Merge branch 'master' into layoutgraph

1 files changed, 140 insertions, 0 deletions

diff --git a/src/org/jsoup/helper/StringUtil.java b/src/org/jsoup/helper/StringUtil.java
new file mode 100644
index 0000000000..071a92c7a5
--- /dev/null
+++ b/src/org/jsoup/helper/StringUtil.java
@@ -0,0 +1,140 @@
+package org.jsoup.helper;
+
+import java.util.Collection;
+import java.util.Iterator;
+
+/**
+ * A minimal String utility class. Designed for internal jsoup use only.
+ */
+public final class StringUtil {
+    // memoised padding up to 10
+    private static final String[] padding = {"", " ", "  ", "   ", "    ", "     ", "      ", "       ", "        ", "         ", "          "};
+
+    /**
+     * Join a collection of strings by a seperator
+     * @param strings collection of string objects
+     * @param sep string to place between strings
+     * @return joined string
+     */
+    public static String join(Collection strings, String sep) {
+        return join(strings.iterator(), sep);
+    }
+
+    /**
+     * Join a collection of strings by a seperator
+     * @param strings iterator of string objects
+     * @param sep string to place between strings
+     * @return joined string
+     */
+    public static String join(Iterator strings, String sep) {
+        if (!strings.hasNext())
+            return "";
+
+        String start = strings.next().toString();
+        if (!strings.hasNext()) // only one, avoid builder
+            return start;
+
+        StringBuilder sb = new StringBuilder(64).append(start);
+        while (strings.hasNext()) {
+            sb.append(sep);
+            sb.append(strings.next());
+        }
+        return sb.toString();
+    }
+
+    /**
+     * Returns space padding
+     * @param width amount of padding desired
+     * @return string of spaces * width
+     */
+    public static String padding(int width) {
+        if (width < 0)
+            throw new IllegalArgumentException("width must be > 0");
+
+        if (width < padding.length)
+            return padding[width];
+
+        char[] out = new char[width];
+        for (int i = 0; i < width; i++)
+            out[i] = ' ';
+        return String.valueOf(out);
+    }
+
+    /**
+     * Tests if a string is blank: null, emtpy, or only whitespace (" ", \r\n, \t, etc)
+     * @param string string to test
+     * @return if string is blank
+     */
+    public static boolean isBlank(String string) {
+        if (string == null || string.length() == 0)
+            return true;
+
+        int l = string.length();
+        for (int i = 0; i < l; i++) {
+            if (!StringUtil.isWhitespace(string.codePointAt(i)))
+                return false;
+        }
+        return true;
+    }
+
+    /**
+     * Tests if a string is numeric, i.e. contains only digit characters
+     * @param string string to test
+     * @return true if only digit chars, false if empty or null or contains non-digit chrs
+     */
+    public static boolean isNumeric(String string) {
+        if (string == null || string.length() == 0)
+            return false;
+
+        int l = string.length();
+        for (int i = 0; i < l; i++) {
+            if (!Character.isDigit(string.codePointAt(i)))
+                return false;
+        }
+        return true;
+    }
+
+    /**
+     * Tests if a code point is "whitespace" as defined in the HTML spec.
+     * @param c code point to test
+     * @return true if code point is whitespace, false otherwise
+     */
+    public static boolean isWhitespace(int c){
+        return c == ' ' || c == '\t' || c == '\n' || c == '\f' || c == '\r';
+    }
+
+    public static String normaliseWhitespace(String string) {
+        StringBuilder sb = new StringBuilder(string.length());
+
+        boolean lastWasWhite = false;
+        boolean modified = false;
+
+        int l = string.length();
+        for (int i = 0; i < l; i++) {
+            int c = string.codePointAt(i);
+            if (isWhitespace(c)) {
+                if (lastWasWhite) {
+                    modified = true;
+                    continue;
+                }
+                if (c != ' ')
+                    modified = true;
+                sb.append(' ');
+                lastWasWhite = true;
+            }
+            else {
+                sb.appendCodePoint(c);
+                lastWasWhite = false;
+            }
+        }
+        return modified ? sb.toString() : string;
+    }
+
+    public static boolean in(String needle, String... haystack) {
+        for (String hay : haystack) {
+            if (hay.equals(needle))
+            return true;
+        }
+        return false;
+    }
+}