From 32eab04d662b0c1128e64a4b91fb81f5f2be5a95 Mon Sep 17 00:00:00 2001
From: Jeroen van Rijn <Kelimion@users.noreply.github.com>
Date: Wed, 1 Dec 2021 03:15:44 +0100
Subject: [xml] Allow multi-line bodies w/o CDATA. Strip trailing whitespace.

---
 core/encoding/xml/debug_print.odin |  9 ++++++---
 core/encoding/xml/tokenizer.odin   | 21 ++++++++++++++++++---
 2 files changed, 24 insertions(+), 6 deletions(-)

(limited to 'core/encoding/xml')

diff --git a/core/encoding/xml/debug_print.odin b/core/encoding/xml/debug_print.odin
index c4d6875cc..65b71e30b 100644
--- a/core/encoding/xml/debug_print.odin
+++ b/core/encoding/xml/debug_print.odin
@@ -27,10 +27,13 @@ print :: proc(writer: io.Writer, doc: ^Document) -> (written: int, err: io.Error
 	}
 
 	written += wprintf(writer, "[Encoding] %v\n", doc.encoding)
-	written += wprintf(writer, "[DOCTYPE]  %v\n", doc.doctype.ident)
 
-	if len(doc.doctype.rest) > 0 {
-	 	wprintf(writer, "\t%v\n", doc.doctype.rest)
+	if len(doc.doctype.ident) > 0 {
+		written += wprintf(writer, "[DOCTYPE]  %v\n", doc.doctype.ident)
+
+		if len(doc.doctype.rest) > 0 {
+		 	wprintf(writer, "\t%v\n", doc.doctype.rest)
+		}
 	}
 
 	if doc.root != nil {
diff --git a/core/encoding/xml/tokenizer.odin b/core/encoding/xml/tokenizer.odin
index 3dcffb0d6..e453552b8 100644
--- a/core/encoding/xml/tokenizer.odin
+++ b/core/encoding/xml/tokenizer.odin
@@ -205,7 +205,7 @@ scan_identifier :: proc(t: ^Tokenizer) -> string {
 	return string(t.src[offset : t.offset])
 }
 
-scan_string :: proc(t: ^Tokenizer, offset: int, close: rune = '<', consume_close := false) -> (value: string, err: Error) {
+scan_string :: proc(t: ^Tokenizer, offset: int, close: rune = '<', consume_close := false, multiline := true) -> (value: string, err: Error) {
 	err = .None
 	in_cdata := false
 
@@ -238,7 +238,7 @@ scan_string :: proc(t: ^Tokenizer, offset: int, close: rune = '<', consume_close
 			}
 
 		case '\n':
-			if !in_cdata {
+			if !(multiline || in_cdata) {
 				error(t, offset, string(t.src[offset : t.offset]))
 				error(t, offset, "[scan_string] Not terminated\n")
 				err = .Invalid_Tag_Value
@@ -256,7 +256,22 @@ scan_string :: proc(t: ^Tokenizer, offset: int, close: rune = '<', consume_close
 		advance_rune(t)
 	}
 
+	/*
+		Strip trailing whitespace.
+	*/
 	lit := string(t.src[offset : t.offset])
+
+	end := len(lit)
+	eat: for ; end > 0; end -= 1 {
+		ch := lit[end - 1]
+		switch ch {
+		case ' ', '\t', '\r', '\n':
+		case:
+			break eat
+		}
+	}
+	lit = lit[:end]
+
 	if consume_close {
 		advance_rune(t)
 	}
@@ -307,7 +322,7 @@ scan :: proc(t: ^Tokenizer) -> Token {
 		case ':': kind = .Colon
 
 		case '"', '\'':
-			lit, err = scan_string(t, t.offset, ch, true)
+			lit, err = scan_string(t, t.offset, ch, true, false)
 			if err == .None {
 				kind = .String
 			} else {
-- 
cgit v1.2.3