...

...

@@ -11,8 +11,7 @@

11

11

<methodsynopsis>

12

12

<type>string</type><methodname>pack</methodname>

13

13

<methodparam><type>string</type><parameter>format</parameter></methodparam>

14

-

<methodparam choice="opt"><type>mixed</type><parameter>args</parameter></methodparam>

15

-

<methodparam choice="opt"><type>mixed</type><parameter>...</parameter></methodparam>

14

+

<methodparam rep="repeat"><type>mixed</type><parameter>values</parameter></methodparam>

16

15

</methodsynopsis>

17

16

<para>

18

17

Pack given arguments into a binary string according to

...

...

@@ -138,10 +137,26 @@

138

137

<entry>float (machine dependent size and representation)</entry>

139

138

</row>

140

139

<row>

140

+

<entry>g</entry>

141

+

<entry>float (machine dependent size, little endian byte order)</entry>

142

+

</row>

143

+

<row>

144

+

<entry>G</entry>

145

+

<entry>float (machine dependent size, big endian byte order)</entry>

146

+

</row>

147

+

<row>

141

148

<entry>d</entry>

142

149

<entry>double (machine dependent size and representation)</entry>

143

150

</row>

144

151

<row>

152

+

<entry>e</entry>

153

+

<entry>double (machine dependent size, little endian byte order)</entry>

154

+

</row>

155

+

<row>

156

+

<entry>E</entry>

157

+

<entry>double (machine dependent size, big endian byte order)</entry>

158

+

</row>

159

+

<row>

145

160

<entry>x</entry>

146

161

<entry>NUL byte</entry>

147

162

</row>

...

...

@@ -151,7 +166,7 @@

151

166

</row>

152

167

<row>

153

168

<entry>Z</entry>

154

-

<entry>NUL-padded string (new in PHP 5.5)</entry>

169

+

<entry>NUL-padded string</entry>

155

170

</row>

156

171

<row>

157

172

<entry>@</entry>

...

...

@@ -164,7 +179,7 @@

164

179

</listitem>

165

180

</varlistentry>

166

181

<varlistentry>

167

-

<term><parameter>args</parameter></term>

182

+

<term><parameter>values</parameter></term>

168

183

<listitem>

169

184

<para>

170

185

</para>

...

...

@@ -194,16 +209,21 @@

194

209

</thead>

195

210

<tbody>

196

211

<row>

197

-

<entry>5.6.3</entry>

212

+

<entry>8.0.0</entry>

198

213

<entry>

199

-

The "q", "Q", "J" and "P" codes were added to enable working with 64-bit numbers.

214

+

This function no longer returns &false; on failure.

200

215

</entry>

201

216

</row>

202

217

<row>

203

-

<entry>5.5.0</entry>

218

+

<entry>7.2.0</entry>

204

219

<entry>

205

-

The "Z" code was added with equivalent functionality to "a" for Perl

206

-

compatibility.

220

+

<type>float</type> and <type>double</type> types supports both Big Endian and Little Endian.

221

+

</entry>

222

+

</row>

223

+

<row>

224

+

<entry>7.0.15,7.1.1</entry>

225

+

<entry>

226

+

The "e", "E", "g" and "G" codes were added to enable byte order support for float and double.

207

227

</entry>

208

228

</row>

209

229

</tbody>

...

...

@@ -235,25 +255,25 @@ $binarydata = pack("nvc*", 0x1234, 0x5678, 65, 66);

235

255

<refsect1 role="notes">

236

256

&reftitle.notes;

237

257

<caution>

238

-

<para>Note that PHP internally stores <type>integer</type> values as

258

+

<para>Note that PHP internally stores <type>int</type> values as

239

259

signed values of a machine-dependent size (C type <literal>long</literal>).

240

260

Integer literals and operations that yield numbers outside the bounds of the

241

-

<type>integer</type> type will be stored as <type>float</type>.

261

+

<type>int</type> type will be stored as <type>float</type>.

242

262

When packing these floats as integers, they are first cast into the integer

243

263

type. This may or may not result in the desired byte pattern.

244

264

</para>

245

265

<para>The most relevant case is when packing unsigned numbers that would

246

-

be representable with the <type>integer</type> type if it were unsigned.

247

-

In systems where the <type>integer</type> type has a 32-bit size, the cast

248

-

usually results in the same byte pattern as if the <type>integer</type> were

266

+

be representable with the <type>int</type> type if it were unsigned.

267

+

In systems where the <type>int</type> type has a 32-bit size, the cast

268

+

usually results in the same byte pattern as if the <type>int</type> were

249

269

unsigned (although this relies on implementation-defined unsigned to signed

250

270

conversions, as per the C standard). In systems where the

251

-

<type>integer</type> type has 64-bit size, the <type>float</type> most

271

+

<type>int</type> type has 64-bit size, the <type>float</type> most

252

272

likely does not have a mantissa large enough to hold the value without

253

273

loss of precision. If those systems also have a native 64-bit C

254

274

<literal>int</literal> type (most UNIX-like systems don't), the only way to

255

275

use the <literal>I</literal> pack format in the upper range is to create

256

-

<type>integer</type> negative values with the same byte representation

276

+

<type>int</type> negative values with the same byte representation

257

277

as the desired unsigned value.

258

278

</para>

259

279

</caution>

...

...

@@ -269,7 +289,6 @@ $binarydata = pack("nvc*", 0x1234, 0x5678, 65, 66);

269

289

</refsect1>

270

290

271

291

</refentry>

272

-

273

292

<!-- Keep this comment at the end of the file

274

293

Local variables:

275

294

mode: sgml

276

295