From 81633c97a89de0f044c1cfdf97f945e21ca17ca7 Mon Sep 17 00:00:00 2001
From: Daniel Jones <admin@danieljon.es>
Date: Sun, 19 Jan 2020 02:08:47 +1030
Subject: rss feed: first image in each post shown in CDATA tag

---
 config.h |  2 ++
 pages.c  | 71 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++------
 pages.h  |  1 +
 rss.txt  |  1 -
 4 files changed, 68 insertions(+), 7 deletions(-)

diff --git a/config.h b/config.h
index 1b13b12..4e7cc5c 100644
--- a/config.h
+++ b/config.h
@@ -100,6 +100,8 @@ static const int 	ignore[] = {}; /* ignore these posts */
 static const char 	*rss_output = "posts/posts.rss";
 static const char 	*rss_template = "rss.txt";
 static const int 	post_count = 10;
+static const char 	*base_url = "https://danieljon.es/posts/";
+static const char 	*author_string = "daniel@danieljon.es";
 
 /* each page to be generated go into this array */
 static const struct page pages[] = {
diff --git a/pages.c b/pages.c
index 1669778..9cc5868 100644
--- a/pages.c
+++ b/pages.c
@@ -762,6 +762,44 @@ char
 	return date;
 }
 
+char
+*getimage(const char *line, char *buff, size_t size)
+{
+	long pos = 0;
+	size_t linesize = strlen(line);
+	while (((line[pos] != '>') && (line[pos] != '\0'))
+				   && pos < linesize
+				   && pos < size)
+	{
+		buff[pos] = line[pos];
+		pos++;
+	}
+	buff[pos++] = '>';
+	buff[pos++] = '\0';
+	/* check src attribute for relative link and make it absolute */
+	char b1[8] = {0};
+	char newbuff[size-pos]; // ensure we dont exceed size, pos bytes already written
+	char finalbuff[size];
+	memset(finalbuff, 0, size);
+	char *src = strstr(buff, "src=\"");
+	pos = 5; // size of src="
+	if (src)
+	{
+		src += pos;
+		strncpy(b1, src, 4);
+		if (strcmp(b1, "http") != 0) // naively assume an http means a url
+		{
+			fprintf(stderr, "%s: relative path in rss feed, fixing..\n", src);
+			snprintf(newbuff, size-pos, "%s/%s", base_url, src);
+			strncpy(finalbuff, buff, src-buff);
+			strncat(finalbuff, newbuff, size-pos);
+			strncpy(buff, finalbuff, size);
+		}
+	}
+
+	return buff;
+}
+
 int
 writerss(FILE *out, int post)
 {
@@ -784,8 +822,10 @@ writerss(FILE *out, int post)
 	char line[4096];
 	char title[512] = {0};
 	char date[512] = {0};
-	char description[512] = {0};
-	char item[4096] = {0};
+	char description[4096] = {0};
+	char item[8195] = {0};
+	char image[512] = {0};
+	int hasimg = 0;
 	int pos = 0;
 	while ((fgets(line, 4096, in)) != NULL)
 	{
@@ -794,22 +834,34 @@ writerss(FILE *out, int post)
 			/* should be the title */
 			strncpy(title, line, 512);
 		}
-		if (pos == 1)
+		else if (pos == 1)
 		{
 			/* should be the date */
 			strncpy(date, line, 512);
 		}
 
-		if (pos >= 2 && (title[0] == '\n' || date[0] == '\n'))
+		else if (pos >= 2 && (title[0] == '\n' || date[0] == '\n'))
 		{
 			fprintf(stderr, "post %s has a broken format, please fix it\n", buff);
 			fclose(in);
 			return 0;
 		}
-		if (pos == 3)
+		else if (pos == 3)
 		{
 			strncpy(description, line, 100);
 		}
+
+		/* try to find images */
+		else if (pos > 3)
+		{
+			char *img;
+			if ((img = strstr(line, "<img")))
+			{
+				hasimg = 1;
+				getimage(img, image, 1024);
+				break;
+			}
+		}
 		pos++;
 	}
 	striphtml(title, strlen(title));
@@ -818,7 +870,14 @@ writerss(FILE *out, int post)
 	strcat(description, " ...");
 	rfc822date(date, 512);
 	//printf("title is: %s date is: %s: desc is: %s\n", title, date, description);
-	snprintf(item, 4096, "<item>\n\t<title>%s</title>\n\t<pubDate>%s</pubDate>\n\t<link>https://danieljon.es/posts/direct/%d</link>\n\t<guid>https://danieljon.es/posts/direct/%d</guid>\n\t<description>%s</description>\n</item>\n", title, date, post, post, description);
+	if (hasimg)
+	{
+		snprintf(item, 8195, "<item>\n\t<title>%s</title>\n\t<pubDate>%s</pubDate>\n\t<author>%s</author>\n\t<link>https://danieljon.es/posts/direct/%d</link>\n\t<guid>https://danieljon.es/posts/direct/%d</guid>\n\t<description>%s\n<![CDATA[%s]]> </description>\n</item>\n", title, date, author_string, post, post, description, image);
+	}
+	else
+	{
+		snprintf(item, 8195, "<item>\n\t<title>%s</title>\n\t<pubDate>%s</pubDate>\n\t<author>%s</author>\n\t<link>https://danieljon.es/posts/direct/%d</link>\n\t<guid>https://danieljon.es/posts/direct/%d</guid>\n\t<description>%s</description>\n</item>\n", title, date, author_string, post, post, description);
+	}
 
 	fprintf(out, "%s", item);
 	fclose(in);
diff --git a/pages.h b/pages.h
index d20ea83..f99a6aa 100644
--- a/pages.h
+++ b/pages.h
@@ -51,6 +51,7 @@ int generaterss(const int *posts, size_t totalposts);
 int writerss(FILE *out, int post);
 char *striphtml(char *str, size_t size);
 char *rfc822date(char *date, size_t size);
+char *getimage(const char *line, char *buff, size_t size);
 
 /* generators (to be put into the pages array) */
 int frontpage(int flags);
diff --git a/rss.txt b/rss.txt
index 8c9c5f2..a23192c 100644
--- a/rss.txt
+++ b/rss.txt
@@ -5,7 +5,6 @@
 	<link>https://danieljon.es/posts</link>
 	<description>Posts made by daniel_j on https://danieljon.es/posts</description>
 	<generator>https://github.com/daniel-Jones/websitegenerator</generator>
-	<author>Daniel Jones (daniel@danieljon.es)</author>
 	<image>
 		<url>https://danieljon.es/media/tux.png</url>
 		<title>danieljon.es posts</title>
-- 
cgit v1.2.3