Another commit with some more progress

author: George Abbott <george@gabbott.dev> 2023-10-09 18:11:41 +0100
committer: George Abbott <george@gabbott.dev> 2023-10-09 18:11:41 +0100
commit: c91061498ca9137c0bb2e65765ab15c1439941fd (patch)
tree: f42b6377f3b8488051836cb5d891e09237860734
parent: 08cec360e16615b1ae157e4926e7317d16765f7e (diff)
4 files changed, 291 insertions, 139 deletions
diff --git a/source/app.d b/source/app.d
index 97630d2..cb57d19 100644
--- a/source/app.d
+++ b/source/app.d
@@ -63,6 +63,8 @@ import core.stdc.stdlib : exit;
 
 import html;
 import rssmeta;
+import entry;
+import sort;
 
 void usage()
 {
@@ -85,141 +87,30 @@ void usage()
 	exit(0);
 }
 
-
-
-// TODO: make ctor so that pubDateAsUnixTimestamp gets populated.
-struct Entry
-{
-	string title;
-	string link;
-	string pubDate;
-	string guid;
-	string description; // The actual entry itself.
-
-private:
-	int pubDateAsUnixTimestamp; // So that we can sort easily.
-}
-
-
-Entry make_entry(DirEntry dirEntry, RssMeta rssmeta)
+// TODO: RssInfo. We need the struct, with the relevant information, and where
+// do we get it from?
+void create_rss(Entry[] entries, RssInfo info)
 {
-	// From the DirEntry, we can get all file info, and also read the full file
-	// since we know its path. We just need to remember, when we read the
-	// entry, to only read between the <body> and </body> tags, whilst
-	// searching the full file for content, and replace all bad characters with
-	// escaped ones.
-	// Or, if <body> does not work, we could try between some comments. 
-	// I say, <!-- bct:beg --> and <!-- bct:end --> would meet those criteria.
-	// COMPLETE
-	string get_title(DirEntry dirEntry, in string text, in RssMeta rssmeta)
-	{
-		// Title is either:
-		// 1) rssmeta:<title>
-		// 2) <!-- rss-title: TITLE --> comment in text.
-		// 3) <title>TITLE</title> from text.
-
-		string title = rssmeta.title_of(dirEntry.name);
-		if (title !is null)
-			return title;
-
-		title = find_rss_comment("title", text);
-		if (title !is null)
-			return title;
-
-		title = find_html_title(text);
-		return title; // even if it's null. 
-	}
-
-	// COMPLETE
-	string get_link(DirEntry dirEntry, in string text, in RssMeta rssmeta)
-	{
-		// Link is either:
-		// 1) rssmeta:<link>
-		// 2) <!-- rss-link: link --> comment in text.
-
-		string link = rssmeta.link_of(dirEntry.name);
-		if (link !is null)
-			return link;
-
-		link = find_rss_comment("link", text);
-		return link;
-	}
-
-	// Output is formatted as ISO-8601.
-	string get_pubdate(DirEntry dirEntry, in string text, in RssMeta rssmeta)
-	{
-		// The date must be formatted YYYY-mm-dd HH:MM. 
-		// pubDate is either:
-		// 1) rssmeta:<pubDate>
-		// 2) <!-- rss-pubDate: TITLE --> comment in text.
-		// 3) the Unix timestamp of the file.
-
-		string pubDate = rssmeta.title_of(dirEntry.name);
-		if (pubDate !is null)
-			return pubDate;
-
-		pubDate = find_rss_comment("pubDate", text);
-		if (pubDate !is null)
-			return pubDate;
-
-		// TODO: add in getting date from the dirEntry.time.
-
-		return pubDate; // even if it's null. 
-	}
-
-	// COMPLETE
-	string get_guid(DirEntry dirEntry, in string text, in RssMeta rssmeta)
-	{
-		// guid is either:
-		// 1) rssmeta:<guid>
-		// 2) <!-- rss-guid: guid --> comment in text.
-		// 3) delegated to get_link as a last resort.
-
-		string guid = rssmeta.guid_of(dirEntry.name);
-		if (guid !is null)
-			return guid;
-
-		guid = find_rss_comment("guid", text);
-		if (guid !is null)
-			return guid;
-
-		return get_link(dirEntry, text, rssmeta);
-	}
-
-	void fail(string what)
+	string ret;
+	ret = 
+		"<rss version="2.0">\n" ~
+		"\t\t<channel>\n" ~
+		"\t\t<title>" ~ info.title ~ "</title>\n" ~
+		"\t\t<description>" ~ info.description ~ "</description>\n" ~
+		"\t\t<language>" ~ info.language ~ "</language>\n" ~
+		"\t\t<link href=\"" ~ info.link ~ "\" rel=\"self\" type=\"application/rss+xml\"/>";
+	
+	for (int i = 0; i < entries.length; ++i)
 	{
-		writeln("Error occurred when retrieving attribute ", what, " in make_entry");
-		exit(-1);
+		// TODO: add in the rest of the entry RSS info.
+		auto entry = entry[i];
 	}
 
-
-	string text    = readText(dirEntry.name);
-	string title   = get_title(dirEntry, text, rssmeta);
-	string link    = get_link(dirEntry, text, rssmeta);
-	string pubDate = get_pubdate(dirEntry, text, rssmeta);
-	string guid    = get_guid(dirEntry, text, rssmeta);
-
-	if (text is null)
-		fail("text");
-
-	if (title is null)
-		fail("title");
-
-	if (link is null)
-		fail("link");
-
-	if (pubDate is null)
-		fail("pubDate");
-
-	if (guid is null)
-		fail("guid");
-
-	Entry ret = Entry(text, title, link, pubDate, guid);
-	return ret;
-
-
+	ret ~= "\n\t</channel>";
+	ret ~= "\n</rss>";
 }
 
+
 // Flags which are passed to -i (--ignore).
 immutable FILE_WO_ENTRY = "file-wo-entry";
 immutable ENTRY_WO_FILE = "entry-wo-file";
@@ -230,11 +121,12 @@ int main(string[] args)
 	/* Flags */
 	string metafile;
 	bool entries_set = false;
-	string entries_dir;
-	string target;
+	string entries_dir = null;
+	string target = null;
 	string[] ignore;
 	bool recursive = false;
 	bool follow_symlinks = false;
+	bool sort_mode = null;
 
 	for (int i = 0; i < args.length; ++i)
 	{
@@ -248,7 +140,30 @@ int main(string[] args)
 		else if (args[i] == "-f" || args[i] == "--follow")
 			follow_symlinks = true;
 		else if (args[i] == "-o" || args[i] == "--order")
-		{}
+		{
+			auto param = args[++i];
+			switch (param)
+			{
+				case "title-alphabetic":
+					sort_mode = SortMode.TitleAlphabetic;
+					break;
+				case "date-reverse":
+				case "date_reverse":
+				case "reverse":
+					sort_mode = SortMode.DateReverse;
+					break;
+				case "date-forward": case "date_forward":
+				case "date-forwards": case "date_forwards":
+				case "forward": case "forwards":
+					sort_mode = SortMode.DateForward;
+					break;
+				default:
+					writeln("Invalid sort mode inputted - valid are" ~
+							" title-alphabetic, reverse, forward");
+					exit(1);
+				
+			}
+		}
 		else if (args[i] == "-i" || args[i] == "--ignore")
 		{
 			auto param = args[++i];
@@ -273,6 +188,17 @@ int main(string[] args)
 		}
 	}
 
+	if (entries_dir == null)
+	{
+		writeln("Please provide a directory containing entries: see --help");
+		exit(1);
+	}
+	if (target == null)
+	{
+		writeln("Please provide a target rss.xml file; see --help");
+		exit(1);
+	}
+
 	// Get RssMeta.
 	// TODO: also make it so it can get the metafile from the current dir 
 	// and rssmeta.xml xor rssmeta.json files.
@@ -298,6 +224,13 @@ int main(string[] args)
 		entries ~= make_entry(dirEntry, rssmeta);
 	}
 
+	// Sort entries. 
+	entries2 = sort(entries, sort_mode); 
+
+	// Iterate over the sorted entries and create the RSS file contents from
+	// them.
+	const result = create_rss(entries2);
+
 
 	// Iterate over all entries in the order as described by -o flag, and 
 	// create the entry for them. This gets spat back out as a big string.
@@ -307,7 +240,5 @@ int main(string[] args)
 
 	// Finally, write back the full RSS XML into the file specified by $2.
 
-
-
 	return 0;
 }
diff --git a/source/entry.d b/source/entry.d
new file mode 100644
index 0000000..01d4e3d
--- /dev/null
+++ b/source/entry.d
@@ -0,0 +1,145 @@
+// TODO: make ctor so that pubDateAsUnixTimestamp gets populated.
+struct Entry
+{
+	string title;
+	string link;
+	string pubDate;
+	string guid;
+	string description; // The actual entry itself.
+
+	this(string title, string link, string pubDate, string guid, string text)
+	{
+		this.title = title;
+		this.link = link;
+		this.pubdate = pubDate;
+		this.guid = guid;
+		this.description = text;
+		pubDateAsUnixTimestamp = unix_timestamp_of_date(pubDate, "%Y-%m-%d %H:%M");
+	}
+
+private:
+	int pubDateAsUnixTimestamp; // So that we can sort easily.
+}
+
+int unix_timstamp_of_date(string date, string format)
+{
+	// TODO: used in Entry ctor.
+}
+
+
+Entry make_entry(DirEntry dirEntry, RssMeta rssmeta)
+{
+	// From the DirEntry, we can get all file info, and also read the full file
+	// since we know its path. We just need to remember, when we read the
+	// entry, to only read between the <body> and </body> tags, whilst
+	// searching the full file for content, and replace all bad characters with
+	// escaped ones.
+	// Or, if <body> does not work, we could try between some comments. 
+	// I say, <!-- bct:beg --> and <!-- bct:end --> would meet those criteria.
+	// COMPLETE
+	string get_title(DirEntry dirEntry, in string text, in RssMeta rssmeta)
+	{
+		// Title is either:
+		// 1) rssmeta:<title>
+		// 2) <!-- rss-title: TITLE --> comment in text.
+		// 3) <title>TITLE</title> from text.
+
+		string title = rssmeta.title_of(dirEntry.name);
+		if (title !is null)
+			return title;
+
+		title = find_rss_comment("title", text);
+		if (title !is null)
+			return title;
+
+		title = find_html_title(text);
+		return title; // even if it's null. 
+	}
+
+	// COMPLETE
+	string get_link(DirEntry dirEntry, in string text, in RssMeta rssmeta)
+	{
+		// Link is either:
+		// 1) rssmeta:<link>
+		// 2) <!-- rss-link: link --> comment in text.
+
+		string link = rssmeta.link_of(dirEntry.name);
+		if (link !is null)
+			return link;
+
+		link = find_rss_comment("link", text);
+		return link;
+	}
+
+	// Output is formatted as ISO-8601.
+	string get_pubdate(DirEntry dirEntry, in string text, in RssMeta rssmeta)
+	{
+		// The date must be formatted YYYY-mm-dd HH:MM. 
+		// pubDate is either:
+		// 1) rssmeta:<pubDate>
+		// 2) <!-- rss-pubDate: TITLE --> comment in text.
+		// 3) the Unix timestamp of the file.
+
+		string pubDate = rssmeta.title_of(dirEntry.name);
+		if (pubDate !is null)
+			return pubDate;
+
+		pubDate = find_rss_comment("pubDate", text);
+		if (pubDate !is null)
+			return pubDate;
+
+		// TODO: add in getting date from the dirEntry.time.
+
+		return pubDate; // even if it's null. 
+	}
+
+	// COMPLETE
+	string get_guid(DirEntry dirEntry, in string text, in RssMeta rssmeta)
+	{
+		// guid is either:
+		// 1) rssmeta:<guid>
+		// 2) <!-- rss-guid: guid --> comment in text.
+		// 3) delegated to get_link as a last resort.
+
+		string guid = rssmeta.guid_of(dirEntry.name);
+		if (guid !is null)
+			return guid;
+
+		guid = find_rss_comment("guid", text);
+		if (guid !is null)
+			return guid;
+
+		return get_link(dirEntry, text, rssmeta);
+	}
+
+	void fail(string what)
+	{
+		writeln("Error occurred when retrieving attribute ", what, " in make_entry");
+		exit(-1);
+	}
+
+
+	string text    = readText(dirEntry.name);
+	string title   = get_title(dirEntry, text, rssmeta);
+	string link    = get_link(dirEntry, text, rssmeta);
+	string pubDate = get_pubdate(dirEntry, text, rssmeta);
+	string guid    = get_guid(dirEntry, text, rssmeta);
+
+	if (text is null)
+		fail("text");
+
+	if (title is null)
+		fail("title");
+
+	if (link is null)
+		fail("link");
+
+	if (pubDate is null)
+		fail("pubDate");
+
+	if (guid is null)
+		fail("guid");
+
+	Entry ret = Entry(text, title, link, pubDate, guid);
+	return ret;
+}
diff --git a/source/rssmeta.d b/source/rssmeta.d
index 6fec5b4..864503c 100644
--- a/source/rssmeta.d
+++ b/source/rssmeta.d
@@ -1,13 +1,39 @@
 // Represents the RssMeta struct, containing the metadata with some fns to 
 // access it easily.
 
+import dxml;
+import entry;
+
 struct RssMeta
 {
-	// TODO: add the magic.
+	// TODO
+	// The RssMeta should be a lookup table: we have a key (the filename given
+	// in rssmeta.xml, etc.; and we have a set of values, e.g. the title, link,
+	// etc. values which form the metadata.
+	string metafile;
+	string[Entry] lookup;
+
+	this(string metafile)
+	{
+
+		this.metafile = metafile;
+		if (metafile[-4:] == ".json") {}
+		else if (metafile[-4:] == ".xml") {}
+		else 
+		{
+			writeln("Cannot deduce whether metafile is json or xml. Please" ~
+					" ensure it is called either rssmeta.xml or rssmeta.json.");
+			exit(1);
+		}
+	}
 
 	// filename: either just name or fully qualified should work.
-	string title_of(string filename) const;
-	string link_of(string filename) const;
-	string pubdate_of(string filename) const;
-	string guid_of(string filename) const;
+	string title_of(string filename) const 
+	{
+		// TODO: we need to sort out the internal magic of RssMeta first, then
+		// we can just do a lookup on the filename. 
+	}
+	string link_of(string filename) const { return ""; }
+	string pubdate_of(string filename) const { return ""; }
+	string guid_of(string filename) const { return ""; }
 }
diff --git a/source/sort.d b/source/sort.d
new file mode 100644
index 0000000..9baa667
--- /dev/null
+++ b/source/sort.d
@@ -0,0 +1,50 @@
+/* For sorting the entries as per the -o flag.
+ * This allows a list of entries to be sorted, 
+ * e.g. title-alphabetic, date-reverse, date-forward.
+ */
+
+enum SortMode
+{
+	/* For instance:
+	 *   TitleAA, TitleZZ, TitleFF
+	 * becomes
+	 *   TitleAA, TitleFF, TitleZZ
+	 */ 
+	TitleAlphabetic;
+
+	/* For instance:
+	 *   2023-10-01, 2023-10-05, 2023-10-03
+	 * becomes
+	 *   2023-10-05, 2023-10-03, 2023-10-01
+	 */
+	DateReverse;
+
+	/* For instance:
+	 *   2023-10-01, 2023-10-05, 2023-10-03
+	 * becomes
+	 *   2023-10-01, 2023-10-03, 2023-10-05
+	 */
+	DateForwards;
+}
+
+// TODO: write sorting functions.
+Entry[] sort_title_alphabetic(Entry[] entries) { return entries; }
+Entry[] sort_date_reverse(Entry[] entries) { return entries; }
+Entry[] sort_date_forward(Entry[] entries) { return entries; }
+
+
+Entry[] sort(Entry[] entries, SortMode sort_mode)
+{
+	switch (sort_mode)
+	{
+		case SortMode.TitleAlphabetic:
+			return sort_title_alphabetic(entries);
+		case SortMode.DateReverse:
+			return sort_date_reverse(entries);
+		case SortMode.DateForward:
+			return sort_date_forward(entries);
+	}
+}
+
+
+
author	George Abbott <george@gabbott.dev>	2023-10-09 18:11:41 +0100
committer	George Abbott <george@gabbott.dev>	2023-10-09 18:11:41 +0100
commit	c91061498ca9137c0bb2e65765ab15c1439941fd (patch)
tree	f42b6377f3b8488051836cb5d891e09237860734
parent	08cec360e16615b1ae157e4926e7317d16765f7e (diff)