Source file src/fmt/doc.go
1 // Copyright 2009 The Go Authors. All rights reserved. 2 // Use of this source code is governed by a BSD-style 3 // license that can be found in the LICENSE file. 4 5 /* 6 Package fmt implements formatted I/O with functions analogous 7 to C's printf and scanf. The format 'verbs' are derived from C's but 8 are simpler. 9 10 # Printing 11 12 The verbs: 13 14 General: 15 16 %v the value in a default format 17 when printing structs, the plus flag (%+v) adds field names 18 %#v a Go-syntax representation of the value 19 (floating-point infinities and NaNs print as ±Inf and NaN) 20 %T a Go-syntax representation of the type of the value 21 %% a literal percent sign; consumes no value 22 23 Boolean: 24 25 %t the word true or false 26 27 Integer: 28 29 %b base 2 30 %c the character represented by the corresponding Unicode code point 31 %d base 10 32 %o base 8 33 %O base 8 with 0o prefix 34 %q a single-quoted character literal safely escaped with Go syntax. 35 %x base 16, with lower-case letters for a-f 36 %X base 16, with upper-case letters for A-F 37 %U Unicode format: U+1234; same as "U+%04X" 38 39 Floating-point and complex constituents: 40 41 %b decimalless scientific notation with exponent a power of two, 42 in the manner of strconv.FormatFloat with the 'b' format, 43 e.g. -123456p-78 44 %e scientific notation, e.g. -1.234456e+78 45 %E scientific notation, e.g. -1.234456E+78 46 %f decimal point but no exponent, e.g. 123.456 47 %F synonym for %f 48 %g %e for large exponents, %f otherwise. Precision is discussed below. 49 %G %E for large exponents, %F otherwise 50 %x hexadecimal notation (with decimal power of two exponent), e.g. -0x1.23abcp+20 51 %X upper-case hexadecimal notation, e.g. -0X1.23ABCP+20 52 53 The exponent is always a decimal integer. 54 For formats other than %b the exponent is at least two digits. 55 56 String and slice of bytes (treated equivalently with these verbs): 57 58 %s the uninterpreted bytes of the string or slice 59 %q a double-quoted string safely escaped with Go syntax 60 %x base 16, lower-case, two characters per byte 61 %X base 16, upper-case, two characters per byte 62 63 Slice: 64 65 %p address of 0th element in base 16 notation, with leading 0x 66 67 Pointer: 68 69 %p base 16 notation, with leading 0x 70 The %b, %d, %o, %x and %X verbs also work with pointers, 71 formatting the value exactly as if it were an integer. 72 73 The default format for %v is: 74 75 bool: %t 76 int, int8 etc.: %d 77 uint, uint8 etc.: %d, %#x if printed with %#v 78 float32, complex64, etc: %g 79 string: %s 80 chan: %p 81 pointer: %p 82 83 For compound objects, the elements are printed using these rules, recursively, 84 laid out like this: 85 86 struct: {field0 field1 ...} 87 array, slice: [elem0 elem1 ...] 88 maps: map[key1:value1 key2:value2 ...] 89 pointer to above: &{}, &[], &map[] 90 91 Width is specified by an optional decimal number immediately preceding the verb. 92 If absent, the width is whatever is necessary to represent the value. 93 Precision is specified after the (optional) width by a period followed by a 94 decimal number. If no period is present, a default precision is used. 95 A period with no following number specifies a precision of zero. 96 Examples: 97 98 %f default width, default precision 99 %9f width 9, default precision 100 %.2f default width, precision 2 101 %9.2f width 9, precision 2 102 %9.f width 9, precision 0 103 104 Width and precision are measured in units of Unicode code points, 105 that is, runes. (This differs from C's printf where the 106 units are always measured in bytes.) Either or both of the flags 107 may be replaced with the character '*', causing their values to be 108 obtained from the next operand (preceding the one to format), 109 which must be of type int. 110 111 For most values, width is the minimum number of runes to output, 112 padding the formatted form with spaces if necessary. 113 114 For strings, byte slices and byte arrays, however, precision 115 limits the length of the input to be formatted (not the size of 116 the output), truncating if necessary. Normally it is measured in 117 runes, but for these types when formatted with the %x or %X format 118 it is measured in bytes. 119 120 For floating-point values, width sets the minimum width of the field and 121 precision sets the number of places after the decimal, if appropriate, 122 except that for %g/%G precision sets the maximum number of significant 123 digits (trailing zeros are removed). For example, given 12.345 the format 124 %6.3f prints 12.345 while %.3g prints 12.3. The default precision for %e, %f 125 and %#g is 6; for %g it is the smallest number of digits necessary to identify 126 the value uniquely. 127 128 For complex numbers, the width and precision apply to the two 129 components independently and the result is parenthesized, so %f applied 130 to 1.2+3.4i produces (1.200000+3.400000i). 131 132 When formatting a single integer code point or a rune string (type []rune) 133 with %q, invalid Unicode code points are changed to the Unicode replacement 134 character, U+FFFD, as in [strconv.QuoteRune]. 135 136 Other flags: 137 138 '+' always print a sign for numeric values; 139 guarantee ASCII-only output for %q (%+q) 140 '-' pad with spaces on the right rather than the left (left-justify the field) 141 '#' alternate format: add leading 0b for binary (%#b), 0 for octal (%#o), 142 0x or 0X for hex (%#x or %#X); suppress 0x for %p (%#p); 143 for %q, print a raw (backquoted) string if [strconv.CanBackquote] 144 returns true; 145 always print a decimal point for %e, %E, %f, %F, %g and %G; 146 do not remove trailing zeros for %g and %G; 147 write e.g. U+0078 'x' if the character is printable for %U (%#U) 148 ' ' (space) leave a space for elided sign in numbers (% d); 149 put spaces between bytes printing strings or slices in hex (% x, % X) 150 '0' pad with leading zeros rather than spaces; 151 for numbers, this moves the padding after the sign 152 153 Flags are ignored by verbs that do not expect them. 154 For example there is no alternate decimal format, so %#d and %d 155 behave identically. 156 157 For each Printf-like function, there is also a Print function 158 that takes no format and is equivalent to saying %v for every 159 operand. Another variant Println inserts blanks between 160 operands and appends a newline. 161 162 Regardless of the verb, if an operand is an interface value, 163 the internal concrete value is used, not the interface itself. 164 Thus: 165 166 var i interface{} = 23 167 fmt.Printf("%v\n", i) 168 169 will print 23. 170 171 Except when printed using the verbs %T and %p, special 172 formatting considerations apply for operands that implement 173 certain interfaces. In order of application: 174 175 1. If the operand is a [reflect.Value], the operand is replaced by the 176 concrete value that it holds, and printing continues with the next rule. 177 178 2. If an operand implements the [Formatter] interface, it will 179 be invoked. In this case the interpretation of verbs and flags is 180 controlled by that implementation. 181 182 3. If the %v verb is used with the # flag (%#v) and the operand 183 implements the [GoStringer] interface, that will be invoked. 184 185 If the format (which is implicitly %v for [Println] etc.) is valid 186 for a string (%s %q %x %X), or is %v but not %#v, 187 the following two rules apply: 188 189 4. If an operand implements the error interface, the Error method 190 will be invoked to convert the object to a string, which will then 191 be formatted as required by the verb (if any). 192 193 5. If an operand implements method String() string, that method 194 will be invoked to convert the object to a string, which will then 195 be formatted as required by the verb (if any). 196 197 For compound operands such as slices and structs, the format 198 applies to the elements of each operand, recursively, not to the 199 operand as a whole. Thus %q will quote each element of a slice 200 of strings, and %6.2f will control formatting for each element 201 of a floating-point array. 202 203 However, when printing a byte slice with a string-like verb 204 (%s %q %x %X), it is treated identically to a string, as a single item. 205 206 To avoid recursion in cases such as 207 208 type X string 209 func (x X) String() string { return Sprintf("<%s>", x) } 210 211 convert the value before recurring: 212 213 func (x X) String() string { return Sprintf("<%s>", string(x)) } 214 215 Infinite recursion can also be triggered by self-referential data 216 structures, such as a slice that contains itself as an element, if 217 that type has a String method. Such pathologies are rare, however, 218 and the package does not protect against them. 219 220 When printing a struct, fmt cannot and therefore does not invoke 221 formatting methods such as Error or String on unexported fields. 222 223 # Explicit argument indexes 224 225 In [Printf], [Sprintf], and [Fprintf], the default behavior is for each 226 formatting verb to format successive arguments passed in the call. 227 However, the notation [n] immediately before the verb indicates that the 228 nth one-indexed argument is to be formatted instead. The same notation 229 before a '*' for a width or precision selects the argument index holding 230 the value. After processing a bracketed expression [n], subsequent verbs 231 will use arguments n+1, n+2, etc. unless otherwise directed. 232 233 For example, 234 235 fmt.Sprintf("%[2]d %[1]d\n", 11, 22) 236 237 will yield "22 11", while 238 239 fmt.Sprintf("%[3]*.[2]*[1]f", 12.0, 2, 6) 240 241 equivalent to 242 243 fmt.Sprintf("%6.2f", 12.0) 244 245 will yield " 12.00". Because an explicit index affects subsequent verbs, 246 this notation can be used to print the same values multiple times 247 by resetting the index for the first argument to be repeated: 248 249 fmt.Sprintf("%d %d %#[1]x %#x", 16, 17) 250 251 will yield "16 17 0x10 0x11". 252 253 # Format errors 254 255 If an invalid argument is given for a verb, such as providing 256 a string to %d, the generated string will contain a 257 description of the problem, as in these examples: 258 259 Wrong type or unknown verb: %!verb(type=value) 260 Printf("%d", "hi"): %!d(string=hi) 261 Too many arguments: %!(EXTRA type=value) 262 Printf("hi", "guys"): hi%!(EXTRA string=guys) 263 Too few arguments: %!verb(MISSING) 264 Printf("hi%d"): hi%!d(MISSING) 265 Non-int for width or precision: %!(BADWIDTH) or %!(BADPREC) 266 Printf("%*s", 4.5, "hi"): %!(BADWIDTH)hi 267 Printf("%.*s", 4.5, "hi"): %!(BADPREC)hi 268 Invalid or invalid use of argument index: %!(BADINDEX) 269 Printf("%*[2]d", 7): %!d(BADINDEX) 270 Printf("%.[2]d", 7): %!d(BADINDEX) 271 272 All errors begin with the string "%!" followed sometimes 273 by a single character (the verb) and end with a parenthesized 274 description. 275 276 If an Error or String method triggers a panic when called by a 277 print routine, the fmt package reformats the error message 278 from the panic, decorating it with an indication that it came 279 through the fmt package. For example, if a String method 280 calls panic("bad"), the resulting formatted message will look 281 like 282 283 %!s(PANIC=bad) 284 285 The %!s just shows the print verb in use when the failure 286 occurred. If the panic is caused by a nil receiver to an Error, 287 String, or GoString method, however, the output is the undecorated 288 string, "<nil>". 289 290 # Scanning 291 292 An analogous set of functions scans formatted text to yield 293 values. [Scan], [Scanf] and [Scanln] read from [os.Stdin]; [Fscan], 294 [Fscanf] and [Fscanln] read from a specified [io.Reader]; [Sscan], 295 [Sscanf] and [Sscanln] read from an argument string. 296 297 [Scan], [Fscan], [Sscan] treat newlines in the input as spaces. 298 299 [Scanln], [Fscanln] and [Sscanln] stop scanning at a newline and 300 require that the items be followed by a newline or EOF. 301 302 [Scanf], [Fscanf], and [Sscanf] parse the arguments according to a 303 format string, analogous to that of [Printf]. In the text that 304 follows, 'space' means any Unicode whitespace character 305 except newline. 306 307 In the format string, a verb introduced by the % character 308 consumes and parses input; these verbs are described in more 309 detail below. A character other than %, space, or newline in 310 the format consumes exactly that input character, which must 311 be present. A newline with zero or more spaces before it in 312 the format string consumes zero or more spaces in the input 313 followed by a single newline or the end of the input. A space 314 following a newline in the format string consumes zero or more 315 spaces in the input. Otherwise, any run of one or more spaces 316 in the format string consumes as many spaces as possible in 317 the input. Unless the run of spaces in the format string 318 appears adjacent to a newline, the run must consume at least 319 one space from the input or find the end of the input. 320 321 The handling of spaces and newlines differs from that of C's 322 scanf family: in C, newlines are treated as any other space, 323 and it is never an error when a run of spaces in the format 324 string finds no spaces to consume in the input. 325 326 The verbs behave analogously to those of [Printf]. 327 For example, %x will scan an integer as a hexadecimal number, 328 and %v will scan the default representation format for the value. 329 The [Printf] verbs %p and %T and the flags # and + are not implemented. 330 For floating-point and complex values, all valid formatting verbs 331 (%b %e %E %f %F %g %G %x %X and %v) are equivalent and accept 332 both decimal and hexadecimal notation (for example: "2.3e+7", "0x4.5p-8") 333 and digit-separating underscores (for example: "3.14159_26535_89793"). 334 335 Input processed by verbs is implicitly space-delimited: the 336 implementation of every verb except %c starts by discarding 337 leading spaces from the remaining input, and the %s verb 338 (and %v reading into a string) stops consuming input at the first 339 space or newline character. 340 341 The familiar base-setting prefixes 0b (binary), 0o and 0 (octal), 342 and 0x (hexadecimal) are accepted when scanning integers 343 without a format or with the %v verb, as are digit-separating 344 underscores. 345 346 Width is interpreted in the input text but there is no 347 syntax for scanning with a precision (no %5.2f, just %5f). 348 If width is provided, it applies after leading spaces are 349 trimmed and specifies the maximum number of runes to read 350 to satisfy the verb. For example, 351 352 Sscanf(" 1234567 ", "%5s%d", &s, &i) 353 354 will set s to "12345" and i to 67 while 355 356 Sscanf(" 12 34 567 ", "%5s%d", &s, &i) 357 358 will set s to "12" and i to 34. 359 360 In all the scanning functions, a carriage return followed 361 immediately by a newline is treated as a plain newline 362 (\r\n means the same as \n). 363 364 In all the scanning functions, if an operand implements method 365 [Scan] (that is, it implements the [Scanner] interface) that 366 method will be used to scan the text for that operand. Also, 367 if the number of arguments scanned is less than the number of 368 arguments provided, an error is returned. 369 370 All arguments to be scanned must be either pointers to basic 371 types or implementations of the [Scanner] interface. 372 373 Like [Scanf] and [Fscanf], [Sscanf] need not consume its entire input. 374 There is no way to recover how much of the input string [Sscanf] used. 375 376 Note: [Fscan] etc. can read one character (rune) past the input 377 they return, which means that a loop calling a scan routine 378 may skip some of the input. This is usually a problem only 379 when there is no space between input values. If the reader 380 provided to [Fscan] implements ReadRune, that method will be used 381 to read characters. If the reader also implements UnreadRune, 382 that method will be used to save the character and successive 383 calls will not lose data. To attach ReadRune and UnreadRune 384 methods to a reader without that capability, use 385 [bufio.NewReader]. 386 */ 387 package fmt 388