htnews.c

用于linux和其他unix下面的

    char *date = NULL;				/* Date string		    */
    char *organization = NULL;			/* Organization string	    */
    char *references = NULL;			/* Hrefs for other articles */
    char *newsgroups = NULL;			/* Newsgroups list	    */
    char *followupto = NULL;			/* Followup list	    */
    char *href = NULL;
    char *p = line;
    char *cp;
    CONST char *ccp;
    BOOL done = NO;

    /*
    **	Read in the HEADer of the article.
    **
    **	The header fields are either ignored,
    **	or formatted and put into the text.
    */
    if (!diagnostic && !rawtext) {
	while (!done) {
	    int ich = NEXT_CHAR;
	    *p++ = (char) ich;
	    if (ich == EOF) {
		if (interrupted_in_htgetcharacter) {
		    interrupted_in_htgetcharacter = 0;
		    CTRACE((tfp, "HTNews: Interrupted on read, closing socket %d\n",
				s));
		    NEWS_NETCLOSE(s);
		    s = -1;
		    return(HT_INTERRUPTED);
		}
		abort_socket();		/* End of file, close socket */
		return(HT_LOADED);	/* End of file on response */
	    }
	    if (((char)ich == LF) || (p == &line[LINE_LENGTH])) {
		*--p = '\0';			/* Terminate the string */
		CTRACE((tfp, "H %s\n", line));

		if (line[0] == '\t' || line[0] == ' ') {
		    int i = 0;
		    while (line[i]) {
			if (line[i] == '\t')
			    line[i] = ' ';
			i++;
		    }
		    if (full_line == NULL) {
			StrAllocCopy(full_line, line);
		    } else {
			StrAllocCat(full_line, line);
		    }
		} else {
		    StrAllocCopy(full_line, line);
		}

		if (full_line[0] == '.') {
		    /*
		    **	End of article?
		    */
		    if (UCH(full_line[1]) < ' ') {
			done = YES;
			break;
		    }
		} else if (UCH(full_line[0]) < ' ') {
		    break;		/* End of Header? */

		} else if (match(full_line, "SUBJECT:")) {
		    StrAllocCopy(subject, HTStrip(strchr(full_line,':')+1));
		    decode_mime(subject);
		} else if (match(full_line, "DATE:")) {
		    StrAllocCopy(date, HTStrip(strchr(full_line,':')+1));

		} else if (match(full_line, "ORGANIZATION:")) {
		    StrAllocCopy(organization,
				 HTStrip(strchr(full_line,':')+1));
		    decode_mime(organization);

		} else if (match(full_line, "FROM:")) {
		    StrAllocCopy(from, HTStrip(strchr(full_line,':')+1));
		    decode_mime(from);

		} else if (match(full_line, "REPLY-TO:")) {
		    StrAllocCopy(replyto, HTStrip(strchr(full_line,':')+1));
		    decode_mime(replyto);

		} else if (match(full_line, "NEWSGROUPS:")) {
		    StrAllocCopy(newsgroups, HTStrip(strchr(full_line,':')+1));

		} else if (match(full_line, "REFERENCES:")) {
		    StrAllocCopy(references, HTStrip(strchr(full_line,':')+1));

		} else if (match(full_line, "FOLLOWUP-TO:")) {
		    StrAllocCopy(followupto, HTStrip(strchr(full_line,':')+1));

		} else if (match(full_line, "MESSAGE-ID:")) {
		    char * msgid = HTStrip(full_line+11);
		    if (msgid[0] == '<' && msgid[strlen(msgid)-1] == '>') {
			msgid[strlen(msgid)-1] = '\0';	/* Chop > */
			msgid++;			/* Chop < */
			HTAnchor_setMessageID(thisanchor, msgid);
		    }

		} /* end if match */
		p = line;			/* Restart at beginning */
	    } /* if end of line */
	} /* Loop over characters */
	FREE(full_line);

	START(HTML_HEAD);
	PUTC('\n');
	START(HTML_TITLE);
	if (subject && *subject != '\0')
	    PUTS(subject);
	else
	    PUTS("No Subject");
	END(HTML_TITLE);
	PUTC('\n');
	/*
	**  Put in the owner as a link rel.
	*/
	if (from || replyto) {
	    char *temp = NULL;
	    StrAllocCopy(temp, author_address(replyto ? replyto : from));
	    StrAllocCopy(href,"mailto:");
	    if (strchr(temp, '%') || strchr(temp, '?')) {
		cp = HTEscape(temp, URL_XPALPHAS);
		StrAllocCat(href, cp);
		FREE(cp);
	    } else {
		StrAllocCat(href, temp);
	    }
	    start_link(href, "made");
	    PUTC('\n');
	    FREE(temp);
	}
	END(HTML_HEAD);
	PUTC('\n');

	START(HTML_H1);
	if (subject && *subject != '\0')
	    PUTS(subject);
	else
	    PUTS("No Subject");
	END(HTML_H1);
	PUTC('\n');

	if (subject)
	    FREE(subject);

	START(HTML_DLC);
	PUTC('\n');

	if (from || replyto) {
	    START(HTML_DT);
	    START(HTML_B);
	    PUTS("From:");
	    END(HTML_B);
	    PUTC(' ');
	    if (from)
		PUTS(from);
	    else
		PUTS(replyto);
	    MAYBE_END(HTML_DT);
	    PUTC('\n');

	    if (!replyto)
		StrAllocCopy(replyto, from);
	    START(HTML_DT);
	    START(HTML_B);
	    PUTS("Reply to:");
	    END(HTML_B);
	    PUTC(' ');
	    start_anchor(href);
	    if (*replyto != '<')
		PUTS(author_name(replyto));
	    else
		PUTS(author_address(replyto));
	    END(HTML_A);
	    START(HTML_BR);
	    MAYBE_END(HTML_DT);
	    PUTC('\n');

	    FREE(from);
	    FREE(replyto);
	}

	if (date) {
	    START(HTML_DT);
	    START(HTML_B);
	    PUTS("Date:");
	    END(HTML_B);
	    PUTC(' ');
	    PUTS(date);
	    MAYBE_END(HTML_DT);
	    PUTC('\n');
	    FREE(date);
	}

	if (organization) {
	    START(HTML_DT);
	    START(HTML_B);
	    PUTS("Organization:");
	    END(HTML_B);
	    PUTC(' ');
	    PUTS(organization);
	    MAYBE_END(HTML_DT);
	    PUTC('\n');
	    FREE(organization);
	}

	/* sanitize some headers - kw */
	if (newsgroups &&
	    ((cp = strchr(newsgroups, '/')) ||
	     (cp = strchr(newsgroups, '(')))) {
	    *cp = '\0';
	}
	if (newsgroups && !*newsgroups) {
	    FREE(newsgroups);
	}
	if (followupto &&
	    ((cp = strchr(followupto, '/')) ||
	     (cp = strchr(followupto, '(')))) {
	    *cp = '\0';
	}
	if (followupto && !*followupto) {
	    FREE(followupto);
	}

	if (newsgroups && HTCanPost) {
	    START(HTML_DT);
	    START(HTML_B);
	    PUTS("Newsgroups:");
	    END(HTML_B);
	    PUTC('\n');
	    MAYBE_END(HTML_DT);
	    START(HTML_DD);
	    write_anchors(newsgroups);
	    MAYBE_END(HTML_DD);
	    PUTC('\n');
	}

	if (followupto && !strcasecomp(followupto, "poster")) {
	    /*
	    **	"Followup-To: poster" has special meaning.
	    **  Don't use it to construct a newsreply link. -kw
	    */
	    START(HTML_DT);
	    START(HTML_B);
	    PUTS("Followup to:");
	    END(HTML_B);
	    PUTC(' ');
	    if (href) {
		start_anchor(href);
		PUTS("poster");
		END(HTML_A);
	    } else {
		PUTS("poster");
	    }
	    MAYBE_END(HTML_DT);
	    PUTC('\n');
	    FREE(followupto);
	}

	if (newsgroups && HTCanPost) {
	    /*
	    **	We have permission to POST to this host,
	    **	so add a link for posting followups for
	    **	this article. - FM
	    */
	    if (!strncasecomp(NewsHREF, "snews:", 6))
		StrAllocCopy(href,"snewsreply://");
	    else
		StrAllocCopy(href,"newsreply://");
	    StrAllocCat(href, NewsHost);
	    StrAllocCat(href, "/");
	    StrAllocCat(href, (followupto ? followupto : newsgroups));
	    if (*href == 'n' &&
		(ccp = HTAnchor_messageID(thisanchor)) && *ccp) {
		StrAllocCat(href, ";ref=");
		if (strchr(ccp, '<') || strchr(ccp, '&') ||
		    strchr(ccp, ' ') || strchr(ccp, ':') ||
		    strchr(ccp, '/') || strchr(ccp, '%') ||
		    strchr(ccp, ';')) {
		    char *cp1 = HTEscape(ccp, URL_XPALPHAS);
		    StrAllocCat(href, cp1);
		    FREE(cp1);
		} else {
		    StrAllocCat(href, ccp);
		}
	    }

	    START(HTML_DT);
	    START(HTML_B);
	    PUTS("Followup to:");
	    END(HTML_B);
	    PUTC(' ');
	    start_anchor(href);
	    if (strchr((followupto ? followupto : newsgroups), ',')) {
		PUTS("newsgroups");
	    } else {
		PUTS("newsgroup");
	    }
	    END(HTML_A);
	    MAYBE_END(HTML_DT);
	    PUTC('\n');
	}
	FREE(newsgroups);
	FREE(followupto);

	if (references) {
	    START(HTML_DT);
	    START(HTML_B);
	    PUTS("References:");
	    END(HTML_B);
	    MAYBE_END(HTML_DT);
	    PUTC('\n');
	    START(HTML_DD);
	    write_anchors(references);
	    MAYBE_END(HTML_DD);
	    PUTC('\n');
	    FREE(references);
	}

	END(HTML_DLC);
	PUTC('\n');
	FREE(href);
    }

    if (rawtext) {
	/*
	 *  No tags, and never do a PUTC. - kw
	 */
	;
    } else if (diagnostic) {
	/*
	**  Read in the HEAD and BODY of the Article
	**  as XMP formatted text. - FM
	*/
	START(HTML_XMP);
	PUTC('\n');
    } else {
	/*
	**  Read in the BODY of the Article
	**  as PRE formatted text. - FM
	*/
	START(HTML_PRE);
	PUTC('\n');
    }

    p = line;
    while (!done) {
	int ich = NEXT_CHAR;
	*p++ = (char) ich;
	if (ich == EOF) {
	    if (interrupted_in_htgetcharacter) {
		interrupted_in_htgetcharacter = 0;
		CTRACE((tfp, "HTNews: Interrupted on read, closing socket %d\n",
			    s));
		NEWS_NETCLOSE(s);
		s = -1;
		return(HT_INTERRUPTED);
	    }
	    abort_socket();	/* End of file, close socket */
	    return(HT_LOADED);	/* End of file on response */
	}
	if (((char)ich == LF) || (p == &line[LINE_LENGTH])) {
	    *p++ = '\0';			/* Terminate the string */
	    CTRACE((tfp, "B %s", line));
#if NEWS_DEBUG	/* 1997/11/09 (Sun) 15:56:11 */
	    debug_print(line);	/* @@@ */
#endif
	    if (line[0] == '.') {
		/*
		**  End of article?
		*/
		if (UCH(line[1]) < ' ') {
		    done = YES;
		    break;
		} else {			/* Line starts with dot */
		    if (rawtext) {
			RAW_PUTS(&line[1]);
		    } else {
			PUTS(&line[1]); /* Ignore first dot */
		    }
		}
	    } else {
		if (rawtext) {
		    RAW_PUTS(line);
		} else if (diagnostic || !scan_for_buried_news_references) {
		    /*
		    **	All lines are passed as unmodified source. - FM
		    */
		    PUTS(line);
		} else {
		    /*
		    **	Normal lines are scanned for buried references
		    **	to other articles.  Unfortunately, it could pick
		    **	up mail addresses as well!  It also can corrupt
		    **	uuencoded messages!  So we don't do this when
		    **	fetching articles as WWW_SOURCE or when downloading
		    **	(diagnostic is TRUE) or if the client has set
		    **	scan_for_buried_news_references to FALSE.
		    **	Otherwise, we convert all "<...@...>" strings
		    **	preceded by "rticle " to "news:...@..." links,
		    **	and any strings that look like URLs to links. - FM
		    */
		    char *l = line;
		    char *p2;

		    while ((p2 = strstr(l, "rticle <")) != NULL) {
			char *q  = strrchr(p2,'>');
			char *at = strrchr(p2, '@');
			if (q && at && at<q) {
			    char c = q[1];
			    q[1] = 0;		/* chop up */
			    p2 += 7;
			    *p2 = 0;
			    while (*l) {
				if (strncmp(l, "news:", 5) &&
				    strncmp(l, "snews://", 8) &&
				    strncmp(l, "nntp://", 7) &&
				    strncmp(l, "snewspost:", 10) &&
				    strncmp(l, "snewsreply:", 11) &&
				    strncmp(l, "newspost:", 9) &&
				    strncmp(l, "newsreply:", 10) &&
				    strncmp(l, "ftp://", 6) &&
				    strncmp(l, "file:/", 6) &&
				    strncmp(l, "finger://", 9) &&
				    strncmp(l, "http://", 7) &&
				    strncmp(l, "https://", 8) &&
				    strncmp(l, "wais://", 7) &&
				    strncmp(l, "mailto:", 7) &&
				    strncmp(l, "cso://", 6) &&
				    strncmp(l, "gopher://", 9)) {
				    PUTC (*l++);
				} else {
				    StrAllocCopy(href, l);
				    start_anchor(strtok(href, " \r\n\t,>)\""));
				    while (*l && !strchr(" \r\n\t,>)\"", *l))
					PUTC(*l++);
				    END(HTML_A);
				    FREE(href);
				}
			    }
			    *p2 = '<';		/* again */
			    *q = 0;
			    start_anchor(p2+1);
			    *q = '>';		/* again */
			    PUTS(p2);
			    END(HTML_A);
			    q[1] = c;		/* again */
			    l = q + 1;
			} else {
			    break;		/* line has unmatched <> */
			}
		    }
		    while (*l) {		/* Last bit of the line */
			if (strncmp(l, "news:", 5) &&
			    strncmp(l, "snews://", 8) &&
			    strncmp(l, "nntp://", 7) &&
			    strncmp(l, "snewspost:", 10) &&
			    strncmp(l, "snewsreply:", 11) &&
			    strncmp(l, "newspost:", 9) &&
			    strncmp(l, "newsreply:", 10) &&
			    strncmp(l, "ftp://", 6) &&
			    strncmp(l, "file:/", 6) &&
			    strncmp(l, "finger://", 9) &&
			    strncmp(l, "http://", 7) &&
			    strncmp(l, "https://", 8) &&
			    strncmp(l, "wais://", 7) &&
			    strncmp(l, "mailto:", 7) &&
			    strncmp(l, "cso://", 6) &&
			    strncmp(l, "gopher://", 9))
			    PUTC (*l++);
			else {
			    StrAllocCopy(href, l);
			    start_anchor(strtok(href, " \r\n\t,>)\""));
			    while (*l && !strchr(" \r\n\t,>)\"", *l))
				PUTC(*l++);
			    END(HTML_A);
			    FREE(href);
			}
		    }
		} /* if diagnostic or not scan_for_buried_news_references */
	    } /* if not dot */
	    p = line;				/* Restart at beginning */
	} /* if end of line */
    } /* Loop over characters */

    if (rawtext)
	return(HT_LOADED);

    if (diagnostic)
	END(HTML_XMP);
    else
	END(HTML_PRE);
    PUTC('\n');
    return(HT_LOADED);
}

/*	Read in a List of Newsgroups
**	----------------------------
**
**  Note the termination condition of a single dot on a line by itself.
**  RFC 977 specifies that the line "folding" of RFC850 is not used,
**  so we do not handle it here.
*/
PRIVATE int read_list ARGS1(char *, arg)
{
    char line[LINE_LENGTH+1];
    char *p;
    BOOL done = NO;
    BOOL head = NO;
    BOOL tail = NO;