src/liblzma/api/lzma/vli.h

   1 /**
   2  * \file        lzma/vli.h
   3  * \brief       Variable-length integer handling
   4  * \note        Never include this file directly. Use <lzma.h> instead.
   5  *
   6  * See ../lzma.h for information about liblzma as a whole.
   7  *
   8  * In the .xz format, most integers are encoded in a variable-length
   9  * representation, which is sometimes called little endian base-128 encoding.
  10  * This saves space when smaller values are more likely than bigger values.
  11  *
  12  * The encoding scheme encodes seven bits to every byte, using minimum
  13  * number of bytes required to represent the given value. Encodings that use
  14  * non-minimum number of bytes are invalid, thus every integer has exactly
  15  * one encoded representation. The maximum number of bits in a VLI is 63,
  16  * thus the vli argument must be less than or equal to UINT64_MAX / 2. You
  17  * should use LZMA_VLI_MAX for clarity.
  18  */
  19
  20 /*
  21  * Author: Lasse Collin
  22  *
  23  * This file has been put into the public domain.
  24  * You can do whatever you want with this file.
  25  */
  26
  27 #ifndef LZMA_H_INTERNAL
  28 #       error Never include this file directly. Use <lzma.h> instead.
  29 #endif
  30
  31
  32 /**
  33  * \brief       Maximum supported value of a variable-length integer
  34  */
  35 #define LZMA_VLI_MAX (UINT64_MAX / 2)
  36
  37 /**
  38  * \brief       VLI value to denote that the value is unknown
  39  */
  40 #define LZMA_VLI_UNKNOWN UINT64_MAX
  41
  42 /**
  43  * \brief       Maximum supported encoded length of variable length integers
  44  */
  45 #define LZMA_VLI_BYTES_MAX 9
  46
  47 /**
  48  * \brief       VLI constant suffix
  49  */
  50 #define LZMA_VLI_C(n) UINT64_C(n)
  51
  52
  53 /**
  54  * \brief       Variable-length integer type
  55  *
  56  * Valid VLI values are in the range [0, LZMA_VLI_MAX]. Unknown value is
  57  * indicated with LZMA_VLI_UNKNOWN, which is the maximum value of the
  58  * underlying integer type.
  59  *
  60  * lzma_vli will be uint64_t for the foreseeable future. If a bigger size
  61  * is needed in the future, it is guaranteed that 2 * LZMA_VLI_MAX will
  62  * not overflow lzma_vli. This simplifies integer overflow detection.
  63  */
  64 typedef uint64_t lzma_vli;
  65
  66
  67 /**
  68  * \brief       Validate a variable-length integer
  69  *
  70  * This is useful to test that application has given acceptable values
  71  * for example in the uncompressed_size and compressed_size variables.
  72  *
  73  * \return      True if the integer is representable as VLI or if it
  74  *              indicates unknown value. False if the integer cannot be
  75  *              represented as VLI.
  76  */
  77 #define lzma_vli_is_valid(vli) \
  78         ((vli) <= LZMA_VLI_MAX || (vli) == LZMA_VLI_UNKNOWN)
  79
  80
  81 /**
  82  * \brief       Encode a variable-length integer
  83  *
  84  * This function has two modes: single-call and multi-call. Single-call mode
  85  * encodes the whole integer at once; it is an error if the output buffer is
  86  * too small. Multi-call mode saves the position in *vli_pos, and thus it is
  87  * possible to continue encoding if the buffer becomes full before the whole
  88  * integer has been encoded.
  89  *
  90  * \param       vli       Integer to be encoded
  91  * \param[out]  vli_pos   How many VLI-encoded bytes have already been written
  92  *                        out. When starting to encode a new integer in
  93  *                        multi-call mode, *vli_pos must be set to zero.
  94  *                        To use single-call encoding, set vli_pos to NULL.
  95  * \param[out]  out       Beginning of the output buffer
  96  * \param[out]  out_pos   The next byte will be written to out[*out_pos].
  97  * \param       out_size  Size of the out buffer; the first byte into
  98  *                        which no data is written to is out[out_size].
  99  *
 100  * \return      Slightly different return values are used in multi-call and
 101  *              single-call modes.
 102  *
 103  *              Single-call (vli_pos == NULL):
 104  *              - LZMA_OK: Integer successfully encoded.
 105  *              - LZMA_PROG_ERROR: Arguments are not sane. This can be due
 106  *                to too little output space; single-call mode doesn't use
 107  *                LZMA_BUF_ERROR, since the application should have checked
 108  *                the encoded size with lzma_vli_size().
 109  *
 110  *              Multi-call (vli_pos != NULL):
 111  *              - LZMA_OK: So far all OK, but the integer is not
 112  *                completely written out yet.
 113  *              - LZMA_STREAM_END: Integer successfully encoded.
 114  *              - LZMA_BUF_ERROR: No output space was provided.
 115  *              - LZMA_PROG_ERROR: Arguments are not sane.
 116  */
 117 extern LZMA_API(lzma_ret) lzma_vli_encode(lzma_vli vli, size_t *vli_pos,
 118                 uint8_t *out, size_t *out_pos, size_t out_size) lzma_nothrow;
 119
 120
 121 /**
 122  * \brief       Decode a variable-length integer
 123  *
 124  * Like lzma_vli_encode(), this function has single-call and multi-call modes.
 125  *
 126  * \param[out]  vli       Pointer to decoded integer. The decoder will
 127  *                        initialize it to zero when *vli_pos == 0, so
 128  *                        application isn't required to initialize *vli.
 129  * \param[out]  vli_pos   How many bytes have already been decoded. When
 130  *                        starting to decode a new integer in multi-call
 131  *                        mode, *vli_pos must be initialized to zero. To
 132  *                        use single-call decoding, set vli_pos to NULL.
 133  * \param       in        Beginning of the input buffer
 134  * \param[out]  in_pos    The next byte will be read from in[*in_pos].
 135  * \param       in_size   Size of the input buffer; the first byte that
 136  *                        won't be read is in[in_size].
 137  *
 138  * \return      Slightly different return values are used in multi-call and
 139  *              single-call modes.
 140  *
 141  *              Single-call (vli_pos == NULL):
 142  *              - LZMA_OK: Integer successfully decoded.
 143  *              - LZMA_DATA_ERROR: Integer is corrupt. This includes hitting
 144  *                the end of the input buffer before the whole integer was
 145  *                decoded; providing no input at all will use LZMA_DATA_ERROR.
 146  *              - LZMA_PROG_ERROR: Arguments are not sane.
 147  *
 148  *              Multi-call (vli_pos != NULL):
 149  *              - LZMA_OK: So far all OK, but the integer is not
 150  *                completely decoded yet.
 151  *              - LZMA_STREAM_END: Integer successfully decoded.
 152  *              - LZMA_DATA_ERROR: Integer is corrupt.
 153  *              - LZMA_BUF_ERROR: No input was provided.
 154  *              - LZMA_PROG_ERROR: Arguments are not sane.
 155  */
 156 extern LZMA_API(lzma_ret) lzma_vli_decode(lzma_vli *vli, size_t *vli_pos,
 157                 const uint8_t *in, size_t *in_pos, size_t in_size)
 158                 lzma_nothrow;
 159
 160
 161 /**
 162  * \brief       Get the number of bytes required to encode a VLI
 163  *
 164  * \param       vli       Integer whose encoded size is to be determined
 165  *
 166  * \return      Number of bytes on success (1-9). If vli isn't valid,
 167  *              zero is returned.
 168  */
 169 extern LZMA_API(uint32_t) lzma_vli_size(lzma_vli vli)
 170                 lzma_nothrow lzma_attr_pure;