zlib_how.html revision 1.1
1 1.1 christos <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" 2 1.1 christos "http://www.w3.org/TR/REC-html40/loose.dtd"> 3 1.1 christos <html> 4 1.1 christos <head> 5 1.1 christos <meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1"> 6 1.1 christos <title>zlib Usage Example</title> 7 1.1 christos <!-- Copyright (c) 2004 Mark Adler. --> 8 1.1 christos </head> 9 1.1 christos <body bgcolor="#FFFFFF" text="#000000" link="#0000FF" vlink="#00A000"> 10 1.1 christos <h2 align="center"> zlib Usage Example </h2> 11 1.1 christos We often get questions about how the <tt>deflate()</tt> and <tt>inflate()</tt> functions should be used. 12 1.1 christos Users wonder when they should provide more input, when they should use more output, 13 1.1 christos what to do with a <tt>Z_BUF_ERROR</tt>, how to make sure the process terminates properly, and 14 1.1 christos so on. So for those who have read <tt>zlib.h</tt> (a few times), and 15 1.1 christos would like further edification, below is an annotated example in C of simple routines to compress and decompress 16 1.1 christos from an input file to an output file using <tt>deflate()</tt> and <tt>inflate()</tt> respectively. The 17 1.1 christos annotations are interspersed between lines of the code. So please read between the lines. 18 1.1 christos We hope this helps explain some of the intricacies of <em>zlib</em>. 19 1.1 christos <p> 20 1.1 christos Without further adieu, here is the program <a href="zpipe.c"><tt>zpipe.c</tt></a>: 21 1.1 christos <pre><b> 22 1.1 christos /* zpipe.c: example of proper use of zlib's inflate() and deflate() 23 1.1 christos Not copyrighted -- provided to the public domain 24 1.1 christos Version 1.2 9 November 2004 Mark Adler */ 25 1.1 christos 26 1.1 christos /* Version history: 27 1.1 christos 1.0 30 Oct 2004 First version 28 1.1 christos 1.1 8 Nov 2004 Add void casting for unused return values 29 1.1 christos Use switch statement for inflate() return values 30 1.1 christos 1.2 9 Nov 2004 Add assertions to document zlib guarantees 31 1.1 christos */ 32 1.1 christos </b></pre><!-- --> 33 1.1 christos We now include the header files for the required definitions. From 34 1.1 christos <tt>stdio.h</tt> we use <tt>fopen()</tt>, <tt>fread()</tt>, <tt>fwrite()</tt>, 35 1.1 christos <tt>feof()</tt>, <tt>ferror()</tt>, and <tt>fclose()</tt> for file i/o, and 36 1.1 christos <tt>fputs()</tt> for error messages. From <tt>string.h</tt> we use 37 1.1 christos <tt>strcmp()</tt> for command line argument processing. 38 1.1 christos From <tt>assert.h</tt> we use the <tt>assert()</tt> macro. 39 1.1 christos From <tt>zlib.h</tt> 40 1.1 christos we use the basic compression functions <tt>deflateInit()</tt>, 41 1.1 christos <tt>deflate()</tt>, and <tt>deflateEnd()</tt>, and the basic decompression 42 1.1 christos functions <tt>inflateInit()</tt>, <tt>inflate()</tt>, and 43 1.1 christos <tt>inflateEnd()</tt>. 44 1.1 christos <pre><b> 45 1.1 christos #include <stdio.h> 46 1.1 christos #include <string.h> 47 1.1 christos #include <assert.h> 48 1.1 christos #include "zlib.h" 49 1.1 christos </b></pre><!-- --> 50 1.1 christos <tt>CHUNK</tt> is simply the buffer size for feeding data to and pulling data 51 1.1 christos from the <em>zlib</em> routines. Larger buffer sizes would be more efficient, 52 1.1 christos especially for <tt>inflate()</tt>. If the memory is available, buffers sizes 53 1.1 christos on the order of 128K or 256K bytes should be used. 54 1.1 christos <pre><b> 55 1.1 christos #define CHUNK 16384 56 1.1 christos </b></pre><!-- --> 57 1.1 christos The <tt>def()</tt> routine compresses data from an input file to an output file. The output data 58 1.1 christos will be in the <em>zlib</em> format, which is different from the <em>gzip</em> or <em>zip</em> 59 1.1 christos formats. The <em>zlib</em> format has a very small header of only two bytes to identify it as 60 1.1 christos a <em>zlib</em> stream and to provide decoding information, and a four-byte trailer with a fast 61 1.1 christos check value to verify the integrity of the uncompressed data after decoding. 62 1.1 christos <pre><b> 63 1.1 christos /* Compress from file source to file dest until EOF on source. 64 1.1 christos def() returns Z_OK on success, Z_MEM_ERROR if memory could not be 65 1.1 christos allocated for processing, Z_STREAM_ERROR if an invalid compression 66 1.1 christos level is supplied, Z_VERSION_ERROR if the version of zlib.h and the 67 1.1 christos version of the library linked do not match, or Z_ERRNO if there is 68 1.1 christos an error reading or writing the files. */ 69 1.1 christos int def(FILE *source, FILE *dest, int level) 70 1.1 christos { 71 1.1 christos </b></pre> 72 1.1 christos Here are the local variables for <tt>def()</tt>. <tt>ret</tt> will be used for <em>zlib</em> 73 1.1 christos return codes. <tt>flush</tt> will keep track of the current flushing state for <tt>deflate()</tt>, 74 1.1 christos which is either no flushing, or flush to completion after the end of the input file is reached. 75 1.1 christos <tt>have</tt> is the amount of data returned from <tt>deflate()</tt>. The <tt>strm</tt> structure 76 1.1 christos is used to pass information to and from the <em>zlib</em> routines, and to maintain the 77 1.1 christos <tt>deflate()</tt> state. <tt>in</tt> and <tt>out</tt> are the input and output buffers for 78 1.1 christos <tt>deflate()</tt>. 79 1.1 christos <pre><b> 80 1.1 christos int ret, flush; 81 1.1 christos unsigned have; 82 1.1 christos z_stream strm; 83 1.1 christos char in[CHUNK]; 84 1.1 christos char out[CHUNK]; 85 1.1 christos </b></pre><!-- --> 86 1.1 christos The first thing we do is to initialize the <em>zlib</em> state for compression using 87 1.1 christos <tt>deflateInit()</tt>. This must be done before the first use of <tt>deflate()</tt>. 88 1.1 christos The <tt>zalloc</tt>, <tt>zfree</tt>, and <tt>opaque</tt> fields in the <tt>strm</tt> 89 1.1 christos structure must be initialized before calling <tt>deflateInit()</tt>. Here they are 90 1.1 christos set to the <em>zlib</em> constant <tt>Z_NULL</tt> to request that <em>zlib</em> use 91 1.1 christos the default memory allocation routines. An application may also choose to provide 92 1.1 christos custom memory allocation routines here. <tt>deflateInit()</tt> will allocate on the 93 1.1 christos order of 256K bytes for the internal state. 94 1.1 christos (See <a href="zlib_tech.html"><em>zlib Technical Details</em></a>.) 95 1.1 christos <p> 96 1.1 christos <tt>deflateInit()</tt> is called with a pointer to the structure to be initialized and 97 1.1 christos the compression level, which is an integer in the range of -1 to 9. Lower compression 98 1.1 christos levels result in faster execution, but less compression. Higher levels result in 99 1.1 christos greater compression, but slower execution. The <em>zlib</em> constant Z_DEFAULT_COMPRESSION, 100 1.1 christos equal to -1, 101 1.1 christos provides a good compromise between compression and speed and is equivalent to level 6. 102 1.1 christos Level 0 actually does no compression at all, and in fact expands the data slightly to produce 103 1.1 christos the <em>zlib</em> format (it is not a byte-for-byte copy of the input). 104 1.1 christos More advanced applications of <em>zlib</em> 105 1.1 christos may use <tt>deflateInit2()</tt> here instead. Such an application may want to reduce how 106 1.1 christos much memory will be used, at some price in compression. Or it may need to request a 107 1.1 christos <em>gzip</em> header and trailer instead of a <em>zlib</em> header and trailer, or raw 108 1.1 christos encoding with no header or trailer at all. 109 1.1 christos <p> 110 1.1 christos We must check the return value of <tt>deflateInit()</tt> against the <em>zlib</em> constant 111 1.1 christos <tt>Z_OK</tt> to make sure that it was able to 112 1.1 christos allocate memory for the internal state, and that the provided arguments were valid. 113 1.1 christos <tt>deflateInit()</tt> will also check that the version of <em>zlib</em> that the <tt>zlib.h</tt> 114 1.1 christos file came from matches the version of <em>zlib</em> actually linked with the program. This 115 1.1 christos is especially important for environments in which <em>zlib</em> is a shared library. 116 1.1 christos <p> 117 1.1 christos Note that an application can initialize multiple, independent <em>zlib</em> streams, which can 118 1.1 christos operate in parallel. The state information maintained in the structure allows the <em>zlib</em> 119 1.1 christos routines to be reentrant. 120 1.1 christos <pre><b> 121 1.1 christos /* allocate deflate state */ 122 1.1 christos strm.zalloc = Z_NULL; 123 1.1 christos strm.zfree = Z_NULL; 124 1.1 christos strm.opaque = Z_NULL; 125 1.1 christos ret = deflateInit(&strm, level); 126 1.1 christos if (ret != Z_OK) 127 1.1 christos return ret; 128 1.1 christos </b></pre><!-- --> 129 1.1 christos With the pleasantries out of the way, now we can get down to business. The outer <tt>do</tt>-loop 130 1.1 christos reads all of the input file and exits at the bottom of the loop once end-of-file is reached. 131 1.1 christos This loop contains the only call of <tt>deflate()</tt>. So we must make sure that all of the 132 1.1 christos input data has been processed and that all of the output data has been generated and consumed 133 1.1 christos before we fall out of the loop at the bottom. 134 1.1 christos <pre><b> 135 1.1 christos /* compress until end of file */ 136 1.1 christos do { 137 1.1 christos </b></pre> 138 1.1 christos We start off by reading data from the input file. The number of bytes read is put directly 139 1.1 christos into <tt>avail_in</tt>, and a pointer to those bytes is put into <tt>next_in</tt>. We also 140 1.1 christos check to see if end-of-file on the input has been reached. If we are at the end of file, then <tt>flush</tt> is set to the 141 1.1 christos <em>zlib</em> constant <tt>Z_FINISH</tt>, which is later passed to <tt>deflate()</tt> to 142 1.1 christos indicate that this is the last chunk of input data to compress. We need to use <tt>feof()</tt> 143 1.1 christos to check for end-of-file as opposed to seeing if fewer than <tt>CHUNK</tt> bytes have been read. The 144 1.1 christos reason is that if the input file length is an exact multiple of <tt>CHUNK</tt>, we will miss 145 1.1 christos the fact that we got to the end-of-file, and not know to tell <tt>deflate()</tt> to finish 146 1.1 christos up the compressed stream. If we are not yet at the end of the input, then the <em>zlib</em> 147 1.1 christos constant <tt>Z_NO_FLUSH</tt> will be passed to <tt>deflate</tt> to indicate that we are still 148 1.1 christos in the middle of the uncompressed data. 149 1.1 christos <p> 150 1.1 christos If there is an error in reading from the input file, the process is aborted with 151 1.1 christos <tt>deflateEnd()</tt> being called to free the allocated <em>zlib</em> state before returning 152 1.1 christos the error. We wouldn't want a memory leak, now would we? <tt>deflateEnd()</tt> can be called 153 1.1 christos at any time after the state has been initialized. Once that's done, <tt>deflateInit()</tt> (or 154 1.1 christos <tt>deflateInit2()</tt>) would have to be called to start a new compression process. There is 155 1.1 christos no point here in checking the <tt>deflateEnd()</tt> return code. The deallocation can't fail. 156 1.1 christos <pre><b> 157 1.1 christos strm.avail_in = fread(in, 1, CHUNK, source); 158 1.1 christos if (ferror(source)) { 159 1.1 christos (void)deflateEnd(&strm); 160 1.1 christos return Z_ERRNO; 161 1.1 christos } 162 1.1 christos flush = feof(source) ? Z_FINISH : Z_NO_FLUSH; 163 1.1 christos strm.next_in = in; 164 1.1 christos </b></pre><!-- --> 165 1.1 christos The inner <tt>do</tt>-loop passes our chunk of input data to <tt>deflate()</tt>, and then 166 1.1 christos keeps calling <tt>deflate()</tt> until it is done producing output. Once there is no more 167 1.1 christos new output, <tt>deflate()</tt> is guaranteed to have consumed all of the input, i.e., 168 1.1 christos <tt>avail_in</tt> will be zero. 169 1.1 christos <pre><b> 170 1.1 christos /* run deflate() on input until output buffer not full, finish 171 1.1 christos compression if all of source has been read in */ 172 1.1 christos do { 173 1.1 christos </b></pre> 174 1.1 christos Output space is provided to <tt>deflate()</tt> by setting <tt>avail_out</tt> to the number 175 1.1 christos of available output bytes and <tt>next_out</tt> to a pointer to that space. 176 1.1 christos <pre><b> 177 1.1 christos strm.avail_out = CHUNK; 178 1.1 christos strm.next_out = out; 179 1.1 christos </b></pre> 180 1.1 christos Now we call the compression engine itself, <tt>deflate()</tt>. It takes as many of the 181 1.1 christos <tt>avail_in</tt> bytes at <tt>next_in</tt> as it can process, and writes as many as 182 1.1 christos <tt>avail_out</tt> bytes to <tt>next_out</tt>. Those counters and pointers are then 183 1.1 christos updated past the input data consumed and the output data written. It is the amount of 184 1.1 christos output space available that may limit how much input is consumed. 185 1.1 christos Hence the inner loop to make sure that 186 1.1 christos all of the input is consumed by providing more output space each time. Since <tt>avail_in</tt> 187 1.1 christos and <tt>next_in</tt> are updated by <tt>deflate()</tt>, we don't have to mess with those 188 1.1 christos between <tt>deflate()</tt> calls until it's all used up. 189 1.1 christos <p> 190 1.1 christos The parameters to <tt>deflate()</tt> are a pointer to the <tt>strm</tt> structure containing 191 1.1 christos the input and output information and the internal compression engine state, and a parameter 192 1.1 christos indicating whether and how to flush data to the output. Normally <tt>deflate</tt> will consume 193 1.1 christos several K bytes of input data before producing any output (except for the header), in order 194 1.1 christos to accumulate statistics on the data for optimum compression. It will then put out a burst of 195 1.1 christos compressed data, and proceed to consume more input before the next burst. Eventually, 196 1.1 christos <tt>deflate()</tt> 197 1.1 christos must be told to terminate the stream, complete the compression with provided input data, and 198 1.1 christos write out the trailer check value. <tt>deflate()</tt> will continue to compress normally as long 199 1.1 christos as the flush parameter is <tt>Z_NO_FLUSH</tt>. Once the <tt>Z_FINISH</tt> parameter is provided, 200 1.1 christos <tt>deflate()</tt> will begin to complete the compressed output stream. However depending on how 201 1.1 christos much output space is provided, <tt>deflate()</tt> may have to be called several times until it 202 1.1 christos has provided the complete compressed stream, even after it has consumed all of the input. The flush 203 1.1 christos parameter must continue to be <tt>Z_FINISH</tt> for those subsequent calls. 204 1.1 christos <p> 205 1.1 christos There are other values of the flush parameter that are used in more advanced applications. You can 206 1.1 christos force <tt>deflate()</tt> to produce a burst of output that encodes all of the input data provided 207 1.1 christos so far, even if it wouldn't have otherwise, for example to control data latency on a link with 208 1.1 christos compressed data. You can also ask that <tt>deflate()</tt> do that as well as erase any history up to 209 1.1 christos that point so that what follows can be decompressed independently, for example for random access 210 1.1 christos applications. Both requests will degrade compression by an amount depending on how often such 211 1.1 christos requests are made. 212 1.1 christos <p> 213 1.1 christos <tt>deflate()</tt> has a return value that can indicate errors, yet we do not check it here. Why 214 1.1 christos not? Well, it turns out that <tt>deflate()</tt> can do no wrong here. Let's go through 215 1.1 christos <tt>deflate()</tt>'s return values and dispense with them one by one. The possible values are 216 1.1 christos <tt>Z_OK</tt>, <tt>Z_STREAM_END</tt>, <tt>Z_STREAM_ERROR</tt>, or <tt>Z_BUF_ERROR</tt>. <tt>Z_OK</tt> 217 1.1 christos is, well, ok. <tt>Z_STREAM_END</tt> is also ok and will be returned for the last call of 218 1.1 christos <tt>deflate()</tt>. This is already guaranteed by calling <tt>deflate()</tt> with <tt>Z_FINISH</tt> 219 1.1 christos until it has no more output. <tt>Z_STREAM_ERROR</tt> is only possible if the stream is not 220 1.1 christos initialized properly, but we did initialize it properly. There is no harm in checking for 221 1.1 christos <tt>Z_STREAM_ERROR</tt> here, for example to check for the possibility that some 222 1.1 christos other part of the application inadvertently clobbered the memory containing the <em>zlib</em> state. 223 1.1 christos <tt>Z_BUF_ERROR</tt> will be explained further below, but 224 1.1 christos suffice it to say that this is simply an indication that <tt>deflate()</tt> could not consume 225 1.1 christos more input or produce more output. <tt>deflate()</tt> can be called again with more output space 226 1.1 christos or more available input, which it will be in this code. 227 1.1 christos <pre><b> 228 1.1 christos ret = deflate(&strm, flush); /* no bad return value */ 229 1.1 christos assert(ret != Z_STREAM_ERROR); /* state not clobbered */ 230 1.1 christos </b></pre> 231 1.1 christos Now we compute how much output <tt>deflate()</tt> provided on the last call, which is the 232 1.1 christos difference between how much space was provided before the call, and how much output space 233 1.1 christos is still available after the call. Then that data, if any, is written to the output file. 234 1.1 christos We can then reuse the output buffer for the next call of <tt>deflate()</tt>. Again if there 235 1.1 christos is a file i/o error, we call <tt>deflateEnd()</tt> before returning to avoid a memory leak. 236 1.1 christos <pre><b> 237 1.1 christos have = CHUNK - strm.avail_out; 238 1.1 christos if (fwrite(out, 1, have, dest) != have || ferror(dest)) { 239 1.1 christos (void)deflateEnd(&strm); 240 1.1 christos return Z_ERRNO; 241 1.1 christos } 242 1.1 christos </b></pre> 243 1.1 christos The inner <tt>do</tt>-loop is repeated until the last <tt>deflate()</tt> call fails to fill the 244 1.1 christos provided output buffer. Then we know that <tt>deflate()</tt> has done as much as it can with 245 1.1 christos the provided input, and that all of that input has been consumed. We can then fall out of this 246 1.1 christos loop and reuse the input buffer. 247 1.1 christos <p> 248 1.1 christos The way we tell that <tt>deflate()</tt> has no more output is by seeing that it did not fill 249 1.1 christos the output buffer, leaving <tt>avail_out</tt> greater than zero. However suppose that 250 1.1 christos <tt>deflate()</tt> has no more output, but just so happened to exactly fill the output buffer! 251 1.1 christos <tt>avail_out</tt> is zero, and we can't tell that <tt>deflate()</tt> has done all it can. 252 1.1 christos As far as we know, <tt>deflate()</tt> 253 1.1 christos has more output for us. So we call it again. But now <tt>deflate()</tt> produces no output 254 1.1 christos at all, and <tt>avail_out</tt> remains unchanged as <tt>CHUNK</tt>. That <tt>deflate()</tt> call 255 1.1 christos wasn't able to do anything, either consume input or produce output, and so it returns 256 1.1 christos <tt>Z_BUF_ERROR</tt>. (See, I told you I'd cover this later.) However this is not a problem at 257 1.1 christos all. Now we finally have the desired indication that <tt>deflate()</tt> is really done, 258 1.1 christos and so we drop out of the inner loop to provide more input to <tt>deflate()</tt>. 259 1.1 christos <p> 260 1.1 christos With <tt>flush</tt> set to <tt>Z_FINISH</tt>, this final set of <tt>deflate()</tt> calls will 261 1.1 christos complete the output stream. Once that is done, subsequent calls of <tt>deflate()</tt> would return 262 1.1 christos <tt>Z_STREAM_ERROR</tt> if the flush parameter is not <tt>Z_FINISH</tt>, and do no more processing 263 1.1 christos until the state is reinitialized. 264 1.1 christos <p> 265 1.1 christos Some applications of <em>zlib</em> have two loops that call <tt>deflate()</tt> 266 1.1 christos instead of the single inner loop we have here. The first loop would call 267 1.1 christos without flushing and feed all of the data to <tt>deflate()</tt>. The second loop would call 268 1.1 christos <tt>deflate()</tt> with no more 269 1.1 christos data and the <tt>Z_FINISH</tt> parameter to complete the process. As you can see from this 270 1.1 christos example, that can be avoided by simply keeping track of the current flush state. 271 1.1 christos <pre><b> 272 1.1 christos } while (strm.avail_out == 0); 273 1.1 christos assert(strm.avail_in == 0); /* all input will be used */ 274 1.1 christos </b></pre><!-- --> 275 1.1 christos Now we check to see if we have already processed all of the input file. That information was 276 1.1 christos saved in the <tt>flush</tt> variable, so we see if that was set to <tt>Z_FINISH</tt>. If so, 277 1.1 christos then we're done and we fall out of the outer loop. We're guaranteed to get <tt>Z_STREAM_END</tt> 278 1.1 christos from the last <tt>deflate()</tt> call, since we ran it until the last chunk of input was 279 1.1 christos consumed and all of the output was generated. 280 1.1 christos <pre><b> 281 1.1 christos /* done when last data in file processed */ 282 1.1 christos } while (flush != Z_FINISH); 283 1.1 christos assert(ret == Z_STREAM_END); /* stream will be complete */ 284 1.1 christos </b></pre><!-- --> 285 1.1 christos The process is complete, but we still need to deallocate the state to avoid a memory leak 286 1.1 christos (or rather more like a memory hemorrhage if you didn't do this). Then 287 1.1 christos finally we can return with a happy return value. 288 1.1 christos <pre><b> 289 1.1 christos /* clean up and return */ 290 1.1 christos (void)deflateEnd(&strm); 291 1.1 christos return Z_OK; 292 1.1 christos } 293 1.1 christos </b></pre><!-- --> 294 1.1 christos Now we do the same thing for decompression in the <tt>inf()</tt> routine. <tt>inf()</tt> 295 1.1 christos decompresses what is hopefully a valid <em>zlib</em> stream from the input file and writes the 296 1.1 christos uncompressed data to the output file. Much of the discussion above for <tt>def()</tt> 297 1.1 christos applies to <tt>inf()</tt> as well, so the discussion here will focus on the differences between 298 1.1 christos the two. 299 1.1 christos <pre><b> 300 1.1 christos /* Decompress from file source to file dest until stream ends or EOF. 301 1.1 christos inf() returns Z_OK on success, Z_MEM_ERROR if memory could not be 302 1.1 christos allocated for processing, Z_DATA_ERROR if the deflate data is 303 1.1 christos invalid or incomplete, Z_VERSION_ERROR if the version of zlib.h and 304 1.1 christos the version of the library linked do not match, or Z_ERRNO if there 305 1.1 christos is an error reading or writing the files. */ 306 1.1 christos int inf(FILE *source, FILE *dest) 307 1.1 christos { 308 1.1 christos </b></pre> 309 1.1 christos The local variables have the same functionality as they do for <tt>def()</tt>. The 310 1.1 christos only difference is that there is no <tt>flush</tt> variable, since <tt>inflate()</tt> 311 1.1 christos can tell from the <em>zlib</em> stream itself when the stream is complete. 312 1.1 christos <pre><b> 313 1.1 christos int ret; 314 1.1 christos unsigned have; 315 1.1 christos z_stream strm; 316 1.1 christos char in[CHUNK]; 317 1.1 christos char out[CHUNK]; 318 1.1 christos </b></pre><!-- --> 319 1.1 christos The initialization of the state is the same, except that there is no compression level, 320 1.1 christos of course, and two more elements of the structure are initialized. <tt>avail_in</tt> 321 1.1 christos and <tt>next_in</tt> must be initialized before calling <tt>inflateInit()</tt>. This 322 1.1 christos is because the application has the option to provide the start of the zlib stream in 323 1.1 christos order for <tt>inflateInit()</tt> to have access to information about the compression 324 1.1 christos method to aid in memory allocation. In the current implementation of <em>zlib</em> 325 1.1 christos (up through versions 1.2.x), the method-dependent memory allocations are deferred to the first call of 326 1.1 christos <tt>inflate()</tt> anyway. However those fields must be initialized since later versions 327 1.1 christos of <em>zlib</em> that provide more compression methods may take advantage of this interface. 328 1.1 christos In any case, no decompression is performed by <tt>inflateInit()</tt>, so the 329 1.1 christos <tt>avail_out</tt> and <tt>next_out</tt> fields do not need to be initialized before calling. 330 1.1 christos <p> 331 1.1 christos Here <tt>avail_in</tt> is set to zero and <tt>next_in</tt> is set to <tt>Z_NULL</tt> to 332 1.1 christos indicate that no input data is being provided. 333 1.1 christos <pre><b> 334 1.1 christos /* allocate inflate state */ 335 1.1 christos strm.zalloc = Z_NULL; 336 1.1 christos strm.zfree = Z_NULL; 337 1.1 christos strm.opaque = Z_NULL; 338 1.1 christos strm.avail_in = 0; 339 1.1 christos strm.next_in = Z_NULL; 340 1.1 christos ret = inflateInit(&strm); 341 1.1 christos if (ret != Z_OK) 342 1.1 christos return ret; 343 1.1 christos </b></pre><!-- --> 344 1.1 christos The outer <tt>do</tt>-loop decompresses input until <tt>inflate()</tt> indicates 345 1.1 christos that it has reached the end of the compressed data and has produced all of the uncompressed 346 1.1 christos output. This is in contrast to <tt>def()</tt> which processes all of the input file. 347 1.1 christos If end-of-file is reached before the compressed data self-terminates, then the compressed 348 1.1 christos data is incomplete and an error is returned. 349 1.1 christos <pre><b> 350 1.1 christos /* decompress until deflate stream ends or end of file */ 351 1.1 christos do { 352 1.1 christos </b></pre> 353 1.1 christos We read input data and set the <tt>strm</tt> structure accordingly. If we've reached the 354 1.1 christos end of the input file, then we leave the outer loop and report an error, since the 355 1.1 christos compressed data is incomplete. Note that we may read more data than is eventually consumed 356 1.1 christos by <tt>inflate()</tt>, if the input file continues past the <em>zlib</em> stream. 357 1.1 christos For applications where <em>zlib</em> streams are embedded in other data, this routine would 358 1.1 christos need to be modified to return the unused data, or at least indicate how much of the input 359 1.1 christos data was not used, so the application would know where to pick up after the <em>zlib</em> stream. 360 1.1 christos <pre><b> 361 1.1 christos strm.avail_in = fread(in, 1, CHUNK, source); 362 1.1 christos if (ferror(source)) { 363 1.1 christos (void)inflateEnd(&strm); 364 1.1 christos return Z_ERRNO; 365 1.1 christos } 366 1.1 christos if (strm.avail_in == 0) 367 1.1 christos break; 368 1.1 christos strm.next_in = in; 369 1.1 christos </b></pre><!-- --> 370 1.1 christos The inner <tt>do</tt>-loop has the same function it did in <tt>def()</tt>, which is to 371 1.1 christos keep calling <tt>inflate()</tt> until has generated all of the output it can with the 372 1.1 christos provided input. 373 1.1 christos <pre><b> 374 1.1 christos /* run inflate() on input until output buffer not full */ 375 1.1 christos do { 376 1.1 christos </b></pre> 377 1.1 christos Just like in <tt>def()</tt>, the same output space is provided for each call of <tt>inflate()</tt>. 378 1.1 christos <pre><b> 379 1.1 christos strm.avail_out = CHUNK; 380 1.1 christos strm.next_out = out; 381 1.1 christos </b></pre> 382 1.1 christos Now we run the decompression engine itself. There is no need to adjust the flush parameter, since 383 1.1 christos the <em>zlib</em> format is self-terminating. The main difference here is that there are 384 1.1 christos return values that we need to pay attention to. <tt>Z_DATA_ERROR</tt> 385 1.1 christos indicates that <tt>inflate()</tt> detected an error in the <em>zlib</em> compressed data format, 386 1.1 christos which means that either the data is not a <em>zlib</em> stream to begin with, or that the data was 387 1.1 christos corrupted somewhere along the way since it was compressed. The other error to be processed is 388 1.1 christos <tt>Z_MEM_ERROR</tt>, which can occur since memory allocation is deferred until <tt>inflate()</tt> 389 1.1 christos needs it, unlike <tt>deflate()</tt>, whose memory is allocated at the start by <tt>deflateInit()</tt>. 390 1.1 christos <p> 391 1.1 christos Advanced applications may use 392 1.1 christos <tt>deflateSetDictionary()</tt> to prime <tt>deflate()</tt> with a set of likely data to improve the 393 1.1 christos first 32K or so of compression. This is noted in the <em>zlib</em> header, so <tt>inflate()</tt> 394 1.1 christos requests that that dictionary be provided before it can start to decompress. Without the dictionary, 395 1.1 christos correct decompression is not possible. For this routine, we have no idea what the dictionary is, 396 1.1 christos so the <tt>Z_NEED_DICT</tt> indication is converted to a <tt>Z_DATA_ERROR</tt>. 397 1.1 christos <p> 398 1.1 christos <tt>inflate()</tt> can also return <tt>Z_STREAM_ERROR</tt>, which should not be possible here, 399 1.1 christos but could be checked for as noted above for <tt>def()</tt>. <tt>Z_BUF_ERROR</tt> does not need to be 400 1.1 christos checked for here, for the same reasons noted for <tt>def()</tt>. <tt>Z_STREAM_END</tt> will be 401 1.1 christos checked for later. 402 1.1 christos <pre><b> 403 1.1 christos ret = inflate(&strm, Z_NO_FLUSH); 404 1.1 christos assert(ret != Z_STREAM_ERROR); /* state not clobbered */ 405 1.1 christos switch (ret) { 406 1.1 christos case Z_NEED_DICT: 407 1.1 christos ret = Z_DATA_ERROR; /* and fall through */ 408 1.1 christos case Z_DATA_ERROR: 409 1.1 christos case Z_MEM_ERROR: 410 1.1 christos (void)inflateEnd(&strm); 411 1.1 christos return ret; 412 1.1 christos } 413 1.1 christos </b></pre> 414 1.1 christos The output of <tt>inflate()</tt> is handled identically to that of <tt>deflate()</tt>. 415 1.1 christos <pre><b> 416 1.1 christos have = CHUNK - strm.avail_out; 417 1.1 christos if (fwrite(out, 1, have, dest) != have || ferror(dest)) { 418 1.1 christos (void)inflateEnd(&strm); 419 1.1 christos return Z_ERRNO; 420 1.1 christos } 421 1.1 christos </b></pre> 422 1.1 christos The inner <tt>do</tt>-loop ends when <tt>inflate()</tt> has no more output as indicated 423 1.1 christos by not filling the output buffer, just as for <tt>deflate()</tt>. In this case, we cannot 424 1.1 christos assert that <tt>strm.avail_in</tt> will be zero, since the deflate stream may end before the file 425 1.1 christos does. 426 1.1 christos <pre><b> 427 1.1 christos } while (strm.avail_out == 0); 428 1.1 christos </b></pre><!-- --> 429 1.1 christos The outer <tt>do</tt>-loop ends when <tt>inflate()</tt> reports that it has reached the 430 1.1 christos end of the input <em>zlib</em> stream, has completed the decompression and integrity 431 1.1 christos check, and has provided all of the output. This is indicated by the <tt>inflate()</tt> 432 1.1 christos return value <tt>Z_STREAM_END</tt>. The inner loop is guaranteed to leave <tt>ret</tt> 433 1.1 christos equal to <tt>Z_STREAM_END</tt> if the last chunk of the input file read contained the end 434 1.1 christos of the <em>zlib</em> stream. So if the return value is not <tt>Z_STREAM_END</tt>, the 435 1.1 christos loop continues to read more input. 436 1.1 christos <pre><b> 437 1.1 christos /* done when inflate() says it's done */ 438 1.1 christos } while (ret != Z_STREAM_END); 439 1.1 christos </b></pre><!-- --> 440 1.1 christos At this point, decompression successfully completed, or we broke out of the loop due to no 441 1.1 christos more data being available from the input file. If the last <tt>inflate()</tt> return value 442 1.1 christos is not <tt>Z_STREAM_END</tt>, then the <em>zlib</em> stream was incomplete and a data error 443 1.1 christos is returned. Otherwise, we return with a happy return value. Of course, <tt>inflateEnd()</tt> 444 1.1 christos is called first to avoid a memory leak. 445 1.1 christos <pre><b> 446 1.1 christos /* clean up and return */ 447 1.1 christos (void)inflateEnd(&strm); 448 1.1 christos return ret == Z_STREAM_END ? Z_OK : Z_DATA_ERROR; 449 1.1 christos } 450 1.1 christos </b></pre><!-- --> 451 1.1 christos That ends the routines that directly use <em>zlib</em>. The following routines make this 452 1.1 christos a command-line program by running data through the above routines from <tt>stdin</tt> to 453 1.1 christos <tt>stdout</tt>, and handling any errors reported by <tt>def()</tt> or <tt>inf()</tt>. 454 1.1 christos <p> 455 1.1 christos <tt>zerr()</tt> is used to interpret the possible error codes from <tt>def()</tt> 456 1.1 christos and <tt>inf()</tt>, as detailed in their comments above, and print out an error message. 457 1.1 christos Note that these are only a subset of the possible return values from <tt>deflate()</tt> 458 1.1 christos and <tt>inflate()</tt>. 459 1.1 christos <pre><b> 460 1.1 christos /* report a zlib or i/o error */ 461 1.1 christos void zerr(int ret) 462 1.1 christos { 463 1.1 christos fputs("zpipe: ", stderr); 464 1.1 christos switch (ret) { 465 1.1 christos case Z_ERRNO: 466 1.1 christos if (ferror(stdin)) 467 1.1 christos fputs("error reading stdin\n", stderr); 468 1.1 christos if (ferror(stdout)) 469 1.1 christos fputs("error writing stdout\n", stderr); 470 1.1 christos break; 471 1.1 christos case Z_STREAM_ERROR: 472 1.1 christos fputs("invalid compression level\n", stderr); 473 1.1 christos break; 474 1.1 christos case Z_DATA_ERROR: 475 1.1 christos fputs("invalid or incomplete deflate data\n", stderr); 476 1.1 christos break; 477 1.1 christos case Z_MEM_ERROR: 478 1.1 christos fputs("out of memory\n", stderr); 479 1.1 christos break; 480 1.1 christos case Z_VERSION_ERROR: 481 1.1 christos fputs("zlib version mismatch!\n", stderr); 482 1.1 christos } 483 1.1 christos } 484 1.1 christos </b></pre><!-- --> 485 1.1 christos Here is the <tt>main()</tt> routine used to test <tt>def()</tt> and <tt>inf()</tt>. The 486 1.1 christos <tt>zpipe</tt> command is simply a compression pipe from <tt>stdin</tt> to <tt>stdout</tt>, if 487 1.1 christos no arguments are given, or it is a decompression pipe if <tt>zpipe -d</tt> is used. If any other 488 1.1 christos arguments are provided, no compression or decompression is performed. Instead a usage 489 1.1 christos message is displayed. Examples are <tt>zpipe < foo.txt > foo.txt.z</tt> to compress, and 490 1.1 christos <tt>zpipe -d < foo.txt.z > foo.txt</tt> to decompress. 491 1.1 christos <pre><b> 492 1.1 christos /* compress or decompress from stdin to stdout */ 493 1.1 christos int main(int argc, char **argv) 494 1.1 christos { 495 1.1 christos int ret; 496 1.1 christos 497 1.1 christos /* do compression if no arguments */ 498 1.1 christos if (argc == 1) { 499 1.1 christos ret = def(stdin, stdout, Z_DEFAULT_COMPRESSION); 500 1.1 christos if (ret != Z_OK) 501 1.1 christos zerr(ret); 502 1.1 christos return ret; 503 1.1 christos } 504 1.1 christos 505 1.1 christos /* do decompression if -d specified */ 506 1.1 christos else if (argc == 2 && strcmp(argv[1], "-d") == 0) { 507 1.1 christos ret = inf(stdin, stdout); 508 1.1 christos if (ret != Z_OK) 509 1.1 christos zerr(ret); 510 1.1 christos return ret; 511 1.1 christos } 512 1.1 christos 513 1.1 christos /* otherwise, report usage */ 514 1.1 christos else { 515 1.1 christos fputs("zpipe usage: zpipe [-d] < source > dest\n", stderr); 516 1.1 christos return 1; 517 1.1 christos } 518 1.1 christos } 519 1.1 christos </b></pre> 520 1.1 christos <hr> 521 1.1 christos <i>Copyright (c) 2004 by Mark Adler<br>Last modified 13 November 2004</i> 522 1.1 christos </body> 523 1.1 christos </html> 524
Indexes created Sat Sep 27 14:09:57 GMT 2025