diff options
Diffstat (limited to 'src/main')
-rw-r--r-- | src/main/model/MyModel.java | 5 | ||||
-rw-r--r-- | src/main/model/css/CssLexer.java | 63 | ||||
-rw-r--r-- | src/main/model/html/HtmlLexer.java | 68 | ||||
-rw-r--r-- | src/main/model/util/AbstractTree.java | 35 | ||||
-rw-r--r-- | src/main/model/util/Lexer.java | 58 |
5 files changed, 224 insertions, 5 deletions
diff --git a/src/main/model/MyModel.java b/src/main/model/MyModel.java deleted file mode 100644 index f9a3dd7..0000000 --- a/src/main/model/MyModel.java +++ /dev/null @@ -1,5 +0,0 @@ -package model; - -public class MyModel { - // delete or rename this class! -} diff --git a/src/main/model/css/CssLexer.java b/src/main/model/css/CssLexer.java new file mode 100644 index 0000000..657d3e1 --- /dev/null +++ b/src/main/model/css/CssLexer.java @@ -0,0 +1,63 @@ +package model.css; + +import java.util.ArrayList; + +/** + * This lexer splits an input by whitespace, brackets, and semicolons. + * Brackets and semicolons are included in the lexed output, whitespace is not. + * <br> + * CSS, thankfully, is far more rigid and less-forgiving of errors than HTMl. + * It also has multiple layers of fallback for errors: ranging from: "ignore this + * property", to "ignore this rule", to "this isn't fucking CSS" and ignore it all. + * <br> + * Still, even though we don't have to deal with garbage like escaped quotes (future edit: whoops, yes we do) and + * what not, we'll still implement our lexer with a for loop instead of split() for future optimizations. + */ +public class CssLexer { + + public static ArrayList<String> lex(String input) { + String token = ""; + ArrayList<String> tokens = new ArrayList<>(); + boolean inSingleQuotes = false; + boolean inDoubleQuotes = false; + char previous = '\0'; + + for (char i : input.toCharArray()) { + // i HATE fallthrough switch statements + switch (i) { + case '{': case '}': case ';': case ':': + case ' ': case '\n': case '\t': + if (!inSingleQuotes && !inDoubleQuotes) { + if (!token.equals("")) { + tokens.add(token); + token = ""; + } + switch (i) { + case '{': case '}': case ';': case ':': + tokens.add(Character.toString(i)); + break; + case ' ': case '\n': case '\t': + break; + } + } else { + token += i; + } + break; + // intentional use of footgun behavior + case '"': + if (previous != '\\') { + inDoubleQuotes = !inDoubleQuotes; + } + case '\'': + if (previous != '\\') { + inSingleQuotes = !inSingleQuotes; + } + default: + token += i; + break; + } + previous = i; + } + return tokens; + } +} 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; + } +} diff --git a/src/main/model/util/AbstractTree.java b/src/main/model/util/AbstractTree.java new file mode 100644 index 0000000..4c74732 --- /dev/null +++ b/src/main/model/util/AbstractTree.java @@ -0,0 +1,35 @@ +package model.util; + +import org.javatuples.*; + +import java.util.*; + +// Utility class for a general tree: we'll be using these a lot +public abstract class AbstractTree<T> { + + // An AbstractTree holds some kind of data; we'll want this to be generic + // e.g. a tag, attributes, a tag and attributes, etc + private T data; + // Since it's a tree every node also has children. + private ArrayList<AbstractTree<T>> children; + + // future implementations may want to consider adding an Optional<> parent; or an Optional<> prevSibling + + public T getData() { + return data; + } + + public ArrayList<AbstractTree<T>> getChildren() { + return children; + } + + // god so much boilerplate + public AbstractTree(T data, ArrayList<AbstractTree<T>> children) { + this.data = data; + this.children = children; + } + + public void addChild(AbstractTree<T> child) { + this.children.add(child); + } +} diff --git a/src/main/model/util/Lexer.java b/src/main/model/util/Lexer.java new file mode 100644 index 0000000..b35caa6 --- /dev/null +++ b/src/main/model/util/Lexer.java @@ -0,0 +1,58 @@ +package model.util; + +import java.util.*; + +// General-purpose Lexer +public class Lexer { + + // private static final Set<String> whitespace = new HashSet<String>(" ", "\n"); + + // unused, helper function for if we implement finding identifers longer than a character + private static int longestDelimiter(Set<String> delimiters) { + int longestDelimiter = 0; + for (String delimiter : delimiters) { + if (delimiter.length() > longestDelimiter) { + longestDelimiter = delimiter.length(); + } + } + return longestDelimiter; + } + + /** + * Lexes a "free-form" language. "free-form" has a specific meaning here that's important to preserve: + * "free-form" means that _additional_ whitespace characters do not affect the language: e.g. two newlines + * instead of one, four spaces instead of two, etc. They are _not_ "whitespace-insensitive", which is usually + * a misnomer. + * The name's a bit of a joke: free-form languages are generally referred to as whitespace-insensitive --> + * insensitive == rude. Jokes are funnier when you have to explain them. + * Also, insensitiveLex() and freeformLex() aren't really that good of names. + * + * NOTE: This lexer only works with single-character deliminators. + * TODO: deduplicate whitespace + */ + // public static ArrayList<String> rudeLex(String input, Set<Character> delimiters) {} + + /** + * We might as well implement a lexer for non-free-form languages, but whatever. We won't use it. + */ + public static ArrayList<String> sensitiveLex(String input, Set<Character> delimiters) { + // int longestDelimiter = longestDelimiter(delimiters); + + ArrayList<String> tokens = new ArrayList<String>(); + String currentToken = ""; + // terrible c-style for loop because we may need to manipulate the index in the future + for (int i = 0; i < input.length(); i++) { + char nextToken = input.charAt(i); + if (delimiters.contains(nextToken)) { + if (!currentToken.equals("")) { + tokens.add(currentToken); + } + tokens.add(Character.toString(nextToken)); + currentToken = ""; + } else { + currentToken += input.charAt(i); + } + } + return tokens; + } +} |