Some text manipulation refactor and fixes (Aurorastation#17683)

* sdfsad

* sdfsda

* fourspaces begone

aurorastation.dme

@@ -109,6 +109,7 @@
 #include "code\__defines\subsystem-priority.dm"
 #include "code\__defines\targeting.dm"
 #include "code\__defines\technomancer.dm"
+#include "code\__defines\text.dm"
 #include "code\__defines\tgs.dm"
 #include "code\__defines\tgui.dm"
 #include "code\__defines\time.dm"

code/__defines/_macros.dm

@@ -23,6 +23,10 @@
 #define SPAN_VOTE(X) ("<span class='vote'>" + X + "</span>")
 #define SPAN_HEAR(X) ("<span class='hear'>" + X + "</span>")
 
+#define SPAN_RED(x) "<span style='color:[COLOR_RED]'>[x]</span>"
+#define SPAN_YELLOW(x) "<span style='color:[COLOR_YELLOW]'>[x]</span>"
+#define SPAN_GREEN(x) "<span style='color:[COLOR_GREEN]'>[x]</span>"
+
 #define SPAN_SIZE(size, text) ("<span style=\"font-size: [size]\">" + text + "</span>")
 
 #define SPAN_HIGHDANGER(X) (FONT_LARGE(SPAN_DANGER(X)))

code/__defines/misc.dm

@@ -477,8 +477,7 @@ example:
 #define CONTAINER_EMPTY		0
 #define CONTAINER_SINGLE	1
 #define CONTAINER_MANY		2
-//Misc text define. Does 4 spaces. Used as a makeshift tabulator.
-#define FOURSPACES "    "
+
 #define CLIENT_FROM_VAR(I) (ismob(I) ? I:client : (isclient(I) ? I : (istype(I, /datum/mind) ? I:current?:client : null)))
 
 // check_items/check_reagents/check_fruits return values

code/__defines/text.dm

@@ -0,0 +1,19 @@
+/// Macro from Lummox used to get height from a MeasureText proc
+#define WXH_TO_HEIGHT(x) text2num(copytext(x, findtextEx(x, "x") + 1))
+
+#define SMALL_FONTS(FONTSIZE, MSG) "<span style=\"font-family: 'Small Fonts'; -dm-text-outline: 1 black; font-size: [FONTSIZE]px;\">[MSG]</span>"
+
+
+//Since we do not have GLOB (yet), this will have to do
+var/regex/html_tags = regex(@"<.*?>", "g")
+var/regex/angular_brackets = regex(@"[<>]", "g")
+var/regex/filename_forbidden_chars = regex(@{""|[\\\n\t/?%*:|<>]|\.\."}, "g")
+
+/// Removes characters incompatible with file names.
+#define SANITIZE_FILENAME(text) (filename_forbidden_chars.Replace(text, ""))
+
+/// Simply removes the < and > characters, and limits the length of the message.
+#define STRIP_HTML_SIMPLE(text, limit) (angular_brackets.Replace(copytext(text, 1, limit), ""))
+
+/// Removes everything enclose in < and > inclusive of the bracket, and limits the length of the message.
+#define STRIP_HTML_FULL(text, limit) (html_tags.Replace(copytext(text, 1, limit), ""))

code/_helpers/text.dm

@@ -1,12 +1,3 @@
-#define SMALL_FONTS(FONTSIZE, MSG) "<span style=\"font-family: 'Small Fonts'; -dm-text-outline: 1 black; font-size: [FONTSIZE]px;\">[MSG]</span>"
-
-/// Macro from Lummox used to get height from a MeasureText proc
-#define WXH_TO_HEIGHT(x) text2num(copytext(x, findtextEx(x, "x") + 1))
-
-#define SPAN_RED(x) "<span style='color:[COLOR_RED]'>[x]</span>"
-#define SPAN_YELLOW(x) "<span style='color:[COLOR_YELLOW]'>[x]</span>"
-#define SPAN_GREEN(x) "<span style='color:[COLOR_GREEN]'>[x]</span>"
-
 /*
  * Holds procs designed to help with filtering text
  * Contains groups:
@@ -31,7 +22,49 @@
  * Text sanitization
  */
 
-//Used for preprocessing entered text
+/**
+ * Same as `strip_html_full`, but readds the newlines after the stripping
+ *
+ * Not SQL escaping safe, do not use alone for SQL sanitizing operations
+ *
+ * Returns a string
+ */
+/proc/strip_html_readd_newlines(text, limit = MAX_MESSAGE_LEN)
+	SHOULD_NOT_SLEEP(TRUE)
+	SHOULD_BE_PURE(TRUE)
+	//For some reason at the first replace the backslash doubles, go figure... Either way, this is how to match it
+	var/static/regex/regex_newlines = regex(@"(\\n)", "g")
+	//And this matches the two newlines after the HTML strip and encoding
+	var/static/regex/regex_newlines_readd = regex(@"(##)", "g")
+
+	//We make them double, so we know the # is not part of the special ascii encoding of characters
+	var/message = regex_newlines.Replace(text, @"\n\n")
+
+	//Strip the HTML from the message
+	message = strip_html_full(message, limit)
+
+	//Bring the newlines back
+	return regex_newlines_readd.Replace(message, "<br>")
+
+/// Runs STRIP_HTML_SIMPLE and sanitize.
+/proc/strip_html(text, limit = MAX_MESSAGE_LEN)
+	return sanitize_tg(STRIP_HTML_SIMPLE(text, limit))
+
+/// Runs STRIP_HTML_FULL and sanitize.
+/proc/strip_html_full(text, limit = MAX_MESSAGE_LEN)
+	return sanitize_tg(STRIP_HTML_FULL(text, limit))
+
+/// Runs byond's html encoding sanitization proc, after replacing new-lines and tabs for the # character.
+/// This is ported from tg, hence the name
+/proc/sanitize_tg(text)
+	var/static/regex/regex = regex(@"[\n\t]", "g")
+	return html_encode(regex.Replace(text, "#"))
+
+/**
+ * DEPRECATED, USE `strip_html_full` / `strip_html` / `sanitize_tg` / `strip_html_readd_newlines` WHERE POSSIBLE
+ *
+ * Used for preprocessing entered text
+ */
 /proc/sanitize(var/input, var/max_length = MAX_MESSAGE_LEN, var/encode = 1, var/trim = 1, var/extra = 1)
 	if(!input)
 		return

code/datums/uplink/announcements.dm

@@ -17,17 +17,17 @@
 	desc = "Causes a falsified [current_map.boss_name] Update. Triggers immediately after supplying additional data."
 
 /datum/uplink_item/abstract/announcements/fake_centcom/extra_args(var/mob/user)
-	var/title = sanitize(input("Enter your announcement title.", "Announcement Title") as null|text)
+	var/title = tgui_input_text(user, "Enter your announcement title.", "Announcement Title", encode = FALSE)
 	if(!title)
 		return
-	var/message = sanitize(input("Enter your announcement message.", "Announcement Title") as null|message)
+	var/message = tgui_input_text(user, "Enter your announcement message.", "Announcement Title", multiline = TRUE, encode = FALSE)
 	if(!message)
 		return
-	return list("title" = title, "message" = message)
+	return list("title" = strip_html_readd_newlines(title), "message" = strip_html_readd_newlines(message))
 
 /datum/uplink_item/abstract/announcements/fake_centcom/get_goods(var/obj/item/device/uplink/U, var/loc, var/mob/user, var/list/args)
-	command_announcement.Announce(args["message"], args["title"], do_newscast=1, do_print=1)
-	return 1
+	command_announcement.Announce(args["message"], args["title"], do_newscast=1, do_print=1, msg_sanitized=TRUE)
+	return TRUE
 
 /datum/uplink_item/abstract/announcements/fake_crew_arrival
 	name = "Crew Arrival Announcement/Records"

code/defines/procs/announce.dm

@@ -40,7 +40,7 @@
 
 	if(!msg_sanitized)
 		message = sanitize(message, extra = 0)
-	message_title = sanitizeSafe(message_title)
+		message_title = sanitizeSafe(message_title)
 
 	var/msg = FormMessage(message, message_title)
 	for(var/mob/M in player_list)

code/modules/modular_computers/computers/modular_computer/interaction.dm

@@ -334,7 +334,7 @@
 		if(!P || (P.program_state == PROGRAM_STATE_KILLED && P.service_state == PROGRAM_STATE_KILLED))
 			return
 		if(P.focused_conv)
-			P.focused_conv.cl_send(P, text, M)
+			P.focused_conv.cl_send(P, html_decode(text), M)
 	registered_message = text
 
 /obj/item/modular_computer/examine(mob/user, distance, is_adjacent)

html/changelogs/fluffyghost-fixantagannouncements.yml

@@ -0,0 +1,45 @@
+################################
+# Example Changelog File
+#
+# Note: This file, and files beginning with ".", and files that don't end in ".yml" will not be read. If you change this file, you will look really dumb.
+#
+# Your changelog will be merged with a master changelog. (New stuff added only, and only on the date entry for the day it was merged.)
+# When it is, any changes listed below will disappear.
+#
+# Valid Prefixes:
+#   bugfix
+#   wip (For works in progress)
+#   tweak
+#   soundadd
+#   sounddel
+#   rscadd (general adding of nice things)
+#   rscdel (general deleting of nice things)
+#   imageadd
+#   imagedel
+#   maptweak
+#   spellcheck (typo fixes)
+#   experiment
+#   balance
+#   admin
+#   backend
+#   security
+#   refactor
+#################################
+
+# Your name.
+author: FluffyGhost
+
+# Optional: Remove this file after generating master changelog.  Useful for PR changelogs that won't get used again.
+delete-after: True
+
+# Any changes you've made.  See valid prefix list above.
+# INDENT WITH TWO SPACES.  NOT TABS.  SPACES.
+# SCREW THIS UP AND IT WON'T WORK.
+# Also, all entries are changed into a single [] after a master changelog generation. Just remove the brackets when you add new entries.
+# Please surround your changes in  double quotes ("), as certain characters otherwise screws up compiling. The quotes will not show up in the changelog.
+changes:
+  - rscadd: "Added text manipulation macros from TG, and a new one to have newlines working."
+  - refactor: "Moved around some macros in the appropriate files."
+  - bugfix: "Antag uplink announcements now shows apostrophes correctly, and can have newlines in them."
+  - refactor: "Antag uplink announcements now uses tgui inputs."
+  - bugfix: "PDA STT now respects the apostrophes and other characters too."