Basic prototypes of HTML/CSS lexers

1 files changed, 68 insertions, 0 deletions

diff --git a/src/main/model/html/HtmlLexer.java b/src/main/model/html/HtmlLexer.java
new file mode 100644
index 0000000..8cad425
--- /dev/null
+++ b/src/main/model/html/HtmlLexer.java
@@ -0,0 +1,68 @@
+package model.html;
+
+import java.util.ArrayList;
+
+/**
+ * We'll tokenize HTML by tags: disregarding the contents of the tag and attributes within the tag.
+ * The file is also considered to be free-form here: whitespace duplicates are disregarded.
+ */
+public class HtmlLexer {
+
+    // Takes a String of raw HTML, and tokenizes it for our parser.
+    public static ArrayList<String> lex(String input) {
+        String token = "";
+        ArrayList<String> tokens = new ArrayList<>();
+        boolean inTag = false;
+        boolean inSingleQuotes = false;
+        boolean inDoubleQuotes = false;
+
+        for (char i : input.toCharArray()) {
+            token += i;
+            switch (i) {
+                case '<':
+                    if (!inSingleQuotes && !inDoubleQuotes) {
+                        inTag = true;
+                        if (!token.equals("<")) {
+                            tokens.add(token.substring(0, token.length() - 1));
+                            token = "<";
+                        }
+                    } else if (inTag) {
+                        System.out.printf("Probably failing parser");
+                    }
+                    break;
+                case '>':
+                    if (!inSingleQuotes && !inDoubleQuotes) {
+                        if (!inTag) {
+                            System.out.printf("Probably failing parser");
+                        }
+                        inTag = false;
+                        tokens.add(token);
+                        token = "";
+                    }
+                    break;
+                case '"':
+                    if (!inSingleQuotes) {
+                        inDoubleQuotes = !inDoubleQuotes;
+                    }
+                    break;
+                case '\'':
+                    if (!inDoubleQuotes) {
+                        inSingleQuotes = !inSingleQuotes;
+                    }
+                    break;
+            }
+        }
+        /**
+         * When lexing invalid HTML: we may end up with trailing garbage: either an unfinished tag or extra text
+         * (those are the only two options since this is just the lex step)
+         */
+        if (!token.equals("")) {
+            if (inTag) {
+                tokens.add(token + ">");
+            } else {
+                tokens.add(token);
+            }
+        }
+        return tokens;
+    }
+}