liblouis/lou_translateString.c

/* liblouis Braille Translation and Back-Translation Library

   Based on the Linux screenreader BRLTTY, copyright (C) 1999-2006 by The
   BRLTTY Team

   Copyright (C) 2004, 2005, 2006 ViewPlus Technologies, Inc. www.viewplus.com
   Copyright (C) 2004, 2005, 2006 JJB Software, Inc. www.jjb-software.com
   Copyright (C) 2016 Mike Gray, American Printing House for the Blind
   Copyright (C) 2016 Davy Kager, Dedicon

   This file is part of liblouis.

   liblouis is free software: you can redistribute it and/or modify it
   under the terms of the GNU Lesser General Public License as published
   by the Free Software Foundation, either version 2.1 of the License, or
   (at your option) any later version.

   liblouis is distributed in the hope that it will be useful, but
   WITHOUT ANY WARRANTY; without even the implied warranty of
   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
   Lesser General Public License for more details.

   You should have received a copy of the GNU Lesser General Public
   License along with liblouis. If not, see <http://www.gnu.org/licenses/>.
*/

/**
 * @file
 * @brief Translate to braille
 */

#include <stdio.h>
#include <stdlib.h>
#include <string.h>

#include "internal.h"

/* additional bits in typebuf */
#define SYLLABLE_MARKER_1 0x2000
#define SYLLABLE_MARKER_2 0x4000
#define CAPSEMPH 0x8000

#define EMPHASIS 0x3fff  // all typeform bits that can be used

/* bits for wordBuffer */
#define WORD_CHAR 0x00000001
#define WORD_RESET 0x00000002
#define WORD_STOP 0x00000004
#define WORD_WHOLE 0x00000008

typedef struct {
	int size;
	widechar **buffers;
	int *inUse;
	widechar *(*alloc)(int index, int length);
	void (*free)(widechar *);
} StringBufferPool;

static widechar *
allocStringBuffer(int index, int length) {
	return _lou_allocMem(alloc_passbuf, index, 0, length);
}

static const StringBufferPool *stringBufferPool = NULL;

static void
initStringBufferPool() {
	static widechar *stringBuffers[MAXPASSBUF] = { NULL };
	static int stringBuffersInUse[MAXPASSBUF] = { 0 };
	StringBufferPool *pool = malloc(sizeof(StringBufferPool));
	pool->size = MAXPASSBUF;
	pool->buffers = stringBuffers;
	pool->inUse = stringBuffersInUse;
	pool->alloc = &allocStringBuffer;
	pool->free = NULL;
	stringBufferPool = pool;
}

static int
getStringBuffer(int length) {
	int i;
	for (i = 0; i < stringBufferPool->size; i++) {
		if (!stringBufferPool->inUse[i]) {
			stringBufferPool->buffers[i] = stringBufferPool->alloc(i, length);
			stringBufferPool->inUse[i] = 1;
			return i;
		}
	}
	_lou_outOfMemory();
	return -1;
}

static int
releaseStringBuffer(int idx) {
	if (idx >= 0 && idx < stringBufferPool->size) {
		int inUse = stringBufferPool->inUse[idx];
		if (inUse && stringBufferPool->free)
			stringBufferPool->free(stringBufferPool->buffers[idx]);
		stringBufferPool->inUse[idx] = 0;
		return inUse;
	}
	return 0;
}

typedef struct {
	int bufferIndex;
	const widechar *chars;
	int length;
} InString;

typedef struct {
	int bufferIndex;
	widechar *chars;
	int maxlength;
	int length;
} OutString;

typedef struct {
	int startMatch;
	int startReplace;
	int endReplace;
	int endMatch;
} PassRuleMatch;

static int
putCharacter(widechar c, const TranslationTableHeader *table,
		const DisplayTableHeader *displayTable, int pos, const InString *input,
		OutString *output, int *posMapping, int *cursorPosition, int *cursorStatus,
		int mode);
static int
passDoTest(const TranslationTableHeader *table, int pos, const InString *input,
		int transOpcode, const TranslationTableRule *transRule, int *passCharDots,
		const widechar **passInstructions, int *passIC, PassRuleMatch *match,
		TranslationTableRule **groupingRule, widechar *groupingOp);
static int
passDoAction(const TranslationTableHeader *table, const DisplayTableHeader *displayTable,
		const InString **input, OutString *output, int *posMapping, int transOpcode,
		const TranslationTableRule **transRule, int passCharDots,
		const widechar *passInstructions, int passIC, int *pos, PassRuleMatch match,
		int *cursorPosition, int *cursorStatus, TranslationTableRule *groupingRule,
		widechar groupingOp, int mode);

static const TranslationTableRule **appliedRules;
static int maxAppliedRules;
static int appliedRulesCount;

static TranslationTableCharacter *
findCharOrDots(widechar c, int m, const TranslationTableHeader *table) {
	/* Look up character or dot pattern in the appropriate
	 * table. */
	static TranslationTableCharacter noChar = { 0, 0, 0, CTC_Space, 32, 32, 32 };
	static TranslationTableCharacter noDots = { 0, 0, 0, CTC_Space, LOU_DOTS, LOU_DOTS,
		LOU_DOTS };
	TranslationTableCharacter *notFound;
	TranslationTableCharacter *character;
	TranslationTableOffset bucket;
	unsigned long int makeHash = _lou_charHash(c);
	if (m == 0) {
		bucket = table->characters[makeHash];
		notFound = &noChar;
	} else {
		bucket = table->dots[makeHash];
		notFound = &noDots;
	}
	while (bucket) {
		character = (TranslationTableCharacter *)&table->ruleArea[bucket];
		if (character->realchar == c) return character;
		bucket = character->next;
	}
	notFound->realchar = notFound->uppercase = notFound->lowercase = c;
	return notFound;
}

static int
checkAttr(const widechar c, const TranslationTableCharacterAttributes a, int m,
		const TranslationTableHeader *table) {
	return (((findCharOrDots(c, m, table))->attributes & a) ? 1 : 0);
}

static int
checkAttr_safe(const InString *input, int pos,
		const TranslationTableCharacterAttributes a, int m,
		const TranslationTableHeader *table) {
	return ((pos < input->length) ? checkAttr(input->chars[pos], a, m, table) : 0);
}

static int
findForPassRule(const TranslationTableHeader *table, int pos, int currentPass,
		const InString *input, int *transOpcode, const TranslationTableRule **transRule,
		int *transCharslen, int *passCharDots, widechar const **passInstructions,
		int *passIC, PassRuleMatch *match, TranslationTableRule **groupingRule,
		widechar *groupingOp) {
	int save_transCharslen = *transCharslen;
	const TranslationTableRule *save_transRule = *transRule;
	TranslationTableOpcode save_transOpcode = *transOpcode;
	TranslationTableOffset ruleOffset;
	ruleOffset = table->forPassRules[currentPass];
	*transCharslen = 0;
	while (ruleOffset) {
		*transRule = (TranslationTableRule *)&table->ruleArea[ruleOffset];
		*transOpcode = (*transRule)->opcode;
		if (passDoTest(table, pos, input, *transOpcode, *transRule, passCharDots,
					passInstructions, passIC, match, groupingRule, groupingOp))
			return 1;
		ruleOffset = (*transRule)->charsnext;
	}
	*transCharslen = save_transCharslen;
	*transRule = save_transRule;
	*transOpcode = save_transOpcode;
	return 0;
}

static int
compareChars(const widechar *address1, const widechar *address2, int count, int m,
		const TranslationTableHeader *table) {
	int k;
	if (!count) return 0;
	for (k = 0; k < count; k++)
		if ((findCharOrDots(address1[k], m, table))->lowercase !=
				(findCharOrDots(address2[k], m, table))->lowercase)
			return 0;
	return 1;
}

static int
makeCorrections(const TranslationTableHeader *table,
		const DisplayTableHeader *displayTable, const InString *input, OutString *output,
		int *posMapping, formtype *typebuf, int *realInlen, int *cursorPosition,
		int *cursorStatus, int mode) {
	int pos;
	int transOpcode;
	const TranslationTableRule *transRule;
	int transCharslen;
	int passCharDots;
	const widechar *passInstructions;
	int passIC; /* Instruction counter */
	PassRuleMatch patternMatch;
	TranslationTableRule *groupingRule;
	widechar groupingOp;
	const InString *origInput = input;
	if (!table->corrections) return 1;
	pos = 0;
	output->length = 0;
	int posIncremented = 1;
	_lou_resetPassVariables();
	while (pos < input->length) {
		int length = input->length - pos;
		int tryThis = 0;
		// check posIncremented to avoid endless loop
		if (!(posIncremented &&
					findForPassRule(table, pos, 0, input, &transOpcode, &transRule,
							&transCharslen, &passCharDots, &passInstructions, &passIC,
							&patternMatch, &groupingRule, &groupingOp)))
			while (tryThis < 3) {
				TranslationTableOffset ruleOffset = 0;
				switch (tryThis) {
				case 0:
					if (!(length >= 2)) break;
					ruleOffset = table->forRules[_lou_stringHash(
							&input->chars[pos], 1, table)];
					break;
				case 1:
					if (!(length >= 1)) break;
					length = 1;
					ruleOffset = findCharOrDots(input->chars[pos], 0, table)->otherRules;
					break;
				case 2: /* No rule found */
					transOpcode = CTO_Always;
					ruleOffset = 0;
					break;
				}
				while (ruleOffset) {
					transRule = (TranslationTableRule *)&table->ruleArea[ruleOffset];
					transOpcode = transRule->opcode;
					transCharslen = transRule->charslen;
					if (tryThis == 1 ||
							(transCharslen <= length &&
									compareChars(&transRule->charsdots[0],
											&input->chars[pos], transCharslen, 0,
											table))) {
						if (posIncremented && transOpcode == CTO_Correct &&
								passDoTest(table, pos, input, transOpcode, transRule,
										&passCharDots, &passInstructions, &passIC,
										&patternMatch, &groupingRule, &groupingOp)) {
							tryThis = 4;
							break;
						}
					}
					ruleOffset = transRule->charsnext;
				}
				tryThis++;
			}
		posIncremented = 1;

		switch (transOpcode) {
		case CTO_Always:
			if (output->length >= output->maxlength) goto failure;
			posMapping[output->length] = pos;
			output->chars[output->length++] = input->chars[pos++];
			break;
		case CTO_Correct: {
			const InString *inputBefore = input;
			int posBefore = pos;
			if (appliedRules != NULL && appliedRulesCount < maxAppliedRules)
				appliedRules[appliedRulesCount++] = transRule;
			if (!passDoAction(table, displayTable, &input, output, posMapping,
						transOpcode, &transRule, passCharDots, passInstructions, passIC,
						&pos, patternMatch, cursorPosition, cursorStatus, groupingRule,
						groupingOp, mode))
				goto failure;
			if (input->bufferIndex != inputBefore->bufferIndex &&
					inputBefore->bufferIndex != origInput->bufferIndex)
				releaseStringBuffer(inputBefore->bufferIndex);
			if (pos == posBefore) posIncremented = 0;
			break;
		}
		default:
			break;
		}
	}

	{  // We have to transform typebuf accordingly
		int k;
		formtype *typebuf_temp;
		if ((typebuf_temp = malloc(output->length * sizeof(formtype))) == NULL)
			_lou_outOfMemory();
		for (k = 0; k < output->length; k++)
			// posMapping will never be < 0 but in theory it could
			if (posMapping[k] < 0)
				typebuf_temp[k] = typebuf[0];  // prepend to next
			else if (posMapping[k] >= input->length)
				typebuf_temp[k] = typebuf[input->length - 1];  // append to previous
			else
				typebuf_temp[k] = typebuf[posMapping[k]];
		memcpy(typebuf, typebuf_temp, output->length * sizeof(formtype));
		free(typebuf_temp);
	}

failure:
	*realInlen = pos;
	if (input->bufferIndex != origInput->bufferIndex)
		releaseStringBuffer(input->bufferIndex);
	return 1;
}

static int
matchCurrentInput(
		const InString *input, int pos, const widechar *passInstructions, int passIC) {
	int k;
	int kk = pos;
	for (k = passIC + 2;
			((k < passIC + 2 + passInstructions[passIC + 1]) && (kk < input->length));
			k++)
		if (input->chars[kk] == LOU_ENDSEGMENT ||
				passInstructions[k] != input->chars[kk++])
			return 0;
	return 1;
}

static int
swapTest(int swapIC, int *pos, const TranslationTableHeader *table, const InString *input,
		const widechar *passInstructions) {
	int p = *pos;
	TranslationTableOffset swapRuleOffset;
	TranslationTableRule *swapRule;
	swapRuleOffset = (passInstructions[swapIC + 1] << 16) | passInstructions[swapIC + 2];
	swapRule = (TranslationTableRule *)&table->ruleArea[swapRuleOffset];
	while (p - *pos < passInstructions[swapIC + 3]) {
		int test;
		if (swapRule->opcode == CTO_SwapDd) {
			for (test = 1; test < swapRule->charslen; test += 2) {
				if (input->chars[p] == swapRule->charsdots[test]) break;
			}
		} else {
			for (test = 0; test < swapRule->charslen; test++) {
				if (input->chars[p] == swapRule->charsdots[test]) break;
			}
		}
		if (test >= swapRule->charslen) return 0;
		p++;
	}
	if (passInstructions[swapIC + 3] == passInstructions[swapIC + 4]) {
		*pos = p;
		return 1;
	}
	while (p - *pos < passInstructions[swapIC + 4]) {
		int test;
		if (swapRule->opcode == CTO_SwapDd) {
			for (test = 1; test < swapRule->charslen; test += 2) {
				if (input->chars[p] == swapRule->charsdots[test]) break;
			}
		} else {
			for (test = 0; test < swapRule->charslen; test++) {
				if (input->chars[p] == swapRule->charsdots[test]) break;
			}
		}
		if (test >= swapRule->charslen) {
			*pos = p;
			return 1;
		}
		p++;
	}
	*pos = p;
	return 1;
}

static int
swapReplace(int start, int end, const TranslationTableHeader *table,
		const InString *input, OutString *output, int *posMapping,
		const widechar *passInstructions, int passIC) {
	TranslationTableOffset swapRuleOffset;
	TranslationTableRule *swapRule;
	widechar *replacements;
	int p;
	swapRuleOffset = (passInstructions[passIC + 1] << 16) | passInstructions[passIC + 2];
	swapRule = (TranslationTableRule *)&table->ruleArea[swapRuleOffset];
	replacements = &swapRule->charsdots[swapRule->charslen];
	for (p = start; p < end; p++) {
		int rep;
		int test;
		int k;
		if (swapRule->opcode == CTO_SwapDd) {
			// A sequence of dot patterns is encoded as the length of the first dot
			// pattern (single widechar) followed by the contents of the first dot pattern
			// (one widechar per cell) followed by the length of the second dot pattern,
			// etc. See the function `compileSwapDots'. Because the third operand of a
			// swapdd rule can only contain single-cell dot patterns, the elements at
			// index 0, 2, ... are "1" and the elements at index 1, 3, ... are the dot
			// patterns.
			for (test = 0; test * 2 + 1 < swapRule->charslen; test++)
				if (input->chars[p] == swapRule->charsdots[test * 2 + 1]) break;
			if (test * 2 == swapRule->charslen) continue;
		} else {
			for (test = 0; test < swapRule->charslen; test++)
				if (input->chars[p] == swapRule->charsdots[test]) break;
			if (test == swapRule->charslen) continue;
		}
		k = 0;
		for (rep = 0; rep < test; rep++)
			if (swapRule->opcode == CTO_SwapCc)
				k++;
			else
				k += replacements[k];
		if (swapRule->opcode == CTO_SwapCc) {
			if ((output->length + 1) > output->maxlength) return 0;
			posMapping[output->length] = p;
			output->chars[output->length++] = replacements[k];
		} else {
			int l = replacements[k] - 1;
			int d = output->length + l;
			if (d > output->maxlength) return 0;
			while (--d >= output->length) posMapping[d] = p;
			memcpy(&output->chars[output->length], &replacements[k + 1],
					l * sizeof(*output->chars));
			output->length += l;
		}
	}
	return 1;
}

static int
replaceGrouping(const TranslationTableHeader *table, const InString **input,
		OutString *output, int transOpcode, int passCharDots,
		const widechar *passInstructions, int passIC, int startReplace,
		TranslationTableRule *groupingRule, widechar groupingOp) {
	widechar startCharDots = groupingRule->charsdots[2 * passCharDots];
	widechar endCharDots = groupingRule->charsdots[2 * passCharDots + 1];
	int p;
	int level = 0;
	TranslationTableOffset replaceOffset =
			passInstructions[passIC + 1] << 16 | (passInstructions[passIC + 2] & 0xff);
	TranslationTableRule *replaceRule =
			(TranslationTableRule *)&table->ruleArea[replaceOffset];
	widechar replaceStart = replaceRule->charsdots[2 * passCharDots];
	widechar replaceEnd = replaceRule->charsdots[2 * passCharDots + 1];
	if (groupingOp == pass_groupstart) {
		for (p = startReplace + 1; p < (*input)->length; p++) {
			if ((*input)->chars[p] == startCharDots) level--;
			if ((*input)->chars[p] == endCharDots) level++;
			if (level == 1) break;
		}
		if (p == (*input)->length)
			return 0;
		else {
			// Create a new string instead of modifying it. This is slightly less
			// efficient, but makes the code more readable. Grouping is not a much used
			// feature anyway.
			int idx = getStringBuffer((*input)->length);
			widechar *chars = stringBufferPool->buffers[idx];
			memcpy(chars, (*input)->chars, (*input)->length * sizeof(widechar));
			chars[startReplace] = replaceStart;
			chars[p] = replaceEnd;
			static InString stringStore;
			stringStore = (InString){
				.chars = chars, .length = (*input)->length, .bufferIndex = idx
			};
			*input = &stringStore;
		}
	} else {
		if (transOpcode == CTO_Context) {
			startCharDots = groupingRule->charsdots[2];
			endCharDots = groupingRule->charsdots[3];
			replaceStart = replaceRule->charsdots[2];
			replaceEnd = replaceRule->charsdots[3];
		}
		output->chars[output->length] = replaceEnd;
		for (p = output->length - 1; p >= 0; p--) {
			if (output->chars[p] == endCharDots) level--;
			if (output->chars[p] == startCharDots) level++;
			if (level == 1) break;
		}
		if (p < 0) return 0;
		output->chars[p] = replaceStart;
		output->length++;
	}
	return 1;
}

static int
removeGrouping(const InString **input, OutString *output, int passCharDots,
		int startReplace, TranslationTableRule *groupingRule, widechar groupingOp) {
	widechar startCharDots = groupingRule->charsdots[2 * passCharDots];
	widechar endCharDots = groupingRule->charsdots[2 * passCharDots + 1];
	int p;
	int level = 0;
	if (groupingOp == pass_groupstart) {
		for (p = startReplace + 1; p < (*input)->length; p++) {
			if ((*input)->chars[p] == startCharDots) level--;
			if ((*input)->chars[p] == endCharDots) level++;
			if (level == 1) break;
		}
		if (p == (*input)->length)
			return 0;
		else {
			// Create a new string instead of modifying it. This is slightly less
			// efficient, but makes the code more readable. Grouping is not a much used
			// feature anyway.
			int idx = getStringBuffer((*input)->length);
			widechar *chars = stringBufferPool->buffers[idx];
			int len = 0;
			int k;
			for (k = 0; k < (*input)->length; k++) {
				if (k == p) continue;
				chars[len++] = (*input)->chars[k];
			}
			static InString stringStore;
			stringStore = (InString){ .chars = chars, .length = len, .bufferIndex = idx };
			*input = &stringStore;
		}
	} else {
		for (p = output->length - 1; p >= 0; p--) {
			if (output->chars[p] == endCharDots) level--;
			if (output->chars[p] == startCharDots) level++;
			if (level == 1) break;
		}
		if (p < 0) return 0;
		p++;
		for (; p < output->length; p++) output->chars[p - 1] = output->chars[p];
		output->length--;
	}
	return 1;
}

static int
doPassSearch(const TranslationTableHeader *table, const InString *input,
		const TranslationTableRule *transRule, int passCharDots, int pos,
		const widechar *passInstructions, int passIC, int *searchIC, int *searchPos,
		TranslationTableRule *groupingRule, widechar groupingOp) {
	int level = 0;
	int k, kk;
	int not = 0;  // whether next operand should be reversed
	TranslationTableOffset ruleOffset;
	TranslationTableRule *rule;
	TranslationTableCharacterAttributes attributes;
	while (pos < input->length) {
		*searchIC = passIC + 1;
		*searchPos = pos;
		while (*searchIC < transRule->dotslen) {
			int itsTrue = 1;  // whether we have a match or not
			if (*searchPos > input->length) return 0;
			switch (passInstructions[*searchIC]) {
			case pass_lookback:
				*searchPos -= passInstructions[*searchIC + 1];
				if (*searchPos < 0) {
					*searchPos = 0;
					itsTrue = 0;
				}
				*searchIC += 2;
				break;
			case pass_not:
				not = !not;
				(*searchIC)++;
				continue;
			case pass_string:
			case pass_dots:
				kk = *searchPos;
				for (k = *searchIC + 2;
						k < *searchIC + 2 + passInstructions[*searchIC + 1]; k++)
					if (input->chars[kk] == LOU_ENDSEGMENT ||
							passInstructions[k] != input->chars[kk++]) {
						itsTrue = 0;
						break;
					}
				*searchPos += passInstructions[*searchIC + 1];
				*searchIC += passInstructions[*searchIC + 1] + 2;
				break;
			case pass_startReplace:
				(*searchIC)++;
				break;
			case pass_endReplace:
				(*searchIC)++;
				break;
			case pass_attributes:
				attributes = (passInstructions[*searchIC + 1] << 16) |
						passInstructions[*searchIC + 2];
				for (k = 0; k < passInstructions[*searchIC + 3]; k++) {
					if (input->chars[*searchPos] == LOU_ENDSEGMENT)
						itsTrue = 0;
					else {
						itsTrue = ((findCharOrDots(input->chars[(*searchPos)++],
											passCharDots, table)
												   ->attributes &
										   attributes)
										? 1
										: 0);
						if (not) itsTrue = !itsTrue;
					}
					if (!itsTrue) break;
				}
				if (itsTrue) {
					for (k = passInstructions[*searchIC + 3];
							k < passInstructions[*searchIC + 4]; k++) {
						if (input->chars[*searchPos] == LOU_ENDSEGMENT) {
							itsTrue = 0;
							break;
						}
						if (!(findCharOrDots(
									  input->chars[*searchPos], passCharDots, table)
											->attributes &
									attributes)) {
							if (!not) break;
						} else if (not)
							break;
						(*searchPos)++;
					}
				}
				not = 0;
				*searchIC += 5;
				break;
			case pass_groupstart:
			case pass_groupend:
				ruleOffset = (passInstructions[*searchIC + 1] << 16) |
						passInstructions[*searchIC + 2];
				rule = (TranslationTableRule *)&table->ruleArea[ruleOffset];
				if (passInstructions[*searchIC] == pass_groupstart)
					itsTrue = (input->chars[*searchPos] ==
									  rule->charsdots[2 * passCharDots])
							? 1
							: 0;
				else
					itsTrue = (input->chars[*searchPos] ==
									  rule->charsdots[2 * passCharDots + 1])
							? 1
							: 0;
				if (groupingRule != NULL && groupingOp == pass_groupstart &&
						rule == groupingRule) {
					if (input->chars[*searchPos] == rule->charsdots[2 * passCharDots])
						level--;
					else if (input->chars[*searchPos] ==
							rule->charsdots[2 * passCharDots + 1])
						level++;
				}
				(*searchPos)++;
				*searchIC += 3;
				break;
			case pass_swap:
				itsTrue = swapTest(*searchIC, searchPos, table, input, passInstructions);
				*searchIC += 5;
				break;
			case pass_endTest:
				if (itsTrue) {
					if ((groupingRule && level == 1) || !groupingRule) return 1;
				}
				*searchIC = transRule->dotslen;
				break;
			default:
				if (_lou_handlePassVariableTest(passInstructions, searchIC, &itsTrue))
					break;
				break;
			}
			if ((!not&&!itsTrue) || (not&&itsTrue)) break;
			not = 0;
		}
		pos++;
	}
	return 0;
}

static int
passDoTest(const TranslationTableHeader *table, int pos, const InString *input,
		int transOpcode, const TranslationTableRule *transRule, int *passCharDots,
		widechar const **passInstructions, int *passIC, PassRuleMatch *match,
		TranslationTableRule **groupingRule, widechar *groupingOp) {
	int searchIC, searchPos;
	int k;
	int not = 0;  // whether next operand should be reversed
	TranslationTableOffset ruleOffset = 0;
	TranslationTableRule *rule = NULL;
	TranslationTableCharacterAttributes attributes = 0;
	int startMatch = pos;
	int endMatch = pos;
	int startReplace = -1;
	int endReplace = -1;
	*groupingRule = NULL;
	*passInstructions = &transRule->charsdots[transRule->charslen];
	*passIC = 0;
	if (transOpcode == CTO_Context || transOpcode == CTO_Correct)
		*passCharDots = 0;
	else
		*passCharDots = 1;
	while (*passIC < transRule->dotslen) {
		int itsTrue = 1;  // whether we have a match or not
		if (pos > input->length) return 0;
		switch ((*passInstructions)[*passIC]) {
		case pass_first:
			if (pos != 0) itsTrue = 0;
			(*passIC)++;
			break;
		case pass_last:
			if (pos != input->length) itsTrue = 0;
			(*passIC)++;
			break;
		case pass_lookback:
			pos -= (*passInstructions)[*passIC + 1];
			if (pos < 0) {
				searchPos = 0;
				itsTrue = 0;
			}
			*passIC += 2;
			break;
		case pass_not:
			not = !not;
			(*passIC)++;
			continue;
		case pass_string:
		case pass_dots:
			itsTrue = matchCurrentInput(input, pos, *passInstructions, *passIC);
			pos += (*passInstructions)[*passIC + 1];
			*passIC += (*passInstructions)[*passIC + 1] + 2;
			break;
		case pass_startReplace:
			startReplace = pos;
			(*passIC)++;
			break;
		case pass_endReplace:
			endReplace = pos;
			(*passIC)++;
			break;
		case pass_attributes:
			attributes = ((*passInstructions)[*passIC + 1] << 16) |
					(*passInstructions)[*passIC + 2];
			for (k = 0; k < (*passInstructions)[*passIC + 3]; k++) {
				if (pos >= input->length) {
					itsTrue = 0;
					break;
				}
				if (input->chars[pos] == LOU_ENDSEGMENT) {
					itsTrue = 0;
					break;
				}
				if (!(findCharOrDots(input->chars[pos], *passCharDots, table)
									->attributes &
							attributes)) {
					if (!not) {
						itsTrue = 0;
						break;
					}
				} else if (not) {
					itsTrue = 0;
					break;
				}
				pos++;
			}
			if (itsTrue) {
				for (k = (*passInstructions)[*passIC + 3];
						k < (*passInstructions)[*passIC + 4] && pos < input->length;
						k++) {
					if (input->chars[pos] == LOU_ENDSEGMENT) {
						itsTrue = 0;
						break;
					}
					if (!(findCharOrDots(input->chars[pos], *passCharDots, table)
										->attributes &
								attributes)) {
						if (!not) break;
					} else if (not)
						break;
					pos++;
				}
			}
			not = 0;
			*passIC += 5;
			break;
		case pass_groupstart:
		case pass_groupend:
			ruleOffset = ((*passInstructions)[*passIC + 1] << 16) |
					(*passInstructions)[*passIC + 2];
			rule = (TranslationTableRule *)&table->ruleArea[ruleOffset];
			if (*passIC == 0 ||
					(*passIC > 0 &&
							(*passInstructions)[*passIC - 1] == pass_startReplace)) {
				*groupingRule = rule;
				*groupingOp = (*passInstructions)[*passIC];
			}
			if ((*passInstructions)[*passIC] == pass_groupstart)
				itsTrue =
						(input->chars[pos] == rule->charsdots[2 * *passCharDots]) ? 1 : 0;
			else
				itsTrue = (input->chars[pos] == rule->charsdots[2 * *passCharDots + 1])
						? 1
						: 0;
			pos++;
			*passIC += 3;
			break;
		case pass_swap:
			itsTrue = swapTest(*passIC, &pos, table, input, *passInstructions);
			*passIC += 5;
			break;
		case pass_search:
			itsTrue = doPassSearch(table, input, transRule, *passCharDots, pos,
					*passInstructions, *passIC, &searchIC, &searchPos, *groupingRule,
					*groupingOp);
			if ((!not&&!itsTrue) || (not&&itsTrue)) return 0;
			*passIC = searchIC;
			pos = searchPos;
		case pass_endTest:
			(*passIC)++;
			endMatch = pos;
			if (startReplace == -1) {
				startReplace = startMatch;
				endReplace = endMatch;
			}
			if (startReplace < startMatch)
				return 0;
			else {
				*match = (PassRuleMatch){ .startMatch = startMatch,
					.startReplace = startReplace,
					.endReplace = endReplace,
					.endMatch = endMatch };
				return 1;
			}
			break;
		default:
			if (_lou_handlePassVariableTest(*passInstructions, passIC, &itsTrue)) break;
			return 0;
		}
		if ((!not&&!itsTrue) || (not&&itsTrue)) return 0;
		not = 0;
	}
	return 0;
}

static int
copyCharacters(int from, int to, const TranslationTableHeader *table,
		const DisplayTableHeader *displayTable, const InString *input, OutString *output,
		int *posMapping, int transOpcode, int *cursorPosition, int *cursorStatus,
		int mode) {
	if (transOpcode == CTO_Context) {
		while (from < to) {
			if (!putCharacter(input->chars[from], table, displayTable, from, input,
						output, posMapping, cursorPosition, cursorStatus, mode))
				return 0;
			from++;
		}
	} else {
		if (to > from) {
			if ((output->length + to - from) > output->maxlength) return 0;
			while (to > from) {
				posMapping[output->length] = from;
				output->chars[output->length] = input->chars[from];
				output->length++;
				from++;
			}
		}
	}

	return 1;
}

static int
passDoAction(const TranslationTableHeader *table, const DisplayTableHeader *displayTable,
		const InString **input, OutString *output, int *posMapping, int transOpcode,
		const TranslationTableRule **transRule, int passCharDots,
		const widechar *passInstructions, int passIC, int *pos, PassRuleMatch match,
		int *cursorPosition, int *cursorStatus, TranslationTableRule *groupingRule,
		widechar groupingOp, int mode) {
	int k;
	TranslationTableOffset ruleOffset = 0;
	TranslationTableRule *rule = NULL;
	int destStartMatch = output->length;
	int destStartReplace;
	int newPos = match.endReplace;

	if (!copyCharacters(match.startMatch, match.startReplace, table, displayTable, *input,
				output, posMapping, transOpcode, cursorPosition, cursorStatus, mode))
		return 0;
	destStartReplace = output->length;

	while (passIC < (*transRule)->dotslen) switch (passInstructions[passIC]) {
		case pass_string:
		case pass_dots:
			if ((output->length + passInstructions[passIC + 1]) > output->maxlength)
				return 0;
			for (k = 0; k < passInstructions[passIC + 1]; ++k)
				posMapping[output->length + k] = match.startReplace;
			memcpy(&output->chars[output->length], &passInstructions[passIC + 2],
					passInstructions[passIC + 1] * CHARSIZE);
			output->length += passInstructions[passIC + 1];
			passIC += passInstructions[passIC + 1] + 2;
			break;
		case pass_groupstart:
			ruleOffset =
					(passInstructions[passIC + 1] << 16) | passInstructions[passIC + 2];
			rule = (TranslationTableRule *)&table->ruleArea[ruleOffset];
			posMapping[output->length] = match.startMatch;
			output->chars[output->length++] = rule->charsdots[2 * passCharDots];
			passIC += 3;
			break;
		case pass_groupend:
			ruleOffset =
					(passInstructions[passIC + 1] << 16) | passInstructions[passIC + 2];
			rule = (TranslationTableRule *)&table->ruleArea[ruleOffset];
			posMapping[output->length] = match.startMatch;
			output->chars[output->length++] = rule->charsdots[2 * passCharDots + 1];
			passIC += 3;
			break;
		case pass_swap:
			if (!swapReplace(match.startReplace, match.endReplace, table, *input, output,
						posMapping, passInstructions, passIC))
				return 0;
			passIC += 3;
			break;
		case pass_groupreplace:
			if (!groupingRule ||
					!replaceGrouping(table, input, output, transOpcode, passCharDots,
							passInstructions, passIC, match.startReplace, groupingRule,
							groupingOp))
				return 0;
			passIC += 3;
			break;
		case pass_omit:
			if (groupingRule)
				removeGrouping(input, output, passCharDots, match.startReplace,
						groupingRule, groupingOp);
			passIC++;
			break;
		case pass_copy: {
			int count = destStartReplace - destStartMatch;
			if (count > 0) {
				memmove(&output->chars[destStartMatch], &output->chars[destStartReplace],
						count * sizeof(*output->chars));
				output->length -= count;
				destStartReplace = destStartMatch;
			}
		}

			if (!copyCharacters(match.startReplace, match.endReplace, table, displayTable,
						*input, output, posMapping, transOpcode, cursorPosition,
						cursorStatus, mode))
				return 0;
			newPos = match.endMatch;
			passIC++;
			break;
		default:
			if (_lou_handlePassVariableAction(passInstructions, &passIC)) break;
			return 0;
		}
	*pos = newPos;
	return 1;
}

static void
passSelectRule(const TranslationTableHeader *table, int pos, int currentPass,
		const InString *input, int *transOpcode, const TranslationTableRule **transRule,
		int *transCharslen, int *passCharDots, widechar const **passInstructions,
		int *passIC, PassRuleMatch *match, TranslationTableRule **groupingRule,
		widechar *groupingOp) {
	if (!findForPassRule(table, pos, currentPass, input, transOpcode, transRule,
				transCharslen, passCharDots, passInstructions, passIC, match,
				groupingRule, groupingOp)) {
		*transOpcode = CTO_Always;
	}
}

static int
translatePass(const TranslationTableHeader *table, const DisplayTableHeader *displayTable,
		int currentPass, const InString *input, OutString *output, int *posMapping,
		int *realInlen, int *cursorPosition, int *cursorStatus, int mode) {
	int pos;
	int transOpcode;
	const TranslationTableRule *transRule;
	int transCharslen;
	int passCharDots;
	const widechar *passInstructions;
	int passIC; /* Instruction counter */
	PassRuleMatch patternMatch;
	TranslationTableRule *groupingRule;
	widechar groupingOp;
	const InString *origInput = input;
	pos = output->length = 0;
	int posIncremented = 1;
	_lou_resetPassVariables();
	while (pos < input->length) { /* the main multipass translation loop */
		// check posIncremented to avoid endless loop
		if (!posIncremented)
			transOpcode = CTO_Always;
		else
			passSelectRule(table, pos, currentPass, input, &transOpcode, &transRule,
					&transCharslen, &passCharDots, &passInstructions, &passIC,
					&patternMatch, &groupingRule, &groupingOp);
		posIncremented = 1;
		switch (transOpcode) {
		case CTO_Context:
		case CTO_Pass2:
		case CTO_Pass3:
		case CTO_Pass4: {
			const InString *inputBefore = input;
			int posBefore = pos;
			if (appliedRules != NULL && appliedRulesCount < maxAppliedRules)
				appliedRules[appliedRulesCount++] = transRule;
			if (!passDoAction(table, displayTable, &input, output, posMapping,
						transOpcode, &transRule, passCharDots, passInstructions, passIC,
						&pos, patternMatch, cursorPosition, cursorStatus, groupingRule,
						groupingOp, mode))
				goto failure;
			if (input->bufferIndex != inputBefore->bufferIndex &&
					inputBefore->bufferIndex != origInput->bufferIndex)
				releaseStringBuffer(inputBefore->bufferIndex);
			if (pos == posBefore) posIncremented = 0;
			break;
		}
		case CTO_Always:
			if ((output->length + 1) > output->maxlength) goto failure;
			posMapping[output->length] = pos;
			output->chars[output->length++] = input->chars[pos++];
			break;
		default:
			goto failure;
		}
	}
failure:
	if (pos < input->length) {
		while (checkAttr(input->chars[pos], CTC_Space, 1, table))
			if (++pos == input->length) break;
	}
	*realInlen = pos;
	if (input->bufferIndex != origInput->bufferIndex)
		releaseStringBuffer(input->bufferIndex);
	return 1;
}

#define MIN(a, b) (((a) < (b)) ? (a) : (b))

static int
translateString(const TranslationTableHeader *table,
		const DisplayTableHeader *displayTable, int mode, int currentPass,
		const InString *input, OutString *output, int *posMapping, formtype *typebuf,
		unsigned char *srcSpacing, unsigned char *destSpacing, unsigned int *wordBuffer,
		EmphasisInfo *emphasisBuffer, int haveEmphasis, int *realInlen,
		int *cursorPosition, int *cursorStatus, int compbrlStart, int compbrlEnd);

int EXPORT_CALL
lou_translateString(const char *tableList, const widechar *inbufx, int *inlen,
		widechar *outbuf, int *outlen, formtype *typeform, char *spacing, int mode) {
	return lou_translate(tableList, inbufx, inlen, outbuf, outlen, typeform, spacing,
			NULL, NULL, NULL, mode);
}

int EXPORT_CALL
lou_translate(const char *tableList, const widechar *inbufx, int *inlen, widechar *outbuf,
		int *outlen, formtype *typeform, char *spacing, int *outputPos, int *inputPos,
		int *cursorPos, int mode) {
	return _lou_translate(tableList, tableList, inbufx, inlen, outbuf, outlen, typeform,
			spacing, outputPos, inputPos, cursorPos, mode, NULL, NULL);
}

int EXPORT_CALL
_lou_translate(const char *tableList, const char *displayTableList,
		const widechar *inbufx, int *inlen, widechar *outbuf, int *outlen,
		formtype *typeform, char *spacing, int *outputPos, int *inputPos, int *cursorPos,
		int mode, const TranslationTableRule **rules, int *rulesLen) {
	// int i;
	// for(i = 0; i < *inlen; i++)
	// {
	// 	outbuf[i] = inbufx[i];
	// 	if(inputPos)
	// 		inputPos[i] = i;
	// 	if(outputPos)
	// 		outputPos[i] = i;
	// }
	// *inlen = i;
	// *outlen = i;
	// return 1;
	const TranslationTableHeader *table;
	const DisplayTableHeader *displayTable;
	InString input;
	OutString output;
	// posMapping contains position mapping info between the initial input and the output
	// of the current pass. It is 1 longer than the output. The values are monotonically
	// increasing and can range between -1 and the input length. At the end the position
	// info is passed to the user as an inputPos and outputPos array. inputPos has the
	// length of the final output and has values ranging from 0 to inlen-1. outputPos has
	// the length of the initial input and has values ranging from 0 to outlen-1.
	int *posMapping;
	int *posMapping1;
	int *posMapping2;
	int *posMapping3;
	formtype *typebuf;
	unsigned char *srcSpacing;
	unsigned char *destSpacing;
	unsigned int *wordBuffer;
	EmphasisInfo *emphasisBuffer;
	int cursorPosition;
	int cursorStatus;
	int haveEmphasis;
	int compbrlStart = -1;
	int compbrlEnd = -1;
	int k;
	int goodTrans = 1;
	if (tableList == NULL || inbufx == NULL || inlen == NULL || outbuf == NULL ||
			outlen == NULL)
		return 0;
	_lou_logMessage(LOU_LOG_ALL, "Performing translation: tableList=%s, inlen=%d",
			tableList, *inlen);
	_lou_logWidecharBuf(LOU_LOG_ALL, "Inbuf=", inbufx, *inlen);

	if (!_lou_isValidMode(mode))
		_lou_logMessage(LOU_LOG_ERROR, "Invalid mode parameter: %d", mode);

	if (displayTableList == NULL) displayTableList = tableList;
	_lou_getTable(tableList, displayTableList, &table, &displayTable);
	if (table == NULL || *inlen < 0 || *outlen < 0) return 0;
	k = 0;
	while (k < *inlen && inbufx[k]) k++;
	input = (InString){ .chars = inbufx, .length = k, .bufferIndex = -1 };
	haveEmphasis = 0;
	if (!(typebuf = _lou_allocMem(alloc_typebuf, 0, input.length, *outlen))) return 0;
	if (typeform != NULL) {
		for (k = 0; k < input.length; k++) {
			typebuf[k] = typeform[k];
			if (typebuf[k] & EMPHASIS) haveEmphasis = 1;
		}
	} else
		memset(typebuf, 0, input.length * sizeof(formtype));

	if ((wordBuffer = _lou_allocMem(alloc_wordBuffer, 0, input.length, *outlen)))
		memset(wordBuffer, 0, (input.length + 4) * sizeof(unsigned int));
	else
		return 0;
	if ((emphasisBuffer = _lou_allocMem(alloc_emphasisBuffer, 0, input.length, *outlen)))
		memset(emphasisBuffer, 0, (input.length + 4) * sizeof(EmphasisInfo));
	else
		return 0;

	if (!(spacing == NULL || *spacing == 'X'))
		srcSpacing = (unsigned char *)spacing;
	else
		srcSpacing = NULL;
	if (outputPos != NULL)
		for (k = 0; k < input.length; k++) outputPos[k] = -1;
	if (cursorPos != NULL && *cursorPos >= 0) {
		cursorStatus = 0;
		cursorPosition = *cursorPos;
		if ((mode & (compbrlAtCursor | compbrlLeftCursor))) {
			compbrlStart = cursorPosition;
			if (checkAttr(input.chars[compbrlStart], CTC_Space, 0, table))
				compbrlEnd = compbrlStart + 1;
			else {
				while (compbrlStart >= 0 &&
						!checkAttr(input.chars[compbrlStart], CTC_Space, 0, table))
					compbrlStart--;
				compbrlStart++;
				compbrlEnd = cursorPosition;
				if (!(mode & compbrlLeftCursor))
					while (compbrlEnd < input.length &&
							!checkAttr(input.chars[compbrlEnd], CTC_Space, 0, table))
						compbrlEnd++;
			}
		}
	} else {
		cursorPosition = -1;
		cursorStatus = 1; /* so it won't check cursor position */
	}
	if (!(posMapping1 = _lou_allocMem(alloc_posMapping1, 0, input.length, *outlen)))
		return 0;
	if (table->numPasses > 1 || table->corrections) {
		if (!(posMapping2 = _lou_allocMem(alloc_posMapping2, 0, input.length, *outlen)))
			return 0;
		if (!(posMapping3 = _lou_allocMem(alloc_posMapping3, 0, input.length, *outlen)))
			return 0;
	}
	if (srcSpacing != NULL) {
		if (!(destSpacing = _lou_allocMem(alloc_destSpacing, 0, input.length, *outlen)))
			goodTrans = 0;
		else
			memset(destSpacing, '*', *outlen);
	} else
		destSpacing = NULL;
	appliedRulesCount = 0;
	if (rules != NULL && rulesLen != NULL) {
		appliedRules = rules;
		maxAppliedRules = *rulesLen;
	} else {
		appliedRules = NULL;
		maxAppliedRules = 0;
	}
	{
		int idx;
		if (!stringBufferPool) initStringBufferPool();
		for (idx = 0; idx < stringBufferPool->size; idx++) releaseStringBuffer(idx);
		idx = getStringBuffer(*outlen);
		output = (OutString){ .chars = stringBufferPool->buffers[idx],
			.maxlength = *outlen,
			.length = 0,
			.bufferIndex = idx };
	}
	posMapping = posMapping1;

	int currentPass = table->corrections ? 0 : 1;
	int *passPosMapping = posMapping;
	while (1) {
		int realInlen;
		switch (currentPass) {
		case 0:
			goodTrans =
					makeCorrections(table, displayTable, &input, &output, passPosMapping,
							typebuf, &realInlen, &cursorPosition, &cursorStatus, mode);
			break;
		case 1: {
			goodTrans = translateString(table, displayTable, mode, currentPass, &input,
					&output, passPosMapping, typebuf, srcSpacing, destSpacing, wordBuffer,
					emphasisBuffer, haveEmphasis, &realInlen, &cursorPosition,
					&cursorStatus, compbrlStart, compbrlEnd);
			break;
		}
		default:
			goodTrans = translatePass(table, displayTable, currentPass, &input, &output,
					passPosMapping, &realInlen, &cursorPosition, &cursorStatus, mode);
			break;
		}
		passPosMapping[output.length] = realInlen;
		if (passPosMapping == posMapping) {
			passPosMapping = posMapping2;
		} else {
			int *prevPosMapping = posMapping3;
			memcpy((int *)prevPosMapping, posMapping, (*outlen + 1) * sizeof(int));
			for (k = 0; k <= output.length; k++)
				if (passPosMapping[k] < 0)
					posMapping[k] = prevPosMapping[0];
				else
					posMapping[k] = prevPosMapping[passPosMapping[k]];
		}
		currentPass++;
		if (currentPass <= table->numPasses && goodTrans) {
			int idx;
			releaseStringBuffer(input.bufferIndex);
			input = (InString){ .chars = output.chars,
				.length = output.length,
				.bufferIndex = output.bufferIndex };
			idx = getStringBuffer(*outlen);
			output = (OutString){ .chars = stringBufferPool->buffers[idx],
				.maxlength = *outlen,
				.length = 0,
				.bufferIndex = idx };
			continue;
		}
		break;
	}
	if (goodTrans) {
		for (k = 0; k < output.length; k++) {
			if (typeform != NULL) {
				if ((output.chars[k] & (LOU_DOT_7 | LOU_DOT_8)))
					typeform[k] = '8';
				else
					typeform[k] = '0';
			}
			if ((mode & dotsIO)) {
				if ((mode & ucBrl))
					outbuf[k] = ((output.chars[k] & 0xff) | LOU_ROW_BRAILLE);
				else
					outbuf[k] = output.chars[k];
			} else {
				outbuf[k] = _lou_getCharFromDots(output.chars[k], displayTable);
				if (!outbuf[k]) {
					// assume that if NUL character is returned, it's because the display
					// table has no mapping for the dot pattern (not because it maps to
					// NUL)
					_lou_logMessage(LOU_LOG_ERROR,
							"%s: no mapping for dot pattern %s in display table",
							displayTableList, _lou_showDots(&output.chars[k], 1));
					return 0;
				}
			}
		}
		*inlen = posMapping[output.length];
		*outlen = output.length;
		// Compute inputPos and outputPos from posMapping. The value at the last index of
		// posMapping is currectly not used.
		if (inputPos != NULL) {
			for (k = 0; k < *outlen; k++)
				if (posMapping[k] < 0)
					inputPos[k] = 0;
				else if (posMapping[k] > *inlen - 1)
					inputPos[k] = *inlen - 1;
				else
					inputPos[k] = posMapping[k];
		}
		if (outputPos != NULL) {
			int inpos = -1;
			int outpos = -1;
			for (k = 0; k < *outlen; k++)
				if (posMapping[k] > inpos) {
					while (inpos < posMapping[k]) {
						if (inpos >= 0 && inpos < *inlen)
							outputPos[inpos] = outpos < 0 ? 0 : outpos;
						inpos++;
					}
					outpos = k;
				}
			if (inpos < 0) inpos = 0;
			while (inpos < *inlen) outputPos[inpos++] = outpos;
		}
	}
	if (destSpacing != NULL) {
		memcpy(srcSpacing, destSpacing, input.length);
		srcSpacing[input.length] = 0;
	}
	if (cursorPos != NULL && *cursorPos != -1) {
		if (outputPos != NULL)
			*cursorPos = outputPos[*cursorPos];
		else
			*cursorPos = cursorPosition;
	}
	if (rulesLen != NULL) *rulesLen = appliedRulesCount;
	_lou_logMessage(LOU_LOG_ALL, "Translation complete: outlen=%d", *outlen);
	_lou_logWidecharBuf(LOU_LOG_ALL, "Outbuf=", (const widechar *)outbuf, *outlen);

	return goodTrans;
}

int EXPORT_CALL
lou_translatePrehyphenated(const char *tableList, const widechar *inbufx, int *inlen,
		widechar *outbuf, int *outlen, formtype *typeform, char *spacing, int *outputPos,
		int *inputPos, int *cursorPos, char *inputHyphens, char *outputHyphens,
		int mode) {
	int rv = 1;
	int *alloc_inputPos = NULL;
	if (inputHyphens != NULL) {
		if (outputHyphens == NULL) return 0;
		if (inputPos == NULL) {
			if ((alloc_inputPos = malloc(*outlen * sizeof(int))) == NULL)
				_lou_outOfMemory();
			inputPos = alloc_inputPos;
		}
	}
	if (lou_translate(tableList, inbufx, inlen, outbuf, outlen, typeform, spacing,
				outputPos, inputPos, cursorPos, mode)) {
		if (inputHyphens != NULL) {
			int inpos = 0;
			int outpos;
			for (outpos = 0; outpos < *outlen; outpos++) {
				int new_inpos = inputPos[outpos];
				if (new_inpos < inpos) {
					rv = 0;
					break;
				}
				if (new_inpos > inpos)
					outputHyphens[outpos] = inputHyphens[new_inpos];
				else
					outputHyphens[outpos] = '0';
				inpos = new_inpos;
			}
		}
	}
	if (alloc_inputPos != NULL) free(alloc_inputPos);
	return rv;
}

static int
hyphenateWord(const widechar *word, int wordSize, char *hyphens,
		const TranslationTableHeader *table) {
	widechar *prepWord;
	int i, k, limit;
	int stateNum;
	widechar ch;
	HyphenationState *statesArray =
			(HyphenationState *)&table->ruleArea[table->hyphenStatesArray];
	HyphenationState *currentState;
	HyphenationTrans *transitionsArray;
	char *hyphenPattern;
	int patternOffset;
	if (!table->hyphenStatesArray || (wordSize + 3) > MAXSTRING) return 0;
	prepWord = (widechar *)calloc(wordSize + 3, sizeof(widechar));
	/* prepWord is of the format ".hello."
	 * hyphens is the length of the word "hello" "00000" */
	prepWord[0] = '.';
	for (i = 0; i < wordSize; i++) {
		prepWord[i + 1] = (findCharOrDots(word[i], 0, table))->lowercase;
		hyphens[i] = '0';
	}
	prepWord[wordSize + 1] = '.';

	/* now, run the finite state machine */
	stateNum = 0;

	// we need to walk all of ".hello."
	for (i = 0; i < wordSize + 2; i++) {
		ch = prepWord[i];
		while (1) {
			if (stateNum == 0xffff) {
				stateNum = 0;
				goto nextLetter;
			}
			currentState = &statesArray[stateNum];
			if (currentState->trans.offset) {
				transitionsArray =
						(HyphenationTrans *)&table->ruleArea[currentState->trans.offset];
				for (k = 0; k < currentState->numTrans; k++) {
					if (transitionsArray[k].ch == ch) {
						stateNum = transitionsArray[k].newState;
						goto stateFound;
					}
				}
			}
			stateNum = currentState->fallbackState;
		}
	stateFound:
		currentState = &statesArray[stateNum];
		if (currentState->hyphenPattern) {
			hyphenPattern = (char *)&table->ruleArea[currentState->hyphenPattern];
			patternOffset = i + 1 - (int)strlen(hyphenPattern);

			/* Need to ensure that we don't overrun hyphens,
			 * in some cases hyphenPattern is longer than the remaining letters,
			 * and if we write out all of it we would have overshot our buffer. */
			limit = MIN((int)strlen(hyphenPattern), wordSize - patternOffset);
			for (k = 0; k < limit; k++) {
				if (hyphens[patternOffset + k] < hyphenPattern[k])
					hyphens[patternOffset + k] = hyphenPattern[k];
			}
		}
	nextLetter:;
	}
	hyphens[wordSize] = 0;
	free(prepWord);
	return 1;
}

static int
doCompTrans(int start, int end, const TranslationTableHeader *table,
		const DisplayTableHeader *displayTable, int *pos, const InString *input,
		OutString *output, int *posMapping, EmphasisInfo *emphasisBuffer,
		const TranslationTableRule **transRule, int *cursorPosition, int *cursorStatus,
		int mode);

static int
for_updatePositions(const widechar *outChars, int inLength, int outLength, int shift,
		int pos, const InString *input, OutString *output, int *posMapping,
		int *cursorPosition, int *cursorStatus) {
	int k;
	if ((output->length + outLength) > output->maxlength ||
			(pos + inLength) > input->length)
		return 0;
	memcpy(&output->chars[output->length], outChars, outLength * CHARSIZE);
	if (!*cursorStatus) {
		if (*cursorPosition >= pos && *cursorPosition < (pos + inLength)) {
			*cursorPosition = output->length;
			*cursorStatus = 1;
		} else if (input->chars[*cursorPosition] == 0 &&
				*cursorPosition == (pos + inLength)) {
			*cursorPosition = output->length + outLength / 2 + 1;
			*cursorStatus = 1;
		}
	} else if (*cursorStatus == 2 && *cursorPosition == pos)
		*cursorPosition = output->length;
	for (k = 0; k < outLength; k++) posMapping[output->length + k] = pos + shift;
	output->length += outLength;
	return 1;
}

static int
syllableBreak(const TranslationTableHeader *table, int pos, const InString *input,
		int transCharslen) {
	int wordStart = 0;
	int wordEnd = 0;
	int wordSize = 0;
	int k = 0;
	char *hyphens = NULL;
	for (wordStart = pos; wordStart >= 0; wordStart--)
		if (!((findCharOrDots(input->chars[wordStart], 0, table))->attributes &
					CTC_Letter)) {
			wordStart++;
			break;
		}
	if (wordStart < 0) wordStart = 0;
	for (wordEnd = pos; wordEnd < input->length; wordEnd++)
		if (!((findCharOrDots(input->chars[wordEnd], 0, table))->attributes &
					CTC_Letter)) {
			wordEnd--;
			break;
		}
	if (wordEnd == input->length) wordEnd--;
	/* At this stage wordStart is the 0 based index of the first letter in the word,
	 * wordEnd is the 0 based index of the last letter in the word.
	 * example: "hello" wordstart=0, wordEnd=4. */
	wordSize = wordEnd - wordStart + 1;
	hyphens = (char *)calloc(wordSize + 1, sizeof(char));
	if (!hyphenateWord(&input->chars[wordStart], wordSize, hyphens, table)) {
		free(hyphens);
		return 0;
	}
	for (k = pos - wordStart + 1; k < (pos - wordStart + transCharslen); k++)
		if (hyphens[k] & 1) {
			free(hyphens);
			return 1;
		}
	free(hyphens);
	return 0;
}

static void
setBefore(const TranslationTableHeader *table, int pos, const InString *input,
		TranslationTableCharacterAttributes *beforeAttributes) {
	widechar before;
	if (pos >= 2 && input->chars[pos - 1] == LOU_ENDSEGMENT)
		before = input->chars[pos - 2];
	else
		before = (pos == 0) ? ' ' : input->chars[pos - 1];
	*beforeAttributes = (findCharOrDots(before, 0, table))->attributes;
}

static void
setAfter(int length, const TranslationTableHeader *table, int pos, const InString *input,
		TranslationTableCharacterAttributes *afterAttributes) {
	widechar after;
	if ((pos + length + 2) < input->length && input->chars[pos + 1] == LOU_ENDSEGMENT)
		after = input->chars[pos + 2];
	else
		after = (pos + length < input->length) ? input->chars[pos + length] : ' ';
	*afterAttributes = (findCharOrDots(after, 0, table))->attributes;
}

static int
brailleIndicatorDefined(TranslationTableOffset offset,
		const TranslationTableHeader *table, const TranslationTableRule **indicRule) {
	if (!offset) return 0;
	*indicRule = (TranslationTableRule *)&table->ruleArea[offset];
	return 1;
}

static int
capsletterDefined(const TranslationTableHeader *table) {
	return table->emphRules[capsRule][letterOffset];
}

static int
validMatch(const TranslationTableHeader *table, int pos, const InString *input,
		formtype *typebuf, const TranslationTableRule *transRule, int transCharslen) {
	/* Analyze the typeform parameter and also check for capitalization */
	TranslationTableCharacter *inputChar;
	TranslationTableCharacter *ruleChar;
	TranslationTableCharacterAttributes prevAttr = 0;
	int k;
	int kk = 0;
	if (!transCharslen) return 0;
	for (k = pos; k < pos + transCharslen; k++) {
		if (input->chars[k] == LOU_ENDSEGMENT) {
			if (k == pos && transCharslen == 1)
				return 1;
			else
				return 0;
		}
		inputChar = findCharOrDots(input->chars[k], 0, table);
		if (k == pos) prevAttr = inputChar->attributes;
		ruleChar = findCharOrDots(transRule->charsdots[kk++], 0, table);
		if ((inputChar->lowercase != ruleChar->lowercase)) return 0;
		if (typebuf != NULL && (typebuf[pos] & CAPSEMPH) == 0 &&
				(typebuf[k] | typebuf[pos]) != typebuf[pos])
			return 0;
		if (inputChar->attributes != CTC_Letter) {
			if (k != (pos + 1) && (prevAttr & CTC_Letter) &&
					(inputChar->attributes & CTC_Letter) &&
					((inputChar->attributes &
							 (CTC_LowerCase | CTC_UpperCase | CTC_Letter)) !=
							(prevAttr & (CTC_LowerCase | CTC_UpperCase | CTC_Letter))))
				return 0;
		}
		prevAttr = inputChar->attributes;
	}
	return 1;
}

static int
insertBrailleIndicators(int finish, const TranslationTableHeader *table, int pos,
		const InString *input, OutString *output, int *posMapping, formtype *typebuf,
		int haveEmphasis, int transOpcode, int prevTransOpcode, int *cursorPosition,
		int *cursorStatus, TranslationTableCharacterAttributes beforeAttributes,
		int *prevType, int *curType, int *prevTypeform, int prevPos) {
	/* Insert braille indicators such as letter, number, etc. */
	typedef enum {
		checkNothing,
		checkBeginTypeform,
		checkEndTypeform,
		checkNumber,
		checkLetter
	} checkThis;
	checkThis checkWhat = checkNothing;
	int ok = 0;
	int k;
	{
		if (pos == prevPos && !finish) return 1;
		if (pos != prevPos) {
			if (haveEmphasis && (typebuf[pos] & EMPHASIS) != *prevTypeform) {
				*prevType = *prevTypeform & EMPHASIS;
				*curType = typebuf[pos] & EMPHASIS;
				checkWhat = checkEndTypeform;
			} else if (!finish)
				checkWhat = checkNothing;
			else
				checkWhat = checkNumber;
		}
		if (finish == 1) checkWhat = checkNumber;
	}
	do {
		const TranslationTableRule *indicRule;
		ok = 0;
		switch (checkWhat) {
		case checkNothing:
			ok = 0;
			break;
		case checkBeginTypeform:
			if (haveEmphasis) {
				ok = 0;
				*curType = 0;
			}
			if (*curType == plain_text) {
				if (!finish)
					checkWhat = checkNothing;
				else
					checkWhat = checkNumber;
			}
			break;
		case checkEndTypeform:
			if (haveEmphasis) {
				ok = 0;
				*prevType = plain_text;
			}
			if (*prevType == plain_text) {
				checkWhat = checkBeginTypeform;
				*prevTypeform = typebuf[pos] & EMPHASIS;
			}
			break;
		case checkNumber:
			if (brailleIndicatorDefined(table->numberSign, table, &indicRule) &&
					checkAttr_safe(input, pos, CTC_Digit, 0, table) &&
					(prevTransOpcode == CTO_ExactDots ||
							!(beforeAttributes & CTC_Digit)) &&
					prevTransOpcode != CTO_MidNum) {
				ok = !table->usesNumericMode;
				checkWhat = checkNothing;
			} else
				checkWhat = checkLetter;
			break;
		case checkLetter:
			if (!brailleIndicatorDefined(table->letterSign, table, &indicRule)) {
				ok = 0;
				checkWhat = checkNothing;
				break;
			}
			if (transOpcode == CTO_Contraction) {
				ok = 1;
				checkWhat = checkNothing;
				break;
			}
			if ((checkAttr_safe(input, pos, CTC_Letter, 0, table) &&
						!(beforeAttributes & CTC_Letter)) &&
					(!checkAttr_safe(input, pos + 1, CTC_Letter, 0, table) ||
							(beforeAttributes & CTC_Digit))) {
				ok = 1;
				if (pos > 0)
					for (k = 0; k < table->noLetsignBeforeCount; k++)
						if (input->chars[pos - 1] == table->noLetsignBefore[k]) {
							ok = 0;
							break;
						}
				for (k = 0; k < table->noLetsignCount; k++)
					if (input->chars[pos] == table->noLetsign[k]) {
						ok = 0;
						break;
					}
				if (pos + 1 < input->length)
					for (k = 0; k < table->noLetsignAfterCount; k++)
						if (input->chars[pos + 1] == table->noLetsignAfter[k]) {
							ok = 0;
							break;
						}
			}
			checkWhat = checkNothing;
			break;

		default:
			ok = 0;
			checkWhat = checkNothing;
			break;
		}
		if (ok && indicRule != NULL) {
			if (!for_updatePositions(&indicRule->charsdots[0], 0, indicRule->dotslen, 0,
						pos, input, output, posMapping, cursorPosition, cursorStatus))
				return 0;
		}
	} while (checkWhat != checkNothing);
	return 1;
}

static int
onlyLettersBehind(const TranslationTableHeader *table, int pos, const InString *input,
		TranslationTableCharacterAttributes beforeAttributes) {
	/* Actually, spaces, then letters */
	int k;
	if (!(beforeAttributes & CTC_Space)) return 0;
	for (k = pos - 2; k >= 0; k--) {
		TranslationTableCharacterAttributes attr =
				(findCharOrDots(input->chars[k], 0, table))->attributes;
		if ((attr & CTC_Space)) continue;
		if ((attr & CTC_Letter))
			return 1;
		else
			return 0;
	}
	return 1;
}

static int
onlyLettersAhead(const TranslationTableHeader *table, int pos, const InString *input,
		int transCharslen, TranslationTableCharacterAttributes afterAttributes) {
	/* Actullly, spaces, then letters */
	int k;
	if (!(afterAttributes & CTC_Space)) return 0;
	for (k = pos + transCharslen + 1; k < input->length; k++) {
		TranslationTableCharacterAttributes attr =
				(findCharOrDots(input->chars[k], 0, table))->attributes;
		if ((attr & CTC_Space)) continue;
		if ((attr & (CTC_Letter | CTC_LitDigit)))
			return 1;
		else
			return 0;
	}
	return 0;
}

static int
noCompbrlAhead(const TranslationTableHeader *table, int pos, int mode,
		const InString *input, int transOpcode, int transCharslen, int cursorPosition) {
	int start = pos + transCharslen;
	int end;
	int p;
	if (start >= input->length) return 1;
	while (start < input->length && checkAttr(input->chars[start], CTC_Space, 0, table))
		start++;
	if (start == input->length ||
			(transOpcode == CTO_JoinableWord &&
					(!checkAttr(input->chars[start], CTC_Letter | CTC_Digit, 0, table) ||
							!checkAttr(input->chars[start - 1], CTC_Space, 0, table))))
		return 1;
	end = start;
	while (end < input->length && !checkAttr(input->chars[end], CTC_Space, 0, table))
		end++;
	if ((mode & (compbrlAtCursor | compbrlLeftCursor)) && cursorPosition >= start &&
			cursorPosition < end)
		return 0;
	/* Look ahead for rules with CTO_CompBrl */
	for (p = start; p < end; p++) {
		int length = input->length - p;
		int tryThis;
		int k;
		for (tryThis = 0; tryThis < 2; tryThis++) {
			TranslationTableOffset ruleOffset = 0;
			TranslationTableRule *testRule;
			switch (tryThis) {
			case 0:
				if (!(length >= 2)) break;
				ruleOffset = table->forRules[_lou_stringHash(&input->chars[p], 1, table)];
				break;
			case 1:
				if (!(length >= 1)) break;
				length = 1;
				ruleOffset = findCharOrDots(input->chars[p], 0, table)->otherRules;
				break;
			}
			while (ruleOffset) {
				const TranslationTableCharacter *character1;
				const TranslationTableCharacter *character2;
				testRule = (TranslationTableRule *)&table->ruleArea[ruleOffset];
				for (k = 0; k < testRule->charslen; k++) {
					character1 = findCharOrDots(testRule->charsdots[k], 0, table);
					character2 = findCharOrDots(input->chars[p + k], 0, table);
					if (character1->lowercase != character2->lowercase) break;
				}
				if (tryThis == 1 || k == testRule->charslen) {
					if (testRule->opcode == CTO_CompBrl ||
							testRule->opcode == CTO_Literal)
						return 0;
				}
				ruleOffset = testRule->charsnext;
			}
		}
	}
	return 1;
}

static int
isRepeatedWord(const TranslationTableHeader *table, int pos, const InString *input,
		int transCharslen, const widechar **repwordStart, int *repwordLength) {
	int start;
	if (pos == 0 || !checkAttr(input->chars[pos - 1], CTC_Letter, 0, table)) return 0;
	if ((pos + transCharslen) >= input->length ||
			!checkAttr(input->chars[pos + transCharslen], CTC_Letter, 0, table))
		return 0;
	for (start = pos - 2;
			start >= 0 && checkAttr(input->chars[start], CTC_Letter, 0, table); start--)
		;
	start++;
	*repwordStart = &input->chars[start];
	*repwordLength = pos - start;
	if (compareChars(*repwordStart, &input->chars[pos + transCharslen], *repwordLength, 0,
				table))
		return 1;
	return 0;
}

static int
checkEmphasisChange(const int skip, int pos, EmphasisInfo *emphasisBuffer,
		const TranslationTableRule *transRule) {
	int i;
	for (i = pos + (skip + 1); i < pos + transRule->charslen; i++)
		if (emphasisBuffer[i].begin || emphasisBuffer[i].end || emphasisBuffer[i].word ||
				emphasisBuffer[i].symbol)
			return 1;
	return 0;
}

static int
inSequence(const TranslationTableHeader *table, int pos, const InString *input,
		const TranslationTableRule *transRule) {
	int i, j, s, match;
	// TODO: all caps words
	// const TranslationTableCharacter *c = NULL;

	/* check before sequence */
	for (i = pos - 1; i >= 0; i--) {
		if (checkAttr(input->chars[i], CTC_SeqBefore, 0, table)) continue;
		if (!(checkAttr(input->chars[i], CTC_Space | CTC_SeqDelimiter, 0, table)))
			return 0;
		break;
	}

	/* check after sequence */
	for (i = pos + transRule->charslen; i < input->length; i++) {
		/* check sequence after patterns */
		if (table->seqPatternsCount) {
			match = 0;
			for (j = i, s = 0; j <= input->length && s < table->seqPatternsCount;
					j++, s++) {
				/* matching */
				if (match == 1) {
					if (table->seqPatterns[s]) {
						if (input->chars[j] == table->seqPatterns[s])
							match = 1;
						else {
							match = -1;
							j = i - 1;
						}
					}

					/* found match */
					else {
						/* pattern at end of input */
						if (j >= input->length) return 1;

						i = j;
						break;
					}
				}

				/* looking for match */
				else if (match == 0) {
					if (table->seqPatterns[s]) {
						if (input->chars[j] == table->seqPatterns[s])
							match = 1;
						else {
							match = -1;
							j = i - 1;
						}
					}
				}

				/* next pattarn */
				else if (match == -1) {
					if (!table->seqPatterns[s]) {
						match = 0;
						j = i - 1;
					}
				}
			}
		}

		if (checkAttr(input->chars[i], CTC_SeqAfter, 0, table)) continue;
		if (!(checkAttr(input->chars[i], CTC_Space | CTC_SeqDelimiter, 0, table)))
			return 0;
		break;
	}

	return 1;
}

static void
for_selectRule(const TranslationTableHeader *table, int pos, OutString output, int mode,
		const InString *input, formtype *typebuf, EmphasisInfo *emphasisBuffer,
		int *transOpcode, int prevTransOpcode, const TranslationTableRule **transRule,
		int *transCharslen, int *passCharDots, widechar const **passInstructions,
		int *passIC, PassRuleMatch *patternMatch, int posIncremented, int cursorPosition,
		const widechar **repwordStart, int *repwordLength, int dontContract,
		int compbrlStart, int compbrlEnd,
		TranslationTableCharacterAttributes beforeAttributes,
		TranslationTableCharacter **curCharDef, TranslationTableRule **groupingRule,
		widechar *groupingOp) {
	/* check for valid Translations. Return value is in transRule. */
	static TranslationTableRule pseudoRule = { 0 };
	int length = ((pos < compbrlStart) ? compbrlStart : input->length) - pos;
	int tryThis;
	int k;
	TranslationTableOffset ruleOffset = 0;
	*curCharDef = findCharOrDots(input->chars[pos], 0, table);
	for (tryThis = 0; tryThis < 3; tryThis++) {
		switch (tryThis) {
		case 0:
			if (!(length >= 2)) break;
			ruleOffset = table->forRules[_lou_stringHash(&input->chars[pos], 1, table)];
			break;
		case 1:
			if (!(length >= 1)) break;
			length = 1;
			ruleOffset = (*curCharDef)->otherRules;
			break;
		case 2: /* No rule found */
			*transRule = &pseudoRule;
			*transOpcode = pseudoRule.opcode = CTO_None;
			*transCharslen = pseudoRule.charslen = 1;
			pseudoRule.charsdots[0] = input->chars[pos];
			pseudoRule.dotslen = 0;
			return;
			break;
		}
		while (ruleOffset) {
			*transRule = (TranslationTableRule *)&table->ruleArea[ruleOffset];
			*transOpcode = (*transRule)->opcode;
			*transCharslen = (*transRule)->charslen;
			if (tryThis == 1 ||
					((*transCharslen <= length) &&
							validMatch(table, pos, input, typebuf, *transRule,
									*transCharslen))) {
				TranslationTableCharacterAttributes afterAttributes;
				/* check before emphasis match */
				if ((*transRule)->before & CTC_EmpMatch) {
					if (emphasisBuffer[pos].begin || emphasisBuffer[pos].end ||
							emphasisBuffer[pos].word || emphasisBuffer[pos].symbol)
						break;
				}

				/* check after emphasis match */
				if ((*transRule)->after & CTC_EmpMatch) {
					if (emphasisBuffer[pos + *transCharslen].begin ||
							emphasisBuffer[pos + *transCharslen].end ||
							emphasisBuffer[pos + *transCharslen].word ||
							emphasisBuffer[pos + *transCharslen].symbol)
						break;
				}

				/* check this rule */
				setAfter(*transCharslen, table, pos, input, &afterAttributes);
				if ((!((*transRule)->after & ~CTC_EmpMatch) ||
							(beforeAttributes & (*transRule)->after)) &&
						(!((*transRule)->before & ~CTC_EmpMatch) ||
								(afterAttributes & (*transRule)->before)))
					switch (*transOpcode) { /* check validity of this Translation */
					case CTO_Space:
					case CTO_Letter:
					case CTO_UpperCase:
					case CTO_LowerCase:
					case CTO_Digit:
					case CTO_LitDigit:
					case CTO_Punctuation:
					case CTO_Math:
					case CTO_Sign:
					case CTO_Hyphen:
					case CTO_Replace:
					case CTO_CompBrl:
					case CTO_Literal:
						return;
					case CTO_Repeated:
						if (dontContract || (mode & noContractions)) break;
						if ((mode & (compbrlAtCursor | compbrlLeftCursor)) &&
								pos >= compbrlStart && pos <= compbrlEnd)
							break;
						return;
					case CTO_RepWord:
						if (dontContract || (mode & noContractions)) break;
						if (isRepeatedWord(table, pos, input, *transCharslen,
									repwordStart, repwordLength))
							return;
						break;
					case CTO_NoCont:
						if (dontContract || (mode & noContractions)) break;
						return;
					case CTO_Syllable:
						*transOpcode = CTO_Always;
					case CTO_Always:
						if (checkEmphasisChange(0, pos, emphasisBuffer, *transRule))
							break;
						if (dontContract || (mode & noContractions)) break;
						return;
					case CTO_ExactDots:
						return;
					case CTO_NoCross:
						if (dontContract || (mode & noContractions)) break;
						if (syllableBreak(table, pos, input, *transCharslen)) break;
						return;
					case CTO_Context:
						// check posIncremented to avoid endless loop
						if (!posIncremented ||
								!passDoTest(table, pos, input, *transOpcode, *transRule,
										passCharDots, passInstructions, passIC,
										patternMatch, groupingRule, groupingOp))
							break;
						return;
					case CTO_LargeSign:
						if (dontContract || (mode & noContractions)) break;
						if (!((beforeAttributes & (CTC_Space | CTC_Punctuation)) ||
									onlyLettersBehind(
											table, pos, input, beforeAttributes)) ||
								!((afterAttributes & CTC_Space) ||
										prevTransOpcode == CTO_LargeSign) ||
								(afterAttributes & CTC_Letter) ||
								!noCompbrlAhead(table, pos, mode, input, *transOpcode,
										*transCharslen, cursorPosition))
							*transOpcode = CTO_Always;
						return;
					case CTO_WholeWord:
						if (dontContract || (mode & noContractions)) break;
						if (checkEmphasisChange(0, pos, emphasisBuffer, *transRule))
							break;
					case CTO_Contraction:
						if (table->usesSequences) {
							if (inSequence(table, pos, input, *transRule)) return;
						} else {
							if ((beforeAttributes & (CTC_Space | CTC_Punctuation)) &&
									(afterAttributes & (CTC_Space | CTC_Punctuation)))
								return;
						}
						break;
					case CTO_PartWord:
						if (dontContract || (mode & noContractions)) break;
						if ((beforeAttributes & CTC_Letter) ||
								(afterAttributes & CTC_Letter))
							return;
						break;
					case CTO_JoinNum:
						if (dontContract || (mode & noContractions)) break;
						if ((beforeAttributes & (CTC_Space | CTC_Punctuation)) &&
								(afterAttributes & CTC_Space) &&
								(output.length + (*transRule)->dotslen <
										output.maxlength)) {
							int p = pos + *transCharslen + 1;
							while (p < input->length) {
								if (!checkAttr(input->chars[p], CTC_Space, 0, table)) {
									if (checkAttr(input->chars[p], CTC_Digit, 0, table))
										return;
									break;
								}
								p++;
							}
						}
						break;
					case CTO_LowWord:
						if (dontContract || (mode & noContractions)) break;
						if ((beforeAttributes & CTC_Space) &&
								(afterAttributes & CTC_Space) &&
								(prevTransOpcode != CTO_JoinableWord))
							return;
						break;
					case CTO_JoinableWord:
						if (dontContract || (mode & noContractions)) break;
						if (beforeAttributes & (CTC_Space | CTC_Punctuation) &&
								onlyLettersAhead(table, pos, input, *transCharslen,
										afterAttributes) &&
								noCompbrlAhead(table, pos, mode, input, *transOpcode,
										*transCharslen, cursorPosition))
							return;
						break;
					case CTO_SuffixableWord:
						if (dontContract || (mode & noContractions)) break;
						if ((beforeAttributes & (CTC_Space | CTC_Punctuation)) &&
								(afterAttributes &
										(CTC_Space | CTC_Letter | CTC_Punctuation)))
							return;
						break;
					case CTO_PrefixableWord:
						if (dontContract || (mode & noContractions)) break;
						if ((beforeAttributes &
									(CTC_Space | CTC_Letter | CTC_Punctuation)) &&
								(afterAttributes & (CTC_Space | CTC_Punctuation)))
							return;
						break;
					case CTO_BegWord:
						if (dontContract || (mode & noContractions)) break;
						if ((beforeAttributes & (CTC_Space | CTC_Punctuation)) &&
								(afterAttributes & CTC_Letter))
							return;
						break;
					case CTO_BegMidWord:
						if (dontContract || (mode & noContractions)) break;
						if ((beforeAttributes &
									(CTC_Letter | CTC_Space | CTC_Punctuation)) &&
								(afterAttributes & CTC_Letter))
							return;
						break;
					case CTO_MidWord:
						if (dontContract || (mode & noContractions)) break;
						if (beforeAttributes & CTC_Letter && afterAttributes & CTC_Letter)
							return;
						break;
					case CTO_MidEndWord:
						if (dontContract || (mode & noContractions)) break;
						if (beforeAttributes & CTC_Letter &&
								afterAttributes &
										(CTC_Letter | CTC_Space | CTC_Punctuation))
							return;
						break;
					case CTO_EndWord:
						if (dontContract || (mode & noContractions)) break;
						if (beforeAttributes & CTC_Letter &&
								afterAttributes & (CTC_Space | CTC_Punctuation))
							return;
						break;
					case CTO_BegNum:
						if (beforeAttributes & (CTC_Space | CTC_Punctuation) &&
								afterAttributes & CTC_Digit)
							return;
						break;
					case CTO_MidNum:
						if (prevTransOpcode != CTO_ExactDots &&
								beforeAttributes & CTC_Digit &&
								afterAttributes & CTC_Digit)
							return;
						break;
					case CTO_EndNum:
						if (beforeAttributes & CTC_Digit &&
								prevTransOpcode != CTO_ExactDots)
							return;
						break;
					case CTO_DecPoint:
						if (!(afterAttributes & CTC_Digit)) break;
						if (beforeAttributes & CTC_Digit) *transOpcode = CTO_MidNum;
						return;
					case CTO_PrePunc:
						if (!checkAttr(input->chars[pos], CTC_Punctuation, 0, table) ||
								(pos > 0 &&
										checkAttr(input->chars[pos - 1], CTC_Letter, 0,
												table)))
							break;
						for (k = pos + *transCharslen; k < input->length; k++) {
							if (checkAttr(input->chars[k], (CTC_Letter | CTC_Digit), 0,
										table))
								return;
							if (checkAttr(input->chars[k], CTC_Space, 0, table)) break;
						}
						break;
					case CTO_PostPunc:
						if (!checkAttr(input->chars[pos], CTC_Punctuation, 0, table) ||
								(pos < (input->length - 1) &&
										checkAttr(input->chars[pos + 1], CTC_Letter, 0,
												table)))
							break;
						for (k = pos; k >= 0; k--) {
							if (checkAttr(input->chars[k], (CTC_Letter | CTC_Digit), 0,
										table))
								return;
							if (checkAttr(input->chars[k], CTC_Space, 0, table)) break;
						}
						break;

					case CTO_Match: {
						widechar *patterns, *pattern;

						if (dontContract || (mode & noContractions)) break;
						if (checkEmphasisChange(0, pos, emphasisBuffer, *transRule))
							break;

						patterns = (widechar *)&table->ruleArea[(*transRule)->patterns];

						/* check before pattern */
						pattern = &patterns[1];
						if (!_lou_pattern_check(
									input->chars, pos - 1, -1, -1, pattern, table))
							break;

						/* check after pattern */
						pattern = &patterns[patterns[0]];
						if (!_lou_pattern_check(input->chars,
									pos + (*transRule)->charslen, input->length, 1,
									pattern, table))
							break;

						return;
					}

					default:
						break;
					}
			}
			/* Done with checking this rule */
			ruleOffset = (*transRule)->charsnext;
		}
	}
}

static int
undefinedCharacter(widechar c, const TranslationTableHeader *table,
		const DisplayTableHeader *displayTable, int pos, const InString *input,
		OutString *output, int *posMapping, int *cursorPosition, int *cursorStatus,
		int mode) {
	/* Display an undefined character in the output buffer */
	if (table->undefined) {
		TranslationTableRule *rule =
				(TranslationTableRule *)&table->ruleArea[table->undefined];

		return for_updatePositions(&rule->charsdots[rule->charslen], rule->charslen,
				rule->dotslen, 0, pos, input, output, posMapping, cursorPosition,
				cursorStatus);
	}

	const char *text = (mode & noUndefined) ? "" : _lou_showString(&c, 1, 1);
	size_t length = strlen(text);
	widechar dots[length];

	for (unsigned int k = 0; k < length; k += 1) {
		widechar c = text[k];
		widechar d = _lou_getDotsForChar(c, displayTable);
		// assume that if d is blank, it's because the display table does not have a
		// mapping to c (not because blank maps to c)
		if (d == LOU_DOTS) d = _lou_charToFallbackDots(c);
		dots[k] = d;
	}

	return for_updatePositions(dots, 1, length, 0, pos, input, output, posMapping,
			cursorPosition, cursorStatus);
}

static int
putCharacter(widechar character, const TranslationTableHeader *table,
		const DisplayTableHeader *displayTable, int pos, const InString *input,
		OutString *output, int *posMapping, int *cursorPosition, int *cursorStatus,
		int mode) {
	/* Insert the dots equivalent of a character into the output buffer */
	const TranslationTableRule *rule = NULL;
	TranslationTableCharacter *chardef = NULL;
	TranslationTableOffset offset;
	widechar d;
	chardef = findCharOrDots(character, 0, table);
	// If capsletter is defined, replace uppercase with lowercase letters. If capsletter
	// is not defined, uppercase letters should be preserved because otherwise case info
	// is lost.
	if ((chardef->attributes & CTC_UpperCase) && capsletterDefined(table))
		chardef = findCharOrDots(chardef->lowercase, 0, table);
	offset = chardef->definitionRule;
	if (offset) {
		rule = (TranslationTableRule *)&table->ruleArea[offset];
		if (rule->dotslen)
			return for_updatePositions(&rule->charsdots[1], 1, rule->dotslen, 0, pos,
					input, output, posMapping, cursorPosition, cursorStatus);
		d = _lou_getDotsForChar(character, displayTable);
		return for_updatePositions(&d, 1, 1, 0, pos, input, output, posMapping,
				cursorPosition, cursorStatus);
	}
	return undefinedCharacter(character, table, displayTable, pos, input, output,
			posMapping, cursorPosition, cursorStatus, mode);
}

static int
putCharacters(const widechar *characters, int count, const TranslationTableHeader *table,
		const DisplayTableHeader *displayTable, int pos, const InString *input,
		OutString *output, int *posMapping, int *cursorPosition, int *cursorStatus,
		int mode) {
	/* Insert the dot equivalents of a series of characters in the output
	 * buffer */
	int k;
	for (k = 0; k < count; k++)
		if (!putCharacter(characters[k], table, displayTable, pos, input, output,
					posMapping, cursorPosition, cursorStatus, mode))
			return 0;
	return 1;
}

// state at the beginning of the current word, used for back-tracking and also for the
// nocont and compbrl rules
typedef struct {
	int inPos;			// begin position of the current word in the input
	int outPos;			// begin position of the current word in the output
	int emphasisInPos;  // position of the next character in the input for which to insert
						// emphasis marks
} LastWord;

static int
doCompbrl(const TranslationTableHeader *table, const DisplayTableHeader *displayTable,
		int *pos, const InString *input, OutString *output, int *posMapping,
		EmphasisInfo *emphasisBuffer, const TranslationTableRule **transRule,
		int *cursorPosition, int *cursorStatus, const LastWord *lastWord,
		int *insertEmphasesFrom, int mode) {
	/* Handle strings containing substrings defined by the compbrl opcode */
	int stringStart, stringEnd;
	if (checkAttr(input->chars[*pos], CTC_Space, 0, table)) return 1;
	if (lastWord->outPos) {
		*pos = lastWord->inPos;
		output->length = lastWord->outPos;
	} else {
		*pos = 0;
		output->length = 0;
	}
	*insertEmphasesFrom = lastWord->emphasisInPos;
	for (stringStart = *pos; stringStart >= 0; stringStart--)
		if (checkAttr(input->chars[stringStart], CTC_Space, 0, table)) break;
	stringStart++;
	for (stringEnd = *pos; stringEnd < input->length; stringEnd++)
		if (checkAttr(input->chars[stringEnd], CTC_Space, 0, table)) break;
	return doCompTrans(stringStart, stringEnd, table, displayTable, pos, input, output,
			posMapping, emphasisBuffer, transRule, cursorPosition, cursorStatus, mode);
}

static int
putCompChar(widechar character, const TranslationTableHeader *table,
		const DisplayTableHeader *displayTable, int pos, const InString *input,
		OutString *output, int *posMapping, int *cursorPosition, int *cursorStatus,
		int mode) {
	/* Insert the dots equivalent of a character into the output buffer */
	widechar d;
	TranslationTableOffset offset = (findCharOrDots(character, 0, table))->definitionRule;
	if (offset) {
		const TranslationTableRule *rule =
				(TranslationTableRule *)&table->ruleArea[offset];
		if (rule->dotslen)
			return for_updatePositions(&rule->charsdots[1], 1, rule->dotslen, 0, pos,
					input, output, posMapping, cursorPosition, cursorStatus);
		d = _lou_getDotsForChar(character, displayTable);
		return for_updatePositions(&d, 1, 1, 0, pos, input, output, posMapping,
				cursorPosition, cursorStatus);
	}
	return undefinedCharacter(character, table, displayTable, pos, input, output,
			posMapping, cursorPosition, cursorStatus, mode);
}

static int
doCompTrans(int start, int end, const TranslationTableHeader *table,
		const DisplayTableHeader *displayTable, int *pos, const InString *input,
		OutString *output, int *posMapping, EmphasisInfo *emphasisBuffer,
		const TranslationTableRule **transRule, int *cursorPosition, int *cursorStatus,
		int mode) {
	const TranslationTableRule *indicRule;
	int k;
	int haveEndsegment = 0;
	if (*cursorStatus != 2 && brailleIndicatorDefined(table->begComp, table, &indicRule))
		if (!for_updatePositions(&indicRule->charsdots[0], 0, indicRule->dotslen, 0, *pos,
					input, output, posMapping, cursorPosition, cursorStatus))
			return 0;
	for (k = start; k < end; k++) {
		TranslationTableOffset compdots = 0;
		/* HACK: computer braille is one-to-one so it
		 * can't have any emphasis indicators.
		 * A better solution is to treat computer braille as its own mode. */
		emphasisBuffer[k] = (EmphasisInfo){ 0 };
		if (input->chars[k] == LOU_ENDSEGMENT) {
			haveEndsegment = 1;
			continue;
		}
		*pos = k;
		if (input->chars[k] < 256) compdots = table->compdotsPattern[input->chars[k]];
		if (compdots != 0) {
			*transRule = (TranslationTableRule *)&table->ruleArea[compdots];
			if (!for_updatePositions(&(*transRule)->charsdots[(*transRule)->charslen],
						(*transRule)->charslen, (*transRule)->dotslen, 0, *pos, input,
						output, posMapping, cursorPosition, cursorStatus))
				return 0;
		} else if (!putCompChar(input->chars[k], table, displayTable, *pos, input, output,
						   posMapping, cursorPosition, cursorStatus, mode))
			return 0;
	}
	if (*cursorStatus != 2 && brailleIndicatorDefined(table->endComp, table, &indicRule))
		if (!for_updatePositions(&indicRule->charsdots[0], 0, indicRule->dotslen, 0, *pos,
					input, output, posMapping, cursorPosition, cursorStatus))
			return 0;
	*pos = end;
	if (haveEndsegment) {
		widechar endSegment = LOU_ENDSEGMENT;
		if (!for_updatePositions(&endSegment, 0, 1, 0, *pos, input, output, posMapping,
					cursorPosition, cursorStatus))
			return 0;
	}
	return 1;
}

static int
doNocont(const TranslationTableHeader *table, int *pos, OutString *output, int mode,
		const InString *input, const LastWord *lastWord, int *dontContract,
		int *insertEmphasesFrom) {
	/* Handle strings containing substrings defined by the nocont opcode */
	if (checkAttr(input->chars[*pos], CTC_Space, 0, table) || *dontContract ||
			(mode & noContractions))
		return 1;
	if (lastWord->outPos) {
		*pos = lastWord->inPos;
		output->length = lastWord->outPos;
	} else {
		*pos = 0;
		output->length = 0;
	}
	*insertEmphasesFrom = lastWord->emphasisInPos;
	*dontContract = 1;
	return 1;
}

static int
markSyllables(const TranslationTableHeader *table, const InString *input,
		formtype *typebuf, int *transOpcode, const TranslationTableRule **transRule,
		int *transCharslen) {
	int pos;
	int k;
	int currentMark = 0;
	int const syllable_marks[] = { SYLLABLE_MARKER_1, SYLLABLE_MARKER_2 };
	int syllable_mark_selector = 0;

	if (typebuf == NULL || !table->syllables) return 1;
	pos = 0;
	while (pos < input->length) { /* the main multipass translation loop */
		int length = input->length - pos;
		int tryThis = 0;
		while (tryThis < 3) {
			TranslationTableOffset ruleOffset = 0;
			switch (tryThis) {
			case 0:
				if (!(length >= 2)) break;
				// memory overflow when pos == input->length - 1
				ruleOffset =
						table->forRules[_lou_stringHash(&input->chars[pos], 1, table)];
				break;
			case 1:
				if (!(length >= 1)) break;
				length = 1;
				ruleOffset = findCharOrDots(input->chars[pos], 0, table)->otherRules;
				break;
			case 2: /* No rule found */
				*transOpcode = CTO_Always;
				ruleOffset = 0;
				break;
			}
			while (ruleOffset) {
				*transRule = (TranslationTableRule *)&table->ruleArea[ruleOffset];
				*transOpcode = (*transRule)->opcode;
				*transCharslen = (*transRule)->charslen;
				if (tryThis == 1 ||
						(*transCharslen <= length &&
								compareChars(&(*transRule)->charsdots[0],
										&input->chars[pos], *transCharslen, 0, table))) {
					if (*transOpcode == CTO_Syllable) {