doc/technical/linebuf.txt

   1
   2 linebuf - a dbuf replacement for the New World Order(tm)
   3
   4 By Adrian Chadd <adrian@creative.net.au>
   5
   6
   7
   8 History
   9 -------
  10
  11 I could probably learn the dbuf history, but basically its evil. The
  12 general idea is that a dbuf holds incoming and outgoing data streams.
  13 The trouble is that well.. it was evil. You can check it out by getting
  14 the old src/dbuf.c and include/dbuf.h files if you really want.
  15
  16
  17 Replacement
  18 -----------
  19
  20 The linebuf system is a replacement for the dbuf code. The general idea here
  21 is that the data should be buffered in "lines" rather than just linearly
  22 like in the dbuf code. This lends to easier manipulation at a later date
  23 (think flushing data lines to a socket, and even "sharing" linebufs to
  24 reduce the copying required for one to many delivery.)
  25
  26 The linebuf system is broken into two structures, the buf_head and buf_line .
  27 buf_head contains the buffer information (queue head/tail, length, allocated
  28 length and the write offset for flushing), and buf_line contains the
  29 line buffer information (buffer and various flags.)
  30
  31 linebuf->terminated is *only* set when a CR/LF combination is received.
  32
  33 linebuf->overflow is set if we get more data than we should, and we simply
  34   truncate the incoming data.
  35
  36 linebuf->flushing is set when we are currently writing the buffer. We should
  37   _NEVER_ be appending to a buffer which we're flushing!
  38
  39 When you get a buffer through linebuf_get() or write one through
  40 linebuf_flush(), it will *always* be terminated with a CR/LF (and a NUL if
  41 its a linebuf_get()).
  42
  43
  44 Linebuf manipulation
  45 --------------------
  46
  47 To use a linebuf, you simply stick a buf_head_t in your structure somewhere.
  48 You then use the following routines:
  49
  50 int
  51 linebuf_parse(buf_head_t *bufhead, char *buf, int len)
  52
  53 Parse the given buf. This routine does some complex manipulation:
  54
  55 - if there is an incomplete buffer at the tail, buf is parsed to try and
  56   fill that incomplete buffer
  57 - a buffer is completed by a CR/LF/CRLF/LFCR. It accepts everything purely
  58   because I wanted to be "liberal in what you accept" ..
  59 - If a buffer is terminated, the linebuf is flagged terminated
  60 - If more data is trying to be squeezed into the buffer than space LEFT
  61   in the buffer, we skip to the next "CRLF", and tag the buffer terminated
  62   _and_ overflowed.
  63 - We treat multiple runs of CR/LF/CRLF/LFCR as a single CRLF. This is just
  64   a little extra goody to stop people sending hundreds of "CRLF"s and creating
  65   unnecessary buffers.
  66 - The number of lines parsed is returned (so you can implement per-line flood
  67   protection ..)
  68
  69
  70 void
  71 linebuf_put(buf_head_t *bufhead, char *buf, int len)
  72
  73 Parse the given buf, ASSUMING it is a single buffer line. This is useful
  74 for server-generated messages where you know you have a single line, and
  75 you don't want to go through the overhead of parsing the data just for
  76 this.
  77
  78
  79 int
  80 linebuf_get(buf_head_t *bufhead, char *buf, int maxlen)
  81
  82 Get a single line from the buffer. This removes data from the head of the
  83 buffer. If the first buffer is empty or is not terminated, 0 is returned
  84 which indicates that there is no data to parse. Terminated buffers are
  85 returned (CR/LF/NUL), and the length INCLUDING the CR/LF/NUL is returned.
  86 The buffer is copied and the linebuf is then deallocated.
  87
  88
  89 int
  90 linebuf_flush(int fd, buf_head_t *bufhead)
  91
  92 Attempt to flush some data to the given socket. bufhead->writeofs tracks
  93 where in the head buffer we currently are. If the buffer is not terminated,
  94 -1 is returned with errno == EWOULDBLOCK to simulate a "retry me" condition.
  95 (See TODO..)
  96
  97 linebuf_flush() returns whatever write() returns, and sets (ie doesn't touch
  98 after write()) errno accordingly.
  99
 100
 101 int
 102 linebuf_len(buf_head_t *bufhead)
 103
 104 Return the length of the buffer, in bytes. This should be used when calculating
 105 how big a buffer is for statistics.
 106
 107
 108 int
 109 linebuf_alloclen(buf_head_t *bufhead)
 110
 111 Return how big the *allocated* space is. This is much more suitable for
 112 anti-flood checking, as someone might be sending a whole bunch of 1-byte
 113 linebufs which might not trigger a recvq / sendq limit but might chew up
 114 way too much memory.
 115
 116
 117
 118 Notes
 119 -----
 120
 121 * Remember that the trailing NUL isn't covered in the string length.
 122
 123
 124 Limitations
 125 -----------
 126
 127 * all the buffers are a fixed size - here they are current 513 bytes
 128   (510 bytes + CR/LF/NUL)
 129
 130
 131 TODO
 132 ----
 133
 134 * linebuf_flush() should be changed a little so if the buffer isn't
 135   terminated, we *dont* retry flushing a buffer until we get more data.
 136
 137 * Implement a reference-friendly linebuf to reduce copies ..
 138