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