Add support for formatted test

Makefile

@@ -12,7 +12,8 @@ C_SRCS = slack.c \
 	 slack-object.c \
 	 slack-json.c \
 	 purple-websocket.c \
-	 json.c
+	 json.c \
+	 markdown.c
 
 # Object file names using 'Substitution Reference'
 C_OBJS = $(C_SRCS:.c=.o)

markdown.c

@@ -0,0 +1,312 @@
+/*
+ *   Markdown library for libpurple
+ *   Copyright (C) 2018 Alyssa Rosenzweig
+ *
+ *   This program is free software: you can redistribute it and/or modify
+ *   it under the terms of the GNU General Public License as published by
+ *   the Free Software Foundation, either version 3 of the License, or
+ *   (at your option) any later version.
+ *
+ *   This program is distributed in the hope that it will be useful,
+ *   but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *   GNU General Public License for more details.
+ *
+ *   You should have received a copy of the GNU General Public License
+ *   along with this program.  If not, see <https://www.gnu.org/licenses/>.
+ */
+
+#include <string.h>
+#include "markdown.h"
+
+/* Markdown test string:
+ *
+ * "<--- \o/ **¯\_(ツ)_/¯**  _italics_ __underline__ *correction right* *italics2* ~~strikethrough~~ ~notstriked~ ~me https://pidgin.im <style>body{background-color:red}</script> &lt;style&gt;body{background-color: red}&lt;/style&gt; <b>notbold</b> &lt;notatag&gt;"
+ *
+ * Checks for:
+ * 	- Correct escaping of lt/gt signs
+ * 	- Aesthetically correct (but non-comformant) handling of escape sequences as found in backslash-containing emoticons
+ * 	- Exhaustive test of syntaxes for italics, underline, strikethrough
+ * 	- Correct escaping for XSS rsisk
+ *
+ * Does not check for (open issues):
+ * 	- Italicised shruggie
+ *
+ */
+
+#define HTML_TOGGLE_OUT(f, a, b) \
+	out = g_string_append(out, f ? b : a); \
+	f = !f;
+
+/* workaround errata in Discord's (users') markdown implementation */
+
+static gboolean
+markdown_underscore_match(const gchar *html, int i)
+{
+	while (html[i] != ' ' && html[i]) {
+		if (html[i++] == '_') {
+			return !html[i] || html[i] == ' ';
+		}
+	}
+
+	return FALSE;
+}
+
+/* Is a character escapable, that is, does it have a special meaning in
+ * Markdown? */
+
+static gboolean
+markdown_is_escapable(char c)
+{
+	switch (c) {
+	case '\\':
+	case '*':
+	case '~':
+	case '_':
+	case '`':
+		return TRUE;
+	default:
+		return FALSE;
+	}
+}
+
+/* Should we interpret a _ as italics?  */
+
+static gboolean
+markdown_should_underscore_italics(const gchar *html, unsigned i, gboolean s_italics)
+{
+	return s_italics || markdown_underscore_match(html, i + 1);
+}
+
+/* Should we interpret _  as special at all? */
+
+static gboolean
+markdown_should_underscore(const gchar *html, unsigned i, gboolean s_italics)
+{
+	return html[i + 1] == '_' || markdown_should_underscore_italics(html, i, s_italics);
+}
+
+static gchar *
+markdown_helper_replace(gchar *html, const gchar *tag, const gchar *replacement)
+{
+	gchar *replace_regex;
+	gchar *replace_with;
+
+	if (tag[0] == '<' && tag[1] == '/') {
+		//closing tag
+		replace_regex = g_strconcat("(\\s*)", tag, NULL);
+		replace_with = g_strconcat(replacement, "\\1", NULL);
+	} else {
+		replace_regex = g_strconcat(tag, "(\\s*)", NULL);
+		replace_with = g_strconcat("\\1", replacement, NULL);
+	}
+
+	GRegex *markdown_replace = g_regex_new(replace_regex, 0, 0, NULL);
+	gchar *temp = g_regex_replace(markdown_replace, html, -1, 0, replace_with, 0, NULL);
+
+	g_free(replace_regex);
+	g_free(replace_with);
+	g_regex_unref(markdown_replace);
+
+	if (temp != NULL) {
+		g_free(html);
+		html = temp;
+	}
+
+	return html;
+}
+
+gchar *
+markdown_convert_markdown(const gchar *html, gboolean escape_html, gboolean discord_hacks)
+{
+	g_return_val_if_fail(html != NULL, NULL);
+
+	guint html_len = strlen(html);
+	GString *out = g_string_sized_new(html_len * 2);
+
+	gboolean s_bold = FALSE;
+	gboolean s_italics = FALSE;
+	gboolean s_strikethrough = FALSE;
+	gboolean s_codeblock = FALSE;
+	gboolean s_codebit = FALSE;
+	gboolean s_smiley = FALSE;
+
+	for (guint i = 0; i < html_len; ++i) {
+		char c = html[i];
+
+		if ((s_codeblock || s_codebit) && c != '`') {
+			out = g_string_append_c(out, html[i]);
+			continue;
+		}
+		if (s_smiley && c != ':') {
+			out = g_string_append_c(out, html[i]);
+			continue;
+		}
+
+		if (c == '\\') {
+			char next_char = html[++i];
+
+			/* If this is an escape-able character, don't print the
+			 * backslash. Otherwise, do because the \ wasn't an
+			 * escape anyway */
+
+			gboolean escapable = markdown_is_escapable(next_char);
+
+			/* Also, if this is an escapable character that would
+			 * not actually -matter-, print it too. Fixes shruggie
+			 * */
+
+			if (next_char == '_' && !markdown_should_underscore(html, i + 1, s_italics) && (escape_html || discord_hacks))
+				escapable = FALSE;
+
+			if (!escapable) {
+				out = g_string_append_c(out, '\\');
+			}
+
+			/* Append the next char regardless */
+			out = g_string_append_c(out, next_char);
+		} else if ((c == '<' || c == '>' || c == '&') && escape_html) {
+			/* These characters lack any particular meaning in
+			 * Markdown, but need to be escaped to prevent getting
+			 * mixed up with HTML. Failing to do so may result in
+			 * valid parts of the message being stripped by
+			 * overzealous sanitizers */
+
+			if (c == '<')
+				out = g_string_append(out, "&lt;");
+			else if (c == '>')
+				out = g_string_append(out, "&gt;");
+			else /*if (c == '&')*/
+				out = g_string_append(out, "&amp;");
+		} else if (c == '*') {
+			HTML_TOGGLE_OUT(s_bold, "<b>", "</b>");
+		} else if (c == '~') {
+			HTML_TOGGLE_OUT(s_strikethrough, "<s>", "</s>");
+		} else if (c == '_') {
+			HTML_TOGGLE_OUT(s_italics, "<i>", "</i>");
+		} else if (c == '`') {
+			if (html[i + 1] == '`' && html[i + 2] == '`') {
+				if (!s_codeblock) {
+#ifdef MARKDOWN_PIDGIN
+					out = g_string_append(out, "<br/><span style='font-family: monospace; white-space: pre'>");
+#else
+					out = g_string_append(out, "<br/><pre>");
+#endif
+				} else {
+#ifdef MARKDOWN_PIDGIN
+					out = g_string_append(out, "</span>");
+#else
+					out = g_string_append(out, "</pre>");
+#endif
+					i += 2;
+				}
+
+				s_codeblock = !s_codeblock;
+			} else {
+#ifdef MARKDOWN_PIDGIN
+				HTML_TOGGLE_OUT(s_codebit, "<span style='font-family: monospace; white-space: pre'>", "</span>");
+#else
+				HTML_TOGGLE_OUT(s_codebit, "<code>", "</code>");
+#endif
+			}
+		} else if (c == ':' && (s_smiley || i == 0 || html[i - 1] == ' ')) {
+			s_smiley = !s_smiley;
+			out = g_string_append_c(out, c);
+		} else {
+			out = g_string_append_c(out, c);
+		}
+	}
+
+	gchar *new_out = purple_strreplace(out->str, "\n", "<br>");
+	g_string_free(out, TRUE);
+	return new_out;
+}
+
+#define REPLACE_TAG(name, repl) \
+	html = markdown_helper_replace(html, "<" name ">", repl); \
+	html = markdown_helper_replace(html, "</" name ">", repl);
+
+gchar *
+markdown_html_to_markdown(gchar *html)
+{
+	REPLACE_TAG("b", "*");
+	REPLACE_TAG("strong", "*");
+	REPLACE_TAG("i", "_");
+	REPLACE_TAG("em", "_");
+	REPLACE_TAG("u", "");
+	REPLACE_TAG("s", "~~");
+	REPLACE_TAG("pre", "```");
+	REPLACE_TAG("code", "`");
+
+	/* Let newlines get passed through as HTML */
+
+	/* Workaround XHTML-IM stuff. TODO: XXX */
+	html = markdown_helper_replace(html, "<span style='font-weight: bold;'>", "**");
+	html = markdown_helper_replace(html, "</span>", "**");
+
+	return html;
+}
+
+gchar *
+markdown_escape_md(const gchar *markdown, gboolean discord_hacks)
+{
+	size_t markdown_len = strlen(markdown);
+	/* Worst case allocation */
+	GString *s = g_string_sized_new(markdown_len * 2);
+
+	gboolean verbatim = FALSE;
+	gboolean code_block = FALSE;
+	gboolean link = FALSE;
+
+	for (guint i = 0; i < markdown_len; ++i) {
+		char c = markdown[i];
+
+		if (c == '`') {
+			if (code_block) {
+				code_block = verbatim = FALSE;
+			} else if (!verbatim) {
+				code_block = verbatim = TRUE;
+			}
+
+			g_string_append_c(s, markdown[i]);
+
+			if (markdown[i + 1] == '`' && markdown[i + 2] == '`') {
+				i += 2;
+				g_string_append_c(s, markdown[i]);
+				g_string_append_c(s, markdown[i]);
+				continue;
+			}
+		}
+
+		if (!verbatim) {
+			if (strncmp(markdown + i, "http://", sizeof("http://") - 1) == 0 ||
+				strncmp(markdown + i, "https://", sizeof("https://") - 1) == 0)
+
+			{
+				link = verbatim = TRUE;
+			}
+		}
+
+		if (link && c == ' ') {
+			link = verbatim = FALSE;
+		}
+
+		if (!verbatim) {
+			if (
+			  (c == '_' && (markdown[i + 1] == ' ' ||
+							markdown[i + 1] == '\0' ||
+							i == 0 ||
+							markdown[i - 1] == ' ' ||
+							markdown[i - 1] == '\0')) ||
+			  (c == '*') ||
+			  (c == '\\' && !(markdown[i + 1] == '_' && (i == 0 || markdown[i - 1] == ' ')) && !discord_hacks) ||
+			  (c == '~' && (markdown[i + 1] == '~'))) {
+				g_string_append_c(s, '\\');
+			}
+		}
+
+		g_string_append_c(s, c);
+	}
+
+	return g_string_free(s, FALSE);
+}

markdown.h

@@ -0,0 +1,30 @@
+/*
+ *   Markdown library for libpurple
+ *   Copyright (C) 2018 Alyssa Rosenzweig
+ *
+ *   This program is free software: you can redistribute it and/or modify
+ *   it under the terms of the GNU General Public License as published by
+ *   the Free Software Foundation, either version 3 of the License, or
+ *   (at your option) any later version.
+ *
+ *   This program is distributed in the hope that it will be useful,
+ *   but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *   GNU General Public License for more details.
+ *
+ *   You should have received a copy of the GNU General Public License
+ *   along with this program.  If not, see <https://www.gnu.org/licenses/>.
+ */
+
+
+#ifndef __MARKDOWN_H
+#define __MARKDOWN_H
+
+#include <purple.h>
+#include <glib.h>
+
+gchar *markdown_convert_markdown(const gchar *html, gboolean escape_html, gboolean markdown_hacks);
+gchar *markdown_escape_md(const gchar *markdown, gboolean markdown_hacks);
+gchar *markdown_html_to_markdown(gchar *html);
+
+#endif

slack-channel.c

@@ -329,10 +329,12 @@ int slack_chat_send(PurpleConnection *gc, int cid, const char *msg, PurpleMessag
 
 	gchar *m = slack_html_to_message(sa, msg, flags);
 	glong mlen = g_utf8_strlen(m, 16384);
-	if (mlen > 4000)
+	if (mlen > 4000) {
+		g_free(m);
 		return -E2BIG;
+	}
 
-	struct send_chat *send = g_new(struct send_chat, 1);
+	struct send_chat *send = g_new0(struct send_chat, 1);
 	send->chan = g_object_ref(chan);
 	send->cid = cid;
 	send->flags = flags;

slack-conversation.c

@@ -93,7 +93,7 @@ void slack_conversation_retrieve(SlackAccount *sa, const char *sid, SlackConvers
 	SlackObject *obj = slack_conversation_lookup_sid(sa, sid);
 	if (obj)
 		return cb(sa, data, obj);
-	struct conversation_retrieve *lookup = g_new(struct conversation_retrieve, 1);
+	struct conversation_retrieve *lookup = g_new0(struct conversation_retrieve, 1);
 	lookup->cb = cb;
 	lookup->data = data;
 	slack_api_get(sa, conversation_retrieve_cb, lookup, "conversations.info", "channel", sid, NULL);

slack-im.c

@@ -161,10 +161,12 @@ int slack_send_im(PurpleConnection *gc, const char *who, const char *msg, Purple
 
 	gchar *m = slack_html_to_message(sa, msg, flags);
 	glong mlen = g_utf8_strlen(m, 16384);
-	if (mlen > 4000)
+	if (mlen > 4000) {
+		g_free(m);
 		return -E2BIG;
+	}
 
-	struct send_im *send = g_new(struct send_im, 1);
+	struct send_im *send = g_new0(struct send_im, 1);
 	send->user = g_object_ref(user);
 	send->msg = m;
 	send->flags = flags;