1 files changed, 285 insertions, 0 deletions
diff --git a/dom/base/nsLineBreaker.h b/dom/base/nsLineBreaker.h
new file mode 100644
index 0000000000..a437bb7846
--- /dev/null
+++ b/dom/base/nsLineBreaker.h
@@ -0,0 +1,285 @@
+/* -*- Mode: C++; tab-width: 8; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
+/* vim: set ts=8 sts=2 et sw=2 tw=80: */
+/* This Source Code Form is subject to the terms of the Mozilla Public
+ * License, v. 2.0. If a copy of the MPL was not distributed with this
+ * file, You can obtain one at http://mozilla.org/MPL/2.0/. */
+
+#ifndef NSLINEBREAKER_H_
+#define NSLINEBREAKER_H_
+
+#include "mozilla/intl/LineBreaker.h"
+#include "mozilla/intl/Segmenter.h"
+#include "nsString.h"
+#include "nsTArray.h"
+
+class nsAtom;
+class nsHyphenator;
+
+/**
+ * A receiver of line break data.
+ */
+class nsILineBreakSink {
+ public:
+  /**
+   * Sets the break data for a substring of the associated text chunk.
+   * One or more of these calls will be performed; the union of all substrings
+   * will cover the entire text chunk. Substrings may overlap (i.e., we may
+   * set the break-before state of a character more than once).
+   * @param aBreakBefore the break-before states for the characters in the
+   * substring. These are enum values from gfxTextRun::CompressedGlyph:
+   *    FLAG_BREAK_TYPE_NONE     - no linebreak is allowed here
+   *    FLAG_BREAK_TYPE_NORMAL   - a normal (whitespace) linebreak
+   *    FLAG_BREAK_TYPE_HYPHEN   - a hyphenation point
+   */
+  virtual void SetBreaks(uint32_t aStart, uint32_t aLength,
+                         uint8_t* aBreakBefore) = 0;
+
+  /**
+   * Indicates which characters should be capitalized. Only called if
+   * BREAK_NEED_CAPITALIZATION was requested.
+   */
+  virtual void SetCapitalization(uint32_t aStart, uint32_t aLength,
+                                 bool* aCapitalize) = 0;
+};
+
+/**
+ * A line-breaking state machine. You feed text into it via AppendText calls
+ * and it computes the possible line breaks. Because break decisions can
+ * require a lot of context, the breaks for a piece of text are sometimes not
+ * known until later text has been seen (or all text ends). So breaks are
+ * returned via a call to SetBreaks on the nsILineBreakSink object passed
+ * with each text chunk, which might happen during the corresponding AppendText
+ * call, or might happen during a later AppendText call or even a Reset()
+ * call.
+ *
+ * The linebreak results MUST NOT depend on how the text is broken up
+ * into AppendText calls.
+ *
+ * The current strategy is that we break the overall text into
+ * whitespace-delimited "words". Then those words are passed to the LineBreaker
+ * service for deeper analysis if they contain a "complex" character as
+ * described below.
+ *
+ * This class also handles detection of which characters should be capitalized
+ * for text-transform:capitalize. This is a good place to handle that because
+ * we have all the context we need.
+ */
+class nsLineBreaker {
+ public:
+  nsLineBreaker();
+  ~nsLineBreaker();
+
+  static inline bool IsSpace(char16_t u) {
+    return mozilla::intl::NS_IsSpace(u);
+  }
+
+  static inline bool IsComplexASCIIChar(char16_t u) {
+    return !((0x0030 <= u && u <= 0x0039) || (0x0041 <= u && u <= 0x005A) ||
+             (0x0061 <= u && u <= 0x007A) || (0x000a == u));
+  }
+
+  static inline bool IsComplexChar(char16_t u) {
+    return IsComplexASCIIChar(u) ||
+           mozilla::intl::NS_NeedsPlatformNativeHandling(u) ||
+           (0x1100 <= u && u <= 0x11ff) ||  // Hangul Jamo
+           (0x2000 <= u && u <= 0x21ff) ||  // Punctuations and Symbols
+           (0x2e80 <= u && u <= 0xd7ff) ||  // several CJK blocks
+           (0xf900 <= u && u <= 0xfaff) ||  // CJK Compatibility Idographs
+           (0xff00 <= u && u <= 0xffef);    // Halfwidth and Fullwidth Forms
+  }
+
+  // Break opportunities exist at the end of each run of breakable whitespace
+  // (see IsSpace above). Break opportunities can also exist between pairs of
+  // non-whitespace characters, as determined by mozilla::intl::LineBreaker.
+  // We pass a whitespace-
+  // delimited word to LineBreaker if it contains at least one character
+  // matching IsComplexChar.
+  // We provide flags to control on a per-chunk basis where breaks are allowed.
+  // At any character boundary, exactly one text chunk governs whether a
+  // break is allowed at that boundary.
+  //
+  // We operate on text after whitespace processing has been applied, so
+  // other characters (e.g. tabs and newlines) may have been converted to
+  // spaces.
+
+  /**
+   * Flags passed with each chunk of text.
+   */
+  enum {
+    /*
+     * Do not introduce a break opportunity at the start of this chunk of text.
+     */
+    BREAK_SUPPRESS_INITIAL = 0x01,
+    /**
+     * Do not introduce a break opportunity in the interior of this chunk of
+     * text. Also, whitespace in this chunk is treated as non-breakable.
+     */
+    BREAK_SUPPRESS_INSIDE = 0x02,
+    /**
+     * The sink currently is already set up to have no breaks in it;
+     * if no breaks are possible, nsLineBreaker does not need to call
+     * SetBreaks on it. This is useful when handling large quantities of
+     * preformatted text; the textruns will never have any breaks set on them,
+     * and there is no need to ever actually scan the text for breaks, except
+     * at the end of textruns in case context is needed for following breakable
+     * text.
+     */
+    BREAK_SKIP_SETTING_NO_BREAKS = 0x04,
+    /**
+     * We need to be notified of characters that should be capitalized
+     * (as in text-transform:capitalize) in this chunk of text.
+     */
+    BREAK_NEED_CAPITALIZATION = 0x08,
+    /**
+     * Auto-hyphenation is enabled, so we need to get a hyphenator
+     * (if available) and use it to find breakpoints.
+     */
+    BREAK_USE_AUTO_HYPHENATION = 0x10
+  };
+
+  /**
+   * Append "invisible whitespace". This acts like whitespace, but there is
+   * no actual text associated with it. Only the BREAK_SUPPRESS_INSIDE flag
+   * is relevant here.
+   */
+  nsresult AppendInvisibleWhitespace(uint32_t aFlags);
+
+  /**
+   * Feed Unicode text into the linebreaker for analysis. aLength must be
+   * nonzero.
+   * @param aSink can be null if the breaks are not actually needed (we may
+   * still be setting up state for later breaks)
+   */
+  nsresult AppendText(nsAtom* aHyphenationLanguage, const char16_t* aText,
+                      uint32_t aLength, uint32_t aFlags,
+                      nsILineBreakSink* aSink);
+  /**
+   * Feed 8-bit text into the linebreaker for analysis. aLength must be nonzero.
+   * @param aSink can be null if the breaks are not actually needed (we may
+   * still be setting up state for later breaks)
+   */
+  nsresult AppendText(nsAtom* aHyphenationLanguage, const uint8_t* aText,
+                      uint32_t aLength, uint32_t aFlags,
+                      nsILineBreakSink* aSink);
+  /**
+   * Reset all state. This means the current run has ended; any outstanding
+   * calls through nsILineBreakSink are made, and all outstanding references to
+   * nsILineBreakSink objects are dropped.
+   * After this call, this linebreaker can be reused.
+   * This must be called at least once between any call to AppendText() and
+   * destroying the object.
+   * @param aTrailingBreak this is set to true when there is a break opportunity
+   * at the end of the text. This will normally only be declared true when there
+   * is breakable whitespace at the end.
+   */
+  nsresult Reset(bool* aTrailingBreak);
+
+  /*
+   * Set word-break mode for line breaker. This is set by word-break property.
+   */
+  void SetWordBreak(mozilla::intl::WordBreakRule aMode) {
+    // If current word is non-empty and mode is changing, flush the breaker.
+    if (aMode != mWordBreak && !mCurrentWord.IsEmpty()) {
+      nsresult rv = FlushCurrentWord();
+      if (NS_FAILED(rv)) {
+        NS_WARNING("FlushCurrentWord failed, line-breaks may be wrong");
+      }
+      // If previous mode was break-all, we should allow a break here.
+      // XXX (jfkthame) css-text spec seems unclear on this, raised question in
+      // https://github.com/w3c/csswg-drafts/issues/3897
+      if (mWordBreak == mozilla::intl::WordBreakRule::BreakAll) {
+        mBreakHere = true;
+      }
+    }
+    mWordBreak = aMode;
+  }
+
+  /*
+   * Set line-break rule strictness mode for line breaker. This is set by the
+   * line-break property.
+   */
+  void SetStrictness(mozilla::intl::LineBreakRule aMode) {
+    if (aMode != mLineBreak && !mCurrentWord.IsEmpty()) {
+      nsresult rv = FlushCurrentWord();
+      if (NS_FAILED(rv)) {
+        NS_WARNING("FlushCurrentWord failed, line-breaks may be wrong");
+      }
+      // If previous mode was anywhere, we should allow a break here.
+      if (mLineBreak == mozilla::intl::LineBreakRule::Anywhere) {
+        mBreakHere = true;
+      }
+    }
+    mLineBreak = aMode;
+  }
+
+  /**
+   * Return whether the line-breaker has a buffered "current word" that may
+   * be extended with additional word-forming characters.
+   */
+  bool InWord() const { return !mCurrentWord.IsEmpty(); }
+
+  /**
+   * Set the word-continuation state, which will suppress capitalization of
+   * the next letter that might otherwise apply.
+   */
+  void SetWordContinuation(bool aContinuation) {
+    mWordContinuation = aContinuation;
+  }
+
+ private:
+  // This is a list of text sources that make up the "current word" (i.e.,
+  // run of text which does not contain any whitespace). All the mLengths
+  // are are nonzero, these cannot overlap.
+  struct TextItem {
+    TextItem(nsILineBreakSink* aSink, uint32_t aSinkOffset, uint32_t aLength,
+             uint32_t aFlags)
+        : mSink(aSink),
+          mSinkOffset(aSinkOffset),
+          mLength(aLength),
+          mFlags(aFlags) {}
+
+    nsILineBreakSink* mSink;
+    uint32_t mSinkOffset;
+    uint32_t mLength;
+    uint32_t mFlags;
+  };
+
+  // State for the nonwhitespace "word" that started in previous text and hasn't
+  // finished yet.
+
+  // When the current word ends, this computes the linebreak opportunities
+  // *inside* the word (excluding either end) and sets them through the
+  // appropriate sink(s). Then we clear the current word state.
+  nsresult FlushCurrentWord();
+
+  void UpdateCurrentWordLanguage(nsAtom* aHyphenationLanguage);
+
+  void FindHyphenationPoints(nsHyphenator* aHyphenator,
+                             const char16_t* aTextStart,
+                             const char16_t* aTextLimit, uint8_t* aBreakState);
+
+  AutoTArray<char16_t, 100> mCurrentWord;
+  // All the items that contribute to mCurrentWord
+  AutoTArray<TextItem, 2> mTextItems;
+  nsAtom* mCurrentWordLanguage;
+  bool mCurrentWordContainsMixedLang;
+  bool mCurrentWordContainsComplexChar;
+  bool mScriptIsChineseOrJapanese;
+
+  // True if the previous character was breakable whitespace
+  bool mAfterBreakableSpace;
+  // True if a break must be allowed at the current position because
+  // a run of breakable whitespace ends here
+  bool mBreakHere;
+  // Break rules for letters from the "word-break" property.
+  mozilla::intl::WordBreakRule mWordBreak;
+  // Line breaking strictness from the "line-break" property.
+  mozilla::intl::LineBreakRule mLineBreak;
+  // Should the text be treated as continuing a word-in-progress (for purposes
+  // of initial capitalization)? Normally this is set to false whenever we
+  // start using a linebreaker, but it may be set to true if the line-breaker
+  // has been explicitly flushed mid-word.
+  bool mWordContinuation;
+};
+
+#endif /*NSLINEBREAKER_H_*/