X-Git-Url: https://diplodocus.org/git/nmh/blobdiff_plain/43ee212f8c79670ecbb55f1f11b53677dd7dd10c..4ac9784480441d2bebe8a2ad9584165bbf2ee345:/sbr/m_getfld.c diff --git a/sbr/m_getfld.c b/sbr/m_getfld.c index 1a1411eb..c99f9e9d 100644 --- a/sbr/m_getfld.c +++ b/sbr/m_getfld.c @@ -159,7 +159,7 @@ names are typically short (~8 char) and the loop that extracts them might terminate on a colon, newline or max width. I considered using a Vax "scanc" to locate the end of the field followed by a - "bcopy" but the routine call overhead on a Vax is too large for this + "memmove" but the routine call overhead on a Vax is too large for this to work on short names. If Berkeley ever makes "inline" part of the C optimiser (so things like "scanc" turn into inline instructions) a change here would be worthwhile. @@ -175,7 +175,7 @@ so message bodies average at least a few hundred characters. Assuming your system uses reasonably sized stdio buffers (1K or more), this routine should be able to remove the body in large - (>500 byte) chunks. The makes the cost of a call to "bcopy" + (>500 byte) chunks. The makes the cost of a call to "memmove" small but there is a premium on checking for the eom in packed maildrops. The eom pattern is always a simple string so we can construct an efficient pattern matcher for it (e.g., a Vax "matchc" @@ -581,6 +581,9 @@ m_getfld (m_getfld_state_t *gstate, char name[NAMESZ], char *buf, int *bufsz, memcpy (buf, name, n - 1); buf[n - 1] = '\n'; buf[n] = '\0'; + /* Indicate this wasn't a header field using a character + that can't appear in a header field. */ + name[0] = ':'; /* The last character read was '\n'. s->bytes_read (and n) include that, but it was not put into the name array in the for loop above. So subtract 1. */