Working insertion of hash values; added a few labels in spec

tools/addlatexhash.dart

Index: tools/addlatexhash.dart
diff --git a/tools/addlatexhash.dart b/tools/addlatexhash.dart
old mode 100644
new mode 100755
index f79a0b32296cffdb983eeb279a2930d41389b341..a3d40559d0be2806b9df31cd435243120d7e38cb
--- a/tools/addlatexhash.dart
+++ b/tools/addlatexhash.dart
@@ -1,3 +1,4 @@
+#!/usr/bin/env dart
 // Copyright (c) 2014, the Dart project authors.  Please see the AUTHORS file
 // for details. All rights reserved. Use of this source code is governed by a
 // BSD-style license that can be found in the LICENSE file.
@@ -6,35 +7,38 @@
 // This is a very specialized tool which was created in order to support
 // adding hash values used as location markers in the LaTeX source of the
 // language specification.  It is intended to take its input file as the
-// first argument and the output file name as the second argument. From
-// docs/language a typical usage would be as follows:
+// first argument, an output file name as the second argument, and a
+// hash listing file name as the third argument. From docs/language a
+// typical usage would be as follows:
 //
-//   dart ../../tools/addlatexhash.dart dartLangSpec.tex tmp.tex
+//   dart ../../tools/addlatexhash.dart dartLangSpec.tex out.tex hash.txt
 //
-// This will yield a normalized variant tmp.tex of the language
-// specification with hash values filled in.  For more details, please
-// check the language specification source itself.
+// This will produce a normalized variant out.tex of the language
+// specification with hash values filled in, and a listing hash.txt of
+// all the hash values along with the label of their textual context
+// (section, subsection, subsubsection, paragraph) .  For more details,
+// please check the language specification source itself.
 //
 // NB: This utility assumes UN*X style line endings, \n, in the LaTeX
 // source file receieved as input; it will not work with other styles.
-//
-// TODO: The current version does not fill in hash values, it only
-// standardizes the LaTeX source by removing comments and normalizing
-// white space.
 
 import 'dart:io';
 import 'dart:convert';
+import '../pkg/utf/lib/utf.dart';
 import '../pkg/crypto/lib/crypto.dart';
 
-// Normalization of the text, i.e., removal or normalization
-// of elements that do not affect the output from latex
+// ----------------------------------------------------------------------
+// Normalization of the text: removal or normalization of parts that
+// do not affect the output from latex, such as white space.
 
-final commentRE = new RegExp(r"[^\\]%.*"); // NB: . does not match \n
+final commentRE = new RegExp(r"[^\\]%.*");  // NB: . does not match \n.
 final whitespaceAllRE = new RegExp(r"^\s+$");
 final whitespaceRE = new RegExp(r"[ \t]{2,}");

[Lasse Reichstein Nielsen, 2014/11/11 08:13:48]

Why is one "whitespace" using \s and the other [ \t]?
At least worth a comment.

[eernst, 2014/11/11 09:04:54]

That's indeed wrong --- the intention is to match white space except
end-of-line.  According to
http://regexlib.com/CheatSheet.aspx?AspxAutoDetectCookieSupport=1 this is
specified as [ \f\t\v], but according to
http://ecma-international.org/ecma-262/5.1/#sec-15.10 (referred from
https://api.dartlang.org/apidocs/channels/stable/dartdoc-viewer/dart:core.RegExp)
we have \s ~ WhiteSpace|LineTerminator, i.e., we want WhiteSpace,

  WhiteSpace :: <TAB>|<VT>|<FF>|<SP>|<NBSP>|<BOM>|<USP>;

which means that we would need to match <NBSP>, <BOM>, <USP> as well.  What is
the recommended approach to these (UTF-8-only) chars in a Dart RegExp context?

Currently I've used [ \f\t\v], and added a short comment.

[Lasse Reichstein Nielsen, 2014/11/11 09:13:23]

There is no special recommendation in a RegExp context, it only depends on the
use-case - i.e., they work, so if you want/need them, you can add them.

The (not really recommended because it's unreadable) RegExp-only way to match
white-space-but-not-line-terminator without listing them would be '(?:(?=\s).)'.
it exactly matches a single character that is both white-space and
non-line-terminator.

If you have already split into lines at line-terminators, you know that the line
contains no further line terminators, so using \s is safe to use on lines, but
not the entire text. A comment is enough in that case.

 
-// normalization steps
-
+/// Removes [match]ing part of [line], adjusting that part with the
+/// given [startOffset] and [endOffset], bounded to be valid indices
+/// into the string if needed, then inserts [glue] where text was
+/// removed.  If there is no match then [line] is returned.
 cutMatch(line, match, {startOffset: 0, endOffset: 0, glue: ""}) {
   if (match == null) return line;
   var start = match.start + startOffset;
@@ -52,6 +56,9 @@ cutRegexp(line, re, {startOffset: 0, endOffset: 0, glue: ""}) {
                   glue: glue);
 }
 
+/// Removes the rest of [line] starting from the beginning of the
+/// given [match], and adjusting with the given [offset].  If there
+/// is no match then [line] is returned.
 cutFromMatch(line, match, {offset: 0, glue: ""}) {
   if (match == null) return line;
   return line.substring(0, match.start + offset) + glue;
@@ -64,124 +71,476 @@ cutFromRegexp(line, re, {offset: 0, glue: ""}) {
 isWsOnly(line) => whitespaceAllRE.firstMatch(line) != null;
 isCommentOnly(line) => line.startsWith("%");
 
+/// Returns the end-of-line character at the end of [line], if any,
+/// otherwise returns the empty string.
 justEol(line) {
   return line.endsWith("\n") ? "\n" : "";
 }
 
+/// Removes the contents of the comment at the end of [line],
+/// leaving the "%" in place.  If no comment is present,
+/// return [line].
+///
+/// NB: it is tempting to remove everything from the '%' and out,
+/// including the final newline, if any, but this does not work.
+/// The problem is that TeX will do exactly this, but then it will
+/// add back a character that depends on its state (S, M, or N),
+/// and it is tricky to maintain a similar state that matches the
+/// state of TeX faithfully.  Hence, we remove the content of
+/// comments but do not remove the comments themselves, we just
+/// leave the '%' at the end of the line and let TeX manage its
+/// states in a way that does not differ from the file from before
+/// stripComment.
 stripComment(line) {
-  // NB: it is tempting to remove everything from the '%' and out,
-  // including the final newline, if any, but this does not work.
-  // The problem is that TeX will do exactly this, but then it will
-  // add back a character that depends on its state (S, M, or N),
-  // and it is tricky to maintain a similar state that matches the
-  // state of TeX faithfully.  Hence, we remove the content of
-  // comments but do not remove the comments themselves, we just
-  // leave the '%' at the end of the line and let TeX manage its
-  // states in a way that does not differ from the file from before
-  // stripComment
   if (isCommentOnly(line)) return "%\n";
   return cutRegexp(line, commentRE, startOffset: 2);
 }
 
-// Reduce a wsOnly line to its eol, remove leading ws
-// entirely, and reduce multiple ws chars to one
+/// Reduces a white-space-only [line] to its eol character,
+/// removes leading ws entirely, and reduces multiple
+/// white-space chars to one.
 normalizeWhitespace(line) {
   var trimLine = line.trimLeft();
   if (trimLine.isEmpty) return justEol(line);
   return trimLine.replaceAll(whitespaceRE, " ");
 }
 
-// Reduce sequences of >1 wsOnly lines to 1, and sequences of >1
-// commentOnly lines to 1; moreover, treat commentOnly lines as
-// wsOnly when occurring in wsOnly line blocks
+/// Reduces sequences of >1 white-space-only lines in [lines] to 1,
+/// and sequences of >1 comment-only lines to 1.  Treats comment-only
+/// lines as white-space-only when they occur in white-space-only
+/// line blocks.
 multilineNormalize(lines) {
-  var afterBlankLines = false; // does 'line' succeed >0 empty lines?
-  var afterCommentLines = false; // .. succeed >0 commentOnly lines?
+  var afterBlankLines = false;    // Does [line] succeed >0 empty lines?
+  var afterCommentLines = false;  // Does [line] succeed >0 commentOnly lines?
   var newLines = new List();
   for (var line in lines) {
     if (afterBlankLines && afterCommentLines) {
-      // can never happen
+      // Previous line was both blank and a comment: not possible.
       throw "Bug, please report to eernst@";
     } else if (afterBlankLines && !afterCommentLines) {
-      // at least one line before 'line' is wsOnly
+      // At least one line before [line] is wsOnly.
       if (!isWsOnly(line)) {
-        // blank line block ended
+        // Blank line block ended.
         afterCommentLines = isCommentOnly(line);
-        // special case: it seems to be safe to remove commentOnly lines
+        // Special case: It seems to be safe to remove commentOnly lines
         // after wsOnly lines, so the TeX state must be predictably right;
         // next line will then be afterCommentLines and be dropped, so
-        // we drop the entire comment block---which is very useful; we can
+        // we drop the entire comment block---which is very useful.  We can
         // also consider this comment line to be an empty line, such that
         // subsequent empty lines can be considered to be in a block of
-        // empty lines; note that almost all variants of this will break..
+        // empty lines.  Note that almost all variants of this breaks.
         if (afterCommentLines) {
-          // _current_ 'line' a commentOnly here
+          // _Current_ 'line' is a commentOnly here.
           afterBlankLines = true;
           afterCommentLines = false;
-          // and do not add 'line'
+          // Omit addition of [line].
         } else {
-          // after blanks, but current 'line' is neither blank nor comment
+          // After blanks, but current 'line' is neither blank nor comment.
           afterBlankLines = false;
           newLines.add(line);
         }
       } else {
-        // blank line block continues, do not add 'line'
+        // Blank line block continues, omit addition of [line].
       }
     } else if (!afterBlankLines && afterCommentLines) {
-      // at least one line before 'line' is commentOnly
+      // At least one line before [line] is commentOnly.
       if (!isCommentOnly(line)) {
-        // comment line block ended
+        // Comment block ended.
         afterBlankLines = isWsOnly(line);
         afterCommentLines = false;
         newLines.add(line);
       } else {
-        // comment line block continues, do not add 'line'
+        // Comment block continues, do not add [line].
       }
     } else {
       assert(!afterBlankLines && !afterCommentLines);
-      // no wsOnly or commentOnly lines preceed 'line'
+      // No wsOnly or commentOnly lines preceed [line].
       afterBlankLines = isWsOnly(line);
       afterCommentLines = isCommentOnly(line);
-      if (!afterCommentLines) newLines.add(line);
-      // else skipping commentOnly line after nonWs, nonComment text
+      if (!afterCommentLines) {
+        newLines.add(line);
+      } else {
+        // skip commentOnly line after nonWs/nonComment text.
+      }
     }
   }
   return newLines;
 }
 
-// Selecting the elements in the pipeline
-
+/// Selects the elements in the normalization pipeline.
 normalize(line) => normalizeWhitespace(stripComment(line));
+
+/// Selects the elements in the significant-spacing block
+/// normalization pipeline.
 sispNormalize(line) => stripComment(line);
 
-// Managing fragments with significant spacing
+// Managing fragments with significant spacing.
+
+final dartCodeBeginRE = new RegExp(r"^\s*\\begin\s*\{dartCode\}");
+final dartCodeEndRE = new RegExp (r"^\s*\\end\s*\{dartCode\}");
+
+/// Recognizes beginning of dartCode block.
+sispIsDartBegin(line) => line.contains(dartCodeBeginRE);
+
+/// Recognizes end of dartCode block.
+sispIsDartEnd(line) => line.contains(dartCodeEndRE);
+
+// ----------------------------------------------------------------------
+// Analyzing the input to point out "interesting" lines
+
+/// Returns the event information for [lines] as determined by the
+/// given [analyzer].  The method [analyzer.analyze] indicates that a
+/// line is "uninteresting" by returning null (i.e., no events here),
+/// and "interesting" lines may be characterized by [analysisFunc] via
+/// the returned event object.
+findEvents(lines, analyzer) {
+  var events = new List();
+  for (var line in lines) {
+    var event = analyzer.analyze(line);
+    if (event != null) events.add(event);
+  }
+  return events;
+}
+
+/// Returns RegExp text for recognizing a command occupying a line
+/// of its own, given the part of the RegExp that recognizes the
+/// command name, [cmdNameRE]
+lineCommandRE(cmdNameRE) =>
+    new RegExp(r"^\s*\\" + cmdNameRE + r"\s*\{.*\}\s*$");
+
+final hashLabelStartRE = new RegExp(r"^\s*\\LMLabel\s*\{");
+final hashLabelEndRE = new RegExp(r"\}\s*$");
+
+final hashMarkRE = lineCommandRE("LMHash");
+final hashLabelRE = lineCommandRE("LMLabel");
+final sectioningRE = lineCommandRE("((|sub(|sub))section|paragraph)");
+final sectionRE = lineCommandRE("section");
+final subsectionRE = lineCommandRE("subsection");
+final subsubsectionRE = lineCommandRE("subsubsection");
+final paragraphRE = lineCommandRE("paragraph");
+
+/// Returns true iff [line] begins a block of lines that gets a hash value.
+isHashMarker(line) => line.contains(hashMarkRE);
+
+/// Returns true iff [line] defines a sectioning label.
+isHashLabel(line) => line.contains(hashLabelRE);
+
+/// Returns true iff [line] is a sectioning command resp. one of its
+/// more specific forms; note that it is assumed that sectioning commands
+/// do not contain a newline between the command name and the '{'.
+isSectioningCommand(line) => line.contains(sectioningRE);
+isSectionCommand(line) => line.contains(sectionRE);
+isSubsectionCommand(line) => line.contains(subsectionRE);
+isSubsubsectionCommand(line) => line.contains(subsubsectionRE);
+isParagraphCommand(line) => line.contains(paragraphRE);
+
+/// Returns true iff [line] does not end a block of lines that gets
+/// a hash value.
+isntHashBlockTerminator(line) => !isSectioningCommand(line);
+
+/// Returns the label text part from [line], based on the assumption
+/// that isHashLabel(line) returns true.
+extractHashLabel(line) {
+  var startMatch = hashLabelStartRE.firstMatch(line);
+  var endMatch = hashLabelEndRE.firstMatch(line);
+  assert(startMatch != null && endMatch != null);
+  return line.substring(startMatch.end, endMatch.start);
+}
+
+// Event classes: Keep track of relevant information about the LaTeX
+// source code lines, such as where \LMHash and \LMLabel commands are
+// used, and how they are embedded in the sectioning structure.
+
+/// Abstract events, enabling us to [setEndLineNumber] on all events.
+abstract class HashEvent {
+  /// For events that have an endLineNumber, set it; otherwise ignore.
+  /// The endLineNumber specifies the end of the block of lines
+  /// associated with a given event, for event types concerned with
+  /// blocks of lines rather than single lines.
+  setEndLineNumber(n) {}
+
+  /// Returns null except for \LMHash{} events, where it returns
+  /// the startLineNumber.  This serves to specify a boundary because
+  /// the preceding \LMHash{} block should stop before the line of
+  /// this \LMHash{} command.  Note that hash blocks may stop earlier,
+  /// because they cannot contain sectioning commands.
+  getStartLineNumber() => null;
+}
+
+class HashMarkerEvent extends HashEvent {
+
+  // Line number of first line in block that gets hashed.
+  var startLineNumber;
+
+  // Highest possible number of first line after block that gets
+  // hashed (where the next \LMHash{} occurs).  Note that this value
+  // is not known initially (because that line has not yet been
+  // reached), so [endLineNumber] will be initialized in a separate
+  // scan.  Also note that the block may end earlier, because a block
+  // ends if it would otherwise include a sectioning command.
+  var endLineNumber;
+
+  HashMarkerEvent(this.startLineNumber);
+
+  setEndLineNumber(n) { endLineNumber = n; }
+  getStartLineNumber() => startLineNumber;
+}
+
+class HashLabelEvent extends HashEvent {
+  var labelText;
+  HashLabelEvent(this.labelText);
+}
+
+class HashAnalyzer {
+  // List of kinds of pending (= most recently seen) sectioning command.
+  // When updating this list, also update sectioningPrefix below.
+  static const PENDING_IS_NONE = 0;
+  static const PENDING_IS_SECTION = 1;
+  static const PENDING_IS_SUBSECTION = 2;
+  static const PENDING_IS_SUBSUBSECTION = 3;
+  static const PENDING_IS_PARAGRAPH = 1;
+
+  var lineNumber = 0;
+  var pendingSectioning = PENDING_IS_NONE;
+
+  HashAnalyzer();
+
+  setPendingToSection() {
+    pendingSectioning = PENDING_IS_SECTION;
+  }
+
+  setPendingToSubsection() {
+    pendingSectioning = PENDING_IS_SUBSECTION;
+  }
+
+  setPendingToSubsubsection() {
+    pendingSectioning = PENDING_IS_SUBSUBSECTION;
+  }
+
+  setPendingToParagraph() {
+    pendingSectioning = PENDING_IS_PARAGRAPH;
+  }
+
+  clearPending() {
+    pendingSectioning = PENDING_IS_NONE;
+  }
+
+  sectioningPrefix() {
+    switch (pendingSectioning) {
+      case PENDING_IS_SECTION: return "sec:";
+      case PENDING_IS_SUBSECTION: return "subsec:";
+      case PENDING_IS_SUBSUBSECTION: return "subsubsec:";
+      case PENDING_IS_PARAGRAPH: return "par:";
+      case PENDING_IS_NONE:
+        throw
+            "\\LMHash{..} should only be used after a sectioning command " +
+            "(\\section, \\subsection, \\subsubsection, \\paragraph)";
+      default:
+        // set of PENDING_IS_.. was extended, but updates here omitted
+        throw "Bug, please report to eernst@";
+    }
+  }
+
+  analyze(line) {
+    var currentLineNumber = lineNumber++;
+    if (isHashMarker(line)) {
+      return new HashMarkerEvent(currentLineNumber);
+    } else if (isHashLabel(line)) {
+      var labelText = sectioningPrefix() + extractHashLabel(line);
+      return new HashLabelEvent(labelText);
+    } else {
+      // No events to emit, but we may need to note state changes
+      if (isSectionCommand(line)) {
+        setPendingToSection();
+      } else if (isSubsectionCommand(line)) {
+        setPendingToSubsection();
+      } else if (isSubsubsectionCommand(line)) {
+        setPendingToSubsubsection();
+      } else if (isParagraphCommand(line)) {
+        setPendingToParagraph();
+      } else {
+        // No state changes.
+      }
+      return null;
+    }
+  }
+}
+
+findHashEvents(lines) {
+  // Create the list of events, omitting endLineNumbers.
+  var events = findEvents(lines, new HashAnalyzer());
+  // Set the endLineNumbers.
+  var currentEndLineNumber = lines.length;
+  for (var event in events.reversed) {
+    event.setEndLineNumber(currentEndLineNumber);
+    var nextEndLineNumber = event.getStartLineNumber();
+    if (nextEndLineNumber != null) currentEndLineNumber = nextEndLineNumber;
+  }
+  return events;
+}
+
+// ----------------------------------------------------------------------
+// Removal of non-normative elements of the text (rationale, commentary).
+
+/// Returns [line] without the command [cmdName] (based on a match
+/// on "\\cmdName\s*{..}") starting at [startIndex]; note that it is
+/// assumed but not checked that [line] contains "\\cmdType\s*{..",
+/// and note that the end of the {..} block is found via brace matching
+/// (i.e., nested {..} blocks are handled), but it may break if '{' is
+/// made an active character etc.etc.
+removeCommand(line, cmdName, startIndex) {
+  const BACKSLASH = 92;     // char code for '\\'.
+  const BRACE_BEGIN = 123;  // char code for '{'.
+  const BRACE_END = 125;    // char code for '}'.
+
+  var blockStartIndex = startIndex + cmdName.length + 1;
+  while (blockStartIndex < line.length &&
+         line.codeUnitAt(blockStartIndex) != BRACE_BEGIN) {
+    blockStartIndex++;
+  }
+  blockStartIndex++;
+  if (blockStartIndex > line.length) {
+    throw "Bug, please report to eernst@";
+  }
+  // [blockStartIndex] has index just after '{'.
+
+  var afterEscape = false;  // Is true iff [index] is just after '{'.
+  var braceLevel = 1;       // Have seen so many '{'s minus so many '}'s.
+
+  for (var index = blockStartIndex; index < line.length; index++) {
+    switch (line.codeUnitAt(index)) {
+      case BRACE_BEGIN:
+        if (afterEscape) {
+          afterEscape = false;
+        } else {
+          braceLevel++;
+        }
+        break;
+      case BRACE_END:
+        if (afterEscape) {
+          afterEscape = false;
+        } else {
+          braceLevel--;
+        }
+        break;
+      case BACKSLASH:
+        afterEscape = true;
+        break;
+      default:
+        afterEscape = false;
+    }
+    if (braceLevel == 0) {
+      return line.substring(0, startIndex) + line.substring(index + 1);
+    }
+  }
+  // Removal failed; we consider this to mean that the input is ill-formed.
+  throw "Unmatched braces";
+}
+
+final commentaryRE = new RegExp(r"\\commentary\s*\{");
+final rationaleRE = new RegExp(r"\\rationale\s*\{");
+
+/// Removes {}-balanced '\commentary{..}' commands from [line].
+removeCommentary(line) {
+  var match = commentaryRE.firstMatch(line);
+  if (match == null) return line;
+  return removeCommentary(removeCommand(line, r"commentary", match.start));
+}
+
+/// Removes {}-balanced '\rationale{..}' commands from [line].
+removeRationale(line) {
+  var match = rationaleRE.firstMatch(line);
+  if (match == null) return line;
+  return removeRationale(removeCommand(line, r"rationale", match.start));
+}
+
+/// Removes {}-balanced '\commentary{..}' and '\rationale{..}'
+/// commands from [line], then normalizes its white-space.
+simplifyLine(line) {
+  var simplerLine = removeCommentary(line);
+  simplerLine = removeRationale(simplerLine);
+  simplerLine = normalizeWhitespace(simplerLine);
+  return simplerLine;
+}
+
+// ----------------------------------------------------------------------
+// Recognition of line blocks, insertion of block hash into \LMHash{}.
 
-final dartCodeBeginRE = new RegExp(r"^\s*\\begin\{dartCode\}");
-final dartCodeEndRE = new RegExp (r"^\s*\\end\{dartCode\}");
+final latexArgumentRE = new RegExp(r"\{.*\}");
 
-sispIs(line, targetRE) {
-  return targetRE.firstMatch(line) != null;
+cleanupLine(line) => cutRegexp(line, commentRE, startOffset: 1).trimRight();
+
+/// Returns concatenation of all lines from [startIndex] in [lines] until
+/// a hash block terminator is encountered or [nextIndex] reached (if so,
+/// the line lines[nextIndex] itself is not included); each line is cleaned
+/// up using [cleanupLine], and " " is inserted between the lines gathered.
+gatherLines(lines, startIndex, nextIndex) =>
+    lines.getRange(startIndex, nextIndex)
+      .takeWhile(isntHashBlockTerminator)
+      .map(cleanupLine)
+      .join(" ");
+
+/// Computes the hash value for the line block starting at [startIndex]
+/// in [lines], stopping just before [nextIndex].  SIDE EFFECT:
+/// Outputs the simplified text and its hash value to [listSink].
+computeHashValue(lines, startIndex, nextIndex, listSink) {
+  final hashEncoder = new SHA1();
+  final gatheredLine = gatherLines(lines, startIndex, nextIndex);
+  final simplifiedLine = simplifyLine(gatheredLine);
+  listSink.write("  % $simplifiedLine\n");
+  hashEncoder.add(encodeUtf8(simplifiedLine));
+  return hashEncoder.close();
 }
 
-sispIsDartBegin(line) => sispIs(line, dartCodeBeginRE);
-sispIsDartEnd(line) => sispIs(line, dartCodeEndRE);
+computeHashString(lines, startIndex, nextIndex, listSink) =>
+    CryptoUtils.bytesToHex(computeHashValue(lines,
+                                            startIndex,
+                                            nextIndex,
+                                            listSink));
 
-// Transform input file into output file
+/// Computes and adds hashes to \LMHash{} lines in [lines] (which
+/// must be on the line numbers specified in [hashEvents]), and emits
+/// sectioning markers and hash values to [listSink], along with
+/// "comments" containing the simplified text (using the format
+/// '  % <text>', where the text is one, long line, for easy grepping
+/// etc.).
+addHashMarks(lines, hashEvents, listSink) {
+  for (var hashEvent in hashEvents) {
+    if (hashEvent is HashMarkerEvent) {
+      var start = hashEvent.startLineNumber;
+      var end = hashEvent.endLineNumber;
+      final hashValue = computeHashString(lines, start + 1, end, listSink);
+      lines[start] =
+          lines[start].replaceAll(latexArgumentRE, "{" + hashValue + "}");
+      listSink.write("  $hashValue\n");
+    } else if (hashEvent is HashLabelEvent) {
+      listSink.write("${hashEvent.labelText}\n");
+    }
+  }
+}
 
+/// Transforms LaTeX input to LaTeX output plus hash value list file.
 main ([args]) {
-  if (args.length != 2) {
-    print("Usage: addlatexhash.dart <input-file> <output-file>");
-    throw "Received ${args.length} arguments, expected two";
+  if (args.length != 3) {
+    print("Usage: addlatexhash.dart <input-file> <output-file> <list-file>");
+    throw "Received ${args.length} arguments, expected three";
   }
 
+  // Get LaTeX source.
   var inputFile = new File(args[0]);
-  var outputFile = new File(args[1]);
   assert(inputFile.existsSync());
-
   var lines = inputFile.readAsLinesSync();
-  // single-line normalization
+
+  // Will hold LaTeX source with normalized spacing etc., plus hash values.
+  var outputFile = new File(args[1]);
+
+  // Will hold hierarchical list of hash values.
+  var listFile = new File(args[2]);
+  var listSink = listFile.openWrite();
+
+  // Perform single-line normalization.
   var inDartCode = false;
-  var newLines = new List();
+  var normalizedLines = new List();
 
   for (var line in lines) {
     if (sispIsDartBegin(line)) {
@@ -190,15 +549,20 @@ main ([args]) {
       inDartCode = false;
     }
     if (inDartCode) {
-      newLines.add(sispNormalize(line + "\n"));
+      normalizedLines.add(sispNormalize(line + "\n"));
     } else {
-      newLines.add(normalize(line + "\n"));
+      normalizedLines.add(normalize(line + "\n"));
     }
   }
 
-  // multi-line normalization
-  newLines = multilineNormalize(newLines);
+  // Perform multi-line normalization.
+  normalizedLines = multilineNormalize(normalizedLines);
+
+  // Insert hash values.
+  var hashEvents = findHashEvents(normalizedLines);
+  addHashMarks(normalizedLines, hashEvents, listSink);
 
-  // output result
-  outputFile.writeAsStringSync(newLines.join());
+  // Produce/finalize output.
+  outputFile.writeAsStringSync(normalizedLines.join());
+  listSink.close();
 }