doc/technical/linebuf.txt

   1 linebuf - a dbuf replacement for the New World Order(tm)
   2
   3 By Adrian Chadd <adrian@creative.net.au>
   4
   5
   6 History
   7 -------
   8
   9 I could probably learn the dbuf history, but basically its evil. The
  10 general idea is that a dbuf holds incoming and outgoing data streams.
  11 The trouble is that well.. it was evil. You can check it out by getting
  12 the old src/dbuf.c and include/dbuf.h files if you really want.
  13
  14
  15 Replacement
  16 -----------
  17
  18 The linebuf system is a replacement for the dbuf code. The general idea here
  19 is that the data should be buffered in "lines" rather than just linearly
  20 like in the dbuf code. This lends to easier manipulation at a later date
  21 (think flushing data lines to a socket, and even "sharing" linebufs to
  22 reduce the copying required for one to many delivery.)
  23
  24 The linebuf system is broken into two structures, the buf_head and buf_line .
  25 buf_head contains the buffer information (queue head/tail, length, allocated
  26 length and the write offset for flushing), and buf_line contains the
  27 line buffer information (buffer and various flags.)
  28
  29 linebuf->terminated is *only* set when a CR/LF combination is received.
  30
  31 linebuf->overflow is set if we get more data than we should, and we simply
  32   truncate the incoming data.
  33
  34 linebuf->flushing is set when we are currently writing the buffer. We should
  35   _NEVER_ be appending to a buffer which we're flushing!
  36
  37 When you get a buffer through linebuf_get() or write one through
  38 linebuf_flush(), it will *always* be terminated with a CR/LF (and a NUL if
  39 its a linebuf_get()).
  40
  41
  42 Linebuf manipulation
  43 --------------------
  44
  45 To use a linebuf, you simply stick a buf_head_t in your structure somewhere.
  46 You then use the following routines:
  47
  48 int
  49 linebuf_parse(buf_head_t *bufhead, char *buf, int len)
  50
  51 Parse the given buf. This routine does some complex manipulation:
  52
  53 - if there is an incomplete buffer at the tail, buf is parsed to try and
  54   fill that incomplete buffer
  55 - a buffer is completed by a CR/LF/CRLF/LFCR. It accepts everything purely
  56   because I wanted to be "liberal in what you accept" ..
  57 - If a buffer is terminated, the linebuf is flagged terminated
  58 - If more data is trying to be squeezed into the buffer than space LEFT
  59   in the buffer, we skip to the next "CRLF", and tag the buffer terminated
  60   _and_ overflowed.
  61 - We treat multiple runs of CR/LF/CRLF/LFCR as a single CRLF. This is just
  62   a little extra goody to stop people sending hundreds of "CRLF"s and creating
  63   unnecessary buffers.
  64 - The number of lines parsed is returned (so you can implement per-line flood
  65   protection ..)
  66
  67
  68 void
  69 linebuf_put(buf_head_t *bufhead, char *buf, int len)
  70
  71 Parse the given buf, ASSUMING it is a single buffer line. This is useful
  72 for server-generated messages where you know you have a single line, and
  73 you don't want to go through the overhead of parsing the data just for
  74 this.
  75
  76
  77 int
  78 linebuf_get(buf_head_t *bufhead, char *buf, int maxlen)
  79
  80 Get a single line from the buffer. This removes data from the head of the
  81 buffer. If the first buffer is empty or is not terminated, 0 is returned
  82 which indicates that there is no data to parse. Terminated buffers are
  83 returned (CR/LF/NUL), and the length INCLUDING the CR/LF/NUL is returned.
  84 The buffer is copied and the linebuf is then deallocated.
  85
  86
  87 int
  88 linebuf_flush(int fd, buf_head_t *bufhead)
  89
  90 Attempt to flush some data to the given socket. bufhead->writeofs tracks
  91 where in the head buffer we currently are. If the buffer is not terminated,
  92 -1 is returned with errno == EWOULDBLOCK to simulate a "retry me" condition.
  93 (See TODO..)
  94
  95 linebuf_flush() returns whatever write() returns, and sets (ie doesn't touch
  96 after write()) errno accordingly.
  97
  98
  99 int
 100 linebuf_len(buf_head_t *bufhead)
 101
 102 Return the length of the buffer, in bytes. This should be used when calculating
 103 how big a buffer is for statistics.
 104
 105
 106 int
 107 linebuf_alloclen(buf_head_t *bufhead)
 108
 109 Return how big the *allocated* space is. This is much more suitable for
 110 anti-flood checking, as someone might be sending a whole bunch of 1-byte
 111 linebufs which might not trigger a recvq / sendq limit but might chew up
 112 way too much memory.
 113
 114
 115
 116 Notes
 117 -----
 118
 119 * Remember that the trailing NUL isn't covered in the string length.
 120
 121
 122 Limitations
 123 -----------
 124
 125 * all the buffers are a fixed size - here they are current 513 bytes
 126   (510 bytes + CR/LF/NUL)
 127
 128
 129 TODO
 130 ----
 131
 132 * linebuf_flush() should be changed a little so if the buffer isn't
 133   terminated, we *dont* retry flushing a buffer until we get more data.
 134
 135 * Implement a reference-friendly linebuf to reduce copies ..
 136