatString, "f") {
 		return newFormatterToFloat(goFormatString), nil
 	}
@@ -164,6 +178,81 @@ func (formatter *formatterToFloat) FormatFloat(floatValue float64) string {
 
 // ----------------------------------------------------------------
 
+func getLanguageTag() language.Tag {
+	v, ok := os.LookupEnv("LANG")
+	if ok {
+		return language.Make(v)
+	} else {
+		return language.Make("en")
+	}
+}
+
+// ----------------------------------------------------------------
+
+type formatterToSeparatedInt struct {
+	goFormatString string
+	printer        *message.Printer
+}
+
+func newFormatterToSeparatedInt(goFormatString string) IFormatter {
+	return &formatterToSeparatedInt{
+		goFormatString: goFormatString,
+		printer:        message.NewPrinter(getLanguageTag()),
+	}
+}
+
+func (formatter *formatterToSeparatedInt) Format(mv *Mlrval) *Mlrval {
+	intValue, isInt := mv.GetIntValue()
+	if isInt {
+		formatted := formatter.printer.Sprintf(formatter.goFormatString, intValue)
+		return TryFromIntString(formatted)
+	}
+	floatValue, isFloat := mv.GetFloatValue()
+	if isFloat {
+		formatted := formatter.printer.Sprintf(formatter.goFormatString, int(floatValue))
+		return TryFromIntString(formatted)
+	}
+	return mv
+}
+
+func (formatter *formatterToSeparatedInt) FormatFloat(floatValue float64) string {
+	return formatter.printer.Sprintf(formatter.goFormatString, int(floatValue))
+}
+
+// ----------------------------------------------------------------
+
+type formatterToSeparatedFloat struct {
+	goFormatString string
+	printer        *message.Printer
+}
+
+func newFormatterToSeparatedFloat(goFormatString string) IFormatter {
+	return &formatterToSeparatedFloat{
+		goFormatString: goFormatString,
+		printer:        message.NewPrinter(getLanguageTag()),
+	}
+}
+
+func (formatter *formatterToSeparatedFloat) Format(mv *Mlrval) *Mlrval {
+	floatValue, isFloat := mv.GetFloatValue()
+	if isFloat {
+		formatted := formatter.printer.Sprintf(formatter.goFormatString, floatValue)
+		return TryFromFloatString(formatted)
+	}
+	intValue, isInt := mv.GetIntValue()
+	if isInt {
+		formatted := formatter.printer.Sprintf(formatter.goFormatString, float64(intValue))
+		return TryFromFloatString(formatted)
+	}
+	return mv
+}
+
+func (formatter *formatterToSeparatedFloat) FormatFloat(floatValue float64) string {
+	return formatter.printer.Sprintf(formatter.goFormatString, floatValue)
+}
+
+// ----------------------------------------------------------------
+
 type formatterToInt struct {
 	goFormatString string
 }