fix: pay attention to comments and processing instructions in DTDs

Closes #19

lib/saxes.js

@@ -24,14 +24,21 @@ const XML_ENTITIES = {
 
 const S_INITIAL = "sInitial"; // initial state
 const S_BEGIN_WHITESPACE = "sBeginWhitespace"; // leading whitespace
+const S_DOCTYPE = "sDoctype"; // <!DOCTYPE
+const S_DOCTYPE_QUOTE = "sDoctypeQuote"; // <!DOCTYPE "//blah
+const S_DTD = "sDTD"; // <!DOCTYPE "//blah" [ ...
+const S_DTD_QUOTED = "sDTDQuoted"; // <!DOCTYPE "//blah" [ "foo
+const S_DTD_OPEN_WAKA = "sDTDOpenWaka";
+const S_DTD_OPEN_WAKA_BANG = "sDTDOpenWakaBang";
+const S_DTD_COMMENT = "sDTDComment"; // <!--
+const S_DTD_COMMENT_ENDING = "sDTDCommentEnding"; // <!-- blah -
+const S_DTD_COMMENT_ENDED = "sDTDCommentEnded"; // <!-- blah --
+const S_DTD_PI = "sDTDPI"; // <?
+const S_DTD_PI_ENDING = "sDTDPIEnding"; // <?hi "there" ?
 const S_TEXT = "sText"; // general stuff
 const S_ENTITY = "sEntity"; // &amp and such
 const S_OPEN_WAKA = "sOpenWaka"; // <
 const S_OPEN_WAKA_BANG = "sOpenWakaBang"; // <!...
-const S_DOCTYPE = "sDoctype"; // <!DOCTYPE
-const S_DOCTYPE_QUOTED = "sDoctypeQuoted"; // <!DOCTYPE "//blah
-const S_DOCTYPE_DTD = "sDoctypeDTD"; // <!DOCTYPE "//blah" [ ...
-const S_DOCTYPE_DTD_QUOTED = "sDoctypeDTDQuoted"; // <!DOCTYPE "//blah" [ "foo
 const S_COMMENT = "sComment"; // <!--
 const S_COMMENT_ENDING = "sCommentEnding"; // <!-- blah -
 const S_COMMENT_ENDED = "sCommentEnded"; // <!-- blah --
@@ -102,7 +109,7 @@ function isQuote(c) {
 const QUOTES = [DQUOTE, SQUOTE];
 
 const DOCTYPE_TERMINATOR = [...QUOTES, OPEN_BRACKET, GREATER];
-const DOCTYPE_DTD_TERMINATOR = [...QUOTES, CLOSE_BRACKET];
+const DTD_TERMINATOR = [...QUOTES, LESS, CLOSE_BRACKET];
 const XML_DECL_NAME_TERMINATOR = [EQUAL, QUESTION, ...S_LIST];
 const ATTRIB_VALUE_UNQUOTED_TERMINATOR = [...S_LIST, GREATER, AMP, LESS];
 
@@ -1027,17 +1034,17 @@ class SaxesParser {
     else if (c) {
       this.doctype += String.fromCodePoint(c);
       if (c === OPEN_BRACKET) {
-        this.state = S_DOCTYPE_DTD;
+        this.state = S_DTD;
       }
       else if (isQuote(c)) {
-        this.state = S_DOCTYPE_QUOTED;
+        this.state = S_DOCTYPE_QUOTE;
         this.q = c;
       }
     }
   }
 
   /** @private */
-  sDoctypeQuoted() {
+  sDoctypeQuote() {
     const { q } = this;
     if (this.captureToChar(q, "doctype")) {
       this.doctype += String.fromCodePoint(q);
@@ -1047,8 +1054,8 @@ class SaxesParser {
   }
 
   /** @private */
-  sDoctypeDTD() {
-    const c = this.captureTo(DOCTYPE_DTD_TERMINATOR, "doctype");
+  sDTD() {
+    const c = this.captureTo(DTD_TERMINATOR, "doctype");
     if (!c) {
       return;
     }
@@ -1057,22 +1064,100 @@ class SaxesParser {
     if (c === CLOSE_BRACKET) {
       this.state = S_DOCTYPE;
     }
+    else if (c === LESS) {
+      this.state = S_DTD_OPEN_WAKA;
+    }
     else if (isQuote(c)) {
-      this.state = S_DOCTYPE_DTD_QUOTED;
+      this.state = S_DTD_QUOTED;
       this.q = c;
     }
   }
 
   /** @private */
-  sDoctypeDTDQuoted() {
+  sDTDQuoted() {
     const { q } = this;
     if (this.captureToChar(q, "doctype")) {
       this.doctype += String.fromCodePoint(q);
-      this.state = S_DOCTYPE_DTD;
+      this.state = S_DTD;
       this.q = null;
     }
   }
 
+  /** @private */
+  sDTDOpenWaka() {
+    const c = this.getCode();
+    this.doctype += String.fromCodePoint(c);
+    switch (c) {
+    case BANG:
+      this.state = S_DTD_OPEN_WAKA_BANG;
+      this.openWakaBang = "";
+      break;
+    case QUESTION:
+      this.state = S_DTD_PI;
+      break;
+    default:
+      this.state = S_DTD;
+    }
+  }
+
+  /** @private */
+  sDTDOpenWakaBang() {
+    const char = String.fromCodePoint(this.getCode());
+    const owb = this.openWakaBang += char;
+    this.doctype += char;
+    if (owb !== "-") {
+      this.state = owb === "--" ? S_DTD_COMMENT : S_DTD;
+      this.openWakaBang = "";
+    }
+  }
+
+  /** @private */
+  sDTDComment() {
+    if (this.captureToChar(MINUS, "doctype")) {
+      this.doctype += "-";
+      this.state = S_DTD_COMMENT_ENDING;
+    }
+  }
+
+  /** @private */
+  sDTDCommentEnding() {
+    const c = this.getCode();
+    this.doctype += String.fromCodePoint(c);
+    this.state = c === MINUS ? S_DTD_COMMENT_ENDED : S_DTD_COMMENT;
+  }
+
+  /** @private */
+  sDTDCommentEnded() {
+    const c = this.getCode();
+    this.doctype += String.fromCodePoint(c);
+    if (c === GREATER) {
+      this.state = S_DTD;
+    }
+    else {
+      this.fail("malformed comment.");
+      // <!-- blah -- bloo --> will be recorded as
+      // a comment of " blah -- bloo "
+      this.state = S_DTD_COMMENT;
+    }
+  }
+
+  /** @private */
+  sDTDPI() {
+    if (this.captureToChar(QUESTION, "doctype")) {
+      this.doctype += "?";
+      this.state = S_DTD_PI_ENDING;
+    }
+  }
+
+  /** @private */
+  sDTDPIEnding() {
+    const c = this.getCode();
+    this.doctype += String.fromCodePoint(c);
+    if (c === GREATER) {
+      this.state = S_DTD;
+    }
+  }
+
   /** @private */
   sComment() {
     if (this.captureToChar(MINUS, "comment")) {

test/dtd.js

@@ -0,0 +1,65 @@
+"use strict";
+
+const { test } = require(".");
+
+describe("dtd", () => {
+  test({
+    name: "DTD with comment containing a quote",
+    xml: `\
+<?xml version="1.0" encoding="UTF-8" ?>
+<!DOCTYPE root [
+<!-- I'm a test. -->
+]>
+<root/>`,
+    expect: [
+      ["text", "\n"],
+      ["doctype", ` root [
+<!-- I'm a test. -->
+]`],
+      ["text", "\n"],
+      ["opentagstart", { name: "root", attributes: {} }],
+      ["opentag", { name: "root", attributes: {}, isSelfClosing: true }],
+      ["closetag", { name: "root", attributes: {}, isSelfClosing: true }],
+    ],
+  });
+
+  test({
+    name: "DTD with processing instruction containing a quote",
+    xml: `\
+<?xml version="1.0" encoding="UTF-8" ?>
+<!DOCTYPE root [
+<? I'm a test. ?>
+]>
+<root/>`,
+    expect: [
+      ["text", "\n"],
+      ["doctype", ` root [
+<? I'm a test. ?>
+]`],
+      ["text", "\n"],
+      ["opentagstart", { name: "root", attributes: {} }],
+      ["opentag", { name: "root", attributes: {}, isSelfClosing: true }],
+      ["closetag", { name: "root", attributes: {}, isSelfClosing: true }],
+    ],
+  });
+
+  test({
+    name: "DTD with ]> in a string",
+    xml: `\
+<?xml version="1.0" encoding="UTF-8" ?>
+<!DOCTYPE root [
+<!NOTATION not1 SYSTEM "]>">
+]>
+<root/>`,
+    expect: [
+      ["text", "\n"],
+      ["doctype", ` root [
+<!NOTATION not1 SYSTEM "]>">
+]`],
+      ["text", "\n"],
+      ["opentagstart", { name: "root", attributes: {} }],
+      ["opentag", { name: "root", attributes: {}, isSelfClosing: true }],
+      ["closetag", { name: "root", attributes: {}, isSelfClosing: true }],
+    ],
+  });
+});